or Connect
AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Using Waterfall and ETC Graphs to analyze room response
New Posts  All Forums:Forum Nav:

Using Waterfall and ETC Graphs to analyze room response - Page 5

post #121 of 162
Thread Starter 
Quote:
Originally Posted by HopefulFred View Post

Is there any smoothing applied to this curve? If so, let's see it with no smoothing. If not, I don't think I would fool with it - but maybe that's me being lazy.
l

No smoothing.
post #122 of 162
Quote:
Originally Posted by AustinJerry View Post

While we are waiting for me to get the courage and energy to tackle the pink fluffy again, here is a waterfall graph and frequency measurement of the bass response after my recent AVR upgrade.  Does it suggest areas to improve?  The 50-75Hz range probably would benefit from some treatments.




Jerry, something looks odd here. I mean maybe these are right, but flat down to 10hz? I am not familiar with your software, but I have never in my life seen a bass curve as flat as yours as deep as yours. And you say its unsmoothed?
post #123 of 162
Quote:
Originally Posted by AustinJerry View Post

While we are waiting for me to get the courage and energy to tackle the pink fluffy again, here is a waterfall graph and frequency measurement of the bass response after my recent AVR upgrade.  Does it suggest areas to improve?  The 50-75Hz range probably would benefit from some treatments.



Your minimum vertical scale for the waterfall is too high Jerry (50 db), obscuring some of the ringing. The graph is also hard to read in its default orientation. Please take a look at the WBF forum thread I posted earlier that showed how to determine the minimum level better and orient the graph to make it easier to analyze.

You have the reverse problem with the SPL graph where you are showing 0 db as its minimum. That tends to flatten the graph visually making it look like it is better than it is smile.gif. As with Jim I also think you may have smoothing applied to that. Double check that it is not that way in REW (Graph->Remove Smoothing).
post #124 of 162
Thread Starter 
Quote:
Originally Posted by amirm View Post


Your minimum vertical scale for the waterfall is too high Jerry (50 db), obscuring some of the ringing. The graph is also hard to read in its default orientation. Please take a look at the WBF forum thread I posted earlier that showed how to determine the minimum level better and orient the graph to make it easier to analyze.
You have the reverse problem with the SPL graph where you are showing 0 db as its minimum. That tends to flatten the graph visually making it look like it is better than it is smile.gif. As with Jim I also think you may have smoothing applied to that. Double check that it is not that way in REW (Graph->Remove Smoothing).

 

OK, Amir.  I have been meaning to read that thread, but haven't had the time yet.  I'll do my homework and re-post.  Thanks.

 

- Jerry

post #125 of 162
Thread Starter 
Quote:
Originally Posted by jim19611961 View Post


Jerry, something looks odd here. I mean maybe these are right, but flat down to 10hz? I am not familiar with your software, but I have never in my life seen a bass curve as flat as yours as deep as yours. And you say its unsmoothed?

 

By not familiar with my software, do you mean not familiar with REW?

 

I am using REW with a Soundblaster X-Fi USB sound card (calibrated within REW, of course), and an EMM-6 microphone (individually calibrated by Spectrum Labs) with its own calibration file loaded into REW.  I am under the impression that my REW kit is good, but I learn something new every day.

 

I have three HSU Research ULS-15 (15" sealed) subs, two at ~10 feet on the front wall, and one immediately behind the MLP (~3 feet).  The specs for the sub are 15Hz +/- 1dB.  I think the graph is good...

 

There is no smoothing.  In REW, if smoothing is applied, it is indicated at the bottom of the graph.

 

 

Edit:  I am using Audyssey MultEQ XT32, which doews a good job of smoothing bass response.  Are you an Audyssey user as well?

post #126 of 162
Quote:
Originally Posted by jim19611961 View Post

Jerry, something looks odd here. I mean maybe these are right, but flat down to 10hz? I am not familiar with your software, but I have never in my life seen a bass curve as flat as yours as deep as yours. And you say its unsmoothed?

Well, he did the same thing you did here: http://www.avsforum.com/t/1432713/using-waterfall-and-etc-graphs-to-analyze-room-response/90#post_22488025

"On a 90dB scale everything looks good." A full range magnitude plot should show about 50-60dB vertically to be comparable.


Jerry,

What is the window size for that graph? If it's too short the effect is similar to smoothing because shortening the window decreases the number of data points.
post #127 of 162
Thread Starter 

I have re-measured using R-30 without the plastic cover.  Here are the full-range measurements (left speaker with and without plastic cover, right speaker with and without plastic cover).  There seems to be a difference without the plastic--I'll let Jim and Markus decide whether it is significant or not.

 

 

 

Here are the actual measurement files for download from Dropbox (file names are descriptive of the measurement):

 

http://dl.dropbox.com/u/78476446/R-30%20no%20plastic%20LEFT%20SPEAKER.mdat

 

http://dl.dropbox.com/u/78476446/R-30%20no%20plastic%20RIGHT%20SPEAKER.mdat

 

http://dl.dropbox.com/u/78476446/R-30%20with%20plastic%20LEFT%20SPEAKER.mdat

 

http://dl.dropbox.com/u/78476446/R-30%20with%20plastic%20RIGHT%20SPEAKER.mdat

 

Here is how I configured the R-30:

 

 

 

I would like to start looking at pre-built treatments for this wall and have appointments next week for phone conversations with both GIK and RealTraps.  Before I meet with them, do the per-octave measurements tell me anything about the type of treatment needed?  The per-octave measurements are in the downloads, and I will be glad to post the graphs if needed.

post #128 of 162
Thread Starter 
Quote:
Originally Posted by markus767 View Post


Well, he did the same thing you did here: http://www.avsforum.com/t/1432713/using-waterfall-and-etc-graphs-to-analyze-room-response/90#post_22488025
"On a 90dB scale everything looks good." A full range magnitude plot should show about 50-60dB vertically to be comparable.
Jerry,
What is the window size for that graph? If it's too short the effect is similar to smoothing because shortening the window decreases the number of data points.

 

If you mean the frequency response graph, the first one was 0-100, the second one 50-90.

post #129 of 162
Quote:
Originally Posted by AustinJerry View Post

By not familiar with my software, do you mean not familiar with REW?

I am using REW with a Soundblaster X-Fi USB sound card (calibrated within REW, of course), and an EMM-6 microphone (individually calibrated by Spectrum Labs) with its own calibration file loaded into REW.  I am under the impression that my REW kit is good, but I learn something new every day.

I have three HSU Research ULS-15 (15" sealed) subs, two at ~10 feet on the front wall, and one immediately behind the MLP (~3 feet).  The specs for the sub are 15Hz +/- 1dB.  I think the graph is good...

There is no smoothing.  In REW, if smoothing is applied, it is indicated at the bottom of the graph.



Edit:  I am using Audyssey MultEQ XT32, which doews a good job of smoothing bass response.  Are you an Audyssey user as well?

Right, not familiar with REW.

Not a Audyssey user, no.

Well, upon looking at you subs, maybe that low extension is real after all. Those subs of your are not cheap, and you have 3 of them? Quite an investment, but IF those graphs are good, then it looks like your getting a very nice response and I wouldnt fool with things.
post #130 of 162
Quote:
Originally Posted by AustinJerry View Post

I have re-measured using R-30 without the plastic cover.  Here are the full-range measurements (left speaker with and without plastic cover, right speaker with and without plastic cover).  There seems to be a difference without the plastic--I'll let Jim and Markus decide whether it is significant or not.





Here are the actual measurement files for download from Dropbox (file names are descriptive of the measurement):

http://dl.dropbox.com/u/78476446/R-30%20no%20plastic%20LEFT%20SPEAKER.mdat

http://dl.dropbox.com/u/78476446/R-30%20no%20plastic%20RIGHT%20SPEAKER.mdat

http://dl.dropbox.com/u/78476446/R-30%20with%20plastic%20LEFT%20SPEAKER.mdat

http://dl.dropbox.com/u/78476446/R-30%20with%20plastic%20RIGHT%20SPEAKER.mdat

Here is how I configured the R-30:




I would like to start looking at pre-built treatments for this wall and have appointments next week for phone conversations with both GIK and RealTraps.  Before I meet with them, do the per-octave measurements tell me anything about the type of treatment needed?  The per-octave measurements are in the downloads, and I will be glad to post the graphs if needed.

I dont think a FR is the most telling test to run to make a determination here. While interesting, it doesnt tell us what the treatment is doing to your reflections.

What I would like to see are ETC 1 octave tests at 500, 1k, 2k, 4k and maybe 6 or 8k.

1) without any treatment
2) with treatment (with plastic)
3) with treatment (without plastic)


Jerry, ive been doing some thinking, and have basically reached a conclusion regarding your situation and how you are wanting to do it.

Since you dont want to DIY, and are going to buy someones product(s) to deal with the reflections in the end anyway, I think I would proceed in this manner.

1) deal with each reflection point you find on a individual basis. Use your R30 (plastic or not) to verify each and log where that spot is being careful to note how large an area you had to cover to deal with it properly. Since R30 isnt going to be your final treatment anyway, for measurement purposes, I dont see that it matters much whether the plastic is on or off. Much of my comments came from forgetting this, and that your ultimate treatment wasnt going to be what you used in your testing.
2) While I understand you wanting to buy a finished product, you have to keep in mind that whatever you measure with the R30 is going to change some when you put something else there.
3) Regarding 1), Each area needs to be considered separately. The wall behind the listening spot I see no other treatment to perform other than absorption. Its too close behind you to use diffusion and probably deflection also. But when you move to other areas, while you can treat each and every reflection point with absorption, you may not want to. Too much absorption can deaden a room and make the presentation lifeless and sterile. But of course, you wont really know until you treat all of it. So I would recommend this. go around your room, find all the reflections you want to deal with and put up temporary absorption panels. Once you have everything covered this way, do some long term listening. See how you like the sound. If it too dead, then other treatments can be considered on a reflection point by reflection point basis.
4) What I dont want to see you do is spend big money on over the counter finished absorption panels, put them all in place, and then discover you dont like it.

You have embarked on a very time consuming and demanding venture here. You owe yourself a lot of credit for taking things this far. You have done a good job. But the finished product may become more complicated than simply finding the early reflections and absorbing them. And this goes back to something said early in this thread about deciding on a model. Its important because what you do here may all go for naught (meaning having to start all over) without a coherent plan and/or basis for deciding what and how to do things.

I did exactly what you are doing when I began. Trying to make my room sound better from just putting something here and there. The problem with this is as you learn, what you already have done can be found to be flawed, and you end up redoing the same spots time and time again. Then you learn about model A, and change everything to suite that criteria, only to find out latter that you like model B better, and have to start over once again. This said, my room has been torn down and refitted at least 5 times. While this is a great way to learn, its very time consuming and expensive (if you dont DIY).

So you need to make choices. Do you want the ultimate room or something that is just a little better than you already have? Are you doing this with wanting to learn the trade in mind, or do you really just want it to sound better? What changes to the room are you willing to make, and which ones are you not? And financially, what are you willing to spend?

How you answer these questions, imo, will determine which way to go and how to get there. Without these questions being answered, we are sort of meandering this way and that, and while we may make things better, we may not.

In a ideal world, you could go to different peoples sound rooms, each modeled a different way, and spend time listening to each. Determine which sound suites you best, and then go to a acoustic designer and tell him what you want. Unfortunately, its not like shopping couches where you can go to showrooms and experience each one before you buy. I wish I could do this myself actually. But you do live in Austin. And unlike me, there are probably a lot of good sounding rooms around that would be worth listening to. Lots of folks that hang around highend audio shops love to showoff their systems. Perhaps you an find some of these people. I dont know. Perhaps you can find people on forums like this that live at a reasonable distance from you that would welcome you to their homes. Again, I am not sure what would work best.

But let me be candid, we really have only touched the surface here. There are many directions you could go. And the further you go, the more subjectivity seems to play a part in things. Blogger A thinks his room sounds awesome, you go to his house and you think it sounds terrible, etcetera.

Anyway, I am not sure what to add at this point. Just give some thought to what endgame your looking for here, and we will try to see how to get there.
Edited by jim19611961 - 10/13/12 at 10:47am
post #131 of 162
Quote:
Originally Posted by AustinJerry View Post

If you mean the frequency response graph, the first one was 0-100, the second one 50-90.

No, I was asking about the window on the impulse response. Go to the "Impulse" tab to see the impulse response and the window time.
post #132 of 162
Thread Starter 
Quote:
Originally Posted by markus767 View Post


No, I was asking about the window on the impulse response. Go to the "Impulse" tab to see the impulse response and the window time.

 

Is this what you mean?

post #133 of 162
Thread Starter 

The R-30 without plastic is a significant improvement over the untreated wall.

 

 

 

 

 

 

 

post #134 of 162
Quote:
Originally Posted by AustinJerry View Post

Is this what you mean?

No, these are the default settings. I meant this:



The window outlined in blue shows which part of the impulse response is used for the calculation of the frequency response.
post #135 of 162
Quote:
Originally Posted by AustinJerry View Post

The R-30 without plastic is a significant improvement over the untreated wall.








I agree smile.gif
post #136 of 162
Quote:
Originally Posted by AustinJerry View Post

Here is how I configured the R-30:

You arranged the absorber differently so your results are not comparable. Will look at the results in detail later.
post #137 of 162
Thread Starter 

@ Jim,

 

Thanks for the guidance.  You are correct, I have been a little short-sighted.  I have been trying to develop my analysis skills by focusing only on improving the treatments on the back wall, expecting to be able to apply the learnings on incremental improvements elsewhere in the room.

 

As for the long-term goal, I think I am looking for achieving incrementally better sound, given the restriction that I do not have a dedicated listening room.  I am perfectly fine with replacing existing treatments, and don't have severe budget limitations.  But I am not likely to end up with a room that looks like some of the examples I see on web sites like RealTraps, with treatments covering the walls, in corners, covering the ceiling, etc. 

 

I also need to assess the type of content that I am listening to.  It doesn't take a perfect room to watch broadcast TV or listen to satellite radio.  If you were to paint a picture of a serious audiophile, perhaps a lover of fine classical or jazz music, played on vinyl or hi-res digital files on a high-end two channel audio system--I'm not that person.  So for me, the law of diminishing returns sets in earlier than for a serious audiophile.

post #138 of 162
Thread Starter 
Quote:
Originally Posted by markus767 View Post


You arranged the absorber differently so your results are not comparable. Will look at the results in detail later.

 

I don't understand, are you talking about the R-30 panels being vertical the first time, and horizontal the second time?  It is not trivial to set up the REW kit, move these panels into place, measure, and then find that the results are not comparable.  

post #139 of 162
Quote:
Originally Posted by markus767 View Post

You arranged the absorber differently so your results are not comparable. Will look at the results in detail later.

Theoretically you are right, but for determining:

1) where a reflection point is
2) if a given treatment improved it

Then Jerry analysis and measurements seem fine. Again, R30 is not what he is going to put there ultimately anyways.
post #140 of 162
Thread Starter 
Quote:
Originally Posted by markus767 View Post


No, these are the default settings. I meant this:

The window outlined in blue shows which part of the impulse response is used for the calculation of the frequency response.

 

post #141 of 162
Quote:
Originally Posted by AustinJerry View Post

@ Jim,

Thanks for the guidance.  You are correct, I have been a little short-sighted.  I have been trying to develop my analysis skills by focusing only on improving the treatments on the back wall, expecting to be able to apply the learnings on incremental improvements elsewhere in the room.

As for the long-term goal, I think I am looking for achieving incrementally better sound, given the restriction that I do not have a dedicated listening room.  I am perfectly fine with replacing existing treatments, and don't have severe budget limitations.  But I am not likely to end up with a room that looks like some of the examples I see on web sites like RealTraps, with treatments covering the walls, in corners, covering the ceiling, etc. 

I also need to assess the type of content that I am listening to.  It doesn't take a perfect room to watch broadcast TV or listen to satellite radio.  If you were to paint a picture of a serious audiophile, perhaps a lover of fine classical or jazz music, played on vinyl or hi-res digital files on a high-end two channel audio system--I'm not that person.  So for me, the law of diminishing returns sets in earlier than for a serious audiophile.

Thats good to know. A sophisticated and technically hard to achieve model for 2 channel listening may not be what you want or a goal then. Much assumption goes into these online technical help blogs.
post #142 of 162
Jerry,

Quickly looked through your data and the plastic seems to be a little bit too thick. It introduces 2 minor spikes. Nothing to worry about but if you can get rid of them then do it. Early reflections coming from similar directions as the direct sound (also from the back) can have a negative effect on perceived timbre.

The 4.5ms reflection is coming from the back wall after it has been reflected by the ceiling. You removed that reflection when you covered a larger area of the back wall with fiberglass (compared to the plastic wrapped configuration):



I would cover the whole back wall floor to ceiling with broadband absorption. Then there's just the ceiling reflection left to be treated. Also try removing the side wall panels. You might get some added spaciousness. Currently I don't have my sidewall treated at all.
post #143 of 162
Quote:
Originally Posted by markus767 View Post

You might get some added spaciousness.

Bammie Wham! I like that idea. The potential caveat is that the side walls should be as symmetrical/alike as possible - in both relative position and material. Even with differences, you may prefer the spaciousness (what Toole refers to as apparent source width).

Perhaps the sidewall treatments you've already purchased can be re-purposed to work on the ceiling.
post #144 of 162
Thread Starter 
Quote:
Originally Posted by markus767 View Post

Jerry,
Quickly looked through your data and the plastic seems to be a little bit too thick. It introduces 2 minor spikes. Nothing to worry about but if you can get rid of them then do it. Early reflections coming from similar directions as the direct sound (also from the back) can have a negative effect on perceived timbre.
The 4.5ms reflection is coming from the back wall after it has been reflected by the ceiling. You removed that reflection when you covered a larger area of the back wall with fiberglass (compared to the plastic wrapped configuration):

I would cover the whole back wall floor to ceiling with broadband absorption. Then there's just the ceiling reflection left to be treated. Also try removing the side wall panels. You might get some added spaciousness. Currently I don't have my sidewall treated at all.

Good input, Markus.  I'll try your suggestions.  I don't have the label that would have shown how thick the plastic was.  It was a roll of drop cloth used by painters, probably thicker than it should have been. 

post #145 of 162
Just to but-in, I have to say that this thread is one of the most fascinating I have read. It's like being back in Chemistry 101 with lots of pretty pictures and mountains of technical info, little of which I understand but still intersting to look at. Whilst I have been a sound engineer and live show mixer and have a pretty good setup at home, I am sooo thankful I am no longer that fanatical about my home theatre sound.
post #146 of 162
Thread Starter 

Welcome, Murbella7.  We are just taking a breather here.  I now need to apply some of the good advice I have received here and report back with the results.

post #147 of 162
While you are taking a breather smile.gif, you may want to read this article I wrote on perceptual aspects of room reflections. It was published in Widescreen Review magazine and it has finally been out long enough in print that I went ahead and put it online: http://www.madronadigital.com/Library/RoomReflections.html.
post #148 of 162
Thread Starter 
Quote:
Originally Posted by amirm View Post

While you are taking a breather smile.gif, you may want to read this article I wrote on perceptual aspects of room reflections. It was published in Widescreen Review magazine and it has finally been out long enough in print that I went ahead and put it online: http://www.madronadigital.com/Library/RoomReflections.html.

I am a WS subscriber, and have read all of the articles you have published. The one you reference was particularly interesting. I have also read the first several lags of your Waterfall thread and will publish updated waterfall graphs as soon as I am sure I understand what I have read.

Thanks for your valued input.
post #149 of 162
Quote:
Originally Posted by jim19611961 View Post

Jerry, take a piece of the plastic covering, can you blow air through it when you put it to your mouth? If not, then thats my point. And if you want, take some of the plastic and hang it in front of your speaker. Does it diminish the sound level noticeably? If it does, then it is also diminishing the content getting to the fiber where it must pass to be absorbed.
Perhaps this is some kind of plastic that does pass sound. I doubt it. But if it does, ill agree with Markus. If it doesnt pass the two tests I set above, then he and I will have to just disagree.

When it comes to plastic film, the mass and compliance is what matters, Not the porosity.
Edit: much of post deleted now -- somehow my browser wasnlt showing the last page or so of posts . . .

Measuring transmission loss thru film, fabric, etc, at the speaker is a valid starting point.
Note, however, that high frequency transmission losses could result from two very different mechanisms - Either sound not getting thru is being blocked, OR it is being absorbed. The latter as a panel covering would make the panel more, not less absorbent.

Obviously, nonporous materials can only block sound, albeit sometimes to a degree that is insignificant. With fabrics it can be complicated, sometimes being both absorptive and reflective depending on frequency.

-- Mark
Edited by Tubamark - 10/16/12 at 6:52pm
post #150 of 162
Thread Starter 
Quote:
Originally Posted by Tubamark View Post

When it comes to plastic film, the mass and flexibility is what matters, Not the porosity. Low mass film with no tension (think of a sheet of cling-wrap) passes sound at all but supersonic frequencies.
many successful commercial examples exist.
If one needs a rule of thumb, aim for 2 mil thick or less. I don't know what mass/unit area that equals, but its too light to accurately weigh at home, unless you have a relatively huge piece to work with.
The film covering the FG in those pix is probably 6 mil thick or more. At low frequencies this makes very little difference as long as the film is limp. It will be reflective beginning in the low midrange.
Anyone still skeptical: Cover a doorway with the film in question, and do a simple RTA (use pink or white noise) with, and without the intervening film present. Experiment w/cheap trash bag, cling wrap, shower curtain, tarp, etc. Very instructive.
Related aside: [Warning: very judgemental comment coming] Those "audiophile" chairs with a leather headrest should be complete nonsense to even the most casual audiophile. Leather is highly reflective. Too many audiophiles hear with their eyes, not ears.
My 2 cents. Thanks!
-- Mark

Mark, the plastic is behind us. The measurements indeed showed a difference with/without the plastic covering (which was a painter's drop cover).

As for the leather chairs, I am not a serious enough audiophile to give up my chairs. However, I understand your point about leather being a reflective surface.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Audio theory, Setup and Chat
AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Using Waterfall and ETC Graphs to analyze room response