or Connect
AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How To Interpret Graphs
New Posts  All Forums:Forum Nav:

Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How... - Page 250

post #7471 of 9494
^@AustinJerry
Sorry, should have explained. I trust audyssey (I am using the pro kit), but in my case my trims are maxed at -12dB on my LCR due to my very efficient JTR T12 speakers. My surrounds are Def Tech ProMonitor 1000s and are set by audyssey at SR 0 & SL -1.5. I am using the internal amps in my Denon AVR4311 and can't use attenuators so I am doing as others suggested in trying to lower the trims by adjusting the MV @ 0 to 80dB and using -5 MV as my reference. I attempted to make these adjustments using my rat shack spl meter and the AVR test tones, but was told that the test tones were pre-audyssey, thus my rather crude attempt at using REW to accomplish this task - I admit I am a complete nubee at REW ( I have just finished reading your guide and setting it all up) and am fairly new at any measuring and Room correction applications.

FWIW - I assumed that audyssey is attempting to set all my speakers to 75dB at 0 MV and so I added 5dB to the SR & SL and measured from MLP using the settings noted in my previous post for the REW generator and got 77~78dB (not sure why it isn't 80dB) so I then adjusted the trims on the rest of my speakers to match the 78dB level.
Edited by HTPCat - 12/20/13 at 7:40pm

Gear mentioned in this thread:

post #7472 of 9494
Quote:
Originally Posted by AustinJerry View Post

You will find more people are using Windows PC's, so the support will be more robust.

OK! Thanks AJ.. PC it is. This will be my first time with REW and a mic. Although I've used a SMS-1 to calibrations for my subs in the past. I've currently now using a Antimode dual core. Which it does have its own screen and graph on the unit it self. I really want to experiment more with calibration of sub placements and time delay with the mains. Antimode DC has many adjustments and eq. Execept the ability to gibe me real time response. So hopefully this journey or new hobby I've took the path to with REW will be interesting and enjoyable. I will be on this thread a lot for lots of guidance and recommendations. Thanks..
post #7473 of 9494
Quote:
Originally Posted by HTPCat View Post

^@AustinJerry
Sorry, should have explained. I trust audyssey (I am using the pro kit), but in my case my trims are maxed at -12dB on my LCR due to my very efficient JTR T12 speakers. My surrounds are Def Tech ProMonitor 1000s and are set by audyssey at SR 0 & SL -1.5. I am using the internal amps in my Denon AVR4311 and can't use attenuators so I am doing as others suggested in trying to lower the trims by adjusting the MV @ 0 to 80dB and using -5 MV as my reference. I attempted to make these adjustments using my rat shack spl meter and the AVR test tones, but was told that the test tones were pre-audyssey, thus my rather crude attempt at using REW to accomplish this task - I admit I am a complete nubee at REW ( I have just finished reading your guide and setting it all up) and am fairly new at any measuring and Room correction applications.

FWIW - I assumed that audyssey is attempting to set all my speakers to 75dB at 0 MV and so I added 5dB to the SR & SL and measured from MLP using the settings noted in my previous post for the REW generator and got 77~78dB (not sure why it isn't 80dB) so I then adjusted the trims on the rest of my speakers to match the 78dB level.

 

OK, I remember your unusual situation, sorry.  Things would be simple if you had an external SPL meter, like the Radio Shack one.  You don't have an SPL meter?

 

Trouble with using the UMIK-1 is you need to make sure it is calibrated.  Supposedly, the calibration file that comes with the mic should have a "sensitivity factor", which REW uses to auto-calibrate the mic.  The sensitivity factor is stored at the beginning of the calibration file, as per this example:

 

 

The problem is that I don't see this sensitivity factor in the calibration files from Cross Spectrum Labs.  Without the sensitivity factor, your mic isn't calibrated.  If the mic isn't calibrated, you can't run the test as you described.  To calibrate the mic, you need an external reference, i.e. a SPL meter.  Then, if you had an external SPL meter, you wouldn't need to do the test using REW.  Vicious circle, isn't it?

 

Recommendation:  invest in an inexpensive SPL meter (you really need one).  Or, better yet, if you have the budget, invest in an external power amp, which would allow you to use line-level attenuators, which would be the elegant, reliable solution to the problem of super-efficient speakers.  An external amp like the Emotiva is quite reasonably priced, and would be a nice edition to your system.

 

Edit:  BTW, PDXRealtor provided this post on how to add the mic's sensitivity factor to the CSL calibration files, if you want to give it a try.  I tried to do that, but even with this sensitivity factor, I found that the UMIK-1 levels were still not accurate.  That is why I always calibrate my mic using an external SPL.


Edited by AustinJerry - 12/20/13 at 8:41pm
post #7474 of 9494
Quote:
Originally Posted by AustinJerry View Post

OK, I remember your unusual situation, sorry.  Things would be simple if you had an external SPL meter, like the Radio Shack one.  You don't have an SPL meter?

Trouble with using the UMIK-1 is you need to make sure it is calibrated.  Supposedly, the calibration file that comes with the mic should have a "sensitivity factor", which REW uses to auto-calibrate the mic.  The sensitivity factor is stored at the beginning of the calibration file, as per this example:




The problem is that I don't see this sensitivity factor in the calibration files from Cross Spectrum Labs.  Without the sensitivity factor, your mic isn't calibrated.  If the mic isn't calibrated, you can't run the test as you described.  To calibrate the mic, you need an external reference, i.e. a SPL meter.  Then, if you had an external SPL meter, you wouldn't need to do the test using REW.  Vicious circle, isn't it?

Recommendation:  invest in an inexpensive SPL meter (you really need one).  Or, better yet, if you have the budget, invest in an external power amp, which would allow you to use line-level attenuators, which would be the elegant, reliable solution to the problem of super-efficient speakers.  An external amp like the Emotiva is quite reasonably priced, and would be a nice edition to your system.

Edit:  BTW, PDXRealtor provided this post on how to add the mic's sensitivity factor to the CSL calibration files, if you want to give it a try.  I tried to do that, but even with this sensitivity factor, I found that the UMIK-1 levels were still not accurate.  That is why I always calibrate my mic using an external SPL.
I do have a rat shack spl digital meter that I can and have used (just wasn't very confident in it's accuracy) I was using REW as a double check because I didn't have much confidence in the meter and audio test disc which would do a sweep of about 15 sec for each speaker I couldn't get consistent measurements. I didn't buy my umik-1 from cross spectrum, I bought it direct from minidsp and the calibration file provided does have the sensitivity factor will this work or do I need to do the calibration as shown in your guide? And if it is calibrated correctly are the settings in the REW generator that I posted the correct ones to use?

FWIW I just sold a Emotiva XPA-5 and a Cinepro 3k6-3 amps as I just wasn't really using them and I always had a hiss coming from my speakers when one of the external amps were in the system which couldn't be heard when playing sounds, but if nothing was playing like when I was using the HTPC for web browsing or just displaying photos I could hear the hiss which drove me crazy eek.gif So for me it has been a vicious circle already cool.gif
Edited by HTPCat - 12/20/13 at 9:46pm
post #7475 of 9494
Quote:
Originally Posted by HTPCat View Post

^@AustinJerry
Sorry, should have explained. I trust audyssey (I am using the pro kit), but in my case my trims are maxed at -12dB on my LCR due to my very efficient JTR T12 speakers.

Does Audyssey even produce valid filters in such a situation?
post #7476 of 9494
Quote:
Originally Posted by AustinJerry View Post
 
Quote:
Originally Posted by HTPCat View Post

I am trying to level match my 5 speakers to 80dB at 0 on MV. I have Audyssey on and DVOL/DEQ off. I am using the SPL meter in REW and the noise generator set up as indicated in this snip - are these the correct settings to use? I am using a USB umik-1 at my MLP on a boom.

 

Well, I am a firm believer that the Audyssey calibration does a very good job at setting each speaker's trim such that a measurement at the MLP would show 75dB.  So, if your objective is to have 80dB at the MLP, simply add 5dB to the speaker trims set by Audyssey and be done with it.  No need to get any more complicated.

 

Of course, one might assume that you don't have confidence in the trims set by Audyssey.  Is this the case?  And why would you want the trims 5dB higher?  If things sound too soft, why not simply raise the master volume?  Come on, tell us the whole story if you want us to help.

 

Hi Jerry,

 

I sent him over here from another thread. He is maxing out the trims on his LCR and is trying to use the procedure outlined in the Audyssey FAQ for compensating for it.

 

It's the procedure under 'satellite speakers maxed out' in this FAQ answer:

 

e)6.   What do I do if my trim levels are at the limits of their adjustment ('maxed out')?

 

EDIT - when will I ever ***ing learn to read ahead!

post #7477 of 9494
Quote:
Originally Posted by markus767 View Post

Does Audyssey even produce valid filters in such a situation?

Exactly my question, and I know of no way to know for sure. It is unfortunate that the OP no longer has the XPA-5.
post #7478 of 9494
Quote:
Originally Posted by AustinJerry View Post

Exactly my question, and I know of no way to know for sure.

You could measure it by varying the input sensitivity of the amp.
post #7479 of 9494
Quote:
Originally Posted by markus767 View Post

Does Audyssey even produce valid filters in such a situation?
According to Chris (ask Audyssey) the only thing that is not valid are the trims (potentially, as -12 might be the level it was supposed to be) I would think all I really need to do is level match all speakers to any number that gets my fronts off -12. This is of course with audyssey engaged and dvol/deq off.

Quote:
Originally Posted by AustinJerry View Post

Exactly my question, and I know of no way to know for sure. It is unfortunate that the OP no longer has the XPA-5.
post #7480 of 9494

Well, it is certainly worth a try.  I am doing some REW stuff this morning, and will provide my input on how you might use REW to set the trims.  Should be a little later this morning.

post #7481 of 9494
Quote:
Originally Posted by HTPCat View Post

According to Chris (ask Audyssey) the only thing that is not valid are the trims (potentially, as -12 might be the level it was supposed to be) I would think all I really need to do is level match all speakers to any number that gets my fronts off -12. This is of course with audyssey engaged and dvol/deq off.

Trust but verify smile.gif Could you measure the preamp output of one of the maxed out channels?
post #7482 of 9494
Quote:
Originally Posted by AustinJerry View Post

Well, it is certainly worth a try.  I am doing some REW stuff this morning, and will provide my input on how you might use REW to set the trims.  Should be a little later this morning.

Simple version:
Play the internal test tones from an AVR and set the trim for each speaker so a SPL meter set to C-weighting/slow reads 75dB SPL at the listening position. For this to work the internal test signal needs to be calibrated to -30dB FS RMS.

Audiophool version:
Send periodic pink noise at -20dB FS RMS from REW to a single speaker:



Put a level-calibrated mic at the listening position and observe its input with REW's RTA set to 1/1 octave resolution (make sure the FFT length in Generator und RTA match otherwise the graph will wiggle around and produce meaningless results):



Adjust the trims of that speaker in the AVR until the 1kHz band reads 75dB. Why 75dB and not 85dB (105dB reference level - 20dB test signal = 85dB)? The RTA has 10 bands and those single bands need to sum to 85dB (75 + 10 log 10 = 85).
Repeat for every speaker.
At the end adjust the sub so its 63Hz band reads the same as the 1kHz band of one of the satellites.
Edited by markus767 - 12/21/13 at 8:21am
post #7483 of 9494
Quote:
Originally Posted by markus767 View Post
 
Quote:
Originally Posted by AustinJerry View Post

Well, it is certainly worth a try.  I am doing some REW stuff this morning, and will provide my input on how you might use REW to set the trims.  Should be a little later this morning.

Simple version:
Play the internal test tones from an AVR and set the trim for each speaker so a SPL meter set to C-weighting/slow reads 75dB SPL at the listening position. For this to work the internal test signal needs to be calibrated to -30dB FS RMS.

Audiophool version:
Send periodic pink noise from REW at -20dB FS RMS to a single speaker:



Put a level-calibrated mic at the listening position and observe it with REW's RTA set to 1/1 octave resolution (make sure the FFT length in Generator und RTA match otherwise the graph will wiggle around and produce meaningless results):



Adjust the trims of that speaker in the AVR until the 1kHz band reads 75dB. Why 75dB and not 85dB? The RTA has 10 bands and those need to sum to 85dB (75 + 10 log 10 = 85). Repeat for every speaker.
At the end adjust the sub so its 63Hz band reads the same as the 1kHz band of one of the satellites.

 

@Jerry...

 

I wonder if this excellent post ought to be an addendum in the Guide, Jerry, as I am sure people will want to refer to it in the future and it will just become lost in the thread if it isn’t immortalised somewhere? I don't mean a link to it because links can rot, but the actual content.  Just a thought...

post #7484 of 9494
Quote:
Originally Posted by markus767 View Post

At the end adjust the sub so its 63Hz band reads the same as the 1kHz band of one of the satellites.

I always wondered about the "correct" frequency for level-matching subs to mains.

But now, of course, I have to ask: why 63Hz?

post #7485 of 9494
Quote:
Originally Posted by kbarnes701 View Post
 

 

@Jerry...

 

I wonder if this excellent post ought to be an addendum in the Guide, Jerry, as I am sure people will want to refer to it in the future and it will just become lost in the thread if it isn’t immortalised somewhere? I don't mean a link to it because links can rot, but the actual content.  Just a thought...

 

Keith, I am not averse to including a useful tip, as long as I understand and agree.  Since I was involved in some REW work this morning, I followed Markus' guidelines.

 

First, the Generator configuration:

 

 

Second, running the RTA as configured per the recommendation:

 

 

IIUC, I need to examine the RTA level at 1K and adjust the speaker trim until that level reads 75dB.  In the example shown above, I am seeing approximately 68dB @ 1K, so this would suggest that I need to raise that speaker's trim by 7dB.

 

However, look at this data:

 

 

This screen capture shows the REW SPL and its logger capturing the internal AVR speaker test tone (which, if the Audyssey calibration is accurate, should read 75dB).  Note that the speaker indeed reads very close to 75dB, using the same mic calibration as the RTA test initially.

 

So, according to the SPL, my speaker trim is spot-on.  According to the RTA, it is 7dB low.  Either I am not following Markus' procedure correctly, or his procedure is flawed.  Let's see what Markus has to say.  In the meantime, I encourage others to conduct this same test.

post #7486 of 9494
Quote:
Originally Posted by AustinJerry View Post
 
Quote:
Originally Posted by kbarnes701 View Post
 

 

@Jerry...

 

I wonder if this excellent post ought to be an addendum in the Guide, Jerry, as I am sure people will want to refer to it in the future and it will just become lost in the thread if it isn’t immortalised somewhere? I don't mean a link to it because links can rot, but the actual content.  Just a thought...

 

Keith, I am not averse to including a useful tip, as long as I understand and agree.  Since I was involved in some REW work this morning, I followed Markus' guidelines.

 

 

Agreed. Good experimental result there, Jerry.

post #7487 of 9494
Quote:
Originally Posted by LastButNotLeast View Post

I always wondered about the "correct" frequency for level-matching subs to mains.
But now, of course, I have to ask: why 63Hz?

The correct way would be to look at the whole in-room frequency response.
post #7488 of 9494
Quote:
Originally Posted by AustinJerry View Post



However, look at this data:




This screen capture shows the REW SPL and its logger capturing the internal AVR speaker test tone (which, if the Audyssey calibration is accurate, should read 75dB).  Note that the speaker indeed reads very close to 75dB, using the same mic calibration as the RTA test initially.

So, according to the SPL, my speaker trim is spot-on.  According to the RTA, it is 7dB low.  Either I am not following Markus' procedure correctly, or his procedure is flawed.  Let's see what Markus has to say.  In the meantime, I encourage others to conduct this same test.

I would like to do this with my internal Denon 4311 test tone with both my RS meter & REW spl meter. I have been told however that the test tones from the AVR are always pre-audyssey is this incorrect? I am going to do it now and see what I get using both. In the end all that really matters is that all my speakers are level matched right? Audyssey has already set all the filters and I just set my volume to a comfortable level usually around -12 to -15 and perhaps set the RLO to 5dB and I should be good.
post #7489 of 9494
Quote:
Originally Posted by AustinJerry View Post

So, according to the SPL, my speaker trim is spot-on.  According to the RTA, it is 7dB low.  Either I am not following Markus' procedure correctly, or his procedure is flawed.

...or the test signal going to your AVR isn't really -20dB FS RMS. Did you verify that the level of the test signal leaving your computer is correct?
post #7490 of 9494
Quote:
Originally Posted by AustinJerry View Post

So, according to the SPL, my speaker trim is spot-on.  According to the RTA, it is 7dB low. 
When you use an RTA, the level in each band will be less than the total SPL. For example, in cinema calibrations, the wideband pink reads 85 dB, but the 1/3d-octave RTA reads 70 dB in each band. And that is correct because it's 1/31th the energy. If you take 10log(1/31) you get -15 dB.

In your case, the bands are 1 octave, so they will read -10 dB compared to wideband level. As your plot shows, the level varies from band to band. THX (and others) recommend covering at least the 2 octaves from 500 Hz to 2 kHz.

Lastly, it looks like the test signal is wideband. But the SPL meter is C weighted. That means the meter is reading a little lower than the RTA because the bass and HF is rolled off. Could account for around 2 dB, which is consistent with the 77.1 RTA SPL.

IMHO, the most accurate way to calibrate speaker levels/balance is with narrow band pink, as found on THX Optimizer and the new Spears and Munsil BD, with an SPL meter. It avoids undue influences of bandwidth and response aberrations at the extremes.
post #7491 of 9494
Ok.. My mic from cross spectrum is on its way. Still need to get a LT. The problem though is the wifey don't want me to get a PC. She wants me to get a Mac Pro. Now I know the Pro lines have USB ports and a HDMI port. I told her I need a windows base LT to do my work with REW. She's a MAC user. And she said I could run a parallel of OS X or Windows. She showed me on her MAC that she could boot to Windows 7. Friggin amazing. Now question, will using a MAC book with Windows work? If so, I guess I still would down load the Windows REW and the ASIO4ALL drivers?
post #7492 of 9494
Quote:
Originally Posted by Roger Dressler View Post

When you use an RTA, the level in each band will be less than the total SPL. For example, in cinema calibrations, the wideband pink reads 85 dB, but the 1/3d-octave RTA reads 70 dB in each band. And that is correct because it's 1/31th the energy. If you take 10log(1/31) you get -15 dB.

In your case, the bands are 1 octave, so they will read -10 dB compared to wideband level. As your plot shows, the level varies from band to band. THX (and others) recommend covering at least the 2 octaves from 500 Hz to 2 kHz.

Lastly, it looks like the test signal is wideband. But the SPL meter is C weighted. That means the meter is reading a little lower than the RTA because the bass and HF is rolled off. Could account for around 2 dB, which is consistent with the 77.1 RTA SPL.

IMHO, the most accurate way to calibrate speaker levels/balance is with narrow band pink, as found on THX Optimizer and the new Spears and Munsil BD, with an SPL meter. It avoids undue influences of bandwidth and response aberrations at the extremes.

But Jerry is sending a -20dB FS RMS signal from within REW. So he should see 85dB (or about 75dB per band) not 77.1dB.
post #7493 of 9494

Back to basics - don't forget the simple stuff!

 

The following graphs were generated by changing only the toe-in of the two front speakers:

 

 

There aren't too many room modifications that are easier than this!

Michael

post #7494 of 9494
I see what looks like seven measurements. Are these before and after and, if so, which is which? BTW, you can include the legend on your screen captures.
post #7495 of 9494

I know. It doesn't really matter, I just wanted to point out that something as minor as the angle of the front speakers can make a dramatic difference. FWIW, I decided on the gold line.

post #7496 of 9494
Quote:
Originally Posted by LastButNotLeast View Post

Back to basics - don't forget the simple stuff!

The following graphs were generated by changing only the toe-in of the two front speakers:




There aren't too many room modifications that are easier than this!
Michael

Are you measuring one speaker at a time?
post #7497 of 9494
Quote:
Originally Posted by AustinJerry View Post

I see what looks like seven measurements. Are these before and after and, if so, which is which? BTW, you can include the legend on your screen captures.

Already moving on? Did you find out what was causing the erroneous regardings you've got yesterday?
There's something wrong with your measuring setup.
Edited by markus767 - 12/22/13 at 8:54am
post #7498 of 9494
When I am looking at the waterfall plots, what sort of decay level (dB) should I be trying to attain in what time? I have read that RT60 isn't practical for home rooms, outside of a mansion's grand ballroom... What "lines" would be significant to me in the RT60 graphs? I am totally ignorant of how to judge these, and also am not sure how to relate them to the waterfalls. Is a reasonable attainable decay time (in a "regular" home room) somewhat different for bass than for the highs? Thanks.
post #7499 of 9494
^
I'm aiming for around 0.2-0.4s. Generally the focus should be on lower frequencies.
post #7500 of 9494
Quote:
Originally Posted by cfraser View Post

When I am looking at the waterfall plots, what sort of decay level (dB) should I be trying to attain in what time? I have read that RT60 isn't practical for home rooms, outside of a mansion's grand ballroom... What "lines" would be significant to me in the RT60 graphs? I am totally ignorant of how to judge these, and also am not sure how to relate them to the waterfalls. Is a reasonable attainable decay time (in a "regular" home room) somewhat different for bass than for the highs? Thanks.

I know you want the decay to be fairly even. Here is an example of poor decay:



As far as how many db down at what timing, I know of no standard recipe. It is typical for the lower registers to decay more slowly that higher ones.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Audio theory, Setup and Chat

Gear mentioned in this thread:

AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How To Interpret Graphs