Simplified REW Setup and Use (USB Mic & HDMI Connection) Including Measurement Techniques and How To Interpret Graphs - Page 485 - AVS | Home Theater Discussions And Reviews
Baselworld is only a few weeks away. Getting the latest news is easy, Click Here for info on how to join the Watchuseek.com newsletter list. Follow our team for updates featuring event coverage, new product unveilings, watch industry news & more!


Forum Jump: 
 414Likes
Reply
Thread Tools
post #14521 of 14549 Old 04-17-2015, 06:27 PM
AVS Special Member
 
toofast68's Avatar
 
Join Date: Nov 2007
Location: NorthEast Ohio, USA
Posts: 1,109
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 98 Post(s)
Liked: 24
Quote:
Originally Posted by AustinJerry View Post
Bass resonance, which is what we are trying to measure, is caused by the low frequency contribution of subs + main speakers. If you want to impress us with your performance, measure center + subs, or left + right + subs.

Center plus subs plus LOUDER - is this decent ?
I know have some super low freq ringing....but that is way down low I think.


toofast68 is online now  
Sponsored Links
Advertisement
 
post #14522 of 14549 Old 04-17-2015, 06:39 PM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 9,451
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 2359 Post(s)
Liked: 1212
^ IMO, looks pretty decent.

BTW, you really should present your graphs with a log scale. Maybe you could correct and re-post?
AustinJerry is online now  
post #14523 of 14549 Old 04-17-2015, 08:59 PM
AVS Special Member
 
toofast68's Avatar
 
Join Date: Nov 2007
Location: NorthEast Ohio, USA
Posts: 1,109
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 98 Post(s)
Liked: 24
Can do I just have to figure out exactly what you mean....
toofast68 is online now  
post #14524 of 14549 Old 04-18-2015, 06:13 AM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 9,451
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 2359 Post(s)
Liked: 1212
Quote:
Originally Posted by toofast68 View Post
Can do I just have to figure out exactly what you mean....
Sure. Here is a graph with a linear scale:



And here is a graph (same measurement) with a log scale (the difference is in the spacing on the horizontal scale):



You can toggle between the two modes by clicking the "Freq Axis" button in the top right corner.

Notice that the two graphs look quite different, and the graph in log format actually makes it much easier to identify issues in the lower frequencies. Going forward, you should always post using a log scale--for our purposes, the linear scale is of no value.
AustinJerry is online now  
post #14525 of 14549 Old 04-18-2015, 06:43 AM
AVS Special Member
 
toofast68's Avatar
 
Join Date: Nov 2007
Location: NorthEast Ohio, USA
Posts: 1,109
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 98 Post(s)
Liked: 24
Quote:
Originally Posted by AustinJerry View Post
Notice that the two graphs look quite different, and the graph in log format actually makes it much easier to identify issues in the lower frequencies. Going forward, you should always post using a log scale--for our purposes, the linear scale is of no value.
Ok, here is the log version..to my untrained eyes (but learning) it does not look quite as good. But I guess not much I can do about the super low stuff anyways....


toofast68 is online now  
post #14526 of 14549 Old 04-18-2015, 08:09 AM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 9,451
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 2359 Post(s)
Liked: 1212
Quote:
Originally Posted by toofast68 View Post
Ok, here is the log version..to my untrained eyes (but learning) it does not look quite as good. But I guess not much I can do about the super low stuff anyways....


First of all, the bass resonance looks to be well under control. Addressing anything under 20Hz is very difficult. Besides, what lives down there? Nothing in music content, which is where resonance is really audible. Probably only LFE effects on movies, and I don't think you will be able to hear much of a difference. Time to move on to the Impulse Response (ETC) graph and look at specular reflections...

BTW, notice how much more revealing the log graph is. In linear mode, you may not have seen that there are slightly different issues at 30Hz vs. 20Hz.
AustinJerry is online now  
post #14527 of 14549 Old 04-18-2015, 08:29 AM
AVS Special Member
 
artur9's Avatar
 
Join Date: Mar 2013
Location: near philly
Posts: 1,027
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 199 Post(s)
Liked: 64
Quote:
Originally Posted by toofast68 View Post
Ok, here is the log version..to my untrained eyes (but learning) it does not look quite as good.
Isn't there a bit of rising resonance as the graph approaches 300Hz? What would cause that?
artur9 is offline  
post #14528 of 14549 Old 04-18-2015, 09:54 AM
AVS Special Member
 
jim19611961's Avatar
 
Join Date: Mar 2011
Location: Texas
Posts: 1,586
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 267 Post(s)
Liked: 98
Click image for larger version

Name:	distortion 128k.jpg
Views:	30
Size:	298.0 KB
ID:	672649

Click image for larger version

Name:	distortion 1M.jpg
Views:	25
Size:	282.7 KB
ID:	672657

These two measurements are EXACTLY the same (one taken right after the other), except one is a 128k sweep (top graph), and the other a 1M sweep (bottom graph). At 21.5hz, the 128k shows 7.6% distortion, the 1M shows 3.13%.

I have observed other differences in data vs sweep length in the low end, notably in waterfalls . Therefore, it may be recommended to do 1M sweeps when finalizing a change.

My Room
My Music
Rega - Apollo, Rega - DAC, Goldpoint Passive, (2) Classe CA-100 bridged power amps (350w)
Jenzen Next
jim19611961 is offline  
post #14529 of 14549 Old 04-18-2015, 10:59 AM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 9,451
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 2359 Post(s)
Liked: 1212
Quote:
Originally Posted by jim19611961 View Post
Attachment 672649

Attachment 672657

These two measurements are EXACTLY the same (one taken right after the other), except one is a 128k sweep (top graph), and the other a 1M sweep (bottom graph). At 21.5hz, the 128k shows 7.6% distortion, the 1M shows 3.13%.

I have observed other differences in data vs sweep length in the low end, notably in waterfalls . Therefore, it may be recommended to do 1M sweeps when finalizing a change.
Yes, but which one is correct, the one which shows the better result? Wishful thinking?
AustinJerry is online now  
post #14530 of 14549 Old 04-18-2015, 11:12 AM
AVS Special Member
 
jim19611961's Avatar
 
Join Date: Mar 2011
Location: Texas
Posts: 1,586
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 267 Post(s)
Liked: 98
Quote:
Originally Posted by AustinJerry View Post
Yes, but which one is correct, the one which shows the better result? Wishful thinking?
No, not wishful thinking. For each doubling of the sweep length, you gain 3db in S/N. JohnPM has said this. So 1M has a better S/N over a 128k by 9db. Its possible that neither is absolutely correct. But my bet is that the 1M is more correct.

My Room
My Music
Rega - Apollo, Rega - DAC, Goldpoint Passive, (2) Classe CA-100 bridged power amps (350w)
Jenzen Next
jim19611961 is offline  
post #14531 of 14549 Old 04-18-2015, 11:35 AM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 9,451
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 2359 Post(s)
Liked: 1212
Quote:
Originally Posted by jim19611961 View Post
No, not wishful thinking. For each doubling of the sweep length, you gain 3db in S/N. JohnPM has said this. So 1M has a better S/N over a 128k by 9db. Its possible that neither is absolutely correct. But my bet is that the 1M is more correct.
Yes, now that you mention it, I do recall that comment by John.

Perhaps we should compare the shorter and longer sweep results. If not much difference, shorter is quicker.
AustinJerry is online now  
post #14532 of 14549 Old 04-18-2015, 11:48 AM
AVS Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 1,122
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 566 Post(s)
Liked: 153
Quote:
Originally Posted by jim19611961 View Post
No, not wishful thinking. For each doubling of the sweep length, you gain 3db in S/N. JohnPM has said this. So 1M has a better S/N over a 128k by 9db. Its possible that neither is absolutely correct. But my bet is that the 1M is more correct.
in room THD measurements, IME, are pretty sensitive to low level environmental noise hence I suspect you're likely to get more robust results from using multiple sweeps.

Is that possible with a USB mic? ISTR multiple sweeps being problematic when multiple clocks are in the picture.
3ll3d00d is offline  
post #14533 of 14549 Old 04-18-2015, 09:50 PM
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 9,651
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 688 Post(s)
Liked: 131
Quote:
Originally Posted by jim19611961 View Post
No, not wishful thinking. For each doubling of the sweep length, you gain 3db in S/N. JohnPM has said this. So 1M has a better S/N over a 128k by 9db. Its possible that neither is absolutely correct. But my bet is that the 1M is more correct.
I assumed it was in the REW User guide, so never asked. But since I see it being talked about here, I wonder... What is the recommended setting then as far as the sweep parameteres/options that can be used?
lovingdvd is online now  
post #14534 of 14549 Old 04-18-2015, 09:52 PM
AVS Special Member
 
lovingdvd's Avatar
 
Join Date: Jul 2002
Posts: 9,651
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 688 Post(s)
Liked: 131
Quote:
Originally Posted by toofast68 View Post
Ok, here is the log version..to my untrained eyes (but learning) it does not look quite as good. But I guess not much I can do about the super low stuff anyways....


This looks excellent! Congratrs. Can you tell us a little about your room setup, where the MLP is, where the sub is placed etc? How about room treatments? I assume you must have some treatments going to have this great of a result (either that or you are sitting on top of your sub ).
lovingdvd is online now  
post #14535 of 14549 Old 04-18-2015, 10:48 PM
Senior Member
 
mattdub1's Avatar
 
Join Date: Nov 2012
Posts: 288
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 147 Post(s)
Liked: 15
^^ I think he already posted about his room treatments a few pages back when you asked him
mattdub1 is offline  
post #14536 of 14549 Old Yesterday, 06:59 AM
AVS Special Member
 
jim19611961's Avatar
 
Join Date: Mar 2011
Location: Texas
Posts: 1,586
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 267 Post(s)
Liked: 98
Quote:
Originally Posted by lovingdvd View Post
I assumed it was in the REW User guide, so never asked. But since I see it being talked about here, I wonder... What is the recommended setting then as far as the sweep parameteres/options that can be used?
IMO, it depends on what you are doing. Short sweeps are fine for FR & ETC. For Waterfalls, Decay & Distortion, id use long ones.

Think of it like making photographs. A 3 Mega Pixel digital image may be fine for a 5"x7" or 8" x 10". But if you want a mural, 10mp would be a lot better. In a similar way, a longer sweep has more data, a more robust image of your measurement. So therefore, you can look at low level detail with more accuracy.

At the end of the day, longer sweeps are always better. But you have to wait.

My Room
My Music
Rega - Apollo, Rega - DAC, Goldpoint Passive, (2) Classe CA-100 bridged power amps (350w)
Jenzen Next
jim19611961 is offline  
post #14537 of 14549 Old Yesterday, 08:40 AM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 9,451
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 2359 Post(s)
Liked: 1212
Quote:
Originally Posted by jim19611961 View Post
IMO, it depends on what you are doing. Short sweeps are fine for FR & ETC. For Waterfalls, Decay & Distortion, id use long ones.

Think of it like making photographs. A 3 Mega Pixel digital image may be fine for a 5"x7" or 8" x 10". But if you want a mural, 10mp would be a lot better. In a similar way, a longer sweep has more data, a more robust image of your measurement. So therefore, you can look at low level detail with more accuracy.

At the end of the day, longer sweeps are always better. But you have to wait.
FYI, the Guide now contains a note explaining the 3dB S/N improvement with each doubling of the sweep length, as well as mention of the "variable smoothing" option which was recently introduced in REW. Still no section on distortion, however, because I am still not comfortable with the topic.
AustinJerry is online now  
post #14538 of 14549 Old Yesterday, 09:54 AM
AVS Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 1,122
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 566 Post(s)
Liked: 153
Quote:
Originally Posted by jim19611961 View Post
IMO, it depends on what you are doing. Short sweeps are fine for FR & ETC. For Waterfalls, Decay & Distortion, id use long ones.

Think of it like making photographs. A 3 Mega Pixel digital image may be fine for a 5"x7" or 8" x 10". But if you want a mural, 10mp would be a lot better. In a similar way, a longer sweep has more data, a more robust image of your measurement. So therefore, you can look at low level detail with more accuracy.

At the end of the day, longer sweeps are always better. But you have to wait.
This is not the complete picture by any stretch. A longer sweep is more demanding of your equipment, possibly unrealistically so, and distortion depends on level so running a longer sweep can also mean a less realistic view of what is going on.
3ll3d00d is offline  
post #14539 of 14549 Old Yesterday, 10:05 AM
AVS Special Member
 
jim19611961's Avatar
 
Join Date: Mar 2011
Location: Texas
Posts: 1,586
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 267 Post(s)
Liked: 98
Quote:
Originally Posted by 3ll3d00d View Post
This is not the complete picture by any stretch. A longer sweep is more demanding of your equipment, possibly unrealistically so, and distortion depends on level so running a longer sweep can also mean a less realistic view of what is going on.
I agree if your talking about SPL's close to your systems capability. But at typical listening levels (maybe 85-95spl), the demand issue is not a problem.

My Room
My Music
Rega - Apollo, Rega - DAC, Goldpoint Passive, (2) Classe CA-100 bridged power amps (350w)
Jenzen Next

Last edited by jim19611961; Yesterday at 10:34 AM.
jim19611961 is offline  
post #14540 of 14549 Old Yesterday, 10:07 AM
AVS Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 1,122
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 566 Post(s)
Liked: 153
Sure but if it's not demanding then what do you learn? Special cases like loose drivers to one side of course. Ultimately that is speaker design territory not in room setup.
3ll3d00d is offline  
post #14541 of 14549 Old Yesterday, 10:45 AM
AVS Special Member
 
jim19611961's Avatar
 
Join Date: Mar 2011
Location: Texas
Posts: 1,586
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 267 Post(s)
Liked: 98
Quote:
Originally Posted by 3ll3d00d View Post
Sure but if it's not demanding then what do you learn? Special cases like loose drivers to one side of course. Ultimately that is speaker design territory not in room setup.
I am not going to pretend we all agree what the best use of all REW's features are. IMO however, on the subject of distortion, whatever the upper reach of ones listening levels are is the level one should look at. If that level is pushing the capability of your system, then I would recommend a more robust system or lower listening levels.

If one is trying to see what the capability of ones system is, that is, when distortion levels start to rise vehemently, then a different approach may be necessary.

As far as what we learn, I would think that knowing things are in order at the levels we listen at is useful. Its seems a little silly to know how your system measures at levels you never hear it at.

My Room
My Music
Rega - Apollo, Rega - DAC, Goldpoint Passive, (2) Classe CA-100 bridged power amps (350w)
Jenzen Next
jim19611961 is offline  
post #14542 of 14549 Old Yesterday, 11:44 AM
AVS Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 1,122
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 566 Post(s)
Liked: 153
The problem is getting repeatable measurements in room at lower levels. At higher levels it is a headroom issue as you say. In either case it is speaker changing event to do something about it hence why I said it is a design time question.
3ll3d00d is offline  
post #14543 of 14549 Old Yesterday, 04:00 PM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 9,451
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 2359 Post(s)
Liked: 1212
I recently posted several waterfalls from measurements taken at various levels from 80dB to 105dB to make a point about how different waterfalls can look depending on measurement level. This same set of measurements also demonstrates how distortion levels vary significantly, increasing as output rises. Here are two examples:





Both measurements are of the center channel + subs with variable smoothing. At 20Hz, the 80dB measurement shows 2.1% THD, while at 105dB, the figure rises to 26.4% THD. At 80dB, there is negligible distortion at 75hZ (.7%), but at 105dB, something significant happens, with a huge distortion bulge (8.8%) developing for an unknown reason.

Interesting...
AustinJerry is online now  
post #14544 of 14549 Old Yesterday, 04:36 PM
AVS Special Member
 
LastButNotLeast's Avatar
 
Join Date: Feb 2007
Location: 08077
Posts: 5,535
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 422
I find stuff in my room starts to rattle at high volumes, which would explain your measured distortion.
Or you have a speaker that's not happy at 105dB.

Did you really need to quote that entire post in your reply?
Downloadable FREE demo discs: Demonstration Blu-Ray Discs (Independently Authored)
Welcome to AVS - Get out while you still can!
LastButNotLeast is offline  
post #14545 of 14549 Old Yesterday, 05:17 PM
AVS Special Member
 
AustinJerry's Avatar
 
Join Date: Jan 2003
Location: Austin, Texas
Posts: 9,451
Mentioned: 18 Post(s)
Tagged: 0 Thread(s)
Quoted: 2359 Post(s)
Liked: 1212
Quote:
Originally Posted by LastButNotLeast View Post
I find stuff in my room starts to rattle at high volumes, which would explain your measured distortion.
Or you have a speaker that's not happy at 105dB.
You could be right. I don't listen at that level anyway. The post was simply to show the relationship between measurment levels and increased distortion.
AustinJerry is online now  
post #14546 of 14549 Old Yesterday, 10:50 PM
AVS Special Member
 
3ll3d00d's Avatar
 
Join Date: Sep 2007
Location: London, UK
Posts: 1,122
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 566 Post(s)
Liked: 153
It could be an artefact of the dips in the 200-300 range. Is that dip in the fundamental at ~225Hz? Rising 3rd harmonic under 20Hz generally means you are driving your sub to distress (approaching xmax so driver is becoming increasingly non linear).
3ll3d00d is offline  
post #14547 of 14549 Old Today, 08:19 AM
AVS Special Member
 
jim19611961's Avatar
 
Join Date: Mar 2011
Location: Texas
Posts: 1,586
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 267 Post(s)
Liked: 98
Quote:
Originally Posted by AustinJerry View Post
Both measurements are of the center channel + subs with variable smoothing. At 20Hz, the 80dB measurement shows 2.1% THD, while at 105dB, the figure rises to 26.4% THD. At 80dB, there is negligible distortion at 75hZ (.7%), but at 105dB, something significant happens, with a huge distortion bulge (8.8%) developing for an unknown reason.

Interesting...
If it were me, I would test the C channel without the subs and look again. Is the C ch. crossed over? Frequency? Maybe a subs only plot also.

IMO, as you go higher in SPL, the distortion curves should increase GRADUALLY. When a sudden spike appears, or the some part of the graph goes up considerably with only a slight increase in SPL, these are indicators of something amuck.

I would also show ALL the later harmonics (4th, 5th, 6th, et cetera) as well as the 3rd. If ALL of these spike at the same time, its a telling sign of something vibrating or rattling.

My Room
My Music
Rega - Apollo, Rega - DAC, Goldpoint Passive, (2) Classe CA-100 bridged power amps (350w)
Jenzen Next

Last edited by jim19611961; Today at 09:51 AM.
jim19611961 is offline  
post #14548 of 14549 Unread Today, 05:49 PM
AVS Special Member
 
LastButNotLeast's Avatar
 
Join Date: Feb 2007
Location: 08077
Posts: 5,535
Mentioned: 4 Post(s)
Tagged: 0 Thread(s)
Quoted: 572 Post(s)
Liked: 422
Jerry, these little hints need to be in the guide (once confirmed, of course ).

Did you really need to quote that entire post in your reply?
Downloadable FREE demo discs: Demonstration Blu-Ray Discs (Independently Authored)
Welcome to AVS - Get out while you still can!
LastButNotLeast is offline  
post #14549 of 14549 Unread Today, 10:26 PM
AVS Special Member
 
tvuong's Avatar
 
Join Date: Dec 2000
Posts: 2,420
Mentioned: 6 Post(s)
Tagged: 0 Thread(s)
Quoted: 584 Post(s)
Liked: 242
Also to note distortion levels rise at a lower frequency for the same measured FR. The THD for my subs went from 1.71% at 18Hz to 6.91% at 15Hz- Not bad, I think for measured at ~110db.



Attached Thumbnails
Click image for larger version

Name:	distortion @18hz.png
Views:	2
Size:	86.5 KB
ID:	676985   Click image for larger version

Name:	distortion @15hz.png
Views:	2
Size:	86.7 KB
ID:	676993  

Last edited by tvuong; Today at 10:33 PM.
tvuong is online now  
Sponsored Links
Advertisement
 
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