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 104

post #3091 of 9583
Quote:
Originally Posted by sdrucker 
^^ +1, Ken tongue.gif

It's intriguing especially since we've become more aware of Audyssey's shortcomings outside of independent channel equalization in the frequency domain, and knowing that there are solutions that can address some of them, as well as customize speaker presentation, but are, ahem, out of reach (even for someone with a Lexus vs, a Bentley LOL) is frustrating.

The Trinnov folks absolutely need to roll out even a less-than-complete version in a consumer product that's priced, say, in the Lumagen VP range, and a platform that's bug tested and at least as stable. But given development costs, the individual level of support needed, and licensing, I can see why it's not apparently in the cards after the R-972 experience.

Anyone here tried DRC-FIR? http://drc-fir.sourceforge.net/

Gear mentioned in this thread:

post #3092 of 9583
Quote:
Originally Posted by AustinJerry View Post

I am seriously thinking about purchasing "Sound Reproduction", which I have yet to read.
Before you buy it, get an idea of what's in it here. If you can get through it and/or like what you read, then it's well worth getting the book.
post #3093 of 9583
Quote:
Originally Posted by jevansoh View Post

Quote:
Originally Posted by AustinJerry View Post

Quote:
Originally Posted by jkasanic View Post

+1 and to this end, is it outside the scope of this thread to start debating the pros and cons of the various models to try and systematically develop an approach one might take to achieve a nice compromise between movies and music (understanding that we're venturing into preference territory)?  More specifically, how the use of tools like REW and various room treatment options can be used to achieve this (e.g. decay times, reflections etc.)?  I realize the purists among us will want to adhere to the strictest of guidelines wrt each model but I'm gathering that if one model fit everyone's goal precisely then we wouldn't be debating things like speech intelligibility of movies vs. music.  It seems Stuart, Jerry and myself are at least a couple members of this thread that are looking for a means to achieve a good compromise that doesn't necessarily fall within the parameters of various other room models (e.g. LEDE, RFZ, ambiechoic, AE, NE etc.) and at least in my case, the abundance of information presented has me confused on next steps for my HT particularly wrt adding room treatments.

I don't think this would be out of bounds at all.  However, I also think that as a starting point, we should establish and agree upon some basic standards.  I have read Nyal's "Acoustics Measurement Standards", and while I think this is a wonderful and useful document, I think some of the standards are unrealistic for those of us who have a mixed-use listening room.  For example, I am currently working on reducing early reflections.  All of my reflections are -15dB or better in the 20ms time frame (except for that damn ceiling fan), and all are -20dB or better after 20ms.  However, I think I have heard that all reflections should be -20dB, regardless of the time frame.  An attainable standard for the purposes of this thread would be useful.

I believe we have already agreed upon a modal decay standard--all resonances below the 40dB noise floor by 450ms in the 40Hz-300Hz range.  This is somewhat less aggressive than Nyal's standard. 

So, I'm not sure I will ever adopt a "model" per se, but I am very interested in understanding the various theories.  I am especially interested in some topics we have not explored in detail, e.g. diffusion (when and where to use it), and how to re-direct reflections so they add spaciousness to the listening room, i.e. more good reflections, less bad reflections (and a clear understanding of what this means).

So much to learn, and so much fun on the journey!

Jerry,

I agree that we need to set some standards and I realize not everyone is going to strictly adhere to a specific acoustic model. I do think one should choose a model and set that as the goal, then try to get as close as possible though.

I have ideas in mind, many of which are close to Nyal's BTW (but there are some key differences which push more towards LEDE/RFZ standards) but I have a big problem right now.

In the community as a whole, most all the experts agree with the long standing research on topics such as early reflections vs late reflections good/bad/what the targets should be but we have ONE poster who is coming in here, derailing us by telling folks they need to pay attention to an RT60 measurement which has no basis in reality for statistically small acoustical spaces, then also telling people to look at only a very specific range between 500hz and 1khz, not setting proper guidelines, getting ahead of the pack, and introducing all kinds of text/references that are irrelevant to the topic and none of his own information or proof of his concepts (peer reviewed and accepted info, for instance) which almost all are in absolute conflict with the proven standards which I aim to teach here.

This is disheartening to me because I personally don't feel like arguing with him, he already stated early on he would leave the thread, but we see that's not true, has personally questioned my credentials and attacked me in another thread, and if he continues I see this thread easily and quickly turning into all the other threads he's been involved in.

I don't want to see this thread go downhill and then fizzle out.

I hope people are learning who to listen to. Localhost, Markus, Roger, Nyal, Jeff Hedback, and a guy who has not graced us with his unlimited knowledge on these subjects by the name of Mark, who goes by "Dragonfyr" on AVS. These folks know what they are talking about (there are several others who I don't mention by name who are active in this thread and many others, and you know who you are talking about....THANK YOU ALL for your participation)

But this one guy is throwing us off, getting us to go down paths that are WAY ahead of where we stand now and aren't even relevant at all.

So.... At first, I suggested everyone pick a room model, list what they were interested in, then i planned to tally up the results and start teaching the most popular model first.

That hasn't happened and it seems as though a lot of folks have multi-purpose rooms and just want to get "close" in this thread.

That's fine. We simply need to decide what we need to focus on and what the majority would like to learn more about first.

I think we need to concentrate on and master one thing at a time.

Where do we begin?

Would you all like to talk more about the ETC, how to use it (and how/why it will give you accurate information that the TOPT feature on the RT60 tab in REW tries and only sometimes comes close to) and what those targets should be, plus of course how to get there?

Or should we concentrate on bass trapping, not overdoing broadband acoustics so we don't overdampen our rooms, etc?

Should we talk more about other methods besides just throwing up insulation to tame reflections, so we can preserve the finite amount of energy in our rooms and properly redirect it then reintroduce it laterally in an exponentially decaying form?

Should we talk about all the different types of room treatments and what they're used for first? Diffusers, 1D, 2D, Poly, QRD, PRD, Helmholtz, BAD panels, Membranes, etc?

Should we just look at one or two specific cases at a time, from helpful and popular posters, such as Jerry and Keith, who each want to accomplish a different model and go at their pace, filling in the information they need to know to accomplish their goals along the way?

This is YOUR thread, not mine. I'm just here to help and try to keep things straight.

What do you all want from me and expect from me? Tell me what you want to know and I'll try to work down the list from most popular to least popular.

BTW, I'm not saying that I know everything and I'm right and everyone else is wrong, but what I AM saying is since this isn't an advanced thread like you'd find over on Gearslutz and we don't have a lot of folks in the industry participating, we need to have a plan, go slow, and teach the methods that are TRIED AND TRUE - Not things from new papers released in the last few years - but methods that have been around for DECADES with proven results.

This isn't the forum/thread to try new methods and develop new room models.

This is a thread dedicated to new users of REW who wish to broaden their horizons and are just starting out. We need to give you the right information the first time so you don't find later that half the things you were told are WRONG or at least contested by most people in the industry.

I need some feedback and a vote. I'll take this wherever everyone wants to go and I'm willing to go wherever you all tell me. biggrin.gif

--Jason

 

I concur heartily, Jason.  My model is NE as you know. If that singles me out as 'odd', compared with the rest of the guys, many of whom want multi-purpose spaces, it is absolutely fine to ignore the NE goals AFAIAC and to concentrate your limited time on things that affect most members. Of course, if time and space permitted, I would be very, very happy to learn more about NE. I am reading about it, but it isn’t as helpful as being able to ask questions.

 

Everything I have read from the posters you single out, and elsewhere, tells me that effort and time expended on RT60 is effort and time wasted (in our context). In fact, I have found, as a useful guide, that more or less anything a certain poster says can be usefully ignored and in that sense alone, his posts serve a sort of 'contrarian' usefulness I guess, although I agree we would be better off without them. Every thread I have seen where this person contributes has ended up going off the rails and I hope that doesn't happen to this thread which -- for me at least -- is one of the most useful because many of us are at the same novice stage yet are being helped by really experienced guys.

post #3094 of 9583
Quote:
Originally Posted by Roger Dressler View Post

That is similar to the question I asked Amir.
Taking any difference in gain out of the equation, my experience has been that the repetition helps.
Quote:
Originally Posted by Roger Dressler View Post

I took it that "the other way" from changing timbre in an "undesirable way" would be to change it in a desirable way. Hence, improve timbre. If you are not saying the timbre improves, we are again in violent agreement.
Right, I'm not saying it adds pleasing/desirable colourations. But if Barron and Toole & Olive are to be believed, then it does make colourations and resonances less audible.

If you want to hear problems with a speaker, listen to one speaker. If you're listening to 5 or 7 of them, it's more difficult to hear those same problems.
Quote:
Originally Posted by Roger Dressler View Post

I'm not sure we can totally absolve early reflections of close-proximity side walls of affecting timbre.
Not talking "total" absolution here, just a matter of degree. From the Johansson and Toole quotes, less chance of colouration when the reflection is angularly/spatially separated from the direct sound than when they're both coming from the same direction.
post #3095 of 9583
Quote:
Originally Posted by Joe0Bloggs View Post

Anyone here tried DRC-FIR? http://drc-fir.sourceforge.net/

No, but to AustinJerry's point, let's keep the thread on topic. Room correction software is only appropriate to discuss in context of REW measurement technique and applications. It's all too easy (even starting out on topic) to move OT, mea culpa as well.
Edited by sdrucker - 5/29/13 at 5:15am
post #3096 of 9583
Quote:
Originally Posted by kbarnes701 View Post

Agree with Jerry - too OT for this thread. If you care to copy and paste it to the Audyssey thread Stuart, I will be happy to tell you what my own researches have discovered.

Me too in retrospect. I deleted my comment - please delete yours as well so there's no record on the thread.
post #3097 of 9583
Quote:
Originally Posted by sdrucker View Post


FYI, I also ran an ETC (envelope) chart on my post-Audyssey measurement of mains (L/R) and subs. The raw plot is, well, reflection happy above -20 db in my untreated room:


 

 

Stuart, my understanding is that the impulse measurements should always be taken for one speaker only.  By measuring the L+R combined, there is no way to determine which speaker is contributing to which reflection, and by how much.  Perhaps you could re-measure and post three graphs, for L, R and C?

post #3098 of 9583
Quote:
Originally Posted by jevansoh View Post


Jerry,

I agree that we need to set some standards and I realize not everyone is going to strictly adhere to a specific acoustic model. I do think one should choose a model and set that as the goal, then try to get as close as possible though.


So.... At first, I suggested everyone pick a room model, list what they were interested in, then i planned to tally up the results and start teaching the most popular model first.

That hasn't happened and it seems as though a lot of folks have multi-purpose rooms and just want to get "close" in this thread.

That's fine. We simply need to decide what we need to focus on and what the majority would like to learn more about first.

I think we need to concentrate on and master one thing at a time.

Where do we begin?

Would you all like to talk more about the ETC, how to use it (and how/why it will give you accurate information that the TOPT feature on the RT60 tab in REW tries and only sometimes comes close to) and what those targets should be, plus of course how to get there?

Or should we concentrate on bass trapping, not overdoing broadband acoustics so we don't overdampen our rooms, etc?

Should we talk more about other methods besides just throwing up insulation to tame reflections, so we can preserve the finite amount of energy in our rooms and properly redirect it then reintroduce it laterally in an exponentially decaying form?

Should we talk about all the different types of room treatments and what they're used for first? Diffusers, 1D, 2D, Poly, QRD, PRD, Helmholtz, BAD panels, Membranes, etc?

Should we just look at one or two specific cases at a time, from helpful and popular posters, such as Jerry and Keith, who each want to accomplish a different model and go at their pace, filling in the information they need to know to accomplish their goals along the way?

This is YOUR thread, not mine. I'm just here to help and try to keep things straight.

What do you all want from me and expect from me? Tell me what you want to know and I'll try to work down the list from most popular to least popular.

BTW, I'm not saying that I know everything and I'm right and everyone else is wrong, but what I AM saying is since this isn't an advanced thread like you'd find over on Gearslutz and we don't have a lot of folks in the industry participating, we need to have a plan, go slow, and teach the methods that are TRIED AND TRUE - Not things from new papers released in the last few years - but methods that have been around for DECADES with proven results.

This isn't the forum/thread to try new methods and develop new room models.

This is a thread dedicated to new users of REW who wish to broaden their horizons and are just starting out. We need to give you the right information the first time so you don't find later that half the things you were told are WRONG or at least contested by most people in the industry.

I need some feedback and a vote. I'll take this wherever everyone wants to go and I'm willing to go wherever you all tell me. biggrin.gif

--Jason

 

I agree, J, we need some focus, and I look forward to your advice and guidance, as well as the contributions from the other knowledge experts you named.  I speak for myself, and hopefully for the others who are learning by participating in this thread, that your contributions are greatly participated.

 

So, here are the areas I would like to see addressed (in no particular order):

 

- How to improve overall bass quality by reducing decay times.  Am I using the right treatments, and are they placed effectively?

- Improving impulse response measurement techniques.  Establishing specific guidelines on how to measure.  Interpreting the results, identifying reflection sources, agreeing on the objectives (whether LEDE/RFZ or NE), and developing the techniques to accomplish the objectives.

- Understanding diffusion--when to use it, where to place diffusion products, which products to consider, etc.

 

Perhaps these objectives are too broad? 

 

Speaking for myself, I feel that the LEDE/RFZ model most closely matches my objectives.  I hear and enjoy the incremental improvements I have made to my listening room most when I am listening to 2-channel music (whether in stereo or PLIIz).  If a change produces undesirable results, I hear it immediately listening to music. 

 

It will be interesting to hear the objectives of the other thread regulars.

post #3099 of 9583
Quote:
Originally Posted by jevansoh View Post



If you are more interested in 2-channel, meaning you more critically listen to it, whether or not you listen to it more than watch movies, then you want to work towards a LEDE/RFZ design.

If you are more interested in movies and mostly watch movies, plus aren't too critical of your 2-channel (don't sit with your head in a vice or buy special recliners with low backs, etc) then you are most likely going to work towards an NE model.

Here's the main difference regarding reflections.

First, reflections ARE NOT BAD!!! We have to quit saying that.

High Gain EARLY reflections (less than about 20ms) are bad. This goes for either room model and in fact ALL common acoustic models.

So, after the direct sound, you should have reflections no higher in gain than -15db and the real goal, especially if going for the specific NE or LEDE model should be -20db.

At around 20ms is where the models change. For NE, this simply continues. It should evenly decay more and more beyond about 20ms to infinity. Where this decays into the noise floor, looking at the ETC (NOT RT60 as that doesn't EXIST in small acoustical spaces) is your decay time and this is how you can figure out how live/dead your room is. (This is a real quick/dumbed down answer but we'll go into more detail later)

So if you mostly listen to movies and aren't too critical of or don't listen to a lot of 2-channel only music, you simply want to make sure you don't have any reflections higher than about -15db in the first 20ms (but really try to make that -20db) and that should simply continue at and after around 20ms to decay more and more, and EVENLY, into the noise floor.


Now, I know this post is jumbled and yet another novel as I don't seem to be able to make short posts, but Jerry, Keith, et al - If you would/could take this information and "pretty it up" so more people can read, understand/interpret, and be able to put this to use in a more clear/concise format I'd be forever grateful.

Thanks a bunch, and let's keep the conversation moving forward...

--J

 

J, this is the kind of information that is especially useful.  Target reflections to be -15dB in the first 20ms--that's simple to understand, and a good starting point.  I'm looking forward to more discussion regarding reflections, especially re-introducing reflection beyond 20ms.

 

As for consolidating these tips in one spot and "prettying it up", I would be happy to give it a try, and think it would be a useful enhancement to the Guide.  Give me a little time to create a draft.  (It would be useful if AVS still had the capability to export a thread to a text file, but they don't.)

post #3100 of 9583
Quote:
Originally Posted by localhost127 View Post


just hold a broadband absorber in the direction/vector that you think the reflection is ingress from and remeasure to see if diminished (eg, the "blocking method"). rinse / repeat.

remember, since you are not utilizing hardware loopback (eg, direct signal arriving at T=0), the total flight path of the indirect energy will not be accurate (although the delta between direct and indirect will be).

 

Local, this is the second time you have advised me regarding the "Use Loopback as Timing Reference" setting for the impulse response measurements, and I appreciate your advice.  For others who are not familiar with where this setting is, open the Analysis tab on the Preferences screen:

 

 

 

I am familiar with how to configure the hardware loopback when using the "legacy" REW gear, but am not sure whether it is even possible when using the USB Mic + HDMI connection to the AVR.  Can you (or anyone else who knows the answer) comment on whether this is possible or not?  If not, does this mean that if we are using the HDMI connection, the impulse measurement is wrong, or meaningless?

 

I'm going to ask this question over on HTS as well.

post #3101 of 9583

BTW, as an interesting exercise (to me at least smile.gif ), and to leverage the new knowledge I have learned recently here in the thread, I plan on conducting the following exercise:

 

- Temporarily removing all of the broadband absorption panels on my walls (easy to do).

- Measuring room response without the panels in place.

- Re-mounting the panels one-by-one, measuring as I go, to assess their effectiveness.

 

I suspect that some of the panels, which were installed several years ago back in my "acoustic dark age", are ineffective, or improperly placed. 

 

I don't plan on doing anything with the bass traps.  These traps are installed in room corners where two walls intersect and, to the best of my knowledge, the placement is correct.  These panels are more difficult to remove temporarily.

post #3102 of 9583
Quote:
Originally Posted by sdrucker View Post

OK - here's some RT60 plots for commentary. If I focus on the 500 to 1000 Hz range, according to the 0.2 to 0.5 'standard' Amir mentioned in his cookbook, am I "OK"?
RT60 plots are run with TOPT, and the full frequency range as per the plot Jerry ran.

Center speaker
With your ~0.45 RT60 time at 500-1000 Hz, you are definitely "OK" n the RT60 department smile.gif. So good job there. Since there is still some concern about this being a contrarian view as Keith is putting it (I love that word smile.gif ), here are a list of experts saying the same:

Acoustical Measurement Standards for Stereo Listening Rooms paper by our own Nyal Mellor (Acoustic Frontiers) and Jeff Hedback (HdAcoustics):
D: Midrange DecayTimes T20, T30, T60 T60 between 0.2 and 0.5s

From another one of our resident acousticians on AVS, Dennis Erskine: http://www.avsforum.com/t/332289/rt60-what-is-a-good-value#post_2993755:
"I try to keep it [RT60] in the range of .35 to .40

From world renowned acoustics and psychoacoustics expert, Dr. Toole in his excellent book, Sound Reproduction, Loudspeakers and Rooms: http://www.amazon.com/Sound-Reproduction-Acoustics-Psychoacoustics-Loudspeakers/dp/0240520092/ref=sr_1_1?ie=UTF8&qid=1363204226&sr=8-1&keywords=toole+sound+reproduction
”The only reason to measure RT in a small room is to be certain that it is not excessively high (over about 0.5 s) to preserve high speech intelligibility or low (under about 0.2 s) to avoid oppressive “deadness.” ... Reverberation times less than 0.5 s are not likely to degrade speech intelligibility. Music and multipurpose rooms might drift toward 0.4 s, and dedicated movie rooms might drift toward 0.2 s.

From private presentation by top acoustic designer, Tony Grimoni:
"Decay Time Guidelines
• Decay time should be .2 to .4 seconds
• Research shows that most people like the same range of reverb time"


Here are some international standards for building listening rooms:
EBU 3276 standard: Listening conditions for the assessment of sound programme material: monophonic and two–channel stereophonic
Reverberation time is frequency–dependent. The nominal value, Tm, is the average of the measured reverberation times in the 1/3–octave bands from 200 Hz to 4 kHz. The nominal reverberation time, Tm, should lie in the range: 0.2 < Tm < 0.4 s”

ITU-R BS.1116-1 standard: METHODS FOR THE SUBJECTIVE ASSESSMENT OF SMALL IMPAIRMENTS IN AUDIO SYSTEMS INCLUDING MULTICHANNEL SOUND SYSTEMS
” 8.2.3.1 Reverberation time
The average value of reverberation, Tm, measured over the frequency range 200 Hz to 4 kHz should be:
Tm = 0,25 (V / V0)^1/3 s where:
V: volume of room
V0 : reference volume of 100 m3”


100 m^3 is ~3,500 cubic feet. For that, you would get 0.25.

You also see some confirmation of why we look at the narrow range/mid-frequencies here. We do that because that is where speech dominates and with respect to RT60 times, one of the goals is to have clear speech. Indeed the default for RT60 if not stated is 500 Hz. Here is a reference on that: Everest and Pohlmann Master Book of Accostics:

A typical reference frequency for reverberation time is 500 Hz, and 125 Hz and 2 kHz are used as well. To be precise, any reverberation time should be accompanied by an indication of frequency. For example, a reverberation time at 125 Hz might be quoted as RT60/125. When there is no frequency designation, the reference frequency is assumed to be 500 Hz.

With respect to this question:
Quote:
Or should I be focusing on absorbing some of these frequencies below 500 Hz, given that I really don't want an NE room (but haven't settled on exactly what my hypothetical model should be)?
RT60 is a measure of overall room response once you have many reflections. At low frequencies, we do not have many reflections and hence, we don't use RT60. Your best tool there is frequency response. It should be smooth and tilting down. Your response pre-Audyssey eq was tilting down but was not super smooth. Post-eq it became smoother but lost its tilting down. Unfortunately that is the characteristics of the Eq you are using and there is no quick solution to it, other than listening and if the bass is too anemic post eq, turn up the sub level. A manual EQ system would let you dial in both but will take a lot more work and knowledge to do it right.

With respect to picking a model, you will never have data of your own to make that evaluation, lest you tell me that you will build half a dozen rooms and do instant AB between them blind smile.gif. So ultimately you have to decide which crowd you want to follow. A bunch of us are in Dr. Toole camp which says those models are antiquated and are devoid of latest research into acoustic science (the parallel discussion on speech intelligibility by Sanjay and I is based on that collective wisdom). Buy his book and read his proof points. I think you will find them superbly impressive, eye opening, and backed up with literally hundreds of research papers. Dr. Toole is a friend and I have had a pleasure of spending a lot of time with him and his team. And he and I share the same knowledge in psychoacoustics. So what he says makes sense to me. If what others sound convincing to you instead, don't let me stop you from going there smile.gif.

If you believe in Dr. Toole's method, then you don't need ETC or any other measure for treatment of early reflections (not to be confused by late reflections represented by RT60). Indeed Dr Toole has pretty harsh words for anyone trusting the accuracy of what they are seeing in ETC:

"It seems obvious to look at reflections in the time domain, in a “reflectogram” or impulse response, a simple oscilloscope-like display of events as a function of time or, the currently popular alternative, the ETC (energy-time curve). In such displays, the strength of the reflection would be represented by the height of the spike. However, the height of a spike is affected by the frequency content of the reflection, and time-domain displays are “blind” to spectrum. The measurement has no information about the frequency content of the sound it represents. Only if the spectra of the sounds represented by two spikes are identical can they legitimately be compared."

You cannot meet the last condition unless all absorption in your room from carpet to furnishings to absorption/diffusion panels have identical response and your speaker is omnidirectional with little change in its frequency response at all angles. Walk around your speaker. If the level of highs changes as I suspect it would, then it is not that way and the reflections from its radiation will have different frequency response and hence, cannot be compared with each other. Likewise if the construction and thickness of everything in your room is not identical (good example: carpet vs 4 inch absorber), then again you are comparing apples vs oranges as you look at different spikes in ETC. The signal processing math is very unforgiving in how it invalidates those results. No amount of disagreeing solves that problem smile.gif.

Since I am in violent agreement with Dr. Toole and other published researchers who have actually measured the above, I will leave it to one of the ETC advocates to walk you through what to do with the ETC measurements. But I plead with you to buy Dr. Toole's book and at least read the so called "contrarian" view smile.gif.

Take care.
post #3103 of 9583

In researching the Hardware Loopback issue I reported earlier, I came across a posting by Localhost127 on the HTS web site in which he referenced a Dec 2010 Gearslutz forum posting by SAC concerning the ETC measurement.  The post is a wonderful summary of ETC, how to use it, how it can be used to achieve the objectives of acoustic models, how to use it to select and place the correct treatment to address different objectives, and other words of ETC wisdom, all in words that are easily understood by beginners, journeymen and masters alike.  It is an interesting read, and a must for any of us who are trying to master using ETC.

 

Enjoy!  http://www.gearslutz.com/board/6133764-post8.html


Edited by AustinJerry - 5/29/13 at 9:21am
post #3104 of 9583
Quote:
Originally Posted by sdurani View Post


Before you buy it, get an idea of what's in it here. If you can get through it and/or like what you read, then it's well worth getting the book.

 

Thanks, Sanjay.  I have that paper in my reference folder, and know I have read it at least once.  However, it's difficult to keep all the diverse theories straight in my head!  You advice is good--read this before buying the book.

post #3105 of 9583
Quote:
Originally Posted by amirm View Post

If you believe in Dr. Toole's method, then you don't need ETC or any other measure for treatment of early reflections (not to be confused by late reflections represented by RT60). Indeed Dr Toole has pretty harsh words for anyone trusting the accuracy of what they are seeing in ETC:

You make it sound like RT would show more precise information than an ETC. That is certainly not the case.
post #3106 of 9583
Wow, quite a lot to digest between Amir and Jason's posts (which go in opposite directiions on RT60). Particularly Amir's, which are a contrarian/heretical view to many of the thought leaders here.

I need to read both of these in more detail, as well as pick up Toole's book and the one Jason recommended, at least, as a reference before making a decision about what's right for me. I'm definitely not in a position for, or have interest in, an NE room. At least a modified LEDE is where I'd start by default.

A couple of general thoughts:
I'm quite aware I have an multipurpose room w/o dedicated trearments, namely our 24x17x8.5 living room, connected to our dining room. IOW, I'm a neophyte barbarian relative to the time, expense, and learning that more senior users like Jerry and Keith have put into creating more dedicated HT spaces. Guilty as charged: but I think I'm not atypical of AVS users of AVRs with room EQ software (even if I'm likely atypical of the active participants on this REW group). I know I'm starting from the bottom, but that's why I'm here. If not for now, the future when we move and I can hopefully have a room better dedicated to HT film or music optimization.

IMO, I have practical problems about where I can put absorption panels due to the room configuration (i.e. you can't block windows or an open part of an entrance with pink fluffy). I described them on Sunday on the thread, but will post pictures so you'll see my point, if the group thinks it's on topic to do so. Or PM them to Jason or anyone else that's curious.

I've relied on Audyssey XT32 EQ and the Pro Kit as a device for correcting the room, as a crutch possibly, but at least it's done a reasonable job in smoothing my FR, albeit (as Amir notes) at the potential cost of 'losing' the tilted-down curve many prefer. I can, as I've considered, use a Behringer unit or MiniDSP with subs to look at this with more control than the Pro Kit's Target Curve Editor, all course subject to REW measurement. OTOH, I actually like the 'flatter' bass. YMMV, naturally, depending on feelings about Audyssey.

Now to Jason's comments:
If you are a proponent of ETC, my room's infected with problematic reflections < 20 ms. I will post separate speaker graphs for L, R and C, all w/o Audyssey. Not that I'd have known without REW!

Jason stated that he wouldn't have been surprised if my CC dialogue had clarity issues (regardless of whether the CC was turned up), and music lyrics were unintelligible given RT60 and IR. When I had no room correction with my old Denon 5803 and Klipsch RF-7 mains, I would agree it was a chronic problem. However, with XT32, two more capable subs, and the Mythos ST+center I have, I disagree strongly, at least for multichannel DVD and BluRay. That's with Audyssey DEQ on or off, BTW.

If I were to take the RT60 plots seriously beyond the range Amir cites (Jason caveats RT60 in general as 'ballpark' estimates), I need not only broadband absorption but also LF absorption to reduce bass issues.The CC does 'worse' on LF than the L/R. But, if anything I have some suckout with bass in the 150 Hz region (more mild with Audyssey, though) and have always needed at least a 90-100 Hz for my CS-8080 HD powered center, which I've attributed to having more capable subs for LF. Are you saying, Jason, that 'too much' bass reflection for the center may be the culprit, in the 'Early reflections more than -15 db < 20 ms are BAD' sense?

Finally, I may not be able to actively do treatments as Jason strongly recommends, given the room setup and WAF. But what I _can_ do I will try, if I'm persuaded it's worth the effort and expense. One thing I started looking into -- Sound Elite acoustic vertical blinds for the windows (curtains are out) --would primarily absorb in the 500 to 2K range, based on the chart on their site. And I'd be spending more on that than I spent on my Denon! So if we go this route, it's non-trivial.

Last but not least: I vote in favor of more knowledge about IR as a priority, and buiding a virtual library of sources for basic and follow-up reading. Assuming that I shouldn't be 'sentenced' to a year of remedial reading and barred from posting until I add extensive treatments, that is rolleyes.gif.
Edited by sdrucker - 5/29/13 at 10:20am
post #3107 of 9583


Black = Absorber, Orange = Diffuser, Gray = Reflector

Those of you heading in a LEDE /RFZ direction may take a look at this. This is what I have done acoustically to get meet the >20ms termination and following "tail". The RED line shows the path of my specular return responsible for my terminator (-6-8db in my case). In a small room (mine is 15.5W x 25.75L), getting a delay of >20ms is no easy task. To merely bounce it off of the rear corner, you would have to have that wall 11.25' away. But by bouncing the signal twice, ive arrived at a 24ms delay.


I am not sure this would work, but since many of you have 7.1, maybe you could situate your surround speakers where they are 120 degrees (RED) and 60 degrees (Green) to each side of you and merely dial in the delay (20-30ms) in order to produce the necessary delay and the right directionality.
Edited by jim19611961 - 5/29/13 at 10:56am
post #3108 of 9583
Jim,

Did you also try a shorter termination?
post #3109 of 9583
Quote:
Originally Posted by sdurani View Post

Originally Posted by kbarnes701 View Post

I choose to kill the reflections. The reason is that all of the ambience I require (for movies only) is already recorded into the soundtrack and reproduced by the array of speakers around the room. I don't want any further influence of the room on the recorded sound.
So you absorb the early reflections of your centre speaker for reasons that have to do with ambience and not intelligibility? Do you think a room can be treated to improve intelligibility or do you think that the improvement can only happen with a volume control? The reason I ask is because if you believe the latter, then discussing treatment vs intelligibility is a moot conversation.

 

No - I was specifically talking there about the issue of ambience or 'spaciousness' wrt to the discussion on reflections. IMO all I require is already engineered into the soundtrack and I do not need reflections of any kind to provide ambiance and spaciousness in a m/ch HT. 

 

Of course I believe that a room can be treated to improve intelligibility - I have treated my own room specifically for that purpose, as I have mentioned several times. But I have targeted the removal of reflections in order to achieve the goal of improved intelligibility. It has worked well here and I have zero issues with dialogue intelligibility.

 

 

Quote:
Quote:
Originally Posted by kbarnes701 View Post

Would you comment on the headphones issue I have mentioned before?  If these reflections are so important for dialogue intelligibility, why are there no problems with the latter when using headphones, where the possibility of reflections influencing anything is zero?
Where did anyone say there would be "problems" without early reflections? You're erecting a strawman. The discussion is whether early reflections aid or hamper intelligibility, not whether dialogue is in unintelligible without reflections.

 

It has been stated by another poster than early reflections are required in order to assist with dialogue intelligibility. It is my contention that they are not required. They may of course help, but are not required. As there are no reflections at all when using headphones, it is clear that reflections are not required. If they were, then dialogue heard via headphones would suffer intelligibility problems, and it does not. 

 

 

Quote:
BTW, do you believe that headphones provide an absolutely equal level of intelligibility as your room or do you think one might be better than the other?

 

Difficult to say for sure as I am not a big user of headphones (other than music on an iPod sometimes, and the headphones -- the in-ear sort -- that I use for that are not of sufficient quality IMO to give a definite answer). But on the whole I have used headphones (proper, over the ear types) on occasion and not found any dialogue intelligibility issues of any kind - but then I don't when listening through speakers anyway. There are too many variables to say with 'absolute certainty' that they are 'absolutely equal' - mostly in the comparative qualities of my headphones and my speakers for example. 

 

 

Quote:
Quote:
Originally Posted by kbarnes701 View Post

I absolutely agree with your general point there (and have been that man myself).
But you don't believe you are that man this time?

 

In the sense that I have absolutely zero problems with dialogue intelligibility, and that I can hear 100% of the words spoken, and that it is impossible to hear more than 100%, I would say, yes, I do not believe I am that man this time. I’d be happy to be proved wrong of course - if somehow I could treat my room further and get dialogue intelligibility that surpasses 100% intelligibility in some way, I'd go for that!  Roger touched on other issues of dialogue quality, in terms of emotional content etc in the reproduced voice, but I also believe I am good to go in that sense too, so it is difficult for me to see a path forward to make significant improvements. I am always open to the possibility that things could be improved - but in this specific case, how?

post #3110 of 9583
Quote:
Originally Posted by markus767 View Post

Jim,

Did you also try a shorter termination?

Yes. I tried a contra-lateral 1st bounce off the sidewall that came in at about 12-14ms. I am not sure if it was the contra-lateral orientation or the much shorter ISD gap, but it never sounded right to me.
post #3111 of 9583
Quote:
Originally Posted by sdrucker View Post

Quote:
Originally Posted by kbarnes701 View Post

Agree with Jerry - too OT for this thread. If you care to copy and paste it to the Audyssey thread Stuart, I will be happy to tell you what my own researches have discovered.

Me too in retrospect. I deleted my comment - please delete yours as well so there's no record on the thread.

 Done.

post #3112 of 9583
Quote:
Originally Posted by AustinJerry View Post

As for consolidating these tips in one spot and "prettying it up", I would be happy to give it a try, and think it would be a useful enhancement to the Guide.  Give me a little time to create a draft.  (It would be useful if AVS still had the capability to export a thread to a text file, but they don't.)

 

I think that would be an enormously useful contribution, Jerry. It would be good to have a single reference point to return to from time to time - currently the info is scattered across 3000 separate posts.

post #3113 of 9583
Quote:
Originally Posted by jim19611961 View Post

Yes. I tried a contra-lateral 1st bounce off the sidewall that came in at about 12-14ms. I am not sure if it was the contra-lateral orientation or the much shorter ISD gap, but it never sounded right to me.

Did you also try a first bounce from the back?
post #3114 of 9583
Quote:
Originally Posted by markus767 View Post

Did you also try a first bounce from the back?

If by "back" you mean 120 - 150 degrees, then yes. If you mean directly behind (180), then no.
post #3115 of 9583
Quote:
Originally Posted by kbarnes701 View Post

Of course I believe that a room can be treated to improve intelligibility - I have treated my own room specifically for that purpose, as I have mentioned several times. But I have targeted the removal of reflections in order to achieve the goal of improved intelligibility.
OK, this really narrows down where you and I differ in opinion. Before reaching for the volume knob, which we both agree can improve intelligibility, the room itself can be treated to have good intelligibility as a starting point, which we also agree on. It then boils down to which treatment: absorbtion, diffusion or reflection. This is where we differ. And I can live with that, considering how much other stuff we do agree on.
Quote:
Originally Posted by kbarnes701 View Post

But on the whole I have used headphones (proper, over the ear types) on occasion and not found any dialogue intelligibility issues of any kind - but then I don't when listening through speakers anyway. There are too many variables to say with 'absolute certainty' that they are 'absolutely equal' - mostly in the comparative qualities of my headphones and my speakers for example.
What happened to Mr. 'some things are absolute' from the previous post? Anyway, you have brought up 3 examples (your heavily treated HT, your untreated music set-up, and headphones) that all have zero intelligibility problems yet sound very different from one another. With that in mind, I hope you are open minded enough to understand that something can sound very different from your HT (e.g., bare side walls) and still not suffer any intelligibility problems.
post #3116 of 9583
amirm lives in a world with magical sidewalls which only reflect the dialog from the center channel and none of the other simultaneous content from the left, right and center channel speakers, thus enhancing intelligibillity.
post #3117 of 9583
Quote:
Originally Posted by markus767 View Post

You make it sound like RT would show more precise information than an ETC. That is certainly not the case.
Well, your point is non-sequitur because we don't use RT60 as a precise measure. It is given with a very wide range of 0.2 to 0.5 so precision is not what it is about or needed in the recommended use. As I have repeated many times, you don't worry about second decimal places and such.

ETC on the hand is used by you all with very precise meaning, looking at db scales and such. Unfortunately in that mode of usage, it can misfire and misfire really big. There is an empirical proof of this in Dr. Toole's book which unfortunately is very hard for a layman to understand. So let me take a shot at simplifying it here to show this point.

The test is a controlled one where the level of a reflection is adjusted to be perceptually the same (i.e. heard the same way) with or without a 2-inch absorber. These are the frequency response and ETC measurements showing before and after:

i-sNvGjhR-M.png

Look at the top graphs first. The one on the left shows the full bandwidth reflection and the one the right, the filtering that occurred > 500 Hz due to that absorber. The level of the spectrum for the reflection with absorber is 3 db higher. It is that way because its level was increased to make it be detectable the same as if that absorber was not there. Since we took part of the spectrum away, it only makes sense that would need to turn up the volume to get to the same audible level. With me so far? We filtered part of the spectrum but then juiced up the level to compensate and make it sound the same as before as far as detection.

Now we perform the analysis on the identical signal in the bottom set of graphs but this time we use ETC. The test manipulated the second reflection so ignore the first tall one and focus on the little guy where the dashed line placed above it. We see that the ETC measurement is telling us that we have reduced the level of that reflection by 20 dB. This is wildly incorrect:

1. The level has not changed perceptually (i.e. what we hear) between the two measurements. So the correct data that would correspond to what we hear, would say 0 dB, not 20 dB. 20 dB error is massive. One looks at that display and thinks that reflection has been dealt with. In reality it is way, way louder.

2. The frequency response/spectrum shows that the electronic level, i.e. what an instrument should have measured as opposed to a human ear, is 3 dB *higher*. ETC unfortunately claims that it is 20 dB lower for a total measurement error of 23 dB.

These are not small errors. And are especially problematic because of the focused way you all look at the ETC measurements.

The reason for massive error is that ETC relies heavily on the power of high frequencies for the amplitude of what it shows. Unfortunately many things in your room impact the level of high frequency reflections from carpeting, to curtains, furnishings, and abosrbers if they are not thick and broadband. The only time ETC can be valid is prior to doing anything to the room. All surfaces must be bare including the floor. Even then, you get hit in the face with the reality that your speaker sends sound waves in different directions with different levels of high frequencies. Sound hitting the ceiling is not going to have the same high frequencies as those hitting the back wall or the side. If you don't believe me, do as I suggested which is get your ear down to the speaker level and circle it. If you hear variations in high frequencies, then your ETC spikes are showing incorrect values that cannot be compared with each other.

All of this leads Dr. Toole to say this:

"The message is that we need to know the spectrum level of reflections to be able to gauge their relative audible effects. This can be done using time-domain representations, like ETC or impulse responses, but it must be done using a method that equates the spectra in all of the spikes in the display, such as bandpass filtering. Examining the “slices” of a waterfall would also be to the point, as would performing FFTs on individual refl ections isolated by time windowing of an impulse response. Such processes need to be done with care because of the trade-off between time and frequency resolution, as explained in Section 13.5. It is quite possible to generate meaningless data. All of this is especially relevant in room acoustics because acoustical materials, absorbers, and diffusers routinely modify the spectra of reflected sounds. Whenever the direct and reflected sounds have different spectra, simple broadband ETCs or impulse responses are not trustworthy indicators of audible effects."

And it is not just him saying this. There is another excellent AES paper which tests for speaker directivity and such and arrives a the same conclusion and cautionary note.

I realize this stuff is hard to understand. It was for me too smile.gif. I must have read the above test half a dozen times until I understood what Dr. Toole was doing and what he was measuring. But difficulty of understanding the science doesn't give one the ticket to ignore how things work.

There is complexity behind every acoustic measurement. We can try to simplify them and put them to good use. RT60 is one. It takes 5 seconds to run and if you look at it in the prescribed manner, you get pretty good insight about your room. It will help calibrate your ears and eyes quickly on what a live or dead room looks like. Over time, you may not even need it. I wish I had 5 second version for ETC but I don't. The math and signal processing and reality of the situation is that it doesn't lend itself to layman usage.

Fortunately there is a modern school of acoustics which does not use ETC at all. Ask a top acoustical engineer to design a room for you. He will create one for you, with all the treatments in there. How can he do that without the room being built and hence lacking ETC measurements??? He is able to do that because he knows the effects of reflections perceptually and uses that model to design a room. He doesn't need a string to find reflection points either since the room does not exist to use the darn thing with it. You too can learn to do it the same way. Maybe we don't feel as manly doing it without a fancy looking measure like ETC smile.gif. I have to grant you that. If the goal however is good audio, then should not matter and the focus should be on tools that don't miss he boatta by 23 dB....
post #3118 of 9583
Quote:
Originally Posted by sdurani View Post

Quote:
Originally Posted by kbarnes701 View Post

Of course I believe that a room can be treated to improve intelligibility - I have treated my own room specifically for that purpose, as I have mentioned several times. But I have targeted the removal of reflections in order to achieve the goal of improved intelligibility.
OK, this really narrows down where you and I differ in opinion. Before reaching for the volume knob, which we both agree can improve intelligibility, the room itself can be treated to have good intelligibility as a starting point, which we also agree on. It then boils down to which treatment: absorbtion, diffusion or reflection. This is where we differ. And I can live with that, considering how much other stuff we do agree on.

 

I think so too. I have said before that I think a lot of this comes down to preference, and the type of room we are modelling and its purpose. You and I probably have different preferences and goals - I, as you know, for example, do not even begin to consider the requirements for a multi-purpose room because I am solely concerned with movies in the HT.  I am sure this is very significant. I also prefer a slightly 'dry' environment and I know others do not. So yes, at last we (well you :)) have identified where we disagree - the route we take to achieve our objectives. I am happy to leave it there - you are the last person I want to be seen, or to appear to be seen, to be 'arguing' with!

 

 

Quote:
Quote:
Originally Posted by kbarnes701 View Post

But on the whole I have used headphones (proper, over the ear types) on occasion and not found any dialogue intelligibility issues of any kind - but then I don't when listening through speakers anyway. There are too many variables to say with 'absolute certainty' that they are 'absolutely equal' - mostly in the comparative qualities of my headphones and my speakers for example.
What happened to Mr. 'some things are absolute' from the previous post? Anyway, you have brought up 3 examples (your heavily treated HT, your untreated music set-up, and headphones) that all have zero intelligibility problems yet sound very different from one another.

 

I'm not sure I ever said that my (formally) untreated music room had zero intelligibility issues - if I did, then I apologise. The truth is, I don't really know if the music room does or does not. The vocal music I listen to is so well known to me you could stuff a rag into Frank's mouth and I'd still know he was singing about strangers exchanging glances, in the night :)  I am also convinced (but could easily be wrong) that I have learned to 'listen through' the room, due to familiarity with the equipment and the content (which hasn’t changed much in decades) and even to do with the familiarity with the room itself, which hasn't significantly changed in at least 7 years. Also, while I sit on the edge of my seat (metaphorically) and concentrate hard on the movies I enjoy so much (there is so much to take in - plot, dialogue, music, acting, cinematography, art direction, set design, production design, editing etc etc), I do tend to listen to music for relaxation - a totally different mind set, often accompanied by a glass of wine and sometimes by herbal enhancement too. This may mean that I have lower expectations and standards from my music room - I probably do in fact. The system in that room is good quality but hasn’t been changed for decades and I am happy to leave it that way.

 

Quote:
 With that in mind, I hope you are open minded enough to understand that something can sound very different from your HT (e.g., bare side walls) and still not suffer any intelligibility problems.

 

Yes, I agree fully with that. All of this started because a poster seemed (to me) to imply that early reflections were required for the proper intelligibility of dialogue. All I am saying really is that they are not required. They may well enhance dialogue intelligibility in some rooms but my own room proves, to me at least, that they are not required. If I have taken the wrong end of that stick, I apologise for doing so.

post #3119 of 9583

Ok, so here's a question on movies vs. music and LEDE /RFZ vs. NE that might influence my direction.  What if some of the time I spend listening to music it's a multi-ch audio track?  Should I really be looking at how much time I spend watching movies, listening to multi-ch audio and finally 2-ch audio when considering which room model suits me better?

post #3120 of 9583
Quote:
Originally Posted by jkasanic View Post

Ok, so here's a question on movies vs. music and LEDE /RFZ vs. NE that might influence my direction.  What if some of the time I spend listening to music it's a multi-ch audio track?  Should I really be looking at how much time I spend watching movies, listening to multi-ch audio and finally 2-ch audio when considering which room model suits me better?

I would think you would want to tune the room to that which you do the most often and/or which is most important to you.

For 2 ch audio, I have a pretty good idea whats needed for good sound, for xx.1 movies, not so much. But I am pretty sure the best room for one isnt for the other.
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