Hi All,

Beta testing has been going well so I've made the latest version of my Room EQ Wizard software available for download from http://homepage.ntlworld.com/john.mulcahy/roomeq/index.html

Room EQ Wizard is a Java application for measuring and correcting room resonances. It includes tools for generating test signals, measuring SPL and frequency responses and automatically adjusting the settings of parametric equalisers to cancel the effects of room modes. It was initially written to help with the setup of the parametric TMREQ filters on TAG McLaren Audio AV32RDP and AV192R, but the latest version also supports the BFD Pro parametric filters (with correct modelling of the effect of the bandwidth control, which is quite different to the description in the BFD manual :)).

Measurement is stepped sine, with local loopback for soundcard response compensation and DFT to isolate the measurement frequency. Log swept sine measurement and impulse response extraction are on the dev list but some way off. The Wizard can also import measurements in the ETF export formats or from text files in a basic comma-delimited format - details are in the help files, which can be browsed online at the site and are included in the program.

The sig gen provides sine waves to 0.1Hz precision, sine sweeps (linear and logarithmic), square waves and various pink noise signals including "full" range (pink spectrum down to just below 10Hz), speaker cal, sub cal and custom filtered to suit your needs.

Note that the app requires V5.0 or later of Sun's Java Runtime Environment (JRE) to be installed, available from http://java.sun.com/j2se/downloads.html

Hope it proves useful.

Best regards,

Very cool! Too bad I just spent too much for an analog parametric, which I was going to use with TrueRTA. At least I haven't bought the latter yet :)

So what's the learning curve and user friendliness EQ Wizard? Those were the reasons for my initial choices.

How long would it take to do the correction for a sub? Does the swept sine measurement mean it takes awhile?

Thanks

Oh, too cool!

Worlds colliding, worlds colliding! (HT and computers).

Hi Noah,

Measurement is stepped rather than swept, and it does indeed take a while as the software waits for the reading to stabilise (SPL display turns green when it is stable) before making a measurement and moving on the the next frequency. How long it takes to settle depends on the room, the frequency and the background noise, but probably averages about 1-2s per measurement so around 2-3 minutes to measure 20-120Hz in 1Hz steps. A bit tedious I know, but fairly reliable :)

You still need an equaliser, this app just helps work out how to set up the filters, so your money certainly wasn't wasted. It's fairly easy to use, and you get immediate visual feedback of the effect your filter tweaks will have on the response so its a quick way to manually tweak if you don't like the app's efforts at choosing filter settings. Easiest is to play around with it and see how you get on - just looking at how the individual filter gain and Q adjustments affect the overall filter frequency response can be useful.

Happy to answer questions as they arise.

Hi John,

"It's fairly easy to use, and you get immediate visual feedback of the effect your filter tweaks will have on the response "

Immediate meaning the 2-3 minutes (which isn't too bad), or do you mean dialing down a peak once they're located? But then I don't know what the BW needs to be.

Or am I not getting something?

Thanks

I mean you get to see on the graph display the frequency response of the gain and bandwidth you are setting and how that will affect the measured response after you apply those corrections in your equaliser, you can easily see what bandwidth you need to set by adjusting the bandwidth setting for the relevant filter until the filter's response mirrors the measured response around that peaks's frequency - to make this easier the Wizard has a check box to draw the filter response inverted, so that it overlays the measurement. Have a look at this help page for more: http://homepage.ntlworld.com/john.mulcahy/roomeq/wizardhelp/help_en-GB/html/filteradjustment.html

Regards,

John,

That's excellent! I'll check it out.

Thanks very much.

did you write the curve fitting piece of the app that originally was on the TAG site or are you layering your work on top of it?

either way is great, but i always thought the underlying curve fitting piece was super slick.

Originally posted by qxlxp
did you write the curve fitting piece of the app that originally was on the TAG site or are you layering your work on top of it?

either way is great, but i always thought the underlying curve fitting piece was super slick.
Yes, I wrote the original app as an interesting way to learn Java. Glad you liked it :)

Very nice. Out of curiosity, have you compared the results to other programs?

Originally posted by Chu Gai
Very nice. Out of curiosity, have you compared the results to other programs?
Do you mean the measurement results? Measurements have been compared with data acquired by ETF, you should get a similar result to the t=0ms slice in an ETF low frequency response if you use a gate time of 300 - 600ms (in smaller rooms 600ms may be too long, the ETF measurement starts to look noisy and the level shifts when gate time is set too long). Since the software is measuring values you can (more or less) read off your SPL meter that also provides a handy check - differences are that the wizard is applying a DFT to exclude frequencies other than the current measurement frequency (noise, harmonic distortion) and if you have told it you have a C-weighted SPL meter (the default setting) it applies the inverse of the C weighting curve to get the true SPL.

wow, this is fantastic! i've been looking for a program like this for ages.

Few updates made, details in the revision history on the web site, app now at V3.17. If anyone uses Behringer's DSP1100 software to look at filter responses, be warned: the bandwidths it uses for its response curve are sqrt(2) narrower than they should be and it appears to be modelled on analog prototype responses rather than actual digital biquads so it gets progressively more wrong compared to the actual unit as the frequency increases.

Well done JohnPM. Have one on me :)

http://www.bottledbeer.co.uk/photos/greatbritishbeer.jpg

John,

Is there anyway to import ETF data?

I understand you're traveling so I patiently wait for your reply. :)

Thanks
Thomas

Hi Thomas,

I haven't left yet, flight is about 14h from now. Yes, you can import data saved from ETF using ETF's export features, Low Freq response data is most useful. More here: http://homepage.ntlworld.com/john.mulcahy/roomeq/wizardhelp/help_en-GB/html/dataimport.html

Regards,

I can confirm there is no problem importing files from ETF. Just make sure you have the graph that you want to export from ETF visible in the ETF screen; the whole file is not exported in one shot.

John's progam is outstanding to use in conjunction with ETF. I would almost consider it mandatory. I wouldn't doubt that the program is excellent on its own but I haven't tried it yet.

This link takes you to a wonderful guide on setting up subs and speakers with ETF:

http://www.avforums.com/forums/misc...EQCaseStudy.pdf

John, in your detailed pdf you make no mention of phase. For those that don't know ETF can produce phase graphs for your sub and speaker and phase can be adjusted by using time delay in your processor or on your sub. I had read that it is important for the phases of the sub and speaker to be fairly close at the crossover point.

What are your views on this? What would be the procedure for aligning phases using ETF?

Many thanks,
George

Hi George,

I haven't used phase responses in setting up the system, if there is a problem with relative phase between sub and mains it will show up in frequency response particuarly in the crossover region between the two, just make some measurements of each main speaker with bass redirection active (assuming the mains are bass limited, otherwise a bit moot) so the sub is contributing and look at the overall measured response. A couple of measurements with the sub phase normal and reversed will give a quick indication which is working best, could then tweak the phase from there. It can be pretty tricky to arrive at a setting that works for all of L, C and R even if your sub is symmetrically placed (which it usually isn't). Have to pick the best compromise.

"John's progam is outstanding to use in conjunction with ETF. I would almost consider it mandatory."

John's program, or ETF?

Thanks

If you have ETF I would consider it mandatory to use John's program in conjunction with ETF.
All the best,
George

What can this software do that ETF cannot?

If I am reading the requirements correctly, my soundcard would need right and left (i.e. two) lines in and out to use this software? Neither my onboard sound, nor my soundcard (Chaintech AV-710) have two lines in or out, but might I be able to use the single lines in and out from both? Or could I split a single line in and a single line out?

Originally posted by kromkamp
What can this software do that ETF cannot?
Work out equaliser settings for you and show you the effect of equaliser settings of your own. There is lots that ETF can do that this software cannot, mind, as stated above the wizard is using a basic stepped sine measurement. If you would like to know more about what the Room EQ Wizard can and cannot do the online help pages are a good place to start.

Regards,

Originally posted by sensibull
If I am reading the requirements correctly, my soundcard would need right and left (i.e. two) lines in and out to use this software? Neither my onboard sound, nor my soundcard (Chaintech AV-710) have two lines in or out, but might I be able to use the single lines in and out from both? Or could I split a single line in and a single line out?
Hi,

Your soundcard has all the I/O you need and more. The right and left are the two halves of a stereo pair, your soundcards inputs and outputs are stereo. You can just use the card's line in and the line out that is normally assigned to left&right channels (as your card actually has 4 stereo line outs). I think that card has connectors that take jack plugs, so you will need a pair of jack plug - to - dual RCA leads and some RCA socket-RCA socket adaptors, there is some more info in the help. The connection requirements are the same as ETF, for example.

Regards,

Thanks John. I thought that might be the case (that's what I meant by "split"), but just wanted to check for sure. All your hard work is much appreciated, as a newbie to parametric EQ. Will certainly take some of the monotony and time expenditure out of the task ahead...

Cheers :-)

It's good to read this. I downloaded and played with the app last night to compare with my ETF measurements. I must have not noticed the default C weighting settings as the low end (especially under 40Hz) seemed distorted compared to my ETF measurements. I'll change this and try again.

Other than needing to find a setting that is faster for the sweeps (the default 10 minute one was painful :) ), the process was pretty good. Great work and thanks for making it available.

Once I check the C-weighting thing, I will follow up with my posts on this on other forums.

Other than needing to find a setting that is faster for the sweeps (the default 10 minute one was painful ), the process was pretty good. Great work and thanks for making it available.

This is very cool software! I was able to spend yesterday afternoon playing with it, and thus far, I would have to say that its a pretty short learning curve, and that coming from someone with little previous experience with ETF, SpectraPLUS etc.

I do have some questions, but I am going to withhold those until I have a chance to run anther session with the software...don't want to expose too much ignorance!:D

When I did my sub measurements, I choose to run sines in 1 Hz increments from 20 to 120 Hz. In my room, it surely did not take 10 minutes.

What did I do that was different from your procedures Brad?

BGL

The only difference I can think of is that I did a 20-200Hz sweep which would take almost twice as long as a 20-120Hz one. If I can tweak a sweep to under one minute or preferably under 30 seconds, that would be good for me (my ETF measurements take 5 seconds, but aren't 1Hz intervals).

Hi Brad,

If you do a Low Frequency response measurement on ETF with the long measurement time selected and export the data you should end up with either 0.67Hz or 0.37Hz (approx) resolution in the exported data - you need to use a higher gate time to get the greater resolution, 600ms or above, though this may be too high for smaller rooms or measurements with poor signal-to-noise. Would pretty much always expect to be using at least 300ms, however.

Room EQ Wizard will eventually have the option for logarithmically swept sine measurements, probably a 5 or a 10s sweep, but that is some months away.

Regards,

Hi all,

I've been waiting a long time for something like this.

I have also run into two problems while using it. I am convinced it's not a problem with the program but rather a problem on my end. First, the computer cannot seem to locate the loopback on the left channel. I have triple-checked the connections and am so far at a loss. Second, Although I can generate test tones that are picked up at the SPL meter I cannot seem to calibrate the program. It says that the input level is set to maximum (1.0) and that I should reduce the settings on the SPL meter. My concern is that the SPL meter is already set pretty low and I am unable to hear anything from the right speaker indicating that there is calibrating even going on.

Any ideas?

-Dave

Originally posted by sensibull
If I am reading the requirements correctly, my soundcard would need right and left (i.e. two) lines in and out to use this software? Neither my onboard sound, nor my soundcard (Chaintech AV-710) have two lines in or out, but might I be able to use the single lines in and out from both? Or could I split a single line in and a single line out?

What state is your 710 in , Mine is flashed with the newer firmware, and does not do analog, only didital.

Thanks

Thanks W4ZOO, but I never flashed mine with the Prodigy firmware, though up 'til now I have only been using the optical out (I get bit-perfect with 3.10a drivers). I was just being an idiot in reading the Room EQ instructions, and not realizing from the get go that I could split the lines in and out into right and left channels.

Extremely neat and self eplanatory program.
I must test this on my Denon AVR-3805 that also has a built in self parametrizing / equalising system using the Denon meassure microphone as in pic below
http://www.denon.de/site/datadir/gr/1_gr_av_AVR3805_inside5.jpg

Then I will make some test from the PC and see if the result adds up or if it is way off - if it is off i can allways tune the AVr-3805 with its 8 band equalizer to get it closer to good. A Pity one aint got all those functionalities that are built into this program on the AVR-3805 though :(

I will start of using the Denon microphone as the base (since it is desgined for the Denon AVR-3805) - will that work fine even if its not a SPL meter it is a meassuring microphone - same thing/will work or not?

Thanks a bunch for your work and for sharing it with us :)

Regards
Boogieman

Originally posted by DavidDahl
Hi all,

I've been waiting a long time for something like this.

I have also run into two problems while using it. I am convinced it's not a problem with the program but rather a problem on my end. First, the computer cannot seem to locate the loopback on the left channel. I have triple-checked the connections and am so far at a loss. Second, Although I can generate test tones that are picked up at the SPL meter I cannot seem to calibrate the program. It says that the input level is set to maximum (1.0) and that I should reduce the settings on the SPL meter. My concern is that the SPL meter is already set pretty low and I am unable to hear anything from the right speaker indicating that there is calibrating even going on.

Any ideas?

-Dave
Hi Dave,

What are the Input Device and Input set to? What soundcard are you using and how are the connectors labelled? From the description likely problem is the input you are connected to is not the one that has been selected, or there may be some other input mixer control that is preventing the wizard picking up the signal. Have you already used this setup with other software such as ETF?

Another possibility is that there is a problem on the signal generation side, again selected vs connected output, or some mixer control preventing the signal getting out. What are the outout device and output set to?

A way to check whether the external connections are the problem can be to select "stereo mix" as the input (assuming that appears in the list), that is a local loopback on the soundcard, with that selected you should be able to see the effect of generating test tones on the various readings in the spl panel if the output side is set up correctly.

Could post some screen captures of the windows mixer settings for record and replay after you have started the wizard up, get to the mixer via the Sounds and Audio Devices item in the control panel and select "Advanced" in the "device volume" box, make one screen cap with properties set to recording and another with properties set to playback and either post them here or email them to me.

Regards,

Hi Boogieman,

I'm not familiar with the Denon mic so not sure if it has mic level or line level output. If mic level it will be a problem, really need line level signals.

Regards,

John,

I finally had time to sit and play with your software. Congratulations on a very easy to use interface. I'm still scouring the help to make sure I'm doing everything correctly, but so far I'm really enjoying it. With your hopes to add midi-connection capability to the PEQ, this is just going to get better.

Thanks!

Reading this thread it and the instructions on soundcard connection, it sounds like the connections should be the same as ETF. I can get ETF to work, but wasn't able to get the RoomEQ Wizard to work. I do get a feedback loop when trying to calibrate. If I have it set to track the SPL, if I speak into the test microphone, it is output through my system. I am using a Creative Soundblaster USB soundcard, Behringer ECM8000 mic and 802 preamp/phantom power. Room EQ settings are input = LINE-IN; output = SPEAKER; windows mixer only has LINE enabled for record and PLAYBACK and WAVE enabled for playback. These settingswork for ETF but for the Wizard, I'm not getting a tone out but if I set the prepro level high enough I get a feedback loop - the mic input is going straight to the output...???
Anyone have any suggestions...???

Thanks, Wyatt

Hi Wyatt,

Only wave should be unmuted for playback, if you unmute Line as well you will get feedback as described. The Wizard mutes all playback sources except for wave when it starts up, did it not do this on your setup?

Regards,

OK, fundamental question....when doing the initial calibration is the RoomEQ generating the test tone or are you using the internal tone on your processor??? If RoomEQ should be generating the tone, what type should it be - white noise???

I get output if I turn on the signal generator section....

Thanks,
Wyatt

When calibrating the SPL meter, use the internal tone on your processor and make sure the cal level that has been set in the Wizard's SPL meter panel is the same as the reading on your handheld SPL meter. That sets the reference level for input signals, which is then used when setting the levels for measurement.

Regards,

John,

Which of the many downloads available here

http://java.sun.com/j2se/1.5.0/download.jsp

should I choose?

Do I correctly presume that I don't need to know Java and it's use is transparent to the user?

Thanks

Correct, no need to know anything about Java. You need to download "JRE 5.0 Update 2".

Regards,

noah katz
It's the "run time" version - JRE which just lets the java programs run. No need to "know" anything at all. Just updating when there are major security updates available is probably not a bad idea.

Thanks!

Getting my feet wet here...

The input device selection window on the lower left shows Avance AV97 Audio, which is what my laptop has.

Does this mean it's suitable (full duplex, etc) for Room EQ Wizard or is it just reporting its presence?

I haven't been able to get my laptop to tell me its sound card capabilities.

Thanks

The wizard lists any soundcards it finds that report they support the required formats. Doesn't necessarily mean it is suitable but most are nowadays.

Thanks, John, I'll give it a try and see.

Forgive me for asking another idiot newbie question, but in your help file you specify to use the AV processor's speaker trim to adjust L, R, or C speaker output until it matches target Cal Level dB, but to adjust the volume during the Measurement Level setting.

While I understand the difference between trim and volume, I'm a little confused about how to hit 75dB using the trim adjustments alone (which on my HK 635, only go from -10db to +10db). Is the assumption that I have previously calibrated my system and already know what volume level should hit 75dB on the test tone, and that I set the volume to that before I begin the SPL calibration, using the trim settings only to fine tune it? Or does it matter whether I use volume alone, or volume + trim during the SPL calibration phase?

While I'm already embarassing myself, I have two further clarification questions:
1. The point of the two calibrations is to match the RadioShack SPL with the Room EQ SPL (as dictated by recording level), and then to determine at what volume your soundcard creates a 75db signal, correct?
2. Does it skew the measurements if you use a lower target SPL? I have twin babies and never run my system very loud. I ran a preliminary test last night and even 80dB left my ears ringing a bit...

Originally posted by sensibull
Forgive me for asking another idiot newbie question, but in your help file you specify to use the AV processor's speaker trim to adjust L, R, or C speaker output until it matches target Cal Level dB, but to adjust the volume during the Measurement Level setting.

While I understand the difference between trim and volume, I'm a little confused about how to hit 75dB using the trim adjustments alone (which on my HK 635, only go from -10db to +10db). Is the assumption that I have previously calibrated my system and already know what volume level should hit 75dB on the test tone, and that I set the volume to that before I begin the SPL calibration, using the trim settings only to fine tune it? Or does it matter whether I use volume alone, or volume + trim during the SPL calibration phase?

While I'm already embarassing myself, I have two further clarification questions:
1. The point of the two calibrations is to match the RadioShack SPL with the Room EQ SPL (as dictated by recording level), and then to determine at what volume your soundcard creates a 75db signal, correct?
2. Does it skew the measurements if you use a lower target SPL? I have twin babies and never run my system very loud. I ran a preliminary test last night and even 80dB left my ears ringing a bit...
Hi,

I'll tweak the help text for those sections to make it a little more clear, it is not the most intuitive part of the process. :)

The input calibration is used to give the Wizard an absolute SPL reference. A signal needs to be generated at a known level (as read off your SPL meter) and the Wizard needs to be told what that level is (via the Cal Level dB control). It will then adjust the input level control so that the signal captured by the soundcard has enough headroom (aims for 18dB) and adjust its own SPL meter reading to show the same figure as your SPL meter.

The process is based on the normal calibration routine for your AV processor, as carried out when you set up your system, using the internal calibration tones on your processor. The level of the calibration tone is usually not affected by your volume control, only by the speaker trims (the processor typically applies an internal volume setting that is appropriate for the calibration process). If the volume control does affect the calibration tone level, set it to whatever level is recommended for calibrating your processor. You will only need to alter the speaker trim if the level on your SPL meter is not at your calibration level - alternatively, just change the Cal Level dB setting on the wizard so that it is the same as the reading on your SPL meter.

Usually the cal level is 75dB, you can use a lower figure as long as you can trim your cal tone output to hit that lower level during calibration.

Setting the measurement level, which does involve adjusting your processor's volume control, is to take account of the output level of the soundcard and the input sensitivity of your processor and set a level for taking measurements that is high enough for accurate results but has enough headroom to allow for the effects of resonances. The target it aims for is to achieve the same level as used for the input calibration. First it sets its own output sig gen level by measuring the level on the local loopback connection, then it generates a speaker cal signal and waits for you to adjust the AV processor's volume to give the desired level measured off the SPL meter.

Regarding keeping the noise down, if you select the "Limit SPL when measuring" box below the Set Target Level button it will automatically reduce the signal levels when resonances are encountered to keep the overall level close to the target level for the speaker - it cuts the excess above the target to 1/4 of what it would otherwise be, e.g. a 12dB resonance would only give a level 3dB above the target. The graph will show the figures that would have resulted without the limiting, i.e. it would still show a 12dB peak. This does extend the measurement time though, as the limiter needs additional settling time. You can even set an artificially low Target Level during the measurement (e.g. 65dB) to reduce levels even further, as long as you remember to set the correct target level before searching for peaks or adjusting filters. However I have just found a bug in the measurement figures that are recorded when the limiter is on :o so wait for tonight's release if you want to use that feature.

Regards,

V3.19 has now been uploaded to the website. The main addition this time is Midi comms
for setting up BFD Pro filters, though you will also need a Midi interface. Details of
the Midi comms are in a new help file here:
http://homepage.ntlworld.com/john.mulcahy/roomeq/wizardhelp/help_en-GB/html/bfdcomms.html

There have been various other updates in the help text to add references to BFD Pro where
appropriate.

There is also a link to download a zip file of the jars used by the wizard so the app
can be tried on Linux or Mac OS X platforms - I haven't tested on other platforms, however.

Summary of V3.19 changes:

BFD Pro can now be set up over a Midi connection
Added a "Comms" menu, put COM port and Midi input and output port selection in there
Tweaks to helptext for unit.html, equaliser.html, soundcard.html, welcome.html (link to download site for J2SE V5.0 for Max OS X), avpcomms.html
Tweaks to the English help text for makingmeasurements.html, inputcal.html, filteradjustment.html
New helptext files added: comms.html, bfdcomms.html
"Generate Soundcard Debug File" added to the Soundcard menu to generate a text file with debug info following reports of problems with some multi-channel soundcards (RME 9632)
When the equaliser is BFD Pro the Aux tab is renamed to Sub2
Fixed a bug that caused wrong measurement values to be recorded when the SPL limiter was active


Regards

Are the widely published correction values for the Radio Shack meter and Room EQ's C-Weighting Compensation adjustments essentially the same thing, or totally unrelated?

In other words, if I am importing measurements manually (with a comma delimited txt file) that were made with a C-weighted Radio Shack meter and that have had the correction values already applied, do I still need to apply the C Weighting Compensation, or would that be redundant?

Also, can Room EQ suggest filters for correcting dips or valleys as well as peaks? I know boosting signals is generally frowned upon, but it may be called for in my situation...

Thanks again for you continued assistance...

Originally posted by sensibull
Are the widely published correction values for the Radio Shack meter and Room EQ's C-Weighting Compensation adjustments essentially the same thing, or totally unrelated?

In other words, if I am importing measurements manually (with a comma delimited txt file) that were made with a C-weighted Radio Shack meter and that have had the correction values already applied, do I still need to apply the C Weighting Compensation, or would that be redundant?

Also, can Room EQ suggest filters for correcting dips or valleys as well as peaks? I know boosting signals is generally frowned upon, but it may be called for in my situation...

Thanks again for you continued assistance...
The various RS meter correction values tend to be a mixture of the required C-weighting compensation and corrections for individual meters that may or may not be valid for others.

If you are importing data to which you have already applied corrections in the range below 200Hz, safest is to say "no" when the Wizard asks if you want C-weighting compensation applied. Alternatively, if you know the corrections that were previously applied you could manually reverse them and then import with C-weighting compensation, but probably not worth bothering.

The automatic EQ routines only search for peaks in the response, and only apply filters to correct those peaks. You can manually apply filters to boost portions of the response, but obviously don't waste any time trying to fill notches/nulls.

Regards,

Thanks John!

I tried out your new version with my BFD, Audigy 2 ZS and RS
analog meter. Had to buy a couple of MIDI cables just to try out the
sending of the filters to the BFD. It all worked! I admit to adding one
more notch and a couple of mild peaking filters as manuals after the
auto optimize initial filter set.

I especially noticed that the left channel feedback calibration really
made a difference in the 10 to 20 Hz measurements with the Creative
Labs Audigy 2 ZS. Just like the similar significant results I got when I
ran TrueRTA's soundcard calibration.

BTW I checked my equalized results with TrueRTA and they seemed to
jive. I probably could have skipped paying $100 for TrueRTA if I had
waited for your Room Eq Wizard software.

Glad to hear all went well. Manual tweaking of filters is very much encouraged, the auto setting is a convenience feature to deal with the main problems, further optimisation is then in the hands of the user.

The Wizard has now reached V3.20, changes this time:

Distortion level displays have been added to the SPL Meter panel, showing the levels of the 2nd and 3rd harmonics of the test frequency. There are separate displays for the local loopback from the soundcard and for the external input - bear in mind that high noise levels will show up as correspondingly high 2nd/3rd harmonic levels. The SPL Meter panel has been rearranged to add the new displays, also rearranged the signal generator panel a little to keep its appearance consistent
After rearranging the SPL Meter panel there was some empty space on the RHS, so I decided to code up some level meters to go there. Don't tell you anything you didn't already know from the numeric peak and rms displays, but does add some colour :D
http://homepage.ntlworld.com/john.mulcahy/roomeq/splpanel.jpg
I've tweaked the SPL limiter loop to reduce settling time
Bug fix: SPL calibration value was not being saved if the soundcard had no input volume control (e.g. some USB cards)
Added help index files in English and Dutch help directories and links to the index and the home page at the bottom of each help file - mainly to provide links back to somewhere central for the web version of the help pages
Added notes on optimiser Q/BW limits in filter adjustment help

Full change history is on the web site here: Revision History (http://homepage.ntlworld.com/john.mulcahy/roomeq/changehistory.html)

Next up: log swept sine measurement

Regards,

John,

I've been too distracted with work and other projects to spend any more time with Room EQ Wizard, but I wanted to tell you much I like the UI - brimming with all the useful info and very eye-appealing - it makes me feel good just to look at it :)

And now even better with meters - great!

Thanks! :)

Good news for anyone who has sat through a 10 minute measurement sweep with the Wizard: now it can all be done in about 10 seconds :)

The default for measurement is now to use a logarithmically swept sine signal, takes 4 seconds for the sweep, about 1.5s of pre and post sampling and, depending on your PC, a few seconds to carry out the FFTs and related processing to generate the responses (my ref for that is a 1.8GHz Pentium M 745 laptop with 1GB RAM). With the extra storage structures needed for the sweep processing the Wizard now peaks at about 55MB RAM usage, so 256MB is the recommended minimum for the PC.

The sweep measurement opens the door for various other interesting features since I now have an impulse response to play with, plenty more to follow :D

Other changes this time:

The graphical meters on the SPL panel looked so nice I added a set to the signal generator panel as well
It is now possible to select the generic OS soundcard drivers (on Windows platforms they are Primary Sound Capture Driver, Primary Sound Driver and Java Sound Audio Engine). I do not recommend using these as they typically do not offer the controls the Wizard needs to select inputs and adjust levels, but if the soundcard's own drivers do not fit the bill these offer a fallback. It will be necessary to manually adjust levels via the soundcard's mixer if using these drivers.
The control panels now scale up as the app is resized, may help to get better display on other platforms where the fonts don't quite correspond to the Windows versions.
The action of the frequency scroll bar has been revised to make it easier to use at small frequency spans
Filter optimisation for BFD Pro has been improved

If anyone is concerned that the measurement results for the sweep might not match those with the stepped sine, don't be: they are identical. The sweep has the benefit of being less likely to suffer from clipping when passing through resonances, but rest assured that it still captures the full effect of the resonance and presents the frequency response accordingly. And it is SO much faster...

There has been some debate about whether room resonances, and in particular the extended decay time they cause, can be fixed with parametric EQ filters. A picture is worth a thousand words, so here is a screen shot I made this afternoon whilst verifying the Wizard's new log swept sine measurements. It shows ETF5 waterfall plots of the subwoofer response in my lounge, with the top plot having a set of correction filters applied via a BFD Pro and the bottom showing the behaviour without any correction. Both plots have the same vertical range, the time axis covers 500ms.

http://homepage.ntlworld.com/john.mulcahy/roomeq/ETF5BeforeAndAfterSmall.jpg

As you can see, without EQ there are some pretty severe resonances between 20 and 30Hz which hang around for quite a while, with another set of problem modes starting at around 47Hz. The EQ filters clean them up very nicely.

For the record, I don't actually use a BFD Pro in my system (my AV processor has parametric filters built in) but I thought it best to use something more people will be familiar with, so spent half an hour or so coming up with some settings for the BFD. The results could be improved further, but I was hungry :D

Thanks for posting those plots John! Very interesting.

Cary

John, I saw the exact same kind of thing when I went back a couple of days ago and measured my sub's response with ETF after using the RoomEQ Wizard. After checking LF response with ETF, I popped up the waterfall plot and was astounded at the reduction in resonances. I quite honestly thought I was setting something in ETF incorrectly, but now I know I wasn't. Thanks for all the hard work you've put into this. It's an amazing tool!

Hello John,

The program has come a long way and is looking very fine indeed.

As a comparison, in the graph where you show the resonance reductions, could you please tell us the frequencies, Q, and amplitudes for the filters you applied?

Many thanks,
George

Hi George,

Here you go:

http://homepage.ntlworld.com/john.mulcahy/roomeq/images/sub_filters.jpg
Regards,

Thank you!

I was curious as to how you were manipulating the gains of the filters. I see that you have chosen to cut all peaks with no boosts. I assume by doing so that you must increase the overall gain of the signal to get back to "reference," 75dB seems to be the usual target.

The other alternative would be to boost some of the valleys and end up using less overall gain.

Is one method preferred over the other? I imagine it could be a bit of juggling act, too many and too deep EQ cuts and the overall volume might have to be set too high causing amplifier clipping, too many or too much EQ boosts and the EQ might clip.

Thanks again,
George

Because the low frequency modes were so strong the peak level after correction was 15dB lower than before, but the shift in the sub cal level was much less because the resonances contribute only part of the overall level, albeit a big part in bad cases.

The worst dips after correction were only 5 or 6dB and quite narrow, just a few Hz. I don't find it useful to try and fill anything beyond a broad dip, if the bandwidth of the dip is less than about 1/6th of an octave I'd leave it alone.

You have to cut the peaks if you want to get rid of the overhang in the time domain, if you were to boost the other parts of the signal instead (the areas between the resonances) you just end up with extended decay times everywhere, which would sound awful.

John - this is an excellent little program and just what I have been looking for for the last few years since I got my BFD. I have been able to significantly improve on the manual method of setting the filters - Thanks!

I've got 2 ideas for enhancements which would be really great to see in new versions:
- add a way to upload multiple seating position recordings and have the tool calculate a best set of filters for say the three primary seats instead of one only (using some sort of interpolation method) - this would help to minimize the one biggest issue of using EQ to resolve resonance issues. Room modes that are excited by front to back wave interference would be best suited for this enhancement as the peaks and valleys would be at more or less at the same frequency for seats in the same row. Though now thinking about it, I suppose I could combine the three input curves via a rms formula before uploading - hmmmmm - gonna try that tonight!
- add the ability to calculate filters that are in the LFE channel yet above the cross over frequency (i.e. input xover freq and slope, and have a "stronger" filter if peak is greater than xover). This would allow for more accurate setting of filters which occur in 80-120 hz range which you would normally not be able to control very well using a BFD and a cross over of 80Hz. I have had some success in doing this by hand but would be really cool to see the tool do it automatically.

At any rate - great tool and thanks!

Dave

Originally posted by DaveBoswell

- add the ability to calculate filters that are in the LFE channel yet above the cross over frequency (i.e. input xover freq and slope, and have a "stronger" filter if peak is greater than xover). This would allow for more accurate setting of filters which occur in 80-120 hz range which you would normally not be able to control very well using a BFD and a cross over of 80Hz. I have had some success in doing this by hand but would be really cool to see the tool do it automatically.
Dave
This is already catered for by the existing Target curve, which incorporate bass management at the specified cutoff (in 10Hz steps from 30Hz-150Hz). The target follows the bass management/LEF cutoff curve so corrections against the target are correspondingly higher for peaks that are above where the bass management roll off starts. If you want more than this do the measuring correcting directly from your PC to the sub via the BFD (so no LFE cutoff from your processor), but still use the target curve for your processor's LFE cutoff rather than a flat target, then you would get the effect you are after - or am I misunderstanding?

Glad you like the app, BTW :)

OK, back to a basic setup question ...

I would like to use this on a laptop that has only the single pin jacks for microphone and speakers. Is this doable? Would I need to get splitter cables to separate the left/right signals to RCA plugs for each single jack? If so, do I then use an aditional RCA to RCA plug to close the loop?

I have a Radio Shack SPL meter, but I'm confused here on the mechanics of the cabling I'll need with this laptop.

Could someone clarify the step by step cabling I'd need for this laptopp? Or maybe even an actual picture would be worth a thousand words.

Much thanks in advance,

Tim

Though it pertains to different software (EFT), there is a visual depiction of the necessary connections about halfway down on this (http://web.archive.org/web/20030201100845/www27.brinkster.com/jmag999/) page. Just substitute your receiver for where they have the BFD, and instead of using the RCA>1/4" connector depicted, you'll need an RCA Y-cable like one of these (http://www.radioshack.com/product.asp?catalog%5Fname=CTLG&product%5Fid=42-2538).

BUT: As per Room EQ's setup instructions, I don't believe a Mic In input will be sufficient. You need a Line In.

Hi,
my laptop doesn't provide the line in. Any recommendation for a usb soundcard providing this which includes midi interface?

thanks Starc

Hi,
my laptop doesn't provide the line in. Any recommendation for a usb soundcard providing this which includes midi interface?

thanks Starc

Don't know about a USB soundcard with MIDI port, but I'm using the SoundBlaster MP3+ (USB), and a separate USB-to-MIDI cable. Works fine!

First I would like to thank you for making this program and sharing it with us.
What I would like to ask if you would consider adding support for the ULTRACURVE PRO DEQ2496 (http://www.behringer.com/DEQ2496/index.cfm?lang=ENG) to do auto calibrating? On the page I linked to it has the "MIDI Implementation Chart" if that is what is needed to do so.

Thanks in advance.


peace

What I would like to ask if you would consider adding support for the ULTRACURVE PRO DEQ2496 (http://www.behringer.com/DEQ2496/index.cfm?lang=ENG) to do auto calibrating?
Sure, can add that to the dev list. Will be several weeks before I can get on to it as I have a few more features to add on the measurement side first, but looks to be a useful unit to support.

Regards,

Sure, can add that to the dev list. Will be several weeks before I can get on to it as I have a few more features to add on the measurement side first, but looks to be a useful unit to support.

Regards,

Sweet, thank you. :cool:


peace

John,

Great software! I am just getting my feet wet, learning to use it, and have not attempted to tweak the system based on the results, but that will be coming soon.

But first a nOOb question...

Is it possible to do a manual measurement using Pink Noise? Looks like whenever I engage a manual session, it runs the sin waves, even if the tone generator is set to Pink Noise.

I would think that using sins would be required for parametric work, but talking a snapshot with pink noise would give you a quick look at overall in-room system performance.

Best Regards,

Brian

Hi Brian,

I guess you are thinking along the lines of an RTA measurement. You will get a faster and much more accurate picture of overall in-room system performance by doing a sweep measurement. The Wizard doesn't provide an RTA display, although it could - can't see the point, though :)

Regards,

Just a couple of quick questions for you guys, first, is there a specific reason why a microphone input on a laptop wouldn't be sufficient? I would imagine it might have something to do with input level or something, but would it not work at all, or just provide inaccurate readings? I would just like to load this onto my laptop for portability's sake, to check out my system's response, as well as a few of my friends' systems. Next question, I am already over budget on my home theater as it is (suprise suprise) and was wondering if there existed some sort of software could open up a soundtrack, apply some equalization to the low end similar to a parametric EQ, then package the soundstream back up to send spdif from my HTPC via optical to my receiver without otherwise molesting the sound, all running in real time. Does anything like this exist, or am I just dreaming?

Thanks for your time guys.

Frequency response on mic in is limited, which is why you need to use line in. If you've ever tried to record a turntable on mic in, you'd understand why you cannot use it in this case. It's pretty much designed for vocals...no real high end, and certainly no low end.

Thanks avdork, that's kinda what I thought, but I didn't really have anything to back it up.

V3.22 is now available on the website.

The main functional change in this build is the addition of support for microphone/SPL meter calibration files. The Wizard can load files containing the gain (and optionally phase) characteristics of the measurement device and subtract those from the displayed responses. The response of the mic/meter alone can be displayed by selecting the Mic/Meter Cal trace on the graph panel. If using a C weighted meter the C weighting compensation will still be applied outside the range of the cal file data - e.g. if the cal file data starts at 500Hz, C weighting compensation can still be applied for the region below 500Hz (and the region above the highest frequency in the cal file).

On the measurement side a "Window" control has been added to select the period over which the impulse response is analysed. The default is 600ms, which should work well in most rooms - however smaller rooms may show noisy responses with a 600ms period, if so try reducing it. Chirp Z Transform is now used on the impulse responses to provide frequency response at intervals independent of the FFT length, for measurement spans below approx 2kHz the interval is 0.1Hz, for a full sweep of 10Hz to 20kHz frequency interval it increases to 1Hz. The recommendation for AV processor input sensitivity (where this is adjustable) has been changed from 2V to 0.5V, should reduce the volume settings required for measurement. There are also two bug fixes, C-weighting compensation was not being applied for sweep measurements and the overall level of the sweep measurement was offset if the measurement level was not -18dB (must try harder).

In the User Interface there have been a few rearrangements, partly to make provision for some features to follow in future releases. The main filter control panel has been moved to the right hand side, swapping places with the signal generator panel, and a couple of the options for filter optimisation have moved into the Equaliser menu. There are also new File menu options to save and load a single channel's filter set and a new submenu for exporting data. Preview capability has been added to the file chooser dialogs so you can see information about files before loading them. The results of Find Peaks are now displayed in a table, with the ability to sort the results by amplitude or frequency by clicking in the table header area.

To help people running the Wizard on Macs, which only have a limited implementation of JavaSound, the wave volume control is now active if the Java Sound Audio Engine has ben selected as the output device.

Quite a few other changes, mostly minor, full details on the Revision History (http://homepage.ntlworld.com/john.mulcahy/roomeq/changehistory.html) page.

Best regards,

Thanks John!

I only recently aquired the necessary equipment (microphone, which I had calibrated so this update is timely), mic preamp, USB soundcard, etc.

I'm really looking forward to trying RoomEQ Wizard this evening. I've read through your documentation and it looks incredible. Thank you.

Cary

Hello John,

I have a calibrated ETF mic and preamp. If I connect them to your program and use your SPL tool, will I then have a very accurate sound pressure meter? I understand that without another calibrated mic my absolute level will not be exact but if I use the RS meter with calibration pink noise at 75dB my absolute error may be very small??

Thanks,
George

Hi George,

In absolute terms the result will be as accurate as the initial reference calibration. Use a normal speaker pink noise cal tone to set the reference level against your RS meter, then your calibrated mic will give you accurate spl readings relative to that inital reference from sine wave test tones across the full range of the mic&preamp's response. Remember to disable the C-weighting correction in the Meter menu.

Note that mic calibration values are not applied to the spl reading shown when reading the level of a pink noise source, however, as in that case the spl figure is being directly calculated from the time domain samples.

Regards,

Any chance you could generate a .pdf 'user's manual' from all the html help pages? I like having
a printed copy of instructions in front of me when I'm twiddling with apps like this.

Any chance you could generate a .pdf 'user's manual' from all the html help pages? I like having
a printed copy of instructions in front of me when I'm twiddling with apps like this.
I don't know of an easy way to do that (i.e. take a set of HTML pages and make a merged document from them) - I guess could just copy and paste each html file into a Word doc? All of the HTML content is also available directly from the Help within the app, so you can have the help window open on the relevant topic while twiddling. When I've got to the stage where I'm not changing and updating the app so often I'll make some case studies that go through the process from start to end, but at the moment the content would go out of date too quickly.

Hello John,
That sounds terrific. Once I calibrate against the RS I should use your program to calibrate the levels for all the speakers and I won't have to worry about the RS correction factor for the sub!
George

Nice program, John! It looks like you are adding many of the features of the standalone measurement programs. Can the current version display the impulse response? That's a big help when tracking down HF reflections. That along with the ability to use short windows to get the quasi-anechoic HF response should handle about everything most users would need.

I don't know of an easy way to do that (i.e. take a set of HTML pages and make a merged document from them) - I guess could just copy and paste each html file into a Word doc?

Adobe Acrobat might be able to do it -- I'll check and see. If I can do it, I'll send you the pdf. I think it may require extensive html editing to convert inter-file links to intra-file links, before converting to pdf or .doc. My html's a little rusty but I can probably do it.


All of the HTML content is also available directly from the Help within the app, so you can have the help window open on the relevant topic while twiddling.

Sorry, I prefer solid paper in hand. Less clutter on the screen. I'm old fashioned that way.

When I've got to the stage where I'm not changing and updating the app so often I'll make some case studies that go through the process from start to end, but at the moment the content would go out of date too quickly.

That will be awesome, thanks.

Krabapple, just print out the web pages. Let John do more important things like adding features. ;)

Can the current version display the impulse response? That's a big help when tracking down HF reflections. That along with the ability to use short windows to get the quasi-anechoic HF response should handle about everything most users would need.
Not yet, but watch this space... :)

Well as long as we are submitting wish lists, I wouldn't mind seeing a simple RTA in the program. Sometimes I am curious as to the spectral content of a test tone, sound, or musical passage. Also I read that an RTA is an excellent way to set the calibration level of your subwoofer channel.

Amazing work John. Surely this thread should be a sticky ?

John (and for those of you playing at home):

OpenOffice (http://www.openoffice.org) creates and edits HTML docs and exports docs as PDF. Freeware. I tested this with a beta version of 2.0 and it worked fine.

--Mike

I have been playing around with REW, and have run into a brick wall in my understanding of how to make it work right. I don't have a BFD or a Tag, so I guess I am not exactly sure if it would be a problem if I had either of those units. Anyway, I did a simple 1/3 octave plot of my main computer's speakers, as seen in the included screen capture. And when I click on "Find peaks", it only finds a peak at 56.8Hz. Under the "Find peaks" button I have 20-20,000 selected. What am I missing?

Screen capture

Oops here it is...

I just found this program and am trying to use it to plot the frequency response for my new SVS sub. However, I am having some problems and can't seem to figure out what is wrong.

The SPL calibration works fine, but I am getting stuck at the Set Measurement Level step. I get a "High Noise Level" error while it is checking the left channel loop. It says the left channel input has a RMS level of -39db, which is too high. I can click OK to disregard the warning, but then I get another error screen stating that it cannot detect the left channel loopback signal.

I can keep going, set the target level, and run a measurment. But during the measurment, it tells me the the left channel is clipping, and the results could be distorted.

I have checked the connections a dozen times but cannot figure out what is wrong. Connected to the soundcard I have a 3.5mm plug to dual RCA plugs for the input, and a 3.5mm plug to dual RCA sockets for the output. The left channels (white) are connected to each other and the right channels (red) are connected to the SPL meter and receiver.

Does anyone have a an idead as to what could be wrong? What am I missing?

For the input, you are using a line in, not the mic in, correct? That would cause the signal to be too hot.

I have been playing around with REW, and have run into a brick wall in my understanding of how to make it work right. I don't have a BFD or a Tag, so I guess I am not exactly sure if it would be a problem if I had either of those units. Anyway, I did a simple 1/3 octave plot of my main computer's speakers, as seen in the included screen capture. And when I click on "Find peaks", it only finds a peak at 56.8Hz. Under the "Find peaks" button I have 20-20,000 selected. What am I missing?
The automatic peak finding is directed at identifying low frequency room modes, the range of Q values and certain characteristics it looks for in the response are tailored accordingly and are typically not met above a few hundred Hz. A full band measurement will also not have the resolution at the lowest frequencies to accurately locate resonances. If you make a measurement over a smaller range (say up to 300Hz) the software will probably find further mode-related peaks. To alter the response further up the range just manually apply some filters and use the on-screen display of the filter responses to tweak them to suit. Additional features to provide EQ suggestions above the modal range will come in a future release.

I just found this program and am trying to use it to plot the frequency response for my new SVS sub. However, I am having some problems and can't seem to figure out what is wrong.

The SPL calibration works fine, but I am getting stuck at the Set Measurement Level step. I get a "High Noise Level" error while it is checking the left channel loop. It says the left channel input has a RMS level of -39db, which is too high. I can click OK to disregard the warning, but then I get another error screen stating that it cannot detect the left channel loopback signal.

I can keep going, set the target level, and run a measurment. But during the measurment, it tells me the the left channel is clipping, and the results could be distorted.

I have checked the connections a dozen times but cannot figure out what is wrong. Connected to the soundcard I have a 3.5mm plug to dual RCA plugs for the input, and a 3.5mm plug to dual RCA sockets for the output. The left channels (white) are connected to each other and the right channels (red) are connected to the SPL meter and receiver.

Does anyone have a an idead as to what could be wrong? What am I missing?
That noise level sounds more typical of a mic input than a line input. What are you using to make the measurements, an SPL meter or a Microphone/preamp? What input is selected? If you use the option in the Soundcard menu to generate a debug file and email me that I can take a look at it.

I am using the Line In on the motherboard (built in sound on NForce2 mobo). Also using the Radio Shack ananlog SPL. The input is LINE_IN and the output is SPEAKER. The Realtek AC97 Audio device is selected for both.

Also tried using the Mic input on the soundcard, but then the calibration didn't want to work and the Set Measurement was getting -35db on the left channel.

John - I am emailing the debug file to you.

Thanks for the quick replies!

The automatic peak finding is directed at identifying low frequency room modes, the range of Q values and certain characteristics it looks for in the response are tailored accordingly and are typically not met above a few hundred Hz. A full band measurement will also not have the resolution at the lowest frequencies to accurately locate resonances. If you make a measurement over a smaller range (say up to 300Hz) the software will probably find further mode-related peaks. To alter the response further up the range just manually apply some filters and use the on-screen display of the filter responses to tweak them to suit. Additional features to provide EQ suggestions above the modal range will come in a future release.

Ok cool thought I was doing something wrong, and not that it matters cause I don't own a TAG or BFD.....but I do have a Rane THX 44 (for front three + sub), DBX 231 (for side surrounds), and a Alesis MEQ something (for the rear two channels).

My HTPC runs win 98, are there any problems with win98 an REW? I know, I should upgrade my OS down there, just money is tight...

My HTPC runs win 98, are there any problems with win98 an REW? I know, I should upgrade my OS down there, just money is tight...
I've heard from one person running the s/w on Win 98, there was an initial glitch with the soundcard which was sorted in the last release, otherwise seemed to work OK.

I, too, have a laptop that I would like to get this program up and running on. A very powerful tool that would be made even more powerful if I could make it portable.
My laptop (like someone already mentioned) only has a mic in, which is apparently unsuitable for use with this program. Can someone recommend a cheap (cheapest that's functional) soundcard that we could order online that would work with this? One or more specific recommendations would be a great big help for myself and, most likely, a good number of people reading this thread.
Thanks in advance for your help.
Lewis

http://www.newegg.com/Product/Product.asp?Item=N82E16829102174 <----maybe something like this???? Just throwing it out there, never used one myself.

Thanks for that suggestion. Looks good, it'd be great if there were a bit less expensive options out there... I should have also specified usb soundcards, seeing as how that seems pretty important for use with laptops (at least for people like me, it's helpful).
Thanks again.

Thanks for that suggestion. Looks good, it'd be great if there were a bit less expensive options out there... I should have also specified usb soundcards, seeing as how that seems pretty important for use with laptops (at least for people like me, it's helpful).
Thanks again.


The card he suggested is a USB card. As an alternative, officedepot.com has the Creative Sound Blaster MP3+ (USB) for $39.99 plus S&H, and you can find it for less on eBay. It's what I use on my laptop with both the RoomEQ Wizard and ETF, and it works perfectly. :)

Wow! What an outstanding piece of software. The help files are fantastic and really help
to get up and running quickly. In no time at all as was able to check out my room and
load the filters into my BFD. The only problems I had came after all that. Since I know
almost nothing about room acoustics, a bit more help with how to tweak the filters would
be nice. But, I suppose that's what these forums are good for ;).

Directing the sound output to the right channel for measurement was a bit of a pain. On
my processor (Parasound 7100), I had to use the 7.1 inputs. It would be really cool if
this could be done from software. Is it possible to generate Dolby 5.1 on the SPDIF port
of the sound card?

Additional icing for the cake would be a before/after view. I found it difficult at times to
decide if my latest change was actually an improvement. Also, and I am not
sure if this request even makes sense, but a mode for suggesting tweaks to an existing
set of filters given both the filters and measurement data seem to me like it would be
helpful. The current peak detection and filter assignment seem to assume that no filters
were used during the measurement.

Overall, I cant even begin to describe how impressed I am with this software. As someone
who is into instant gratification, I cant imagine trying to EQ a room by taking manual measurements and then programing an equalizer by hand. Turn-around from measurement
to filter assignment back to measurement is around 1 minute. Outstanding!

I do have one set-up quesiton for the forum. Regarding correction data for the RS SPL meter, is it needed or not? I got the impression that most of the correction values were
were to compensate for C weighting which is already covered by the wizard. Consequently,
I have not applied a correction file. Should I?

I too was wondering about an overlay type function so a comparison could be done.

Hi dBeau,

Thanks for the compliments, glad you like the software.

Is it possible to generate Dolby 5.1 on the SPDIF port
of the sound card?
Possible, but far from trivial, particularly as the compression might make a bit of a mess of the test signals, but worth considering. Not any time soon though, sorry.

Additional icing for the cake would be a before/after view. I found it difficult at times to decide if my latest change was actually an improvement. Also, and I am not sure if this request even makes sense, but a mode for suggesting tweaks to an existing set of filters given both the filters and measurement data seem to me like it would be helpful. The current peak detection and filter assignment seem to assume that no filters were used during the measurement.
Agreed. You can make measurements using the tabs for other channels to provide a reference for comparison, but it is not as convenient as it could be. I have rearranged the graph operation in the current dev build to provide various groups of traces, so it is now easy to view all measurements at once, but also on the dev list is a form of iterative correction that will take the result of a set of filters and work out improvements.

Regarding correction data for the RS SPL meter, is it needed or not? I got the impression that most of the correction values were were to compensate for C weighting which is already covered by the wizard. Consequently, I have not applied a correction file. Should I?
For low frequency measurements you only need the Wizard's C weighting compensation. If you are measuring above a kHz or so with the RS meter it does need additional correction, if you look in the leaflet that comes with it you will see a typical response curve. I don't know how consistent, if at all, the meter is up there - best not to rely on it for much beyond a few hundred Hz.

I'm finishing off the changes to add display of the impulse responses at the moment, I'll make that the next release. Then I'll add an RTA display mode, probably followed by some plots to show the modal decay. After that I'll probably get back to improving the correction side of things.

Best regards,

Is it possible to generate Dolby 5.1 on the SPDIF port
of the sound card?

This would be extremely difficult, and impossible to do for free as Dolby would charge for their codec and licensing.

I believe "on the fly" DD 5.1 (i.e. DD stream generation) is only supported by a couple of soundcards. The legendary but discontinued nVidia nForce 2 MCP was one of them, I think there's another one out there that's very difficult to obtain.

Quote:
Originally Posted by dBeau
"Is it possible to generate Dolby 5.1 on the SPDIF port of the sound card?"

Possible, but far from trivial, particularly as the compression might make a bit of a mess of the test signals, but worth considering. Not any time soon though, sorry.
rovements.

This would be a very, very nice feature. I've always felt that a 1.4Mbps DTS bit stream could be fashioned that had very little lossy compression accross all channels. But, not having done any DTS or Dolby Digital compression, I wouldn't know the difficulty of creating a multi-channel test tone and whether it is possible to minimize or eliminate lossy compression for a DTS bit stream (for selected combinations of channels).

This would be extremely difficult, and impossible to do for free as Dolby would charge for their codec and licensing.

I believe "on the fly" DD 5.1 (i.e. DD stream generation) is only supported by a couple of soundcards. The legendary but discontinued nVidia nForce 2 MCP was one of them, I think there's another one out there that's very difficult to obtain.

Since these are test sequences, you could have many, many of them pre-generated as bit-streams, that way the encoding would not be part of the test harness.

Since these are test sequences, you could have many, many of them pre-generated as bit-streams, that way the encoding would not be part of the test harness.

Wouldn't you still owe licensing costs though?

Not sure on this.

But to record the streams in the first place, you'd need an encoder.

Wouldn't you still owe licensing costs though?

Not sure on this.

But to record the streams in the first place, you'd need an encoder.

I'm not sure either. I'm going off the assumption that if you're not providing a general purpose encoder, that you wouldn't incur an encoding license cost or a minimal one at most.

cool program!

Am I understanding dorrectly if I guess that this is the equivalent to using a tone generator and a spl meter, only done in software?

what about implementing more refined measurement methods allowing IR to be measured? Then windowing could be used, and correction files could be made for general FIR filtering, not only minimum-phase equalizers?

best regards
Knut Inge

I am using the Line In on the motherboard (built in sound on NForce2 mobo). Also using the Radio Shack ananlog SPL. The input is LINE_IN and the output is SPEAKER. The Realtek AC97 Audio device is selected for both.

Also tried using the Mic input on the soundcard, but then the calibration didn't want to work and the Set Measurement was getting -35db on the left channel.

John - I am emailing the debug file to you.

Thanks for the quick replies!

John

I have the same problem, connections have been double checked. Did you find an answer in the debug file?

Thank you

John

I have the same problem, connections have been double checked. Did you find an answer in the debug file?

Thank you
All looked OK with Elkcaps debug file, the Realtek chipset is fairly common (I have a laptop that uses it which works OK). Elkcap was going to try another PC also, but I didn't hear any more after that.

Worth pulling up the Windows mixer to make sure the playback and recording mixer settings have been applied OK (i.e. Line In is selected, output is not muted etc).

You could just go through the process, acknowledging any warning messages, then afterwards manually set the measurement level to -18dB, make sure the wave volume is set to 1.0, then use the sig gen to play a sine wave at -18dB and see what level appears on the left channel. Could also email me a screen capture taken while using the sig gen to generate a sine wave at 1k, so I can see the various levels and control settings, might spot something from that.

Regards,

Am I understanding dorrectly if I guess that this is the equivalent to using a tone generator and a spl meter, only done in software?

what about implementing more refined measurement methods allowing IR to be measured? Then windowing could be used, and correction files could be made for general FIR filtering, not only minimum-phase equalizers?

Hi Knut,

The Wizard also performs log swept sine measurements and extracts impulse responses from those, the next release will include display of the impulse responses.

Just sharing my results of using the BFD and this great software.

Before Calibration (http://www.bitrealm.com/junk/before-bfd.jpg)

After Calibration (http://www.bitrealm.com/junk/after-bfd.jpg)

I could play with this more, but it is flat enough. I can change the curve quite dramatically by simply moving the SPL meter around the room. I got rid of that nasty boom at 40hz, and moved the sub around a bit here and there, adjusted some bass controls on the amp to get rid of the 55hz dip. I DID add +5db at that frequency using the BFD. People seem to frown on it, but the BFD barely blinks the yellow at ear-bleeding levels, so I shouldn't be clipping any signals.

I've spent some time playing around with this. Must say what others have said--very, very cool program.
Couple questions. I am using a BFD and a Denon 3802 receiver crossed over at 80hz. I ended up plugging the right output of the soundcard into the receiver (the Help section says to plug it into the BFD) so that it would go through the receiver's crossover (then the BFD) and I could tweak things the most accurately to make sub response equal target curve. I got a flatter response in the graph (undesirable I think) beyond 80 if I plugged the soundcard into into the BFD. Did I do this correctly or there another reason to plug into BFD first?
I seem to remember not being able to go below 20hz on measurement or sweep when everything was set up and I was working on it. I'd like to be able to do this. Is this normal? I notice that, just opening the program and playing with it now, I can adjust the sweep range window below 20hz. Fwiw, I was using a Soundblaster mp3+ external usb soundcard. Also, what range do people go up to in terms of sweep and in terms of filters? I ended sweep at 300hz and highest filter was around 250hz.
Wish I could show you guys what the before and after graphs look like as another testimonial for this program, but I don't really know how to do screenshots or host or whatever. Oh, well--I'll just say that the cure looks a whole lot closer to ideal now than it did before. Also makes me appreciate the contribution of room to the acoustical equation.
Thanks.

Lower limit for measurement and display is 10Hz. When measuring you can connect to your receiver or directly to the BFD, the main difference is that going through the receiver will mean the receiver's bass management is active which is what contributes the roll off you noticed above 80Hz - If connecting directly to the BFD the target curve depends on your sub's roll-off and the settings of the sub's filters.

JohnPM said;

going through the receiver will mean the receiver's bass management is active which is what contributes the roll off you noticed above 80Hz

Regarding going through the receiver, it is my understanding that a receivers sub crossover also rolls off (slopes) BELOW the crossover setting. If so, should we set our receiver crossover higher than normal for EQ purposes with your software, or just leave it at the setting we normally use?

Regarding going through the receiver, it is my understanding that a receivers sub crossover also rolls off (slopes) BELOW the crossover setting. If so, should we set our receiver crossover higher than normal for EQ purposes with your software, or just leave it at the setting we normally use?
There is no need to change anything on the receiver, just select your receiver crossover frequency in the "cutoff" box for the speaker being measured, the target response curve will then show the corresponding crossover characteristic.

John,
I am using ETF alongwith their mic and preamp and Soundblaster USB MP3+. When I get all the mixer levels right in ETF and then use the Room EQ Wizard the Wizard changes the mixer level settings, sometimes causing an overload in ETF. Is the trick to calibrate the Wizard first or how do I use the same mixer levels for both the Wizard and ETF?
Thanks,
George

I am using ETF alongwith their mic and preamp and Soundblaster USB MP3+. When I get all the mixer levels right in ETF and then use the Room EQ Wizard the Wizard changes the mixer level settings, sometimes causing an overload in ETF. Is the trick to calibrate the Wizard first or how do I use the same mixer levels for both the Wizard and ETF?
The Wizard applies its cal settings to the mixer on startup (or when you calibrate) and restores the original settings when it closes. If you have ETF and the Wizard open at the same time, change the Wave volume setting on the Wizard to 0.25 and you will typically find the level is then OK for an ETF measurement, then change it back to 1.0 for a Wizard measurement.

I feel as though I have a really dumb question. I do know how to use my BFD and have used the snapbug website. I have no experience with any PC based EQ. In fact I have to go get an external soundcard to make this all work. My prepro is the AVM-20.

With respect to connections, I don't understand what is meant by signal routing facility. If I connect the R channel out from the soundcard to the AVM20 via the R channel of any stereo line input(let's say "DVD"), how can the signal come out of the center speaker, if I wanted to measure that speaker's response? I guess I'm supposed to use the 6ch analog inputs to do this and manually use the appropriate channel's input, right? And if this is true, can I simply use the sub input of the 6ch input to acheive this? Why even bother with "right channel of an input on your AV processor"? Unless.....(i'm trying to answer my own question here).....

I wanted to see the response of the transition between the R speaker used as "small" and the sub.


I think I just want to EQ the sub alone and leave the mains outta this as I learn this. I don't want to mess with the volume control on sub and since it is a Servo-15, I have to do the X-over in the preamp. What is the easiest connection for this?


Sorry for the long winded question. I have a billion and a half more, but I'm going to fiddle with this first.

Hi Bing,

The TMA AV32R DP and AV192R processors allow any input to be used as a test signal source and routed to any speaker (or any pair of speakers). Don't know if any other processors have a similar facility, probably not.

To measure/EQ the sub whilst applying the processor's own bass management you will probably need to use a R channel input, set the R speaker as small and disconnect the R speaker (unless/until you want to measure the integration between sub and main). Typically the 6ch analog inputs do not go through bass management, but if they do (which would imply they are digitised) you could simply connect to the sub analog input.

Regards,

My Meridian 568 does not have that really neat TMA test feature, a very clever concept. The Meridian also digitizes all incoming analog signals. To test my 7.1 setup I set the processor to PLIIx and physically connect only the speaker(s) in my 7.1 setup that I wish to test. Is this the correct methodology for my case?

Since I cannot access Dolby Digital with the Wizard or ETF, does using PLIIx give me accurate results for DD, assuming my Dolby Digital crossover is set the same as the PLIIx?

Thanks one more time,
George

George,

You should not use PLIIx when making measurements as it applies processing to all channels which will distort the results. Need to set the mode to plain stereo or the equivalent to avoid any processing beyond bass management, then to test the centre or surround speakers etc you would need to connect them to (say) the right channel output temporarily and if necessary change the right channel's bass management crossover setting to correspond to the usual setting for the speaker you want to test.

The results will be valid for any processing mode, unless the mode changes the bass management crossover settings.

John - first off I just want to say this program looks fantastic and I can't wait to get it working correctly. If you did this just as an exercise to teach yourself java, I'd love to see the output if you really set your mind to something. ;)

I'm getting the 2 errors I see described here most frequently. I have a BFD (not that it is affecting anything yet, but it's connected), a sound blaster audigy 2 zs, and a rat shack meter. I was using 71 dB as my target level (mostly because 75 is an inconvenient measurement on the rat shack meter). I may try 80 dB tonight if the overall input level should be higher.

Oh, and I have a receiver that constantly changes speakers when outputting the test tones. I presume I can use a test CD to give me a more constant output signal for measurement (specifically the input calibration step on your program) The receiver also won't output speaker test tones in "stereo" (e.g. no surround mode selected). What a PITA.

1) Can't detect loopback. I've checked and double checked the connections, the cables and the adaptors. I'm using the Lin_in on the sound card, and the speaker out to the loopback and the receiver. I have two mini-plug to stereo RCA adaptors connected (properly, I checked them about 15 times) I couldn't get it to work and ignored the message.

2) Can't remeber the exact text but I'm also getting the mic calibration error that the input levels are set to max, please adjust the output range on the spl meter. I've tried changing the cables, changing the range on the rat shack meter, changing cables again and nothing works.

If I ignore these two errors and run the sweep anyway, I get no result (just the same curve as the input signal), which leads me to think that the program isn't "seeing" any input from the sound meter. I hooked the sound meter up to my receiver directly to test it out, forgot to turn the volume down and scared the bejeesus out of my dog with some feedback, so I know that's working OK (cable and meter, that is) I also know the sound card is working because I had the line in connected to my DVR and the line out connected to my receiver, and I had sound in and sound out.

The only real solutions I have found so far are "make sure you aren't using the mic input" and "ignore the error messages" but those aren't working for me yet.

Any and all tips and hits are welcome - thanks!

I'm getting the 2 errors I see described here most frequently.
{SNIP}
1) Can't detect loopback.
{SNIP}
2) Can't remeber the exact text but I'm also getting the mic calibration error that the input levels are set to max, please adjust the output range on the spl meter. I've tried changing the cables, changing the range on the rat shack meter, changing cables again and nothing works.

The errors are probably both caused by not being able to access the input signal. A few things that should help in tracking down what is going on:

- please use the option in the soundcard menu to generate a debug file and email that to me (john_mulcahy@hotmail.com)
- let me know what the SPL meter rms levels are when there is nothing connected to the soundcard inputs
- try connecting the soundcard outputs directly to the inputs (L -> L, R-> R), set the signal generator running a 1kHz sine wave at -20dB, turn on the wizard's SPL meter, grab a screenshot and email that to me, might spot something from there.

In the meantime you might be able to get things working by selecting "Primary sound capture driver" as the input device and using the windows volume control to manually select line in and adjust the input volume level (make sure the Audigy is set as the default device for playback and recording in the control panel Sounds and Audio Devices properties).

Best regards,

1) Can't detect loopback. I've checked and double checked the connections, the cables and the adaptors. I'm using the Lin_in on the sound card, and the speaker out to the loopback and the receiver. I have two mini-plug to stereo RCA adaptors connected (properly, I checked them about 15 times) I couldn't get it to work and ignored the message.




Just a quick thanks to John for this excellent program. I finally got it up and running last night and am thoroughly impressed. The filter response for the BFD is almost spot on to the calculated values.

Trojan...I had the same issue at first. Try setting your output to something other than speaker. I had my input set to LINE_IN and my output set to SPEAKER also...seemed intuitive since the port I was plugged into was labled FRONT SPEAKERS. When I changed the output to DIGITAL SOMETHINGOROTHER and left the cable connections alone it was able to detect loopback without a problem. It's worth a shot if you have another option on the output pulldown menu.

I can't recall exactly what I had for options on the output, but I think it was only speakers. I'm pretty sure I tried using all the inputs available, all channels, as well as outputs. (default windows sound driver, Audigy etc)

John has been emailing me suggestions and I think he's nailed the problem, but I need to get home and check (and try his advice). He thinks it's a conflict with the creative mixer. In my experience, if there's a way to muck up your system, creative will find it, so that might be it. I usually don't even install their "applications" that come with the sound card, not sure why I did this time. Oh well.

Thanks for the tip and encouragement! Now if only I could find a driver for my old midi connection (parallel port, if you can believe that. What's an IRQ again? LOL)

The only problem I seem to have now is popping in my main speaker. If I hook my R output to a R input on my receiver and then try to run a sweep I get poppoing in my main speaker but not my sub. I have everything calibrated and there are no issues in any of the setup steps. The SPL never gets above 70dB with the master volume setting I am using so I don't think it's the speaker. The really strange thing is that every once in a while I can get a "clean" sweep with no popping.

John - I think your presumptions were spot on. I changed the record settings in the creative mixer (to analog mix from "what you hear") and only enabled the line in on the creative mixer.

I still got the no loopback, but everything else worked great. Except the results, holy crap do I have some work to do.

Anyway, I'll email you the screen shots, and I have a few other things I want to try, like disabling the creative controls altogether. I'll include the resulting setup in an email. Hopefully I can get my midi interface working, if not it's off to the shop for another (this one is from the days of windoze 3.11)

All I can say now is your program ROCKS! Hats off to your effort and dedication, room EQ wizard is extremely slick.

All I can say now is your program ROCKS! Hats off to your effort and dedication, room EQ wizard is extremely slick.

Amen to that.

I apologize if this is a stupid question. There is much discussion about the radio shack spl meter correction tables. In the room eq program there is a graphic in the help section, meter menu section that shows the effect of c-weighting. Are rs correction and c-weighting the same? It seems that the radio shack correction tables are spot on with the c-weighting graph. If I apply the c-weighting in the room eq program, does that effectively take care of the correction that people talk about with the meter, or do I apply those corrections on top of the c-weighting? I've run the program with the c-weighting and am wondering if I need to do it again with another set of corrections on top (such as adding ANOTHER 7.5 db at 20hz, etc.). Hope I'm being clear. Thanks.
Lewis

Lewis,

The RS corrections are the same as the C weighting corrections. No need to add it again.

Thank you very much.

does anyone know where i can get db meter software for the pc using a normal mic.

As mentioned earlier, the mic-in is frequency limited, you must use the line-in input for your sound card. You also need a device reasonably capable of measuring the entire range. A cheap pc microphone isn't going to cut it. The RS meter with it's c-weighting ability is about the cheapest way to go, although it isn't perfect. I believe that the people who sell the ETF software also sell a lab-certified microphone, with the calibration file specific for the mic. It's over $250 just for the mic.

I installed the software and it looks truely amazing.
I have a question however.
Is it possible to have all those resultant corrections applied via software to the output signal to a simple receiver instead of having to purchase an expensive parametric equalizer?

I currently use a Home theater PC for all my HT experience and use the analogs out of the Revolution 7.1 to go to a Sherwood Newcastle receiver which in turn goes to multiple amplifiers.

It would be nice to be able to use your software to create the filters, but then be able to correct the sound via software...

Anybody knows?

Thank you,
Arno

$99 isn't what I would consider expensive. That's what the BFD sells for on partsexpress.com.

That's an interesting question though. The way I have mine set up, I only want it affecting the subwoofer, and since there is overlap between the mains and the sub, it might be difficult to accomplish what you describe.

If you are doing the bass management in your PC and then sending 8 analog channels to your receiver, you could probably get an EQ and assign it to that channel. Does your sound card come with a EQ ap you could use?

I agree that $99 is reasonable, but I have 6 channels to feed and that would mean $300, which is not as inexpensive... ;)

I don't think that the drivers for my soundcard will do eq per channel unfortunately...

Arno

I agree that $99 is reasonable, but I have 6 channels to feed and that would mean $300, which is not as inexpensive... ;)


Generally it's most desireable to EQ the bass frequencies. Therefore you only need one for the subwoofer line.

Bass frequencies have the largest room interactions by far.

I understand by my rooms is actually my living room.
About 400 sqf, tile on the ground, cathedral ceiling...

All the good stuff... ;)

I wish that I could some eq to correct a lot of the problems that I think I have.
$300 is not unreasonable is that boosts the quality by 20 to 30% at least...

Software could be free...

Arno

Can someone clearly explain why it's eq'ing bass is more helpful/worthwhile vs. eq'ing the rest of the frequency spectrum? I have chosen to do it and have a vague form of an idea about why it is thus (wavelengths being close to room dimensions so some of them create noticeable nodes and variations in frequency response or something). I just feel like I don't have a concrete understanding of what's going on...

$300 is not unreasonable is that boosts the quality by 20 to 30% at least...

The BFD is fine to use on your sub but most people wouldn't consider it transparent enough to use on your main channels. If your source is an HTPC, look into software EQ solutions.

Poke around on this site lewdog... http://www.harman.com/wp/index.jsp?articleId=default

Basically, your room affects low frequencies more than high frequencies (simplistic).

Catapult - I think that was his original question. Anybody know any good software solutions?

Yes, good software solutions would be nice to have and of course more flexible...
Of course, I do not want to hijack this thread and I apologize if I already did so.
Let me know.

Thank you,
Arno

if someone can confirm the following for me:

I will be using a laptop to do some testing. Will get a USB sound card which has a line in and line out. The line input will be used to go from the RS meter output to the notebook. The line output will go into the analog input to the receiver. Assuming first I can do the left and right channel, then I can move the right channel into let's say the center to test it and the sub, etc?

Ultimately I will be able to adjust the receiver built in eq to as flat response as I can with the limited settings. Can also adjust the phase on the SVS to get a flat as possible response without using a BFD. Now with a BFD I will be able to correctly tune the sub frequencies.

The BFD has two ouputs which I could use for a sub / another channel? And somoene said that the BFD should not be used to tune regular speakers?

Thanks!

Read the help file for connecting your equipment to the sound card, but you will need a loopback connection as well (ideally, I think you can not connect it but the results are probably better with it). You will basically only connect one output channel to your receiver, the other goes back to the line in.

Just use the BFD for your sub, and yes, you could use it for another sub. The BFD will help you with peaks (and nulls, really) and playing with the phase can help you fix interaction issues near the crossover.

Read the help file for connecting your equipment to the sound card, but you will need a loopback connection as well (ideally, I think you can not connect it but the results are probably better with it). You will basically only connect one output channel to your receiver, the other goes back to the line in.

Just use the BFD for your sub, and yes, you could use it for another sub. The BFD will help you with peaks (and nulls, really) and playing with the phase can help you fix interaction issues near the crossover.

Ok I understand, now the output channel will also be a Y - cable going initially into the right and left analog inputs of the receiver. When I want to graph the center, do I just move the right input in the receiver to the center and leave the left alone? Or do I unplug the left and only utilize teh center input?

You only connect one channel to the receiver, the other channel is connected back into the sound card, so it's not really a Y.

If you want to test different channels, just plus a different speaker into the R or L (I can't remember which channel you are getting from the sound card) speaker connections.

I suppose you could use the analog inputs on your receiver for 6.1 and just move the input around when you want to test different speakers. Just make sure your receiver isn't doing any sound processing (i.e. turn off any sound fields & just use stereo)

Some required steps for Audigy 1 and Audigy 2 ZS and ZS Pro soundcard users here at http://audio.rightmark.org/download.shtml
You'll see the appropriately named bottom two document links. And also the documents are from Creative.

Well I get a few problems trying to use an Audigy 1 with Audigy 4 YouP drivers. Also using latest program version, J2SE 5.0 update 4, winxp sp2 .
http://driverheaven.net/showthread.php?t=78049
They are 'adjusted' to run on an Audigy 1, 2, so on with the newest drivers released.
I really don't want to go back to the 'official' drivers since Creative doesn't update their older product lines.

1st problem/step: when trying to do the SPL Calibration this program mutes the 'Line In'. Have to manually unmute.
Took me a while to figure that one out considering that TrueRTA worked just fine.

Next step/problem (Set Measurement level): This program puts the 'Rec' slider at 50% along with the 'Volume' which is causing the program to come up with the appropriate error which is basically not enough signal level. I have to quickly slide up the 'Rec' slider real quick before it finishes. Not good.

3rd problem/step: Now the read out for the GUI SPL are too low. I'm comparing the GUI with what is on the meter (RS Analog). Usually the UI would read 75 dB but the meter would read 70 or below after adjusting the amp per the instructions.

I have a funny feeling that this program doesn't work well with Creative products. I'm not blaming the program. I'm blaming Creative.
I guess I'll try with my onboard sound next (Abit kv8pro).

Edit #1
Using onboard sound:
1st step: Does the same thing as Audigy. Mutes the "Line In" at program startup.

That's good feedback, thanks. I'll see if I can figure out how/why that might be happening - will be a while before I can get to it though as I'm away on business for the next week or so.

If you want to manually align the GUI figure to the meter you can tweak the registry setting "last/Spl/Offset" in HKEY_CURRENT_USER\Software\JavaSoft\Prefs\room eq wizard, change the value by the amount the reading differs from the SPL meter. Do this while the Wizard is not running.

Of course, it is not advisable to directly edit the registry if you are not comfortable doing that as it is easy to wreak unintended havoc.

Regards,

I haven't read this whole thread, so I'm sorry if it was asked, if you only have a mic in on your laptop, I don't see how software RTA is better than a standalong RTA on a cost basis considering you not just need the mic, you also need a USB preamp. All three are more than most stand alone units.

Also, how do you know your line in is realloy that accurate? :confused:

I went with the Phonic PAA-2 for $300 delivered.

videobruce,

Good questions, and here's my thoughts.

You are correct that a laptop with only mic in is not adequate, but you can buy a USB sound card with line in for as cheap as $50. A Radio Shack SPL meter (for use as a mic) is at most $50, and will be plenty accurate for sub woofer equalization. Buy these two things and use this free software, and you have only invested $100. (this software has the correction factors for the RS spl meter built in.)

If you want to equalize above the sub-woofer frequencies you are correct that one should invest in a separate mic and pre-amp, but this should still be less than $300 if you shop around at all. (of course this all assumes you already have a laptop computer, and won't buy one just for this! ;) )

Even if it does cost you $300 (or more), you will have a much more accurate setup with what we have been discussing here than the Phonic PAA-2. From what I have read about it, the Phonic PAA-2 will only display in 1/3 octave increments. Most people would say you need readings no less than 1/6 octave, and 1/12 is better.

That said, the Phonic PAA-2 does appear to be a simpler to use and easier to set up approach. If it works for you, that's great!

Hello,

I went ahead and bought a Creative USB external sound card for the latop to give it a go. First I bought a couple mini-jack to RCA converter. I plugged one end into the Line In and Speakout out on the soundcard. The Right channel from the Line In went into the SPL Meter. The Right channel from the Speaker out went into a Analog input into my receiver. The left channel from the Spaker out and Line out I connected together.

Initially I have all setting muted in the Creative panel except for Line In. I fire up the EQ WIzard and hit "Configure SPL Input". Proceed to pay a test tone so the meter itself reads 75dB. Then I adjust the Creative Line-In Volume so the SPL Meter in EQ Wizard also read's 75Db.

The I go to proceed to "Set Measurement Level" for the tones. I un-mute the volume / speaker out and what happens is I get feedback. The SPL meter is picking up the tone and generates a annoying high pitch noise. I mute the output and all goes back to normal. What am I missing here?

Hello,

I went ahead and bought a Creative USB external sound card for the latop to give it a go. First I bought a couple mini-jack to RCA converter. I plugged one end into the Line In and Speakout out on the soundcard. The Right channel from the Line In went into the SPL Meter. The Right channel from the Speaker out went into a Analog input into my receiver. The left channel from the Spaker out and Line out I connected together.

Initially I have all setting muted in the Creative panel except for Line In. I fire up the EQ WIzard and hit "Configure SPL Input". Proceed to pay a test tone so the meter itself reads 75dB. Then I adjust the Creative Line-In Volume so the SPL Meter in EQ Wizard also read's 75Db.

The I go to proceed to "Set Measurement Level" for the tones. I un-mute the volume / speaker out and what happens is I get feedback. The SPL meter is picking up the tone and generates a annoying high pitch noise. I mute the output and all goes back to normal. What am I missing here?
Sounds like Line In is not muted in the playback mixer, hence the feedback from input to output. The basic setup should be to have Line In as the only selected source in the record control, and Wave as the only unmuted source in the playback control.

Regards,

Can someone clearly explain why it's eq'ing bass is more helpful/worthwhile vs. eq'ing the rest of the frequency spectrum? I have chosen to do it and have a vague form of an idea about why it is thus (wavelengths being close to room dimensions so some of them create noticeable nodes and variations in frequency response or something). I just feel like I don't have a concrete understanding of what's going on...

It's because bass is the hardest thing to get right, mids and highs, with good speakers and some room treatment, are easy by comparson.

Think of bass as needing 70-80% of the work to get right while the other whole rest of the spectrum only needs 20-30%.

In my HT just to cover the 15-80hz area I'm spending twice the amount that was spent on 80-20k, and that's just equipment and doesn't include the massive amount of work in room construction, traps ect.

Many thanks for this great tool John!

Is it possible to smooth out the FR chart to e.g. 1/6 or 1/3 octave average? I've made some sweeps and although the detailed response graph is great it would be nice to be able to see the big picture. Maybe I'm missing something?

Many thanks for this great tool John!

Is it possible to smooth out the FR chart to e.g. 1/6 or 1/3 octave average? I've made some sweeps and although the detailed response graph is great it would be nice to be able to see the big picture. Maybe I'm missing something?
Haven't put in any smoothing features yet (so many features to add, so little time :) ), you can get some smoothing of the low end by reducing the gate time to (say) 200ms or so, otherwise have to wait until those reach the top of the request list.

Regards,

Sounds like Line In is not muted in the playback mixer, hence the feedback from input to output. The basic setup should be to have Line In as the only selected source in the record control, and Wave as the only unmuted source in the playback control.

Regards,

I had a similar problem with feedback with ETF. Turns out the default sound card on my computer was active at the same time as the Soundblaster. I found the default sound card's controls, turned it off, and everything worked fine. This has happened to me several times, once in the middle of a measuring session when everything was running fine and then feedback out of nowhere.

John, I like the program, but I gotta tell you I think it would be helpful to have an option to manually calibrate & set the measurement levels without quite so much automation. I would prefer an option to not have RoomEQ change my mixer settings and I would also like to understand what is happening during the SPL measurement setting phase so I could try to figure out the problems I have with "too much noise on loopback". I think you may find that the wide variety of soundcards and mixer drivers is going to make reliably doing the calibration in a completely automatic mode a real challenge - and frustrating for those of us who don't fit the mold.

I like the more automated approach.
Surely if you want to do it manually and you are using it to set up TMREQ on tag kit then using the older software that is still available from IAG and using ETF will give you all the manual fun you can handle....
-Scott

I agree with BillyBob. I have had difficulty getting it setup and at least part of it is because I don't know all of what it is trying to do when automatically setting up my audio for the tests. Therefore, I don't know specifically what is going wrong and can't adjust the setting manually (shooting in the dark) and the messages are cryptic and not very helpful to the end user. I believe some descriptions of what the software is trying to do behind the scenes and how to set it up manually would be very helpful for those having issues.

There used to be lots of information on using the early versions to set-up the TAG hardware, take a look at the IAG site to see if any is still there. At Johns site there is a reasonable amount of information.
I think that if you try to use the older versions of the software you will appreciate the work done on these later versions. If you get your head around setting it up in its early itterations then seems heaven sent.
I would rather see more time spent on developing the software at the moment rather than documenting it more fully in its current state. The option is to wait till its more finished!
Scott

It's easy for those who have it working to say more features. But for those of us who don 't...well... it should be just as easy to see why we would want to just have the current capabilities working for us before moving ahead.

Don't get me wrong I can appreciate the automation. But being in software development myself I also understand that when the automation is not 100% you need to provide work arounds. That is all I am suggesting at this point.

The software is John's and provided freely so he can obviously take it in any direction he wishes. Even though I have followed all the instructions it is unfortunately not working for me. Therefore, my preference would be to iron out these problems first.

smithb understood - My point is not about how it is working with the PEQ - I was talking about the SPL & measurement level calibration set-up. If you have the soundcard HW that works without issue, you are golden - but, if you are getting messages about the noise level being too high, you are pretty much stuck with no way to figure out what is wrong. I can't manually adjust the mixer levels to see if that would help the noise issue - and John's response to set-up issues reported in this thread was to edit the registry and/or send him a debug log. That's OK for a mature application where you know it will work with the vast majority of HW - but I have an 18 month old mainstream Dell desktop with the standard Dell soundcard HW (SoundMax) and the RS SPL meter - and the app is having issues with my set-up. I can't tell if I'm doing something dumb (likely) or if my HW is the issue. Maybe having a list of HW that is known to work and/or have issues would help. Typically, one of the reasons for releasing a beta is to get feedback on compatibility issues from the unwashed masses. :)

I think John has done a tremendous job on the app - maybe that's why I'm frustrated with the set-up - I WANT to use it. I have an m-audio mobilepre USB soundcard coming, I'll try again when it arrives.

Here's an outline of the calibration processes:

On startup the app selects the input that has been chosen in the spl panel (you should see this in the Windows Recording Control). The input volume setting that was in use when the app was last shut down is applied. It also selects the output that has been chosen in the sig gen panel, unmutes the Wave source and mutes all other output sources (you should see this in the windows playback control). The wave volume and output volume that were in use when the app was last shut down are applied. The Wizard saves all its settings (on Windows platforms) in the registry under the HKEY_CURRENT_USER\Software\JavaSoft\Prefs\room eq wizard key. The names of settings are fairly self-explanatory, note that names that have capital letters in them are saved with a / in front of each capital.


Calibrating SPL Input
The aims of SPL calibration are to find an input volume setting that produces a -18dB rms level (relative to digital full scale) on the selected input when a speaker cal signal at the desired cal level (usually 75dB) is being played and to work out the offset between the actual rms level and the desired cal level, so that the Wizard knows what to add to the rms value to show the corresponding SPL value. Typically this offset would be around 93dB, since you need to add 93 to -18 to get to 75.

The process starts by calculating an initial input volume setting based on the current setting and the measured level, then does a binary search to refine that setting based on whether the level (as displayed on the bars) is below or above the target. Once the level is within 0.5dB of the -18dB target the search stops. The exact rms value at that point is then subtracted from the desired cal level to produce the spl offset.

If at the end of the spl cal search the input volume is at or close to the minimum volume setting (which on most mixers means less than 0.05) a warning message is displayed - if the measured level is still above -18dB with the volume this low then it may be because the SPL meter is set to a very low range (i.e. a high sensitivity) that is producing a signal that is overwhelming the input, so the message suggests reducing the SPL meter range and trying again. Other possibilities would be selecting an input that is unsuitable (such as mic) or a soundcard which does not allow direct selection of the required input via the windows mixer and which requires a further setting in the card's own mixer, with the result that the input volume changes are having no effect.

If the volume setting is close to or at the maximum a warning message is displayed as there may be a missing connection, or again the soundcard may not allow direct selection of the input via the Windows mixer so the desired input is not actually being listened to, or the SPL meter being used may be set to a very high range (low sensitivity).

Setting Measurement Level
The aims of the measurement level setting are to establish the signal level required to achieve -18dB via the loopback input and then the AV Processor (or subwoofer) volume control setting to achieve the same via the SPL Meter input.

The process begins by checking the input level when the signal generator is off. The loopback level should be at the noise floor of the sound card input being used. If the level is above -45dB the Wizard generates a warning, as most cards will be better than this (i.e. the level with no signal will be lower than -45dB). This might be due to a poorly specified soundcard, or trying to use a mic input, or to some other sound source being active on the PC and not muted in the playback control, which could occur with some soundcards that do not provide full control via the Windows mixer. The process can be continued by OK'ing the warning.

The next step is to generate a 1kHz sine wave at -20dB on the loopback channel. The wizard checks that this increases the measured level on the left (loopback) input by at least 20dB - typically it would go from the noise level of the card (perhaps -80dB or so) to around -20dB, giving a 60dB change. If the level changes by less than 20dB a warning is given as there may be an error in the connections.

If the level has changed as expected, the generator rms level is adjusted by the difference between the measured value and that target of -18dB. For example, if the measured level were -22dB the generator level would be increased by 4dB. After changing the generator level the loopback level is checked again, warning messages are generated if the loopback level was so far from the target than the generator level has hit its upper or lower endstops.

If the loopback level is OK the measurement level is remembered so that it can be applied for subsequent measurements and the loopback level is remembered for comparison with subsequent measurements to check for roll off in the soundcard.

Once the measurement level has been set for the left loopback channel, a speaker cal signal at that level is generated on the right output and the user is asked to adjust the AV Processor (or subwoofer) volume to achieve the desired cal level - this should result in the target level of -18dB being observed on the SPL meter right input if the input cal has been done OK.


Hope all that helps. The main problems seem to be with multi-channel cards, where the Windows mixer does not support much of the card's functionality so the manufacturers have provided only limited access via the Windows mixer and encourage use of their own mixer. The debug log shows all the mixer settings, which sources are muted/unmuted etc and so can provide clues if something is not working as expected. I'll aim to put in some manual controls for the cal process, but I'm a bit overloaded at the moment so the Wizard is taking a back seat. I hope to be on the case again soon.

Regards,

John,

Thanks for the follow-up of what is being attempted behind the scenes. Hopefully, it will help me work around my issues and thanks again for providing software such as this to help many of us try and tame the bass in our HTs.

Brad

John - Thank you!!! OK, I think the high noise level on my loop-back channel may be due to a slight hum I'm getting on my Aux channel. It only seems to happen when I crank the receiver up high enough to get to 75 dB on the internal reference tones and it is connected to my soundcard with the long cable I need to use to reach from my desktop to the receiver. There is no hum when the internal tones are used on the Aux channel without the long cable attached. I'm going to try the whole thing again after I get the m-audio usb soundcard for my laptop - which I will be able to use with the short cable. Also, any ground loop issues should also be helped by using the laptop instead of the desktop.

Once again John - congratulations on an excellent program. It is a excellent product idea that you have implemented with an excellent application. As someone who has also slung a fair amount of code in his day - I salute you!

I'm looking to use the BFD/RoomEQ purely for sub tweaking. Step #2 of the "SPL Input Calibration Procedure" where is has you outputting a L/C/R channel signal from the "AV processor" - is this referring to the BFD or the receiver? Can I also assume that whichever is used, the test sub tone should be used for calibration?

I'm looking to use the BFD/RoomEQ purely for sub tweaking. Step #2 of the "SPL Input Calibration Procedure" where is has you outputting a L/C/R channel signal from the "AV processor" - is this referring to the BFD or the receiver? Can I also assume that whichever is used, the test sub tone should be used for calibration?
Refers to the receiver, best to use one of the main speaker test tones rather than the sub tone for setting the level as the sub tone is fairly variable (just the nature of pink noise over the narrow band used for sub cal).

Refers to the receiver, best to use one of the main speaker test tones rather than the sub tone for setting the level as the sub tone is fairly variable (just the nature of pink noise over the narrow band used for sub cal).

I had my first go at using RoomEQ to calibrate my sub last night in preparation for setting up my BFD. I found that when Setting the Measurement Level by connecting the right output to my SVS sub that I was unable to reach a 75db reading on the sound meter even when turning the gain on the sub almost full up. I turned the sub gain down a little but still way past the normal setting and when I tried an Automatic Measurement the windows nearly blew out. :eek: I did miss the Set Target Level step, which would have warned me of the high volume :o , but, I think there is a slight flaw in the process for sub calibration. My guess is that the signal generated during the Set Measurement level step is too high a frequency for the sub to generate the necessary SPL. I found that if I left the sub at its usual gain setting during the Set Measurement Level step and then adjusted the volume when Setting the Target Level then everything seemed fine. I am slightly concerned that I have damaged my brand new sub but it seems ok. I just wanted to warn others about a potential pitfall or find out what I am doing wrong.

David

The test tone generated is just white noise, not even going to hit the sub. I think you're doing something woefully wrong. Before you start, you should use test tones generated by your amp to set the speaker levels to roughly the same value. It might be 75db, it might be 68db...depends on the sensitivity of your speakers.

To recap:

Calibrate your system without using RoomEQ first, this will set the sub value roughly where it needs to be, as well as your surrounds and mains.

Run RoomEQ and follow the steps. The input is just left or right into your amp and the test tone will not be in your subwoofer at all. Essentially, RoomEQ asks you to adjust your volume on your AMP until the meter reads 75db, not the sub.

Got it?

Anyone have a clue how this software compares to what comes with the Onix/Rocket RE-DES ?

Terrific program! Thanks.

What's the process for getting measurement data out of TrueRTA and into yours?

Anyone have a clue how this software compares to what comes with the Onix/Rocket RE-DES ?

They are very different at the moment. The R-DES software does not currently have an auto-measure or auto-EQ function. The software it comes with is very user friendly for setting the curves though.

The R-DES software can only upload the curves to an R-DES box. Room EQ Wizard can only upload to the TAG hardware or a BFD.

The test tone generated is just white noise, not even going to hit the sub. I think you're doing something woefully wrong. Before you start, you should use test tones generated by your amp to set the speaker levels to roughly the same value. It might be 75db, it might be 68db...depends on the sensitivity of your speakers.

To recap:

Calibrate your system without using RoomEQ first, this will set the sub value roughly where it needs to be, as well as your surrounds and mains.

Run RoomEQ and follow the steps. The input is just left or right into your amp and the test tone will not be in your subwoofer at all. Essentially, RoomEQ asks you to adjust your volume on your AMP until the meter reads 75db, not the sub.

Got it?

For sub calibration the wizard directs you to connect directly to the sub, or through the BFD if you are using one, and use the gain on the sub to set the spl. Connecting this way then the test tone generated by RoomEQ seems inappropriate to achieve an SPL equivalent to that set during the calibration of the SPL input.

For sub calibration the wizard directs you to connect directly to the sub, or through the BFD if you are using one, and use the gain on the sub to set the spl. Connecting this way then the test tone generated by RoomEQ seems inappropriate to achieve an SPL equivalent to that set during the calibration of the SPL input.
Indeed, the comment is wrong for the way the Wizard runs the level cal - it is intended for use with one of the main channels so generates noise that is band limited to 500Hz .. 2kHz. If your system setup is such that you are connecting directly to the sub via the BFD your best bet is to just OK the remaining stages of the cal without altering anything, then turn on the sign gen in Pink Noise mode, select Sub Cal as the pink noise type, make sure the signal level is at the measurement level value, turn on the SPL meter with the SPL Update set to V Slow and then adjust the sub volume for your ref level (e.g. 75dB). There are some problems with using the sub for setting the measurement level due to the significant contribution of room modes, so it is better to use one of the main channels if possible. I'll give some thought to how things should best be handled when using the sub alone.

Regards,

Terrific program! Thanks.

What's the process for getting measurement data out of TrueRTA and into yours?
Use File -> Export Data... on TrueRTA to save your measurement as a text file then use File -> Import Measured Data... on the Wizard to import it.

Thanks. Found out that TrueRTA exports the data that's live on the screen (the contents of the "active workbench".

New Question: When I tried the Wizzard it would only generate peaks/filters below about 340hz or so. If I set the range from, say, 500 to 1000hz it would find zero. Is there some way to get the code to generate filters at higher frequencies?

Thanks.

ps -- I know it's not "best practice" to EQ at higher frequencies, but, I'm trying to find the few frequencies that I need to deal with ..... whether that's via EQ, passive devices, or something else. I'm taking four 1/24th measurements at the listening position, averaging those, and then looking for "peaks" in that. Having done that, will go back and do the same for the other three seats and address those peaks that the four seats share in common.

Thanks.

New Question: When I tried the Wizzard it would only generate peaks/filters below about 340hz or so. If I set the range from, say, 500 to 1000hz it would find zero. Is there some way to get the code to generate filters at higher frequencies?
In that range you are no longer correcting individual modes, so need to focus on trends in the response and set filters to deliver the desired shape. The Wizard needs some filtering tools to smooth out the response above the modal region so that the shape is easier to see, they are on the dev list but a long way off. In the meantime, you could just manually tweak filters based on imported TrueRTA measurements to get the response shape you want.

I do not have a laptop to take the measurements but is there a way to manually input the information so that I can use this software with my BFD? I think I read something threw this thread that you can input with a txt. file....is this correct. This is probably a stupid question but if it is possible to upload a txt file how exactly do I do this.

Thanks

I do not have a laptop take the measurements but is there a way to manually input the information so that I can use this software with my BFD? I think I read some where threw this thread that you can input with a txt. file....is this correct. This is probably a stupid question but if it is posable to upload a txt file how exactly do I do this.
Yes, you can do that, as described in the program's help or online at the website here: Importing Measurement Data (http://homepage.ntlworld.com/john.mulcahy/roomeq/wizardhelp/help_en-GB/html/dataimport.html)

Thanks John, I have not read threw the help yet as I am just know doing the measurements on my room and picking up my BFD today.

Thanks for this great tool.

I just got taking my measurements and have a question. I know it is not a great ideal to try to raise your nulls to much because of the power consumption but I have two big null one at 53.9Hz and one at 81.0 Hz. To fix it I had to raise the gain 16 on the 53.9Hz and 14 at 81Hz. Do you think it will be ok to do this. I think the 81Hz will be ok but not sure about the 53.9Hz. Please help

Hey John,

First, thanks for your wonderful program. I have a laptop with the standard "mic in" and "headphone out" jacks on the back. I read that mic level inputs are not suitable for the EQ wizard. However, I have an old Mackie 1604 mixer, and I was hoping that I could use one of my microphones as an SPL meter and send the signal into the computer from the Mackie. I assume that line level is higher than mic level, but I wondered if I could tone down my mixer's output enough to get it to mic level. Could this work? Thanks for your help.

John

Hey John,

First, thanks for your wonderful program. I have a laptop with the standard "mic in" and "headphone out" jacks on the back. I read that mic level inputs are not suitable for the EQ wizard. However, I have an old Mackie 1604 mixer, and I was hoping that I could use one of my microphones as an SPL meter and send the signal into the computer from the Mackie. I assume that line level is higher than mic level, but I wondered if I could tone down my mixer's output enough to get it to mic level. Could this work? Thanks for your help.

John
Not really, sorry. A local loopback is required from line out to line in to provide a reference signal for the measurements, and mic inputs have high noise levels and often have very restricted low frequency extension.

JohnPM, what cheap USB soundcard would you (or anyone else) recommend that would suffice for the EQ wizard? I'd like to keep this as cheap as possible, since I probably won't use it for anything else. Thanks.

John

I have the same question

JohnPM, what cheap USB soundcard would you (or anyone else) recommend that would suffice for the EQ wizard? I'd like to keep this as cheap as possible, since I probably won't use it for anything else. Thanks.

John
For the cheapest option try ebay for the old Creative Sound Blaster MP3+ USB card.

John - thanks for making this tool available - the software is very cool.

I took an initial measurement and was hoping someone had an idea of what I might be doing wrong (see image). My readings are all well above the Baseline measurement.

http://dotnet.baseline.com/measurement.jpg

Here are the steps I took:
* I calibrated the Radio Shack SPL to 75 db with my front center speaker.
* Set measurement level
* Connected via the analog SW input on my Denon 3803 to my HSU VTF-3
* Ran automatic measurement sweep

Thanks for any input!

John - thanks for making this tool available - the software is very cool.

I took an initial measurement and was hoping someone had an idea of what I might be doing wrong (see image). My readings are all well above the Baseline measurement.
Two things: use the "Set Target Level" button to establish the baseline level for that setup, and since the SW input does not seem to have any crossover filtering (at least within the range of that plot) may as well set the speaker type to "Flat" to get a ref trace that is a flat line rather than with an LP rolloff. You can also manually adjust the level of the target trace if preferred.

Regards,

Hi John,

Has anyone reported to you that your software works with the new model of the FBD which is called the FBQ2496? this model has 20 equaliser per channel.

Thanks
Oliver

What software do pro audio calibration specialist use? Just wondered.

Hi John,

Has anyone reported to you that your software works with the new model of the FBD which is called the FBQ2496? this model has 20 equaliser per channel.

Not yet Oliver, but it will soon.

Ok I need some help. I just got my laptop and was going to try this software this way. I am using a usb sound card (sound blaster). When I got everything setup and ready to calibrate I get a pop up screen that says my input level is to high. It does not matter what level I have it set at I still get the same screen. Could someone shed some light on what I am doing wrong.

Ok I need some help. I just got my laptop and was going to try this software this way. I am using a usb sound card (sound blaster). When I got everything setup and ready to calibrate I get a pop up screen that says my input level is to high. It does not matter what level I have it set at I still get the same screen. Could someone shed some light on what I am doing wrong.
Hi,

What is the exact text of the message? Also if you email me the file that is produced by the "Generate Soundcard Debug File..." option in the Soundcard menu that may shed some light.

Regards,

Not yet Oliver, but it will soon.

Sweet.

From what I can tell the FBQ-2496 is a very nice piece of equipment for its price. While my main application is bass EQ, I plan to see if it will be ok for HT full spectrum EQ too.

We need to make you another plaque.

Is there a FAQ on making the correct connection? I have a laptop and disabled the onboard and using the USB SoundBlaster. Maybe a step by step on using the software including making the proper connections, etc?

Thanks!

The help file available within the application has a step by step procedure. I printed it out, and keep it in a small three ring binder. It's very handy to have.

Brian

I'd appreciate comments on using Room EQ Wizard in combination with the KX-driver's equilizer.

I have a HTPC which drives my entire system with an Audigy4 card and the Tannoy digital speakers (6D's and 8D's). So the digital output from the Audigy4 drives the speakers; the Tannoy's then do the D/A and amplification. Although the Tannoy's have DIP switches for the DSP's D/A equalization, I'd prefer to do it in software on the computer since that gives much more flexibility and is easier to set.

The KX-driver has an equilizer and I can put one on each signal line and set the speakers differently in order to do room correction. My speaker locations are fixed in the room -- I got to select nice speakers, but my wife got to decide where they go. I can move them around only about a foot or so. The front-left, front-right, and center are on top of the entertainment center so they are basically fixed. The back-left and back-right are behind the couch mounted in the corners. The sub is down-firing and is under the end table.

Is KX-driver's equilizer the best choice for this kind of solution to room equilization, or is there some other high-band software equilizer that people would recommend?

Thanks for any insights.

Scott Nelson

I have the jar files & the Java 5.0 on my Mac. Now what? how do I open the application? I've clicked on the various jar files but all I get is an error message telling me to check the "Console for errors". Any Mac advice?

I have the jar files & the Java 5.0 on my Mac. Now what? how do I open the application? I've clicked on the various jar files but all I get is an error message telling me to check the "Console for errors". Any Mac advice?
Afraid I'm not a Mac person but I know some folk have run the Wizard on the Mac. Hopefully one of them will chip in, in the meantime all I can offer is as I put in the readme file in the jars zip:

The Wizard uses classes from 3 jars in addition to its own RoomEQ_Wizard_obf.jar jar file.
These are: jh.jar, comm.jar and TableLayout.jar.
All must be in the same directory as the wizard jar.

To run the Wizard use the command:

java -jar RoomEQ_Wizard_obf.jar

Regards,

Okay, I'm a little confused. Can I use my Audigy 2 soundcard with this program?

These are it's connectors on page 16: http://files.americas.creative.com/manualdn/Manuals/TSD/797/English.pdf

Can I not use the Audigy 2 because it doesn't have separate left/right line outs?

Edit: Re-reading the help files I guess I need those to be separate. Why did I think that the Audigy would work? Damnit.

Can I not use the Audigy 2 because it doesn't have separate left/right line outs?

I thought the same thing when I first tried the program (and actually asked the same question (http://www.avsforum.com/avs-vb/showthread.php?p=5481935&&#post5481935) in this thread near the beginning) but all you need are two 1/8" > RCA splitters like these (http://www.radioshack.com/product/index.jsp?productId=2102972&cp) (like what you would use to hook up a portable CD player to your stereo.)

Now that's what I call a great response to a question, thank you very much :)

I guess my brain disengaged when I tried reading this thread (and searching) for an answer. It happens right? ;)

Just want to see if I am doing something wrong. After I send the filters to the BFD and run the measurement it changes a little but it does not match what it shows it should. I can't or don't know how to capture a screen shoot or I would show you.

Just want to see if I am doing something wrong. After I send the filters to the BFD and run the measurement it changes a little but it does not match what it shows it should. I can't or don't know how to capture a screen shoot or I would show you.
There's not really enough info there to offer much help. Best bet might be to wait until Friday evening when the next release will come out, the setup has been simplified and the help files contain additional trouble-shooting information to make sure all has been set up correctly and sort out problems that may arise.

Regards,

John, thanks again for providing such a robust, useful program - and especially for keeping it up to date. I look forward to trying the new release.

After a very lengthy gap, the next version of the Room EQ Wizard (V3.23) is now on the website (http://homepage.ntlworld.com/john.mulcahy/roomeq/index.html) for download. There have been quite a few changes, which are detailed in the Revision History (http://homepage.ntlworld.com/john.mulcahy/roomeq/changehistory.html), but the main points are:

The graph panel has been revised to have several groups of graph traces, e.g. one for doing Filter Adjustment, one for all Measured responses etc. The mechanism is fairly flexible so if there is a grouping that would be useful, let me know.
Amongst the new graph groups are impulse response displays, including control over the pre- and post-impulse window types and durations. Note that for a true impulse response it is necessary to carry out a sweep over the full frequency range (10Hz to 20kHz in the Wizard). The Y axis can be set to % or dB, use dB when adjusting window durations. The impulse responses can be exported as text files.
The new BFD Pro, FBQ2496, is now supported. Currently the Wizard only configures and downloads 12 of the 20 filters available per channel, I'll extend this to all 20 in a future release (need to tweak the UI a bit to handle them nicely).
The left channel loopback is no longer required. Instead a mechanism has been added to make a reference measurement of the soundcard by looping back the right (meaurement) channel and saving the result as a cal file that is subtracted from subsequent measurements. The SPL calibration and measurement level setting processes have also been revised and simplified, and the help file includes an explanation of how to work with manual soundcard mixer control if the Wizard's automatic settings do not have the desired result.
The help files have been extensively revised to go along with the changes. Some checking post-release revealed a few broken links in the help files (items that have moved between chapters), they'll get fixed in the next release. Please let me know if you come across other errors or omissions in the help files. The Dutch help files are now too far out of date so they have been omitted from this release.
A few bugs have been fixed, including the one that prevented exporting measured data as text from sweep measurements.


Internally the changes have been fairly extensive, so the previous version is still available for download just in case :)

I'm going to be away for the next 10 days and will only have occasional access to the web and email, so bear with me if problems or questions arise, I'll answer as I get a chance.

Regards,

I'm very excited to try this new version out. You the man.

I've played with this a little in conjunction with my BFD and it looks to be a great tool.

I have bought the midi leads to enable the REW to automatically set the filters for me but can someone please explain how/what I should set the BFD to, prior to and after receiving the filters from the REW.

Thanks

can someone please explain how/what I should set the BFD to, prior to and after receiving the filters from the REW
The wizard automatically sets up all it can, you need to enable Midi comms which is explained in the help file, "Communicating with the BFD Pro", beyond that nothing else to do. For more general info on using the BFD check out http://bfdguide.ws/

Thanks John, I already had comms turned on as explained and I observed that it looked like the REW was simply overwriting the pervious settings. I was just unsure if there was a specific parameter to set after this the filters had been introduced particularly after a re-powerup following the unit being turned off?

I just loaded the new version on top of the previous, and I note that the "find peaks" command does not seem to be working with previously saved measurement data that I loaded.

The peaks window opens, but no peaks are identified. I uninstalled the newest version, and reinstalled the previous version, and "find peaks" does in fact work as I would expect.

Am I missing something in terms of procedures with the new version?

Brian

Thanks John, I already had comms turned on as explained and I observed that it looked like the REW was simply overwriting the pervious settings. I was just unsure if there was a specific parameter to set after this the filters had been introduced particularly after a re-powerup following the unit being turned off?
The only thing that needs re-enabling after power-off on the BFD is store enable, the wizard pops up a help box about that once per session. Otherwise all should be OK.

Regards,

I just loaded the new version on top of the previous, and I note that the "find peaks" command does not seem to be working with previously saved measurement data that I loaded.

The peaks window opens, but no peaks are identified. I uninstalled the newest version, and reinstalled the previous version, and "find peaks" does in fact work as I would expect.

Am I missing something in terms of procedures with the new version?

Brian
Could be broken! Please email me the measurement data file and I'll see what I can work out.

Whilst checking out V3.23 on the plane I found a bug in the release with the Set Target Level function, only works once per session and only if the input level or measurement level are not cal'd (rats!). I fixed it before landing but I'm having some probs uploading the fixed files to the ftp site (doesn't allow connections from other ISPs!). I'm on the case, but a workaround is to select another input, then reselect the original input (which resets the cal states of the input and output) and then just OK the warning message about not being calibrated when you run set target level (or make a measurement) - the message only comes up once per session.

I had a problem with the new version...I run a sweep and save the data...then I try and re open it and it says " not eqwiz data"....I can open data I created with previsious version..but not new data.

This seems like a great program, but damnit I can't get it to work. I'm really confused and frustrated.

With minor hiccups, I made it through setting the input volume and calibrating the spl reading. But...

1) While using "set measurement level" it says it can't achieve it's target of -18db and sets it to -3db.

2) I figured wtf and tried measuring to calibrated the soundcard, but the loopback connection isn't detected. I have the wires connected (line-out to line-in one cable, or pair of y-splitters with one cable inbetween).

I'm really pissed and feeling like an idiot. Why the hell is this so hard? I've tried checking all the mute-type settings and selections but I didn't help anything. If anyone is feeling like a good samaritan and would like to PM me to help me trouble shoot I would really appreciate it. I need to walk away from this for awhile.

Edit - also, it looked like settings in my "creative surround mixer" were changing. Weird.

Hi Jon,

On the creative mixer you need to select analog mix as the input and then select line in in the creative mixer source tab for analog mix, and ensure no others are selected - there is some info about this in the new help file, along with some other troubleshooting tips.

Regards,

Hi Jon,

On the creative mixer you need to select analog mix as the input and then select line in in the creative mixer source tab for analog mix, and ensure no others are selected - there is some info about this in the new help file, along with some other troubleshooting tips.

Regards,

Hi John,

Thank you very much for the response. I apologize for my anger above - it's been a bad week for me. :-/ I don't want my short temper today to reflect on your very good program.

Anyhow, I was using the help file step by step, and unless they changed on me, I did have the correct settings selected in the mixer as you specified.

Like I said, I'm just going to come back to it later. Maybe today, maybe not :) It's one of those weeks. (But being single again, now I have more time to play with Room EQ....ha!)

Thanks again...I'll chime in again later if I can or can't figure it out.

jonnyozero3, I have a USB creative 24 bit soundcard, and I had the exact same problem. Try setting the Room EQ input to "microphone", and not "what you hear" (whatever the hell that means). "Microphone" is counterintuitive, as JohnPM states that a microphone input will not work for this program. However, the "microphone" option in Room EQ evidently monitors both the soundcard's line and mic level inputs, and, at least for me, solved my problem. Leave only the "wave" unmuted in the creative mixer (this should be the default setting when Romm EQ boots up).

Also, I think the newest version of this program no longer uses a feedback loop (left output to left input), although I'll have to say that all my experience so far is based on the older version. I just downloaded the new one but have yet to try it out. I hope this works for ya.

John

V3.25 is now on the website, fixes the problem Ray spotted that stopped saved measurement data sets being reloaded (obfuscation drop-off in the release build) and the problem Brian found with finding peaks on loaded measurement sets. Thanks guys.

Thanks for the fix....thought I was doing something wrong :)

RayJr

This seems like a great program, but damnit I can't get it to work. I'm really confused and frustrated.

With minor hiccups, I made it through setting the input volume and calibrating the spl reading. But...

1) While using "set measurement level" it says it can't achieve it's target of -18db and sets it to -3db.

2) I figured wtf and tried measuring to calibrated the soundcard, but the loopback connection isn't detected. I have the wires connected (line-out to line-in one cable, or pair of y-splitters with one cable inbetween).

I'm really pissed and feeling like an idiot. Why the hell is this so hard? I've tried checking all the mute-type settings and selections but I didn't help anything. If anyone is feeling like a good samaritan and would like to PM me to help me trouble shoot I would really appreciate it. I need to walk away from this for awhile.

Edit - also, it looked like settings in my "creative surround mixer" were changing. Weird.

Excellent program, thanks John.

I had the same problem (w/ a different sound card - aureal audio), I was able to
solve the problem by cranking up the receiver volume. It turned out that the
amp/receiver volume should have been left at where it was, though the pop up
suggest to change the volume to "your typical listening level" - I lowered the
volume and got the "can't achieve target -18db" error.

I did have another problem when measuring soundcard response, though. It
complained about the right channel input of -41db being too high than it should
be with no signal present, tried several times and checked my soundcard
connection (R-LINEIN to R-LINEOUT) and even replaced the cable without
success, had to ignore the warning. A later popup also complained about not
able to detect right channel loopback, but was able to just ignore the warnings
and do a measurement - not sure though, if the warnings are so critical as to
affect the accuracy of the measurement.


- Lei

I did have another problem when measuring soundcard response, though. It complained about the right channel input of -41db being too high than it should be with no signal present, tried several times and checked my soundcard connection (R-LINEIN to R-LINEOUT) and even replaced the cable without success, had to ignore the warning. A later popup also complained about not able to detect right channel loopback, but was able to just ignore the warnings and do a measurement - not sure though, if the warnings are so critical as to affect the accuracy of the measurement.
- Lei
Some PC internal soundcards can have noise levels that are as high as -40dB (even saw -38dB on one PC). The software will still work OK but the measurements will not be as clean as with a better soundcard.

Hi,

I love this program! Thanks so much for making it available. I've read the whole thread but am still confused about the C-weighting compensation. Why should it be turned off for the Digital SPL meter that I have? I thought that both the Analog and Digital SPL meters had approximately the same response and C-weighting ability, and therefore both should be compensated in the Wizard. Should I turn off C-weighting on the SPL meter and apply compensation in the software?

I want to get this right because the correction factors are large at low frequencies, and no one seems to know for sure if the Analog and Digital meters read differently.

The help page in question is the 'Getting Started' page, in the 'SPL Meter Range' section.


Also, should I always leave my SPL meter set to 'slow' as opposed to 'fast'?

Thanks again!

Hi,

I love this program! Thanks so much for making it available. I've read the whole thread but am still confused about the C-weighting compensation. Why should it be turned off for the Digital SPL meter that I have? I thought that both the Analog and Digital SPL meters had approximately the same response and C-weighting ability, and therefore both should be compensated in the Wizard. Should I turn off C-weighting on the SPL meter and apply compensation in the software?

I want to get this right because the correction factors are large at low frequencies, and no one seems to know for sure if the Analog and Digital meters read differently.

The help page in question is the 'Getting Started' page, in the 'SPL Meter Range' section.


Also, should I always leave my SPL meter set to 'slow' as opposed to 'fast'?

Thanks again!
Slow vs fast does not matter as it does not affect the analog output, so can leave it on slow.

Reportedly (see Ilka's thread on using True RTA) the digital RS meter does not apply the C weighting to the analog outpout, whereas the analog meter does. Try running a test tone at a low frequency (say 20Hz or so) and compare the meter reading with the Wizard reading, if the Wizard reads higher with the C weighting off then can leave it off.

2) I figured wtf and tried measuring to calibrated the soundcard, but the loopback connection isn't detected. I have the wires connected (line-out to line-in one cable, or pair of y-splitters with one cable inbetween).

When I tried to calibrate my sound card with the latest version, I had to flip-flop the connections on my Y-cables to get it to measure. I believe the instructions said, left-out to left-in, right-out to right-in, but REW prints an error that says it can't find the calibration signal.

I fliped it so right-out to left-in and left-out to right-in and it completed the calibration without any errors. I am not sure if it calibrated appropriately though. :p

Ron

John,

I picked up Creative lab sound blaster live external 24 bit on the weekend and tried your program. It was very good and I enjoyed using it. It took me about 1.5 hour to get it to work due to soundcard. When things does not work during "Set Input Volume" or "Set Measurement Level", I found that leaving "Mixer control" open helps diagnostic the problem. I can see where things got muted when it should not be muted. (This is not Room EQ Wizard problem).

I had trouble with using receiver internal test tone to generate enough sound level during the calibration and my receiver (Yamaha and Sony) does not have option to cycle the test tone manually. I ended up making pink noise audio CD for L/R and use it to calibrate instead. My next step is to make pink noise DVD so I can test all channels. I'll be happy to share ISO files if it does not viloate Dolby copyright for using their encoder.

Later on I found out that I can use Room EQ Wizard Signal Generator to generate pink noise instead. It is much more convenient than using receiver internal test tone. Do you see any problems with that ?

Thanks again for this great gem.

-Nick

John,

I picked up Creative lab sound blaster live external 24 bit on the weekend and tried your program. It was very good and I enjoyed using it. It took me about 1.5 hour to get it to work due to soundcard. When things does not work during "Set Input Volume" or "Set Measurement Level", I found that leaving "Mixer control" open helps diagnostic the problem. I can see where things got muted when it should not be muted. (This is not Room EQ Wizard problem).

I had trouble with using receiver internal test tone to generate enough sound level during the calibration and my receiver (Yamaha and Sony) does not have option to cycle the test tone manually. I ended up making pink noise audio CD for L/R and use it to calibrate instead. My next step is to make pink noise DVD so I can test all channels. I'll be happy to share ISO files if it does not viloate Dolby copyright for using their encoder.

Later on I found out that I can use Room EQ Wizard Signal Generator to generate pink noise instead. It is much more convenient than using receiver internal test tone. Do you see any problems with that ?

Thanks again for this great gem.

-Nick
Hi Nick,

Would you let me know what was getting muted when it should not be, please? The Wizard does attempt to mute all replay sources except for Wave when it starts up to prevent feedback loops (if the Line In feed to the Line Out is not muted) or interference from other sources. Any experiences you can pass on will be added to the help files so others can avoid the same problems in future, and perhaps I can modify the Wizard to prevent them altogether. A screen capture of correct mixer settings would be very useful.

There is no problem using the Wizard's pink noise signal, you may just have to make some additional volume adjustments as the receiver's built-in signal is internally generated to the correct level to be adjusted for ref level replay, whereas the Wizard's signal has to be digitised by the receiver and the sensitivity of the analog input will alter the required volume settings - but it should all work OK.

Glad you like the program, there's much more on the way :)

Regards,

Ok guys, I don't know what it is, but I can't seem to get REW to work. I am using a Echo Gina 2in/8out sound card using a mono 1/4" from the right input to the output on the Radio Shack SPL Meter. I have a1/4" mono connected from the #4(R) out to the a BFD, which plays through my Sony 2 channel receiver. I just want to measure the response played through my system with the spl meter. No surround sound to worry about. If I play pink noise, etc. through REW, I can hear it. However, when I turn the spl meter on I get serious feedback unless I turn the meter up to say at least the 80db reading. When I select input volume on REW, I get an error saying REW can't access the input volume for the sound card-I have to do it manually, I guess. Basically, what am I doing wrong? Even when I do a sound card test, I get a graph up at the 120db scale range. Is this normal? Please help, because I can't believe I am having trouble with this!

Ok guys, I don't know what it is, but I can't seem to get REW to work. I am using a Echo Gina 2in/8out sound card using a mono 1/4" from the right input to the output on the Radio Shack SPL Meter. I have a1/4" mono connected from the #4(R) out to the a BFD, which plays through my Sony 2 channel receiver. I just want to measure the response played through my system with the spl meter. No surround sound to worry about. If I play pink noise, etc. through REW, I can hear it. However, when I turn the spl meter on I get serious feedback unless I turn the meter up to say at least the 80db reading. When I select input volume on REW, I get an error saying REW can't access the input volume for the sound card-I have to do it manually, I guess. Basically, what am I doing wrong? Even when I do a sound card test, I get a graph up at the 120db scale range. Is this normal? Please help, because I can't believe I am having trouble with this!
Would guess that the Line In is not muted on the soundcard's playback mixer, hence the feedback loop - check that only Wave is unmuted as a playback source. Can also email me the soundcard debug file (generated via the option in the soundcard menu) if you like.

Regards,

Ok, trying to run the 'measurement level' part of setup. It keeps telling me the spl metere reading is such and such, which is too hgih for no sgnal. However, the meter doesn't show anything! What am I doing wrong? Wizard keeps setting the rms level db meter to "-3" and the measurement level at "-3". Also, during the input volume setup, I had to click the red button on the Wizard's input level meter to get a reading. Is this normal? It doesn't say to do this in the help file. I also noticed that doing a sound card measurement, there was a ton of ripple on the graph. Something can't be right because the sound card (Echo Gina) has excellent frequency response. Please get me in the right direction.

I'm just trying the program too and it looks fabulous.

However, I seem to be having a similar problem as cr8250.

I'm using a Creative SB Live Value card and have that selected as the device and "line-in" as the source. Same for the output side, the card as the device and "speaker" as the source.

Using an analog RS meter as the mic and some Monster 201 cables with 1/8" mini to RCA Y-splitters. Have the mic input to the sound card right channel line-in and the output to the pre/pro using the right channel speaker out.

When I do the input level wizard, the routine says the level can't be set to -18 with the volume at max, but the bargraph shows it at -3. And if I adjust the input level manually from 1.000 to something like 0.350, the bargraph drops to about -18.

At this point I go to the SPL calibration sequence and REqW program is showing like 2-3dB on the SPL digital readout. Then, I set the SPL value in the pop-up window to 75dB like the RS meter is showing, and REqW display then matches the RS meter. However, if I make noise, the RS meter increases accordingly, but the REqW display never changes much. Even if I unplug the RS meter input from the sound card, the REqW display only drops to 64dB even though the small signal level displays to the left of the bargraph shows signal levels at like -70dB or -80dB (from memory only) and the bargraph is basically showing nothing.

Was that clear enough to suggest what might be going on? It's probably just a loose connection between the seat and the keyboard, but help figuring it out would be appreciated.

On some of the Creative cards the line in needs further configuration through the creative mixer, some of the required controls are not exposed to Windows - for example, on Audigy 2 selecting the Line In is done by selecting "Analog Mix" in the Record panel of the Basic tab of the Creative Surround Mixer then going to the Source panel and muting all the sources except for Line In. If the Wizard is not able to handle the soundcard's mixer properly an option is to do it manually. For that, select the input device but do not select the actual input (leave it set to Choose Input... ). You can then use either the Windows volume controls or the Soundcard's own mixer controls to select the line in and adjust the record level so that you get about -18dB rms on the input channel when you play a speaker cal signal at about 75dB SPL.

I've made some further simplifications of the cal process in the current dev build and altered the sequence so it should become a bit easier to get things going, hope to have this build out in a few weeks.

John, Thanks for the response. It seems I can get the input channel calibrated, but the SPL display seems like it stays way above the noise floor of the room.

For instance, I get the input level set to about -18dB and this is 0.376 on the input level control with the 75dB cal tone from the pre/pro playing. At this same time, the SPL display is reading around 75dB as well. However, when I stop the 75dB cal tone, the bargraph drops to the point no segments are showing, but the SPL display drops only to 64dB while the RS meter shows the noise floor in the room to be about 50dB or less (needle barely ever wiggles on the meter's 60dB scale measuring the noise floor).

Even if I unplug the RS meter entirely, the SPL display in the program still shows a relatively high value. If I recall, when I first start the program the SPL display is around 0, but it seems like it is getting stuck or hung up on some value once it's used.

I'll try to post a screen shot and picture later this evening.

As a possible elimination of mixer issues, I uninstalled all the Creative software so only the Windows driver and mixer are left from what I can tell. It seems to have made no change in what I'm seeing.

The Read Me mentions WinXP but I'm running Win2K Pro. Any issues with that OS and the REqW software?

What level do the numeric RMS and peak displays show when the meter is unplugged?

I'm shooting from memory, but seems they are like -70 or so. I'll be able to say for sure later on this evening.

When I do the input level wizard, the routine says the level can't be set to -18 with the volume at max, but the bargraph shows it at -3. And if I adjust the input level manually from 1.000 to something like 0.350, the bargraph drops to about -18.
I've tracked down a timing problem that is responsible for the input volume setting going wrong sometimes, fixed for the next release. Workaround for the current release is to go through the procedure then adjust the input volume control manually to get the RMS level to around -18dB. No idea what might be causing the SPL display problems, however.

Hey John, I got the REQW working, using the C-Media mixer. This program is a lot of fun! I can't believe how fast using this program instead of doing it manually with sine waves, etc.. I have a question about using the analog RS SPL Meter. How accurate are the readings at the higher frequencies? I noticed on my system, the response tends to rise above say 4khz, which I don't necessarily hear as exaggerated in everyday use. If those frequencies aren't trustworthy, at what freq. would you recommend as an upper cutoff for testing?
Thanks for the great program!

Jeff

I've tracked down a timing problem that is responsible for the input volume setting going wrong sometimes, fixed for the next release. Workaround for the current release is to go through the procedure then adjust the input volume control manually to get the RMS level to around -18dB. No idea what might be causing the SPL display problems, however.

Thanks again.

I figured out the SPL display issue... Hardly want to say, I should have seen it earlier.

The 1/8" mini y-adapter is wired wrong so the L/R are reversed. The whole thing seems to work much better when the input is on the expected channel. :)

Apologies for the false alarm.

Hey John, I got the REQW working, using the C-Media mixer. This program is a lot of fun! I can't believe how fast using this program instead of doing it manually with sine waves, etc.. I have a question about using the analog RS SPL Meter. How accurate are the readings at the higher frequencies? I noticed on my system, the response tends to rise above say 4khz, which I don't necessarily hear as exaggerated in everyday use. If those frequencies aren't trustworthy, at what freq. would you recommend as an upper cutoff for testing?
Glad to hear it is all going OK. The RS meter is fine for low frequency measurements (up to a few hundred Hz) which is the primary focus for the Wizard's EQ, it needs correction above that and there is likely to be variation between units. If you look at the SPL Meter correction Tables sticky on the subwoofer forum, or Ilkka's thread on TrueRTA for dummies, or the ETF website, there are correction files for the RS meter.