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 73

post #2161 of 9495
Quote:
Originally Posted by amirm View Post

You will get the same phase display either way but in GD you don't see your frequency response. To understand what is going on, we need to see the excess phase in relation to our frequency response. That way we can correlate the frequency response variations with the excessive phase. If there is none, and we have response that is not smooth, we can then dial it out. Just looking at the phase alone won't enable that. So you need to use SPL+Phase.
No, the unrwap button gets rid of those sawtooth bounces in the phase display. The phase jumps 360 degrees at those spots. Unwrap simply unwinds that sudden jump. We are not interested in phase reversals here. Getting rid of them makes it a lot easier to interpret the graph.

The unsmoothing comment relates to leaving your *frequency response* measurement unfiltered. That way you see all the variations in it, including that of phase. So select Graph->Remove Smoothing when looking at this.

Wow, that's true. In GD I can only see the group delay graph and not the frequency response. Had me going back and forth between SPL and GD windows for a while which is no fun, and there is no step-by-step in the REW Help file to explain this. Thank you so much, amirm.

Gear mentioned in this thread:

post #2162 of 9495
Quote:
Originally Posted by Pres2play View Post

Wow, that's true. In GD I can only see the group delay graph and not the frequency response. Had me going back and forth between SPL and GD windows for a while which is no fun, and there is no step-by-step in the REW Help file to explain this. Thank you so much, amirm.
You are welcome. smile.gif I really think Room Eq Wizard is no Wizard smile.gif. If a true Wizard was added to it, it would be wonderful. As it is, it generates lots and lots of raw results which requires significant knowledge of acoustics, signal process and psychoacoustics to properly generate and use. Much of that complexity can be hidden and the right data presented.

I think feature creep is part of the problem in that features are added on top of many other features resulting in duplicate data as you observe.

Mind you, it is a superb implementation of these techniques and considering that it is free, it is a true gift to all of us. It is hard to imagine that it exist the way it does. Just does not target its audience as the name "wizard" implies.
post #2163 of 9495
Quote:
Originally Posted by Pres2play View Post


I'll make some calls and see. Also need to explore what's out there, the 5509 looks tempting too.

 

If you are not restricted by a budget constraint, then there are better options than the 818, IMHO.  The 818 does not have SubEQ HT, that component of Audyssey XT32 that independently equalizes two sub channels.  The 5509 is a fine system, as well as the Denon 4520, or the Denon 4311 (if you can find one), or the Marantz Pre-pro.  As you narrow your choice, there are many people here on AVS in the specific AVR threads who will be glad to answer your questions.

post #2164 of 9495
Quote:
Originally Posted by turboman123 View Post


hello Jerry
I am a beginner with REW. So I am trying to understand what I am measuring. I do not understand what you do with "mic calibrated to xdB".
So let me explain what I see in my graphs. The issue is not the level of sub compared to mains(above crossover). What I am wondering about is the level of LFE compared to mains(redirected to sub below crossover).
Here are some more measuremets I did. Mains set to smal with crossover at 110Hz. For I signal I sent to the sub, I lowered the AVR volume by 10dB. Despite this, the level of the measurement of signal to sub, is the same as the mains redirected to sub below crossover.
Graph: no smoothing.

woof.jpg 97k .jpg file
According to the link I included in my previous post, that is how it should be for an all HDMI signal because the receiver adds 10dB to a signal sent direct to the subwoofer. As I lowered the volume in my AVR by 10dB, the measurement is then the same as a redirected main below crossover, because the AVR adds 10dB to the signal sent directly to sub over HDMI..

That is why I raised the question: in order to compare sub measurement and main measurement, using HDMI, I have to lower the volume for the sub by 10dB in the AVR. Do you agree?

About the dips, I will investigate that once I have a correct interpretation of the measurements. Presently, the sub is in a corner.

 

Jkasanic provided a good response to your questions.  Do I agreed that you need to lower the volume by 10dB?  No, I already answered that question.  For the time being, forget the LFE channel and focus on measuring the response of left+subs, right+subs, and center+subs by outputting the REW signal to the left channel, then the right channel (both while in stereo mode), and finally both the REW left and right channels while in PLII Cinema mode (which will measure the center speaker).  As Keith and I have both requested, publish each of these three measurements (separate graphs, please), and let us see how your system is performing.

 

Collectively, the members of this thread have developed the REW guide to help beginners like yourself--please follow it for the time being.

post #2165 of 9495
Quote:
Originally Posted by amirm View Post


You are welcome. smile.gif I really think Room Eq Wizard is no Wizard smile.gif. If a true Wizard was added to it, it would be wonderful. As it is, it generates lots and lots of raw results which requires significant knowledge of acoustics, signal process and psychoacoustics to properly generate and use. Much of that complexity can be hidden and the right data presented.

I think feature creep is part of the problem in that features are added on top of many other features resulting in duplicate data as you observe.

Mind you, it is a superb implementation of these techniques and considering that it is free, it is a true gift to all of us. It is hard to imagine that it exist the way it does. Just does not target its audience as the name "wizard" implies.

 

+1, Amir!

post #2166 of 9495
Quote:
Originally Posted by AustinJerry View Post

Quote:
Originally Posted by Pres2play View Post


I'll make some calls and see. Also need to explore what's out there, the 5509 looks tempting too.

 

If you are not restricted by a budget constraint, then there are better options than the 818, IMHO.  The 818 does not have SubEQ HT, that component of Audyssey XT32 that independently equalizes two sub channels.  The 5509 is a fine system, as well as the Denon 4520, or the Denon 4311 (if you can find one), or the Marantz Pre-pro.  As you narrow your choice, there are many people here on AVS in the specific AVR threads who will be glad to answer your questions.

 

There's no doubt in my mind that the 5509 is a superior unit* to the 818 (it costs three times as much!) but the lack of SubEQ is not such a big deal IMO, especially if the user has only one sub (in which case the ability to set two subs is irrelevant of course).

 

SubEQ doesn't really 'equalise' two sub channels independently - XT32 equalises the two subs as one but SubEQ does set the sub distances and levels independently.

 

If the user has two subs, all that SubEQ does is set the levels and distances (delays) separately - it then EQs the subs together as you know. It is a trivial matter to set the sub levels independently before running Audyssey and Audyssey rarely seems to get the distance settings right for two subs anyway - hence the usefulness of the 'sub distance tweak' procedure. If one is going to alter the sub distances after running Audyssey, it hardly matters what Audyssey sets them to - although this does assume some measuring gear is available.  If one has two subs and they are identical and equidistant from the MLP, then SubEQ does more or less nothing. I think SubEQ's abilities are overrated and for less than a grand, the 818 offers XT32 which is the real biggy.

 

Of course, if money is no object, then I’d go for the Onkyo 5509 or the Marantz 8801 as pure prepros or the Denon 4520/4311, as you suggest, for AVRs with preouts.

 

*Declaration of interest: I have a 5509 and am very happy with it :)


Edited by kbarnes701 - 3/25/13 at 2:46pm
post #2167 of 9495

Good points, Keith.  Fully agree.

post #2168 of 9495
Quote:
Originally Posted by AustinJerry View Post

Good points, Keith.  Fully agree.

 

:)  I wasn't trying to be argumentative, Jerry (I re-read my post and it seemed as if I might come across that way) so I am glad you interpreted my intent properly. The 818 is a remarkable unit at the price and the cheapest way into XT32 (in a new unit) which is why I often suggest it for consideration. And it has preouts too. I know you will agree with me that XT32 is well worth having!!  But yes, if money is not an issue, I'd probably aim higher.

post #2169 of 9495
Quote:
Originally Posted by AustinJerry View Post

Jkasanic provided a good response to your questions.  Do I agreed that you need to lower the volume by 10dB?  No, I already answered that question.  For the time being, forget the LFE channel and focus on measuring the response of left+subs, right+subs, and center+subs by outputting the REW signal to the left channel, then the right channel (both while in stereo mode), and finally both the REW left and right channels while in PLII Cinema mode (which will measure the center speaker).  As Keith and I have both requested, publish each of these three measurements (separate graphs, please), and let us see how your system is performing.

Hello Jerry,
Thanks for your reply. Sorry, now I am getting really confused.

The guide says to take measurements (page 65) 1, 3, 7 and 9: left, right, center and sub (I have no Audyssey, so I skipped measurement 2,4, 8 and 10).
Now you are instructing to forget sub(LFE) and to focus on left+sub, right+sub and center+sub. But left+sub and right+sub is not in the guide, and center+sub only as the last measurement.

Some more questions: for left+right you say to use PLIICinema. But the guide says to use PLIICinema for non-HDMI or if HDMI does not support multichannel. My HDMI supports multichannel however.

For left+sub and right+sub, you say to put the AVR in stereo mode. What is the reason for this? Is that needed when my HDMI supports multichannel?

Can you clarify these items please? Thanks in advance for your patience.

I am also struggling with guideline 2 on the same page: output at least 80dB in order to have enough headroom above the noise floor, assuming that is 40dB. Yesterday, my noise floor was around 60dB. I have a UMIK-1. I thought my room was reasonable quiet enough. I will measure again today and report back.
post #2170 of 9495
Quote:
Originally Posted by turboman123 View Post


Hello Jerry,
Thanks for your reply. Sorry, now I am getting really confused.

The guide says to take measurements (page 65) 1, 3, 7 and 9: left, right, center and sub (I have no Audyssey, so I skipped measurement 2,4, 8 and 10).
Now you are instructing to forget sub(LFE) and to focus on left+sub, right+sub and center+sub. But left+sub and right+sub is not in the guide, and center+sub only as the last measurement.

Some more questions: for left+right you say to use PLIICinema. But the guide says to use PLIICinema for non-HDMI or if HDMI does not support multichannel. My HDMI supports multichannel however.

For left+sub and right+sub, you say to put the AVR in stereo mode. What is the reason for this? Is that needed when my HDMI supports multichannel?

Can you clarify these items please? Thanks in advance for your patience.

I am also struggling with guideline 2 on the same page: output at least 80dB in order to have enough headroom above the noise floor, assuming that is 40dB. Yesterday, my noise floor was around 60dB. I have a UMIK-1. I thought my room was reasonable quiet enough. I will measure again today and report back.

 

OK, I can understand part of your confusion.  Since I have only two-channel support (not 8-channel), I was giving you the wrong instructions.

 

So, let's start again.  Take these measurements:

 

- Single REW output to the left channel.  Assuming your left speaker is set to small, the full sweep 15-20,000Hz will include the bass response because of bass management.

- Single REW output to the right channel.

- Single REW output to the center channel.

- Single REW output to the LFE channel, 15-300Hz.  (Without any 10dB correction, please!)

 

Publish the first three measurements, 15-20,000Hz, 1/6 smoothing, separate graphs.  Publish the fourth measurement, 15-300Hz, no smoothing.

 

Stick with 80Hz as the calibration point for now.  60dB is not likely your noise floor, it is the lowest that the UMIK-1 mic will measure.  There was considerable discussion earlier in the thread regarding the UMIK-1 and the UMM-6 USB mics, and the fact that they will not measure as low as legacy mics like the EMM-6, which will go down to 40dB and below.  For your initial set of graphs, don't let this be any concern.  Just take the measurements and post them.


Edited by AustinJerry - 3/26/13 at 8:32am
post #2171 of 9495
Quote:
Originally Posted by AustinJerry View Post

OK, I can understand part of your confusion.  Since I have only two-channel support (not 8-channel), I was giving you the wrong instructions.

So, let's start again.  Take these measurements:

- Single REW output to the left channel.  Assuming your left speaker is set to small, the full sweep 15-20,000Hz will include the bass response because of bass management.
- Single REW output to the right channel.
- Single REW output to the center channel.
- Single REW output to the LFE channel, 15-300Hz.  (Without any 10dB correction, please!)

Publish the first three measurements, 15-20,000Hz, 1/6 smoothing, separate graphs.  Publish the fourth measurement, 15-300Hz, no smoothing.

Stick with 80Hz as the calibration point for now.  60dB is not likely your noise floor, it is the lowest that the UMIK-1 mic will measure.  There was considerable discussion earlier in the thread regarding the UMIK-1 and the UMM-6 USB mics, and the fact that they will not measure as low as legacy mics like the EMM-6, which will go down to 40Hz and below.  For your initial set of graphs, don't let this be any concern.  Just take the measurements and post them.
Thanks Jerry.
Here is the result for left, right and center, 1/6 smoothing, crossover 110Hz (I checked several crossovers, and 110Hz gave the most level response around the crossover point. GoldenEar recommends 120Hz).

leftrightcenter.jpg 76k .jpg file

And sub, without 10dB correction, no smoothing:

sub(lfe).jpg 63k .jpg file

What about the level. The around 80dB is Ok, but headroom for front was 50dB, which is not according to guideline. Noise floor was 60dB.
post #2172 of 9495
Quote:
Originally Posted by AustinJerry View Post

60dB is not likely your noise floor, it is the lowest that the UMIK-1 mic will measure.  There was considerable discussion earlier in the thread regarding the UMIK-1 and the UMM-6 USB mics, and the fact that they will not measure as low as legacy mics like the EMM-6, which will go down to 40Hz and below.  For your initial set of graphs, don't let this be any concern.  Just take the measurements and post them.

 

40dB you meant to type, Jerry :)

post #2173 of 9495
My previous measurements showed some strong peak and dips for the subwoofer. So I did some more measurements.
First, I wanted to check out some controls I have on my sub B&W ASW608. Following shows (with sub in corner), 1/6 smoothing:
- bass extension C & EQ A
- bass extension A & EQ A
- bass extension C & EQ B

Bass extension A caused a rattle lowndown. Effect of EQ was minor. So I left it at bass extension C & EQ A.

Next, I wanted to check another position of the sub, which was at the wall but more in the middle of the room, 1/6 smoothing:

I also checked waterfalls (graph up to 200Hz, not according to guideline, but I did not make front+sub at that stage).
First the corner position:

Then midposition:

Based on this, sub in middle of wall looked better than the corner.

But the guideline shows a waterfall of main+sub. So I made measurements of FrontLeft + sub with 110Hz crossover, 1/6 smoothing:

And the waterfalls Frontleft+subcorner (now with proper 300Hz upper limit):

And waterfall for sub middle position:

My conclusion is that the sub in the middle of the wall position is better than the corner position for future optimization (crossover, PEQ, DRC ...). Any comments on this conclusion?
Edited by turboman123 - 3/26/13 at 4:05am
post #2174 of 9495
Quote:
Originally Posted by turboman123 View Post

Thanks Jerry.
Here is the result for left, right and center, 1/6 smoothing, crossover 110Hz (I checked several crossovers, and 110Hz gave the most level response around the crossover point. GoldenEar recommends 120Hz).

leftrightcenter.jpg 76k .jpg file

And sub, without 10dB correction, no smoothing:

sub(lfe).jpg 63k .jpg file

What about the level. The around 80dB is Ok, but headroom for front was 50dB, which is not according to guideline. Noise floor was 60dB.

 

Those front left, front right and centre are just those speakers?  It looks as if the sub is playing too. To get a reading for the main speakers only it is necessary to physically switch off the sub or bass management sends redirected bass to the sub.  In the preference pane you will have Audio Output 1 set for the Left speaker, 2 for the right speaker, 3 for the centre speaker - in the drop down where I have circled in blue on the screen cap below.

 

 

 

For the LCR measurements ONLY, the drop-down I have circled in red should be left blank, as in the screenie. The sub should be physically disconnected for these LCR measurements.

 

To measure the sub only, select Output 4 in the drop-down circled in blue and nothing in the drop-down box circled in red. 

 

To measure L or R or C with the sub, the sub needs to be connected and powered on and then you select 1 for L, 2 for R, 3 for C in the drop-down circled in blue and NOTHING in the drop-down circled in red.

post #2175 of 9495
Quote:
Originally Posted by kbarnes701 View Post

Those front left, front right and centre are just those speakers?  It looks as if the sub is playing too. To get a reading for the main speakers only it is necessary to physically switch off the sub or bass management sends redirected bass to the sub.  In the preference pane you will have Audio Output 1 set for the Left speaker, 2 for the right speaker, 3 for the centre speaker - in the drop down where I have circled in blue on the screen cap below.

For the LCR measurements ONLY, the drop-down I have circled in red should be left blank, as in the screenie. The sub should be physically disconnected for these LCR measurements.
To measure the sub only, select Output 4 in the drop-down circled in blue and nothing in the drop-down box circled in red. 

To measure L or R or C with the sub, the sub needs to be connected and powered on and then you select 1 for L, 2 for R, 3 for C in the drop-down circled in blue and NOTHING in the drop-down circled in red.

hello Keith
Thanks for looking at the data. The data are with speakers small with crossover and bass management, as Jerry had proposed in his last post.
About the red drop down box, I cannot remember what was in it when I made the measurements. I did not select anything for that box. I did not look at it because it was not described in the help guide.
What does the red box do?
post #2176 of 9495
Quote:
Originally Posted by turboman123 View Post

Quote:
Originally Posted by kbarnes701 View Post

Those front left, front right and centre are just those speakers?  It looks as if the sub is playing too. To get a reading for the main speakers only it is necessary to physically switch off the sub or bass management sends redirected bass to the sub.  In the preference pane you will have Audio Output 1 set for the Left speaker, 2 for the right speaker, 3 for the centre speaker - in the drop down where I have circled in blue on the screen cap below.

For the LCR measurements ONLY, the drop-down I have circled in red should be left blank, as in the screenie. The sub should be physically disconnected for these LCR measurements.
To measure the sub only, select Output 4 in the drop-down circled in blue and nothing in the drop-down box circled in red. 

To measure L or R or C with the sub, the sub needs to be connected and powered on and then you select 1 for L, 2 for R, 3 for C in the drop-down circled in blue and NOTHING in the drop-down circled in red.

hello Keith
Thanks for looking at the data. The data are with speakers small with crossover and bass management, as Jerry had proposed in his last post.
About the red drop down box, I cannot remember what was in it when I made the measurements. I did not select anything for that box. I did not look at it because it was not described in the help guide.
What does the red box do?

 

It's important to label your tests very clearly because when you look back at them later I guarantee you will not remember what you did!  In your labels on the first of the graphs I commented on, you had the speakers labelled as 'Front Left" etc whereas they are actually 'Front Left + Sub'. It is easy to see just by looking at the graphs that those measurements include the sub, but there will be many occasions when you cannot easily tell what you measured simply by looking at the graph - so accurate and descriptive labels are really important.

 

The 'blue' section and the 'red' section:

 

Blue is where you select the channel (speaker) you are measuring. 1 = LF, 2 = RF, 3=C, 4=Sub(s). If you are measuring only one channel, then select it in the 'blue' box. In this case, there should be nothing selected in the 'red' box. In addition, to measure the L, R or C channel ONLY, the sub will need to be physically turned off or disconnected. If you want to measure L, R or C channels WITH the sub, then there is still nothing in the 'red' box but the sub needs to be physically powered on and/or connected.

 

The red box comes into play when you want to measure two channels together, usually the L & R. In that case, you will select Audio Output 1 in the 'blue' box and Audio Output 2 in the 'red' box. Again, if you want these measurements with and/or without the sub, you will need to physically turn it on/off as required. When you have a channel selected in both the 'blue' and the 'red' box, then both will play, and be measured, at the same time. 

 

If you want to measure the sub only, then select Audio Output 4 in the 'blue' box and nothing in the 'red' box (making sure the sub is physically turned on of course).

 

It is important to realise that you do not select the sub (Audio Output 4) in the 'red' box when you are trying to measure the L, R or C channels plus subs. The sub will play regardless due to bass management, as described above.

 

If you read through the above it can be confusing at first but it will click into place and will become intuitive after a while.

post #2177 of 9495

@turboman123:  After reading your posts, I think I may understand some of your confusion.  It's not necessary to select the sub channel in the REW dropdown for it to be present in your measurements.  As Keith noted, if the sub is powered on and you have your speakers set to small then the sub will dominate the FR below the XO point.  If you wan to take full sweep measurements of your L, C or R channel without the sub then set your speakers to large and power off the sub.

 

I believe the red dropdown allows you to choose another output channel but I'm not 100% sure.

 

EDIT:  That d*mn KBarnes...I would swear he's on a mission to get to 10k! wink.gif

post #2178 of 9495
Quote:
Originally Posted by turboman123 View Post

My previous measurements showed some strong peak and dips for the subwoofer. So I did some more measurements.
First, I wanted to check out some controls I have on my sub B&W ASW608. Following shows (with sub in corner), 1/6 smoothing:
- bass extension C & EQ A
- bass extension A & EQ A
- bass extension C & EQ B

Bass extension A caused a rattle lowndown. Effect of EQ was minor. So I left it at bass extension C & EQ A.

Next, I wanted to check another position of the sub, which was at the wall but more in the middle of the room, 1/6 smoothing:

 

The bass response looks rather 'ragged'. The main issue though is that huge dip centred on 50Hz, which seems to be a room null. If so this would indicate that you perhaps need to experiment more with sub location to see if you can eliminate or at least improve it. If you could improve or remove that null, the blue trace above looks promising.

 

With Audio Output 4 in the 'blue' box and nothing in the 'red' box, put the sub somewhere in the room and take a measurement. Then move the sub to another location (maybe only a foot or so away from the first one) and measure again. Repeat for every possible location of the sub in the room. I suggest you put the sub, for experimentation, even in places it cannot live permanently as this will still yield useful information. Be sure to label each sub position carefully - maybe use masking tape or something on the carpet. When you have finished, compare all the graphs looking for the on with the best (smoothest) response - this is where the sub should live, if at all possible. If not, choose the next best and so on.

 

You may find that no matter what you do, you cannot find an acceptable, practical position for the sub. In that case, the addition of a second (preferably identical) sub will be a good idea.

 

By way of comparison here is my current unsmoothed subs measurement, both subs together (I have two Seaton Submersive F2s). I am still working to improve this around the XO region with the Mains but it sounds excellent as it stands TBH. The boost from about 60Hz up is intentional (my 'house curve') as I like my bass!

 

 

post #2179 of 9495
Quote:
Originally Posted by jkasanic View Post

@turboman123:  After reading your posts, I think I may understand some of your confusion.  It's not necessary to select the sub channel in the REW dropdown for it to be present in your measurements.  As Keith noted, if the sub is powered on and you have your speakers set to small then the sub will dominate the FR below the XO point.  If you wan to take full sweep measurements of your L, C or R channel without the sub then set your speakers to large and power off the sub.

 

I believe the red dropdown allows you to choose another output channel but I'm not 100% sure.

 

EDIT:  That d*mn KBarnes...I would swear he's on a mission to get to 10k! wink.gif

 

LOL. I hadn't noticed how close I am to 10k!  I must try harder this next couple of days.... :)

 

EDIT:

 

Blimey - I was top poster on the entire AVS site over the last 7 days - even beating JD!

 

 

 

 

 

But I drop to third overall if you look at 30 days:

 

 

 

 

 

But the real test is over 'all time'.... where I am a nobody... those two guys are unassailable if they are still active (which they may not be as they don't figure in the Top 10 over 30 days). So JD is the one to watch :)

 

 

 


Edited by kbarnes701 - 3/26/13 at 6:36am
post #2180 of 9495
Keith,

You wrote
Quote:
put the sub somewhere in the room and take a measurement. Then move the sub to another location (maybe only a foot or so away from the first one) and measure again. Repeat for every possible location of the sub in the room

Shouldn't the traditional "sub crawl" work? i.e place the subwoofer in the primary listening position and place the microphone in various locations to find the position with the best response. Then put the sub there. You're much less likely to pull a muscle.
post #2181 of 9495
Quote:
Originally Posted by Selden Ball View Post

Keith,

You wrote
Quote:
put the sub somewhere in the room and take a measurement. Then move the sub to another location (maybe only a foot or so away from the first one) and measure again. Repeat for every possible location of the sub in the room

Shouldn't the traditional "sub crawl" work? i.e place the subwoofer in the primary listening position and place the microphone in various locations to find the position with the best response. Then put the sub there. You're much less likely to pull a muscle.

 

Yes, good thinking Selden - it would work in the same way but be far less physically demanding!

post #2182 of 9495

If moving the sub is cumbersome then try putting it in the MLP and walk around the room with your mic at the same level as the woofer (aka sub crawl).  When you find a better response, trying moving the sub there and measuring the response at the MLP.

 

EDIT:  Geez, is it Monday morning?  I'm a day late and a dollar short...d*mn day job! wink.gif

post #2183 of 9495
Quote:
Originally Posted by turboman123 View Post



What about the level. The around 80dB is Ok, but headroom for front was 50dB, which is not according to guideline. Noise floor was 60dB.

 

Nice progress, Turboman!  You have received some good advice from Keith and Jkasanic.  Regarding the 80dB question, I assume you mean you calibrated the microphone for an 80dB level, which is fine.  However, when you do the measurement sweeps, the headroom should be much higher than 50dB, if that is what you mean.  So, just testing for understanding, on Step 9 of the mic calibration (p. 47), there is a screen that shows the maximum SPL that you can measure.  What is this value?  Is it in the 90-120dB range as the guide states?

 

I suspect there is still something amiss with the mic calibration, but you are still getting useful measurements, so we can address this further after you answer my questions.

 

I agree that the sub placement in the center looks better, but I also think there is a lot of room for improvement.  IIRC, you have only one sub, and even by finding the best spot for one sub, I predict that there will still be some unevenness to the frequency response below 100Hz.  If you have the space and the budget, my recommendation would be to consider a second sub, ideally one that matches your current sub.  As you may have already read in the countless discussions of achieving smooth bass response, having at least two subs is probably the most important step.  The order of priority:

 

1.  More than one sub (matched, if at all possible).

2.  Careful placement of the subs using measurements.

3.  Room treatments to address issues placement cannot correct.

4.  DSP room correction (sets proper delays and trims, irons out remaining response issues).

 

And finally, there is no "correct" set of measurements.  The important point is to learn how to use which measurements to analyze specific problems in your listening area.  For example, a good starting point is a set of measurements as I listed in my last post.  If, for example, the Left+sub or Right+sub exhibits choppy response in the low frequencies, then a second set of measurements that isolates the Left main and Right main speakers without the subs, and a measurement of the sub only, allows you to analyze whether it's the mains or the sub that is largely responsible for the response issues.  Being able to select the right measurement for a specific issue is a sign that you have progressed from being an REW novice to a journeyman, like the rest of us.  Expert status takes a few more years....  wink.gif

post #2184 of 9495
Quote:
Originally Posted by kbarnes701 View Post

 

40dB you meant to type, Jerry :)

 

Oops, of course.  Edited.  Hate it when that happens.  Getting old.

post #2185 of 9495
Quote:
Originally Posted by AustinJerry View Post

Oops, of course.  Edited.  Hate it when that happens.  Getting old.

Congratulations on breaking 3K, Jerry! I'm still a relative newbie, aiming for breaking four digits....but we're following in Keith's footsteps LOL....,
post #2186 of 9495
Quote:
Originally Posted by amirm View Post

I suggest only looking at the sub response for this purpose. And ignore the straight phase. Instead plot excess phase. Steps are:

1. Click on Controls to see the panel.
2. Click on "unwrap phase."
3. Click on " Estimate IR delay." Click on Shift IR when the dialog comes up.
4. Clock on "Generate Minimum Phase."
5. On the graph checkboxes, only enable your frequency response and "excess phase."

Now you should have one black graph over your measurement. Scale the excess phase graph so that it fits on the same display. Try to align it so that it goes right through your sub response. Use the "offset" in the control to shift the graphs.

Now you are ready to analyze smile.gif. Anywhere the excess phase is flat, EQ can be applied safely because the response variation can be inverted properly.

In general you will find that resonant peaks are minimum phase so the above analysis is not mandatory but is helpful. It is also helpful in determining if you have made things better with multiple subs or not.


Why does my excess phase graph come out this way? The black line goes downward instead of across.

How exactly do I scale just the excess graph?

post #2187 of 9495
Quote:
Originally Posted by Pres2play View Post

Why does my excess phase graph come out this way? The black line goes downward instead of across.

How exactly do I scale just the excess graph?

You ignored my advice to only look at sub frequencies smile.gif. Please set the scale so that it stops at 150 to 200 Hz.

On scaling, hover your mouse over the graph and then on top right you see +- controls. Click on "-" and it will zoom out. Likewise, you can use the scroll bar on the right to position it.
post #2188 of 9495
I'll try it when I get home. Thanks!
post #2189 of 9495
Quote:
Originally Posted by jkasanic View Post

If moving the sub is cumbersome then try putting it in the MLP and walk around the room with your mic at the same level as the woofer (aka sub crawl).  When you find a better response, trying moving the sub there and measuring the response at the MLP.

EDIT:  Geez, is it Monday morning?  I'm a day late and a dollar short...d*mn day job! wink.gif

What is the recommendation for integrating a second sub after optimizing the position for the first sub?
post #2190 of 9495
Gain match the two subs, then place the second sub in all available locations and choose the one with the smoothest response. If you have the second sub a considerable distance from the first sub, some tweaking with the phase and/or distance setting may be in order.
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