Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How To Interpret Graphs - Page 268 - AVS Forum
Forum Jump: 
 263Likes
Reply
 
Thread Tools
post #8011 of 12676 Old 01-13-2014, 05:48 PM
Advanced Member
 
artur9's Avatar
 
Join Date: Mar 2013
Location: near philly
Posts: 803
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 84 Post(s)
Liked: 45
Quote:
Originally Posted by sdrucker View Post

The latter sounds like the Bag End E-trap "anti-subwoofer":
http://www.bagend.com/bagend/downloads/E-Trap.pdf

Yeah, pretty much the same thing. The interesting bit is that the black hole is based off some crazy idea Nelson Pass had, if you know the name. wink.gif
Quote:
For that much money, you could get another sub, or just try to achieve better sub placement, if you were looking for an electronic solution vs. more hands-on approaches.

Oh, sure. But I think the idea is if those other solutions did not/can not/will not work.
Quote:

Now here's esoteric:
Dirac is working on something called Dirac Unity, which is supposed to open the door to active room treatment as an addition to their Dirac Live software. It was demo'd at CES last week.

But from what I just read on Computer Audiophile on Google, the demands are so high (meaning the more speakers, subs, and channels you have, the better), and they're already mentioning the "substantial price tag" and restriction to units like the five figure Datasat, that even when it hits the market in the intermediate term future, it's not practical outside of the 1%.

Darn 1% get all the toys.
artur9 is offline  
Sponsored Links
Advertisement
 
post #8012 of 12676 Old 01-13-2014, 05:51 PM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 7,954
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 1076 Post(s)
Liked: 911
Quote:
Originally Posted by artur9 View Post

If it doesn't could one just add it? Assuming a device to calibrate against, of course.

Yes, the instructions on how to do this have been discussed several times. Perhaps I should include the instruxpctions in the guide?
AustinJerry is online now  
post #8013 of 12676 Old 01-13-2014, 05:59 PM
AVS Special Member
 
fotto's Avatar
 
Join Date: Nov 2003
Location: N.E. OH
Posts: 1,869
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 12 Post(s)
Liked: 19
Quote:
Originally Posted by AustinJerry View Post

OK, once again I need guidance on the waterfall method of determining the noise floor.

Here is the RTA measurement:




Here is the waterfall measurement with sound turned off (first with 40dB lower limit, second with 30dB lower limit):







Based on the two waterfalls, what is my noise floor?

Here is an overlay:




I had to set the lower limit at 30dB for the overlay to appear.  What does this tell me?  It certainly looks quite different from the waterfall I am used to publishing:




I'm confused.  Markus, you have been extraordinarily quiet WRT using the waterfall to measure noise floor.  Care to weigh in?
Quote:
Originally Posted by AustinJerry View Post

OK, once again I need guidance on the waterfall method of determining the noise floor.

Here is the RTA measurement:




Here is the waterfall measurement with sound turned off (first with 40dB lower limit, second with 30dB lower limit):







Based on the two waterfalls, what is my noise floor?

Here is an overlay:




I had to set the lower limit at 30dB for the overlay to appear.  What does this tell me?  It certainly looks quite different from the waterfall I am used to publishing:




I'm confused.  Markus, you have been extraordinarily quiet WRT using the waterfall to measure noise floor.  Care to weigh in?
Based on the guidance in this thread, I'd have to say your noise floor is around 38dB (taking out the ultra low freq). Now, I have been curious about this measurement criteria for quite a while now, and would love to hear commentary on your question.
fotto is offline  
post #8014 of 12676 Old 01-13-2014, 06:44 PM
AVS Special Member
 
Cowboys's Avatar
 
Join Date: Jun 2012
Posts: 1,346
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 106
Kbarnes,
I read your Auddessey subwoofer distance tweak write up and am amazed what you were able to do. All this time, I've been concentrated on flatten my two subs and trust Auddessey distance settings. I can't wait to do your distance tweak with center and subs. I just got my UMIK-1 a few days ago and still learning REW first to find best possible position for my subs then I will worry about integrating my subs with the front 3 later. Without REW, tweaking subs distances is impossible (at least for me). Again I can't believe tweaking the subs distances makes a HUGE different and it shows on graph. Which version of Auddessey were you running in that write up? I guest it is okay to add 6 foot in addition to what Auddessey sets for distance. Thanks so much. There are so much to do smile.gif
Cowboys is offline  
post #8015 of 12676 Old 01-13-2014, 07:02 PM
AVS Club Gold
 
sdrucker's Avatar
 
Join Date: Feb 2006
Location: Chicago, IL
Posts: 2,271
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 181 Post(s)
Liked: 160
Quote:
Originally Posted by artur9 View Post

Yeah, pretty much the same thing. The interesting bit is that the black hole is based off some crazy idea Nelson Pass had, if you know the name. wink.gif
Oh, sure. But I think the idea is if those other solutions did not/can not/will not work.
Darn 1% get all the toys.

Not all of them...I'm only in the 10% LOL.

Anyway, I did a comparison of pre and post-Trinnov for my center channel (crossed at 150 Hz due to my multiple subs setup), and found that at least for measuring the room effects, the results I saw a week ago aren't just an anomaly. I hope. Again, this is based on the R-972's implementation of Trinnov I've been working with.

I did a measurement at 85 db as per the SPL reading in REW, pre and post-Trinnov (Flat, no remapping, no Cinema "X-curve" style rolloff), and ran an IR plot, using the % FS scale that Igor suggested to capture pure impulse response on both a positive and negative scale.

IMO it clearly illustrates that there's an impulse response reduction occurring due to Trinnov in parts of the range up to 40 ms, with generally more correction occurring earlier. The measurements were taken about 15 minutes apart (roughly 12:30 AM pre, a little under 12:45 AM post on 1/11), with the mic in exactly the same position at my MLP, as I measure it with my tape measure, with NO variation. Everybody else was asleep, and there were no additional electronics being run in the room aside from my measuring and A/V equipment.

See the plot below:



Blue is without Trinnov, Grey is with Trinnov. I'm assuming that shrinkage toward 0% from either positive or negative % FS numbers is a good thing. If so, the Trinnov calibration is reducing the "blue" measurement to the levels in grey in much of the range (particularly below 15 ms). There's still significant early reflections, but what's happening IMO is non-trivial throughout the range.

However, I will note that for the worst IR peaks, the energy seems to be reversed (meaning extreme positive->extreme negative at about 3 and 7 ms), and I would want to see by measuring where the worst peaks are (i.e. what hard surfaces) and ideally apply treatment there - or figure out what needs to be moved.

I'm also not familiar with the appropriate scale for % FS charts or what "good" means here vs. a db measurement. And I still would owe Igor pre-out measurements, which I didn't do because I didn't have the RCA adaptors handy.

Finally, two other plots. One is where I took my MLP measurement and deliberately went "off-axis". I didn't have a way of measuring formal angles off-axis, but what I did is did a measurement 18" to the right of MLP and tilted at an angle about 6" forward of where I'd normally sit, aimed at the right speaker.



Grey is the original MLP measurement, and black the "off axis" one. Unless I'm losing my mind, there's variation from the MLP measurement (grey is MLP, black is "off axis"), but there's considerably less variation than the "pre" Trinnov measurement. Whether that's stable enough to consider credible is a judgment call. Of course I wouldn't sit that far off-axis in reality, and slouch to the right like this..

I don't have it accessible right now, but I'll note that the matching full-range frequency plot is, as one would expect, less flat than the MLP version when I did the "off-axis" plot.

And last, a full range correction plot, at 1/6th smoothing. Green is Trinnov, yellow is without Trinnov.


The results could be flatter at 200 Hz, but overall, the response is +/- 3 db of 85 db, and plots visually flatter, from roughly the crossover point to about 6 kHz before falling into the speaker's natural roll-off pattern (which I've seen in Audyssey). I do think there's some overcorrecting at about 5 kHz to 9khz (meaning the frequency response is a bit more rolled off than I think it could be), and there's a little "boost" to slightly flatten the response at about 14 kHz which may or may not be HF noise, but YMMV. I also wonder about the noise getting picked up below 20 Hz, but it's below my roughly low 50's db noise floor.

Would it be worth it to filter the IR plot and look at specific frequencies, to see where the correction is taking place (whether I show as % FS or an ETC plot)? Or is this simply a waste of the thread's time, and I'll return you to discussion of the usual subjects? You tell me.

You can call this wishful thinking, or placebo effect if you may, but my hypothesis is that the reduction of reflection on me pre/post IR chart is playing a role in why I think the sound is "clearer" and I can pick out more detail, plus hear more natural male voicing coming from the center channel on movies with REQ engaged vs not (and placebo effect or not, more pleasing than Audyssey). And I'm aware that I don't have a multi-position REQ with my version of Trinnov, let alone the ability to conduct a spatially averaged REW plot, but it's a start.

Stuart

 

Denon 4311 with XT32 and Audyssey Pro

Oppo 93 and 103

Panasonic VT50

Sherwood R-972 with its version of the Trinnov Optimizer

MiniDSP 10x10 HD

PSB Imagine T2, Center, and Surrounds (as of 5/2014); HSU ULS-15 subs (2)

 

The Audyssey FAQ Guide can be found here:

http://www.avsforum.com/...

sdrucker is online now  
post #8016 of 12676 Old 01-13-2014, 11:32 PM
AVS Special Member
 
Cowboys's Avatar
 
Join Date: Jun 2012
Posts: 1,346
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 106
Quote:
Originally Posted by chadwick537 View Post

Cowboys, you need to make sure you download your UMIK-1 calibration file from the MiniDSP website. Open the calibration file in Wordpad, copy the top line of it (sensitivity info and serial number) and paste that into whatever calibration file you are using from Cross Spectrum.
I went to the website and enter a serial number (just made up a number as I am not home to check my serial number. I will when I get home) to see what it looks like and it shows
"Sens Factor =-5.846dB, SERNO: 7001234"
10.054 -5.3955
10.179 -5.2271
And a whole bunch of lines with similar format as the two above
Do I copy the very first line including the double quote " and paste it to the very top line of my cross spectrum file (i.e narrow_band_response_90_degree)? Thanks.
Cowboys is offline  
post #8017 of 12676 Old 01-14-2014, 12:17 AM
Senior Member
 
Audionut11's Avatar
 
Join Date: May 2008
Posts: 360
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 82 Post(s)
Liked: 52
Quote:
Originally Posted by Cowboys View Post

Do I copy the very first line including the double quote " and paste it to the very top line of my cross spectrum file (i.e narrow_band_response_90_degree)? Thanks.

Yes.
Audionut11 is offline  
post #8018 of 12676 Old 01-14-2014, 01:22 AM
AVS Special Member
 
markus767's Avatar
 
Join Date: Apr 2009
Posts: 5,518
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 954 Post(s)
Liked: 375
Quote:
Originally Posted by AustinJerry View Post

Markus, you have been extraordinarily quiet WRT using the waterfall to measure noise floor.  Care to weigh in?

Hi Jerry,

A waterfall generated from an impulse response with the measuring signal muted does not show the noise floor. Use the RTA to examine the noise floor. However, it won't tell you where the noise floor of a waterfall really is. There is no direct relationship between those two measuring methods. The swept sine method doesn't measure sound levels, it measures the transfer function of a system. For example, if you increase the sweep length, the noise level is reduced.

Markus

"In science, contrary evidence causes one to question a theory. In religion, contrary evidence causes one to question the evidence." - Floyd Toole
markus767 is online now  
post #8019 of 12676 Old 01-14-2014, 02:49 AM
Member
 
Scott27's Avatar
 
Join Date: Apr 2013
Posts: 24
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 12 Post(s)
Liked: 10
Hi all,

Im pretty new to all this, and have recently set up REW on a Macbook Pro to take some measurements of my simple home theatre room. I have B&W CM5's for front and rear and a B&W centre, plus an SVS SB12 NSD sub in a ~3000 cubic room (on a suspended floor). I have taken the following sub FR using REW and a UMIK1 mic, this is after Ive run Auddessey and the sub is run 3dB hot. But I'm not sure what's good and bad about the results, and input would be greatly appreciated smile.gif

Scott27 is offline  
post #8020 of 12676 Old 01-14-2014, 04:14 AM
AVS Addicted Member
 
kbarnes701's Avatar
 
Join Date: Jan 2008
Location: Main Listening Positon
Posts: 19,522
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 2419 Post(s)
Liked: 2195
Quote:
Originally Posted by sdrucker View Post
 
Quote:
Originally Posted by kbarnes701 View Post

Yes - the third one is just prettier, that's all. If I could do something about that issue at about 50Hz I'd be content with it. I am comforted though by Jim's remark that he thought it wasn't causing too much of an audible problem. Good thing because no matter what I try I can't make much difference to it.

So you are actually implying that you're effective "done"? And just here for the social company? biggrin.gif

 

Haha. You are leading the witness, Sir.... :)  

 

I don't think I could ever say 'done'. But I do seem to have reached a stage where I can't find anything else which I can do that is making any audible (or measurable for that matter) improvement. This may be a limitation of the room, or of me, or of both. In that last waterfall I posted, the main problem is whatever is going on around 50Hz - the rest of the graph is pretty good for a small, square room I think. I have tried everything I know to help with the 50Hz issue - additional treatments, extra gubbins, speaker placement being the most obvious - but nothing has made very much difference.

 

Of course, if I can’t hear a problem at 50Hz (and I am not sure that I can) then whatever I do doesn't matter all that much anyway as I am not interested in just producing better looking graphs.

 

The one thing I have yet to try at around 50Hz is some nice PEQ - I am waiting for the nanoAVR to come along, with some user-experience of it and reviews, as this might be a nice tool for me. Failing that, I will invoke the presently unused Behringer, which sits like an ugly sister at the ball at the top of my rack.  Because I am not hearing a dreadful problem, it is a disincentive to fiddle I must admit.

kbarnes701 is offline  
post #8021 of 12676 Old 01-14-2014, 04:24 AM
AVS Addicted Member
 
kbarnes701's Avatar
 
Join Date: Jan 2008
Location: Main Listening Positon
Posts: 19,522
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 2419 Post(s)
Liked: 2195
Quote:
Originally Posted by Cowboys View Post

Kbarnes,
I read your Auddessey subwoofer distance tweak write up and am amazed what you were able to do. All this time, I've been concentrated on flatten my two subs and trust Auddessey distance settings. I can't wait to do your distance tweak with center and subs. I just got my UMIK-1 a few days ago and still learning REW first to find best possible position for my subs then I will worry about integrating my subs with the front 3 later. Without REW, tweaking subs distances is impossible (at least for me). Again I can't believe tweaking the subs distances makes a HUGE different and it shows on graph. Which version of Auddessey were you running in that write up? I guest it is okay to add 6 foot in addition to what Auddessey sets for distance. Thanks so much. There are so much to do smile.gif

 

Thanks for the kind words, which are appreciated. The Sub Distance Tweak Guide is largely the work of several other members, all of whom are acknowledged in the Guide. All I did really was to pull the disparate bits of information together into one handy place. But yes, it is a useful Guide and I have achieved substantial improvements by putting its advice into practice.  I am using Audyssey XT32 and Audyssey Pro.

 

The main problem that Audyssey encounters wrt to optimising phase around the splice is that at no time does Audyssey measure the combined response of the mains and subs together. So its assessment of the correct delay setting for the subs is only a calculated one and not a measured one. It seems that the calculation is not done all that well unfortunately because 99% of the people who have followed the practice outlined in the Guide have seen an improvement at the crossover region (splice) with quite a few seeing a substantial improvement. The adjustment is pretty easy to make, as you have seen, but it is only possible with measuring gear - it cannot really be done 'by ear'.

 

You can see why adjusting the delay makes such a difference - at the crossover frequency, both the mains and the subs are playing the same content, so it is essential that they are in phase with each other. Good luck with trying it, and please do a 'before' and 'after' graph and post it here so we can see your results.

kbarnes701 is offline  
post #8022 of 12676 Old 01-14-2014, 04:30 AM
AVS Addicted Member
 
kbarnes701's Avatar
 
Join Date: Jan 2008
Location: Main Listening Positon
Posts: 19,522
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 2419 Post(s)
Liked: 2195
Quote:
Originally Posted by Scott27 View Post

Hi all,

Im pretty new to all this, and have recently set up REW on a Macbook Pro to take some measurements of my simple home theatre room. I have B&W CM5's for front and rear and a B&W centre, plus an SVS SB12 NSD sub in a ~3000 cubic room (on a suspended floor). I have taken the following sub FR using REW and a UMIK1 mic, this is after Ive run Auddessey and the sub is run 3dB hot. But I'm not sure what's good and bad about the results, and input would be greatly appreciated smile.gif

 

What crossover frequency are you using?  The plot doesn't look too bad at all. Also, is the graph unsmoothed (as it should be for sub-only)?

kbarnes701 is offline  
post #8023 of 12676 Old 01-14-2014, 04:33 AM
AVS Addicted Member
 
kbarnes701's Avatar
 
Join Date: Jan 2008
Location: Main Listening Positon
Posts: 19,522
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 2419 Post(s)
Liked: 2195
Quote:
Originally Posted by artur9 View Post
 
Quote:
Originally Posted by kbarnes701 View Post

If I could do something about that issue at about 50Hz I'd be content with it. I am comforted though by Jim's remark that he thought it wasn't causing too much of an audible problem. Good thing because no matter what I try I can't make much difference to it.

I'm hoping you'd be willing to try some of the more esoteric devices like AustinJerry did a while ago. And let us know if it worked.

e.g a Helmhotz resonator (as an aside, here's a page with a resonator you're supposed to put in your ear eek.gif )

Or one of the black hole devices?

 

My problem is that I am so limited in what I can do by the room itself, which is too small by far. This restricts me in all sorts of ways. If I was starting from scratch, with a decent sized room, I would be very interested in using tuned traps.

kbarnes701 is offline  
post #8024 of 12676 Old 01-14-2014, 06:16 AM
AVS Special Member
 
Cowboys's Avatar
 
Join Date: Jun 2012
Posts: 1,346
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 106
Quote:
Originally Posted by kbarnes701 View Post

Good luck with trying it, and please do a 'before' and 'after' graph and post it here so we can see your results.
Definitely will but here my very first attempt graph with crossover at 90hz with my dual fv15hps. What do y'all think? I am still playing with finding the best location for them. This graph represents where my dual has been sitting which is upfront between my main LR. I am using hdmi out from my htpc and chose hdmi out 4 in REW preference setting which generates my subs sweep. Is it true that this is only LFE signal which is limited by my 120hz avr setting? Thanks again.

Cowboys is offline  
post #8025 of 12676 Old 01-14-2014, 07:00 AM
AVS Addicted Member
 
kbarnes701's Avatar
 
Join Date: Jan 2008
Location: Main Listening Positon
Posts: 19,522
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 2419 Post(s)
Liked: 2195
Quote:
Originally Posted by Cowboys View Post
 
Quote:
Originally Posted by kbarnes701 View Post

Good luck with trying it, and please do a 'before' and 'after' graph and post it here so we can see your results.
Definitely will but here my very first attempt graph with crossover at 90hz with my dual fv15hps. What do y'all think? I am still playing with finding the best location for them. This graph represents where my dual has been sitting which is upfront between my main LR. I am using hdmi out from my htpc and chose hdmi out 4 in REW preference setting which generates my subs sweep. Is it true that this is only LFE signal which is limited by my 120hz avr setting? Thanks again.

 

Congrats on getting REW working well and on your first graph. That response doesn't look too bad at all from 15Hz upwards. There's a bit of a hump at 30-35Hz. The two dips at 100Hz and 180Hz are probably nulls caused by the room but they are both quite narrow so shouldn't be an issue. And in any event, the one at 180Hz isn't all that relevant as the mains will have taken over by then. Might be interesting to see a plot of R+L+Subs 15Hz to 20kHz (1/6th smoothed for the full range plot). 

 

The next thing I would do is to ensure that your subs are optimally placed in the room. For that, we'd need room dimensions and a diagram to start with. Then suggestions can be made, using a room mode calculator, to see how your subs might best interact with the room. They may or may not be in the optimum place as things stand. You may be able to achieve a smoother response, across a wider area of the room, by relocating the subs.  Or, you can do it by trial and error, moving the subs to what appear to be good candidate positions and measuring again (without moving the mic between measurements). For example you might try at the halfway points on opposing walls, or in the two rear corners, if practical.

kbarnes701 is offline  
post #8026 of 12676 Old 01-14-2014, 07:10 AM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 7,954
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 1076 Post(s)
Liked: 911

Too add to what Keith has said, I would also look at an unsmoothed 15-300Hz measurement that includes the left and right speakers as well.  A signal on HDMI4 is sending a signal to the sub LFE channel only.  You should output a signal to HDMI1 and HDMI2, with left and right speakers set to small and crossover at 90 Hz.  This will allow you to see how well the subs and mains integrate around the crossover point.

AustinJerry is online now  
post #8027 of 12676 Old 01-14-2014, 07:54 AM
AVS Special Member
 
Cowboys's Avatar
 
Join Date: Jun 2012
Posts: 1,346
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 106
Keith, Thanks for your comment and suggestion. My room is seal rectangular with 14x19x9.5' dimension. Unfortunately due to WAF, my next position for them is one upfront just a bit to the right of my Left speaker and one in the right back corner behind my seat approximately 4' from MLP. I will measure RL and subs together after rerunning Auddessey. Is it okay to just run 1 mic Auddessey position and measure instead of doing all 8 for REW measuring purpose?

Austin, Thanks for your commend and suggestion. I will post an unsmooth and a 1/6 one. If I want to measure just subs, is LFE hdmi 4 ok? Or should I just measure hdmi 1 with my left speaker crossovered real high like 200hz? I wish someday, you can drive down to Dallas and help me integrating my system together smile.gif
Thanks again.
Cowboys is offline  
post #8028 of 12676 Old 01-14-2014, 08:02 AM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 7,954
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 1076 Post(s)
Liked: 911
Do not alter the crossover--the splice between sub and mains is what we are interested in. The subs receive a signal by virtue of bass management. Depending on what you want to examine, you can measure left+subs, right+subs, or left+right+subs. The last measurement is most useful for examining the splice. For measuring the subs alone, HDMI4 is fine.
AustinJerry is online now  
post #8029 of 12676 Old 01-14-2014, 08:09 AM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 7,954
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 1076 Post(s)
Liked: 911
Quote:
Originally Posted by markus767 View Post

Hi Jerry,

A waterfall generated from an impulse response with the measuring signal muted does not show the noise floor. Use the RTA to examine the noise floor. However, it won't tell you where the noise floor of a waterfall really is. There is no direct relationship between those two measuring methods. The swept sine method doesn't measure sound levels, it measures the transfer function of a system. For example, if you increase the sweep length, the noise level is reduced.

Fotto suggested my noise floor is 37dB. I am not sure how he derived that value from the graphs I posted, but it seems to be quite low. And TBH, Markus, I don't understand your response.

So, I am no closer to understanding how to actually determine the noise floor and, once it understand the proper method, how to use the noise floor to filter out noise in waterfall graphs. Perhaps I am being overly dense, but unless I understand it, I am unlikely to make any additions or changes to the Guide.
AustinJerry is online now  
post #8030 of 12676 Old 01-14-2014, 08:16 AM
AVS Addicted Member
 
kbarnes701's Avatar
 
Join Date: Jan 2008
Location: Main Listening Positon
Posts: 19,522
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 2419 Post(s)
Liked: 2195
Quote:
Originally Posted by AustinJerry View Post
 

Too add to what Keith has said, I would also look at an unsmoothed 15-300Hz measurement that includes the left and right speakers as well.  A signal on HDMI4 is sending a signal to the sub LFE channel only.  You should output a signal to HDMI1 and HDMI2, with left and right speakers set to small and crossover at 90 Hz.  This will allow you to see how well the subs and mains integrate around the crossover point.

 

+1

kbarnes701 is offline  
post #8031 of 12676 Old 01-14-2014, 08:22 AM
AVS Addicted Member
 
kbarnes701's Avatar
 
Join Date: Jan 2008
Location: Main Listening Positon
Posts: 19,522
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 2419 Post(s)
Liked: 2195
Quote:
Originally Posted by Cowboys View Post

Keith, Thanks for your comment and suggestion. My room is seal rectangular with 14x19x9.5' dimension. Unfortunately due to WAF, my next position for them is one upfront just a bit to the right of my Left speaker and one in the right back corner behind my seat approximately 4' from MLP. I will measure RL and subs together after rerunning Auddessey. Is it okay to just run 1 mic Auddessey position and measure instead of doing all 8 for REW measuring purpose?
 

 

Good news is a sealed, rectangular room works well with the room mode calculators.

 

I would use at least 3 Audyssey positions, around the MLP. But you don't need to use Audyssey at all to find the best place for the subs. Just measure with Audyssey off, then move the subs and then measure again and rinse and repeat. At this stage you're not after the best you can get - you are just looking for the best locations for the subs, each position relative to the next. Once you have optimised sub locations, then you can do a full Audyssey calibration using all 8 points and then measure again for the final result. If possible don't move the mic at all during measurement runs when optimizing sub locations. Despite WAF, you may still find that relocating the subs gives a smoother response. Sometimes moving them just a foot or two is all that is needed.

kbarnes701 is offline  
post #8032 of 12676 Old 01-14-2014, 08:26 AM
AVS Addicted Member
 
kbarnes701's Avatar
 
Join Date: Jan 2008
Location: Main Listening Positon
Posts: 19,522
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 2419 Post(s)
Liked: 2195
Quote:
Originally Posted by AustinJerry View Post
 
Quote:
Originally Posted by markus767 View Post

Hi Jerry,

A waterfall generated from an impulse response with the measuring signal muted does not show the noise floor. Use the RTA to examine the noise floor. However, it won't tell you where the noise floor of a waterfall really is. There is no direct relationship between those two measuring methods. The swept sine method doesn't measure sound levels, it measures the transfer function of a system. For example, if you increase the sweep length, the noise level is reduced.

Fotto suggested my noise floor is 37dB. I am not sure how he derived that value from the graphs I posted, but it seems to be quite low. And TBH, Markus, I don't understand your response.

So, I am no closer to understanding how to actually determine the noise floor and, once it understand the proper method, how to use the noise floor to filter out noise in waterfall graphs. Perhaps I am being overly dense, but unless I understand it, I am unlikely to make any additions or changes to the Guide.

 

We now seem to have an impasse, with two highly respected and knowledgeable members saying diametrically opposed things. So, you are not the only one who is now confused. 

kbarnes701 is offline  
post #8033 of 12676 Old 01-14-2014, 09:17 AM
AVS Special Member
 
jim19611961's Avatar
 
Join Date: Mar 2011
Location: Texas
Posts: 1,426
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 144 Post(s)
Liked: 89



Here is my methodology.

The red curve is an averaged response with no signal. The green is with a sine sweep. At this measured spl, I have about 35db of S/N in the bass region. When I look at the waterfall measured at this spl, I just ignore all the data below the S/N for whatever band of the FR I am looking at.

Although it is a worthy endeavor, trying to make a waterfall absent of the noise in the room I dont think is possible. I agree with Marcus in that waterfalls are not designed to measure noise floors. I suspect, but don't know for certain, that waterfalls depend on the sweep to derive the "echos" in a time domain. And therefore are useless when no sweep is present.

My Room
My Music
Rega - Apollo, Rega - DAC, Goldpoint Passive, (2) Classe CA-100 bridged power amps (350w)
Jenzen Next
jim19611961 is offline  
post #8034 of 12676 Old 01-14-2014, 09:37 AM
AVS Special Member
 
fotto's Avatar
 
Join Date: Nov 2003
Location: N.E. OH
Posts: 1,869
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 12 Post(s)
Liked: 19
Quote:
Originally Posted by jim19611961 View Post


Here is my methodology.

The red curve is an averaged response with no signal. The green is with a sine sweep. At this measured spl, I have about 35db of S/N in the bass region. When I look at the waterfall measured at this spl, I just ignore all the data below the S/N for whatever band of the FR I am looking at.

Although it is a worthy endeavor, trying to make a waterfall absent of the noise in the room I dont think is possible. I agree with Marcus in that waterfalls are not designed to measure noise floors. I suspect, but don't know for certain, that waterfalls depend on the sweep to derive the "echos" in a time domain. And therefore are useless when no sweep is present.

I think it would be interesting to see a waterfall generated from a sweep with no source signal and compare that to the red response, to further understand if there is correlation between the two,
fotto is offline  
post #8035 of 12676 Old 01-14-2014, 09:47 AM
Senior Member
 
BCRSS's Avatar
 
Join Date: Aug 2013
Posts: 389
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 91 Post(s)
Liked: 70
I recently got my UMIK-1 from Minidsp and downloaded the .cal file. What I noticed is that if I would have purchased from Cross Spectrum Labs, I would have received a 0 and 90 degree .cal file. Question I have while I wait for my mic stand is "will there be any noticeable difference in the way I postion the UMIK-1 beside pointing at the ceiling for a 90 .cal file and pointing toward the speaker with the 0 .cal file?"
Or can I get a 90 .cal file for my mic from another source?

Tux's Elusive 1099 (L/C/R)---Ascend CBM-170 SE (surrounds 9.2)---SVS PB 12 Plus---SVS PC 12 Plus---2 DIY 18's sealed--AVR Denon 4520CI---TV Vizio 70---110" Motorized Screen---Epson 8350 Projector---Unraid 32TB Network Storage---Gateway SX 2370 HTPC with 120 SSD
BCRSS is online now  
post #8036 of 12676 Old 01-14-2014, 09:48 AM
AVS Special Member
 
jim19611961's Avatar
 
Join Date: Mar 2011
Location: Texas
Posts: 1,426
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 144 Post(s)
Liked: 89
Another aspect of this equation so far ignored is the S/N of the mic itself.



http://www.fritzsoundlab.com/info.html

The Dayton EMM6 has a S/N of about 40-45db in the bass region. So assuming you are trying to stay 10db above the noise floor, anything below -30db to -35db on a waterfall in the bass region is useless.

My Room
My Music
Rega - Apollo, Rega - DAC, Goldpoint Passive, (2) Classe CA-100 bridged power amps (350w)
Jenzen Next
jim19611961 is offline  
post #8037 of 12676 Old 01-14-2014, 10:29 AM
AVS Addicted Member
 
kbarnes701's Avatar
 
Join Date: Jan 2008
Location: Main Listening Positon
Posts: 19,522
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 2419 Post(s)
Liked: 2195
Quote:
Originally Posted by jim19611961 View Post




Here is my methodology.

The red curve is an averaged response with no signal. The green is with a sine sweep. At this measured spl, I have about 35db of S/N in the bass region. When I look at the waterfall measured at this spl, I just ignore all the data below the S/N for whatever band of the FR I am looking at.

Although it is a worthy endeavor, trying to make a waterfall absent of the noise in the room I dont think is possible. I agree with Marcus in that waterfalls are not designed to measure noise floors. I suspect, but don't know for certain, that waterfalls depend on the sweep to derive the "echos" in a time domain. And therefore are useless when no sweep is present.

 

Very interesting Jim. I hadn't thought of using my OM kit for this but OM makes it real easy. In fact, I get the noise floor read 'automatically' while waiting for the test tone disc to load up in the player... never even thought of that before.

kbarnes701 is offline  
post #8038 of 12676 Old 01-14-2014, 10:30 AM
AVS Addicted Member
 
kbarnes701's Avatar
 
Join Date: Jan 2008
Location: Main Listening Positon
Posts: 19,522
Mentioned: 9 Post(s)
Tagged: 0 Thread(s)
Quoted: 2419 Post(s)
Liked: 2195
Quote:
Originally Posted by fotto View Post
 
Quote:
Originally Posted by jim19611961 View Post


Here is my methodology.

The red curve is an averaged response with no signal. The green is with a sine sweep. At this measured spl, I have about 35db of S/N in the bass region. When I look at the waterfall measured at this spl, I just ignore all the data below the S/N for whatever band of the FR I am looking at.

Although it is a worthy endeavor, trying to make a waterfall absent of the noise in the room I dont think is possible. I agree with Marcus in that waterfalls are not designed to measure noise floors. I suspect, but don't know for certain, that waterfalls depend on the sweep to derive the "echos" in a time domain. And therefore are useless when no sweep is present.

I think it would be interesting to see a waterfall generated from a sweep with no source signal and compare that to the red response, to further understand if there is correlation between the two,

 

+1

kbarnes701 is offline  
post #8039 of 12676 Old 01-14-2014, 11:53 AM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 7,954
Mentioned: 5 Post(s)
Tagged: 0 Thread(s)
Quoted: 1076 Post(s)
Liked: 911
Quote:
Originally Posted by BCRSS View Post

I recently got my UMIK-1 from Minidsp and downloaded the .cal file. What I noticed is that if I would have purchased from Cross Spectrum Labs, I would have received a 0 and 90 degree .cal file. Question I have while I wait for my mic stand is "will there be any noticeable difference in the way I postion the UMIK-1 beside pointing at the ceiling for a 90 .cal file and pointing toward the speaker with the 0 .cal file?"
Or can I get a 90 .cal file for my mic from another source?

 

Look in the Guide, page 3, Note 3.

AustinJerry is online now  
post #8040 of 12676 Old 01-14-2014, 11:53 AM
AVS Special Member
 
Cowboys's Avatar
 
Join Date: Jun 2012
Posts: 1,346
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 1 Post(s)
Liked: 106
Quote:
Originally Posted by AustinJerry View Post

Too add to what Keith has said, I would also look at an unsmoothed 15-300Hz measurement that includes the left and right speakers as well.  A signal on HDMI4 is sending a signal to the sub LFE channel only.  You should output a signal to HDMI1 and HDMI2, with left and right speakers set to small and crossover at 90 Hz.  This will allow you to see how well the subs and mains integrate around the crossover point.
Honestly I am blindly playing with the crossover point and don't know if 90 is good. My main are Klipsch legend Klf10 (Port design)that per spec it goes down to 32hz +-3db which Auddessey sets to full range. My center is Klipsch klf-7 (seal design) which per spec can go down to 75hz and Audessey sets it to small/60hz. How do I determine the best crossover points for my main and center? That is what I probably need to do first before tweaking subs distance to achieve the best graph with them. Thanks.
Cowboys is offline  
Reply Audio Theory, Setup, and Chat

Tags
Dayton , Dayton Audio , Room Equilizer Wizard Rew

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off