or Connect
AVS › AVS Forum › Audio › Audio theory, Setup and Chat › Are audio companies all involved in a huge conspiracy?
New Posts  All Forums:Forum Nav:

Are audio companies all involved in a huge conspiracy? - Page 23

post #661 of 3048
I agree with Amir, mostly. For minimum phase phenomenon, which covers low frequency response in small rooms, time and frequency are directly coupled. Improve frequency response, you improve time domain behavior. The only caveat is that improvements at one location may be detrimental in another location. Every peak in one location has a corresponding null in another. If you are unlucky enough to have equipment and listening positions so arranged that at a particularly problematic frequency you have listening positions in both conditions, you may not be able to do much with eq. This is particularly true if you have only one sub. The more you add, the more likely you will be able to use eq to optimize for multiple seating positions. Especially if you have flexibility in bother sub and seating locations.

But there is no guarantee that Amir's experiment can be duplicated for both ends of your couch or both rows of your seating. If you have one sub, can't move it much, and are unlucky with room and seating layout, you may be stuck. You can usually improve the response at least a bit, but may not make substantial improvements for multiple locations.

On the other hand, acoustic treatments work at the source of the problem and affect all locations in the room. Big and bulky, sure. No one denies that. Effective? They certainly can be. Can do some things eq cannot? Yep.

Like many things in life, the best results are often obtained from a balanced approach. Room dimensions that aren't obvious red flags, multiple capable subs, sub and listening positions carefully tweaked, acoustic products to smooth response further and perhaps attack a troublesome mode, and eq/gain/timing tweaks as icing to flatten response as much as possible.

I hope people don't mistake Amir's post fir proof that simple eq with a single sub is a panacea for modal problems any more than they mistake Ethan's tutorials for proof that a few corner traps are a modal panacea.
post #662 of 3048
Quote:
Originally Posted by Bigus View Post

Like many things in life, the best results are often obtained from a balanced approach. Room dimensions that aren't obvious red flags, multiple capable subs, sub and listening positions carefully tweaked, acoustic products to smooth response further and perhaps attack a troublesome mode, and eq/gain/timing tweaks as icing to flatten response as much as possible.

Yes absolutely, as I've stated many times, room acoustics are dynamic and all of the above is considered room treatment in my books.... and then some.
post #663 of 3048
Quote:
Originally Posted by Originally Posted by arnyk 
There are other kinds of errors, such as timing problems. They are often handled in other ways.
Quote:
Originally Posted by goneten 
I think I'm a little confused. I assumed (incorrectly) that error correction was used to prevent timing issues from becoming audible. If you hear it, correction can't help. But from the above (as I understand it) you are saying that error correction handles issues not related to timing. So error correction can't help with timing at all? What then helps with timing, if not error correction? Sorry, I'm not terribly well versed in this, but I am interested to learn more. Thanks for the help.

Just in case it got glossed over. smile.gif
post #664 of 3048
I read this from another "expert" in the field :

"The biggest audible effect jitter has is in smearing the stereo image because of the time based distortions introduced by high levels of jitter. I can probably spend hours serving up proof of this (time I don't have right now), but I'll leave it up to you to go check a little yourself".

confused.gif
post #665 of 3048
Moderator

respectfully request members avoid attacking other members to make a point: please take the high road in every post


thanks
post #666 of 3048
Quote:
Originally Posted by hd_newbie View Post

Ethan:
What do you think about claims made on this URL:
http://www.madronadigital.com/Showroom/HomeTheater.html

That system has so much invested in it in terms of equipment and design staff that it has zero relevance to virtually every audiophile who reads posts here.

Just for grins, what do you think would happen if we saw the invoice for the services described below, as applied to a real world audiophile's listening room?

http://keithyates.com/services.htm

As far as the Harman gear end of the equation goes, the room correction processors, amplifiers, and speakers probably add up to at least someplace in the middle-5-figures.

In the context applied here, being the kind of work and products that a good honest room room acoustics product manufacturer/retailer provides, it looks like a straw man argument to me.

It's like saying that an ALMS Porsche 911 is a better race car than the next 911 that rolls off of Porsche's Zuffenhausen Factory line.

In other words, what is the comparison between:

http://www.autoweek.com/article/20111104/ALMS/111109915

versus:

http://911nation.com/2010/04/ultimate-factories-inside-the-house-of-porsche-911/

I say, there is no meaningful comparison. Two completely different worlds. Using one to criticize the other is so unfair as to be totally illogical. Why would anybody do such a thing?
Edited by arnyk - 10/10/12 at 6:43am
post #667 of 3048
Quote:
Originally Posted by goneten View Post

I read this from another "expert" in the field :
"The biggest audible effect jitter has is in smearing the stereo image because of the time based distortions introduced by high levels of jitter. I can probably spend hours serving up proof of this (time I don't have right now), but I'll leave it up to you to go check a little yourself".
confused.gif

I can't find the sentence above with google searching, but I'd like to know who said it just for grins and giggles. ;-)

It is false because jitter generally affects all channels that are being played at one time in the identical same way. Since all of the channels are affected the same way, imaging would be the last thing to change. The first things to change audibly would be (and in the real world are) things like the tone and clarity of all of the channels.

If you want to read something real world and practical about what jitter is and sounds like, please try this:

http://nwavguy.blogspot.com/2011/02/jitter-does-it-matter.html
post #668 of 3048
Quote:
Originally Posted by Bigus View Post

For minimum phase phenomenon, which covers (very simple situations like) low frequency response in small rooms, time and frequency are directly coupled. Improve frequency response, you (can) improve time domain behavior (by means of digital or analog processing applied to the signal path).
(items in parenthesis added by me to make the statement agreeable to me)

My first point would be that acoustical performance in most listening rooms isn't minimum phase. This is widely known, easy to observe, and generally agreed upon.

Even when the room is small, you have problems using signal processing to control sound quality at middle and high audio frequencies. As the room gets larger, your ability to even just control sound quality of low frequencies with one subwoofer faces stiff challenges.
Quote:
The only caveat is that improvements at one location may be detrimental in another location.

I don't think that is the only caveat. ;-)

To some degree room acoustics breaks down into a problem in multivariate calculus. You have so many degrees of freedom, and so it is axiomatic that in order to control a certain number of degrees of freedom you need to have an equal or greater number of relevant parameters to change. For even that potentially huge level of complexity to be effective, there has to be an absence of of non-linearity and interdependence in the room. There is a kind of uncertainty principle that poses an ultimate limit to how far you can go, no matter what you try to do. All the kings horses and all the kings men can't rescue a pathological room.
Quote:
Every peak in one location has a corresponding null in another.

There is one kind of room that breaks that rule - an anechoic chamber.
Quote:
If you are unlucky enough to have equipment and listening positions so arranged that at a particularly problematic frequency you have listening positions in both conditions, you may not be able to do much with eq.

Exactly.
Quote:
This is particularly true if you have only one sub. The more you add, the more likely you will be able to use eq to optimize for multiple seating positions. Especially if you have flexibility in bother sub and seating locations.

One thought came into my mind. By using multiple subs we are trying to turn a large room into a collection of small rooms.
Quote:
But there is no guarantee that Amir's experiment can be duplicated for both ends of your couch or both rows of your seating. If you have one sub, can't move it much, and are unlucky with room and seating layout, you may be stuck. You can usually improve the response at least a bit, but may not make substantial improvements for multiple locations.

The general application of any particular experiment is always in question, particularly if it is an experiment that is designed by an advocate of a particular viewpoint.
Quote:
On the other hand, acoustic treatments (can) work at the source of the problem and (may even) affect all locations in the room. Big and bulky, sure. No one denies that. Effective? They certainly can be. Can do some things eq cannot? Yep.

Exactly! +1!
Quote:
Like many things in life, the best results are often obtained from a balanced approach. Room dimensions that aren't obvious red flags, multiple capable subs, sub and listening positions carefully tweaked, acoustic products to smooth response further and perhaps attack a troublesome mode, and eq/gain/timing tweaks as icing to flatten response as much as possible.

I hope people don't mistake Amir's post fir proof that simple eq with a single sub is a panacea for modal problems any more than they mistake Ethan's tutorials for proof that a few corner traps are a modal panacea.

Well said!
post #669 of 3048
Quote:
Originally Posted by Bigus View Post

I agree with Amir, mostly. For minimum phase phenomenon, which covers low frequency response in small rooms, time and frequency are directly coupled. Improve frequency response, you improve time domain behavior. The only caveat is that improvements at one location may be detrimental in another location. Every peak in one location has a corresponding null in another. If you are unlucky enough to have equipment and listening positions so arranged that at a particularly problematic frequency you have listening positions in both conditions, you may not be able to do much with eq. This is particularly true if you have only one sub. The more you add, the more likely you will be able to use eq to optimize for multiple seating positions. Especially if you have flexibility in bother sub and seating locations.
But there is no guarantee that Amir's experiment can be duplicated for both ends of your couch or both rows of your seating. If you have one sub, can't move it much, and are unlucky with room and seating layout, you may be stuck. You can usually improve the response at least a bit, but may not make substantial improvements for multiple locations.
On the other hand, acoustic treatments work at the source of the problem and affect all locations in the room. Big and bulky, sure. No one denies that. Effective? They certainly can be. Can do some things eq cannot? Yep.
Like many things in life, the best results are often obtained from a balanced approach. Room dimensions that aren't obvious red flags, multiple capable subs, sub and listening positions carefully tweaked, acoustic products to smooth response further and perhaps attack a troublesome mode, and eq/gain/timing tweaks as icing to flatten response as much as possible.
I hope people don't mistake Amir's post fir proof that simple eq with a single sub is a panacea for modal problems any more than they mistake Ethan's tutorials for proof that a few corner traps are a modal panacea.

That was my first question. How many positions did the mic sit in. You can EQ for a spot. You can throw up absorption and get the entire room.
post #670 of 3048
Why is Dragon not posting? I really hope he was not banned due to his abrupt style. His absence will be a net loss to this forum
post #671 of 3048
Quote:
Originally Posted by hd_newbie View Post

Ethan:
What do you think about claims made on this URL:
http://www.madronadigital.com/Showroom/HomeTheater.html

There are way too many claims there to comment on, but I can tell you a few impressions:

I see a lot of claims made that are not backed up with frequency response and ringing graphs, or data showing that a statistically significant number of listeners prefer what is offered for sale there. Without seeing the in-room response at high resolution for a number of listening positions, it seems more like flowery prose than audio science.

I do agree that bi-amped speakers have the potential to offer the best results. But excellent bi-amped speakers are a commodity, and can be had for a very reasonable cost without needing a "consultant" or design service to tell you what to buy.

--Ethan
post #672 of 3048
Quote:
Originally Posted by goneten View Post

Just in case it got glossed over. smile.gif
I thought you asked a question about lack of error correct in HDMI and now can't find it. Was that the edited post? I thought it was a good question so here is the answer smile.gif.

HDMI only has error correction for the "data island" and not the video data. The data island is part of the invisible portion of the video scan lines and holds the audio data. It is packetized with a subpacket of 64 bits, 8 of which is forward error correction bits. Protecting the data/audio segment was a wise choice as if there is a problem there, it will be quite annoying to the user. A single bit error in a compressed bit stream such as dobly digial can cause many milliseconds of audio data to be lost. In contrast, a single bit error in the video will likely not even be noticed as it will likely not be there in the next video frame. BTW, "forward" error correction means that the sender tries to send you extra bits that the receiver can use to correct some number of failing bits. This is in contrast to error correction over a network such as LAN or Internet where if data is corrupted the receiver can request for it to be resent. HDMI does not have any provisions. If what is there is not enough to correct the error, that is that.

The error correction gives pretty high resilience to the data/audio channel. By the time you get enough errors to make your audio fail, likely the video segment will be in far worse shape (since it has no error correction) so you will know you have a bad situation. If video is perfect then there is no reason to worry about audio channel having errors in it. The video is the canary in the coal mine so to speak.

In the context of discussion in this thread, none of this is material. Error correction does nothing to repair clock timing that drives the DAC since that signal is "analog" in nature. I can have high jitter driving my audio DAC while life is perfect in HDMI land. Samples could arrive all correct and you can still have distortion coming out of your DAC because the clock that is derived from HDMI is not clean. So the whole discussion of error correction is orthogonal to audio jitter. The assumption is that you have a working HDMI system and hence samples are arrive error free. Once there, you still have to worry about high jitter values which seem to be prevalent in consumer gear and at any rate not spec'ed by the manufacturer.

BTW, here is a graph that I have stashed away from a paper by Dolby and listening tests they did on Jitter:

i-qwRxQtT-XL.png

This is a test of pure tones. Staying with that for a minute, we see that listener 2 was able to detect jitter as low as 3 nanoseconds. We also see that different listeners have different thresholds here. Listener 6 for example, needed 20 nanoseconds for his minimum threshold. So one has to be careful in generalizing their own personal test results for others. Back to pure tone nature of this test, they also tested music and found thresholds far higher than this due to masking effect of music signals. The excitation remained a single non-data-variant pure jitter sinewave though. If you take one look at the Paull Miller data I showed earlier, you see that the actual situation is not anything like that but a storm of jitter sources at all different levels. And of course you are likely not listening to the same music as the Dolby people did (music selection also contributed to the detection level of jitter).

Net, net, you are not going to get to the bottom of this using simplified tests. Best path in my opinion is hygiene. Yes, you read that right smile.gif. When you go to bathroom, do you wash your hands? For how long? Do you use soap? Will you get sick for sure if you didn't? Likely you wash your hands often or all the time since the cost in doing so is not that high and benefits of not getting sick is. No matter how unlikely it might be. Same here. You can pick devices or solutions that have very low jitter. Simple thing like using S/PDIF instead of HDMI between same two devices at a cost of next to nothing would gain you potentially an order magnitude less jitter. To advocate and jump up and down that you shouldn't is tantalum to insisting that someone must not wash their hands lest they can show double blind tests that they did get sick by doing so!

So set a high bar here. Demand that manufacturers give you performant devices. Nothing goes wrong from picking a device with less jitter. It is all goodness anyway you look at it.
post #673 of 3048
Quote:
Originally Posted by Bigus View Post

For minimum phase phenomenon, which covers low frequency response in small rooms, time and frequency are directly coupled. Improve frequency response, you improve time domain behavior. The only caveat is that improvements at one location may be detrimental in another location.

This is the key. It's impossible for EQ to improve more than one or two very small locations in a room. You can fine tune an EQ and get pretty close to perfection for one square inch. But move the measuring microphone even one or two inches and it all goes out the window. This is why you never see high resolution waterfall data for EQ at more than one location. Either only one place is shown, or several places are shown but only the raw response and they're 1/3 octave averaged. Related, we don't hear the way a single microphone picks up sound waves. If you clamp your head in a vise such that one ear remains in the "corrected" area, the other ear still hears all the ringing.

Not only are other locations not improved, many are made worse. Often much worse. People that sell EQ systems never show you that either. A room is not "minimum phase" in the usual sense because a room has three filter poles, with one each for length, width, and height. All of these poles interact.

Also, understand that when I tested the Audyssey system, it was the main contender and gushed over the most in the popular press. It's not my fault that it doesn't work very well, or as well as newer systems out now. At that time it was wildly expensive! And you had to hire a professional to install and set it up, and then hire them again every time you moved anything in the room like your speakers or sub or couch.

But the real issue is physics. It's impossible for any EQ system, no matter how sophisticated, to improve the response for all locations in a room. Versus bass traps that do exactly that. And bass traps rarely make things worse elsewhere (though it can happen in rare situations and is easily corrected). Further, nulls are usually far more damaging than peaks, at least in small rooms. There are many more threads in this forum complaining about lack of bass than too much bass. Only bass traps can substantially improve nulls. With enough good bass traps, 30 dB nulls are easily reduced to only 10 dB, which is a huge improvement that EQ can never achieve.

--Ethan
post #674 of 3048
Quote:
Originally Posted by Ethan Winer View Post

This is the key. It's impossible for EQ to improve more than one or two very small locations in a room. You can fine tune an EQ and get pretty close to perfection for one square inch.

Too true. I eq for my listening spot (just HT, my 2.0 is careful speaker placement). If other people don't like it they can go out and get their own system.
Quote:
Originally Posted by Ethan Winer View Post

With enough good bass traps, 30 dB nulls are easily reduced to only 10 dB, which is a huge improvement that EQ can never achieve.
--Ethan

Another sage piece of wisdom: You can't EQ your way out of a severe null. You'll run out of amp and speaker travel. EQ is a great tool not to have to use.
post #675 of 3048
Quote:
Originally Posted by Ethan Winer View Post

There are way too many claims there to comment on, but I can tell you a few impressions:
I see a lot of claims made that are not backed up with frequency response and ringing graphs, or data showing that a statistically significant number of listeners prefer what is offered for sale there. Without seeing the in-room response at high resolution for a number of listening positions, it seems more like flowery prose than audio science.
I do agree that bi-amped speakers have the potential to offer the best results. But excellent bi-amped speakers are a commodity, and can be had for a very reasonable cost without needing a "consultant" or design service to tell you what to buy.
--Ethan
I was disheartened by not seeing any DBTs to back up the assertions.
post #676 of 3048
Here is the frequency response at multiple seating positions covered in my article online: http://www.madronadigital.com/Library/BassOptimization.html

LowFrequencyRoomImpact.png

It has 30 db variation from lowest to highest. Here it is now with 4 subwoofers in each corner:

Multiple-Subwoofers.png

Huge amount of variations were taken out together with the null.

We apply eq and delay/level optimizations to the fours subs instead of driving them equally (Sound Filed Management or SFM). Now let's look at seat to seat variations:

Sound-Field-Management-Relative-Levels.png

Here is a more dramatic example of what electronic correction can do from my WSR article. Here is four subs before correction:

i-t4j3C5m-L.gif

Now with SFM:

i-r3FxB9s-L.gif

The remaining variations can easily be smoothed using EQ now and across that wide set of seats. No 2 inches.

Zero acoustic material was used for any of this. Science is our friend here. It says we can cancel room modes with smart placement and addition of subwoofers. It says that we can driving subs independently and using computer optimization get to far better results across large seating area. Importantly there is no risk of making your room too dead from putting a ton of bass absorbers in your room. Don't live in 1970s acoustic world before all of this was discovered. I don't know of any major theater designer who would put a single sub in a room and then try to patch things up with fiberglass. Multiple sub techniques and eq are standard fair now.

Not trying to tell you to not use acoustic products. But don't buy the monocle view that acoustic products is where it is at. Celebrate all the research and science which gives us powerful new tools. You don't believe this stuff works? You can easily verify by turning things on and off electrically. Try doing that with acoustic products!
post #677 of 3048
Quote:
Originally Posted by amirm View Post

The remaining variations can easily be smoothed using EQ now and across that wide set of seats. No 2 inches.
Zero acoustic material was used for any of this. Science is our friend here. It says we can cancel room modes with smart placement and addition of subwoofers. It says that we can driving subs independently and using computer optimization get to far better results across large seating area. Importantly there is no risk of making your room too dead from putting a ton of bass absorbers in your room. Don't live in 1970s acoustic world before all of this was discovered. I don't know of any major theater designer who would put a single sub in a room and then try to patch things up with fiberglass. Multiple sub techniques and eq are standard fair now.
Not trying to tell you to not use acoustic products. But don't buy the monocle view that acoustic products is where it is at. Celebrate all the research and science which gives us powerful new tools. You don't believe this stuff works? You can easily verify by turning things on and off electrically. Try doing that with acoustic products!

Before you start off on another tangent no one is discounting multiple subs aren't an answer. But EQ is minimally doing room correction in this scenario. The real heavy lifting is the four subs. Now if you're happy with two subs or one sub and some simple traps can smooth out the response, well that is the straight and simple path vs more active components.

And you wouldn't even need computer optimization. If the subs have the right controls you could do this with a SPL meter. Done that before.
Edited by Jinjuku - 10/10/12 at 12:47pm
post #678 of 3048
But - its easier, more effective, aesthetically superior, and probably cheaper to use multiple subs than enough bass trapping to even come close to those results above - so what's the upside to using bass traps instead? About the only scenario I can see is if you are using an infinite baffle sub which is large and limited in placement.
post #679 of 3048
Quote:
Originally Posted by kromkamp View Post

But - its easier, more effective, aesthetically superior, and probably cheaper to use multiple subs than enough bass trapping to even come close to those results above - so what's the upside to using bass traps instead? About the only scenario I can see is if you are using an infinite baffle sub which is large and limited in placement.

You simply use the most effective measure that gets the job done. If you can't place 4 subs.
post #680 of 3048
I guess what I'm saying is that, if you are interested in a flat bass frequency response for multiple seats, I can't envision a scenario where you *can* place sufficient bass trapping, but you *cant* place multiple subs. Not from an aesthetics or cost POV. I suppose anything is possible though....
post #681 of 3048
Quote:
Originally Posted by amirm View Post

Importantly there is no risk of making your room too dead from putting a ton of bass absorbers in your room.
Looks like you are not familiar with bass absorbers, a. k. a. bass traps. rolleyes.gif You can put tons of membrane bass traps and not make the room too dead. If you don't know what membrane bass trap is, try Google search.
post #682 of 3048
I think SPL might be the missing link in this debate, and I come at this from practical (subjective) observation only. But I just moved my gear from a well treated room with good bass-trapping using Audyssey XT32 Pro, to an untreated room, with what should be pathetic acoustics, and still using Audyssey.
I'm finding that Audyssey is doing a very good job, the bass is articulate and top end refined with good sound staging. Basically it's not too far off of my treated room, until I crank up the volume pretty loud and then I hear it loose a bit of imaging and separation. In my previous room, there wasn't any degradation regardless of how loud I played the material and I'd venture it easily got to rock concert levels and then some.
The difference to me with Audyssey vs none in both rooms is a no-brainer, if it's doing any damage I'm not hearing it, but what it does right is plainly obvious to me. Toggling between it on and off is quite dramatic. Again just a preference based observation and not a DBT....

So I guess that's a long winded way of saying I don't believe that either product falls into what I consider an "audio conspiracy accessory" to be, in that they both make a difference.
Edited by rnrgagne - 10/10/12 at 4:25pm
post #683 of 3048
In fact with all the talk about jitter, room treatements and correction there's an irony in that the best 2ch set up I've experienced to date was an all digital (including amp) TaCT setup with modest Totem Forrests using TaCT's proprietary RC in a small well treated room. wink.gif
post #684 of 3048
Quote:
Originally Posted by hd_newbie View Post

I was disheartened by not seeing any DBTs to back up the assertions.

Got any solutions to the physical problems related to doing DBTs involving room treatments?

Note that there is only one facility in the world (Harman's) that provides a good means for doing speaker DBTs, and they are far smaller and easier to work with than room treatments.
post #685 of 3048
Quote:
Originally Posted by arnyk View Post

Ever check out a performance space with adjustable acoustics?
I've worked in 2 and visited a third One is a high school auditorium at Kettering High School in Waterford, Michigan. Another is a rectal hall in Varner Hall at Oakland University in Rochester, Michigan. The third is part of the Orchestra Hall Complex in Detroit.
A possible criticism is that the range of adjustment is too limited. The Kettering room is supposed to be adjustable from drama to band to orchestra to choir. It is a little too live when set up to be a room for drama, and little too dead when set to max reverb to make every choir director happy, even with a shell added.
The basic methodologies for implementation don't seem to be impossibly expensive.
One involves the use of motorized sound absorbing drapes and the other involves absorptive panels that slide behind reflective panels. The sound absorbing materials can be backed into pockets faced with reflective materials for nearly 0% effect. When fully extended they might cover 80% or more of the surface that they are part of.

Sorry for the late replay, I'm way behind in the thread. The screening room at Dolby Labs has adjustable acoustics, and has the widest adjustment range I personally have experienced.
These aren't my pix, and a few years old, but when I visited it was pretty much like this. During my visit Ioan explained there were many "settings" for the room which could be for screening films, performance, recording, lectures, etc.
http://www.film-tech.com/warehouse/wareview.php?id=1689&category=1

Adjustable, yes. Practical...well it was Dolby, so unlikely anyone else is doing this to this extent.
post #686 of 3048
Quote:
Originally Posted by arnyk View Post

Got any solutions to the physical problems related to doing DBTs involving room treatments?
Note that there is only one facility in the world (Harman's) that provides a good means for doing speaker DBTs, and they are far smaller and easier to work with than room treatments.
Not really. I was hoping Amir can enlighten us since he has become a huge fan of DBTs lately.
post #687 of 3048
Quote:
I was hoping Amir can enlighten us since he has become a huge fan of DBTs lately.
That's funny. But then so is this:
Quote:
I was hoping Amir can enlighten us
post #688 of 3048
Quote:
Originally Posted by hd_newbie View Post

Not really. I was hoping Amir can enlighten us since he has become a huge fan of DBTs lately.

well he might. He might say something like 'Well, if you all agree that dbt's are the standard in audio as elsewhere, and if you all agree it is mighty hard for us to do it successfully, then it might be prudent to at least consider the results from well respected researchers who have done sterling work in the past and who are capable of conducting properly set up dbt's and have done so with room treatment rpoducts"

If he did, is there any reason you can come up with that explains why those findings should be rejected? I dunno, your own contrary findings from your own dbts, that would be a valid point from which to argue.

Still, this is all conjecture about what he might or might not say.
post #689 of 3048
Quote:
Originally Posted by Jinjuku View Post

That was my first question. How many positions did the mic sit in. You can EQ for a spot. You can throw up absorption and get the entire room.

There is a technique that results in generalized EQ for more than one position. The idea is you take detailed RTA measurements like 1/12 octave or better, then average multiple points. What happens is you get a general trend, a curve that represents an average of all positions measured. This is essentially the THX Cinema technique, though they do it with four measurement mics running into a multiplexer, integrated over time. Another way to do it is with one mic on an isolated mount moved through the space continuously, while running the integration. That's essentially thousands-of-positions average.

The problem with this technique is that the average includes everything, including odd data excursions that are extreme, but highly localized. In essence, the average is a reduction in resolution, a smoothing of data. Proponents of this technique would say that the curve you get this way is sufficient to apply and be an improvement everywhere, while preventing the "tail chasing" of trying to eq a single point, then moving the mic. Opponents would say that the EQ curve you end up with isn't one that's good everywhere, it's one that's good nowhere. Both ideas are right, as it's generally good everywhere, but specifically not perfect anywhere. The question would be, is a generalized correction better than no correction? I'd say in many cases yes. I use the moving mic, time integrating RTA technique for generalized EQ of distributed sound systems for lecture halls, and the like, where the speakers are many and cheap. The trend is actually useful in that case, and it's surprising how good a $20 ceiling speaker can sound once EQ has been applied.

One method to increase the precision of the multi-position average technique is to not use an RTA at all, but use a swept-sine FFT, and pick and choose your time window, or display a waterfall. Then, multiple points can again be averaged. Perhaps petter precision is obtainable that way, but in practice it's much slower.

Another method of increasing the precision of this technique is to not just average the measurements, but rather to reject the extreme excursions that appear on only one measurement, and average the excursions that are approximately in the same general direction. The resulting average represents the trend of the system, and includes only the data that is similar at each point, without being adversely influenced by the one-off anomalies. Thus it could be said that an EQ curve derived from the process would be appropriate to all measurement points, though not a perfect reciprocal curve for any of them. In fact, this is essentially what the "fuzzy clustering" process found in Audyssey does. The theory is that if you improve the precision of the average (by not using a strict average at all), you can produce a more accurate generalized EQ the would follow the trend of many positions.

The entire system leaves plenty for room for the things not correctable by EQ, like deep response nulls, ringing, etc. It can't duplicate what room treatment does, and doesn't try, but then it does do things room treatment doesn't too. So the two concepts can and should work together. The one thing that many of us old-time room-treatment guys have a hard time with is anyone claiming that any of the auto-eq devices are a substitute for room treatment. Sorry, still can't swallow that. But I will say that if you can't treat a room, there is a lot of good that can be done with EQ. And conversely, I've seen Audyssey, in specific, do very little and be nearly inaudible when applied to a room that was treated properly and already performing well.

So, if you're willing to accept a general EQ curve as an improvement, it's possible to have that applied to many locations with a degree of benefit to each one. If that "general curve" is processed in a way that results in higher precision, then it is more appropriately applied in those same positions.
post #690 of 3048
Quote:
Originally Posted by arnyk 
My first point would be that acoustical performance in most listening rooms isn't minimum phase. This is widely known, easy to observe, and generally agreed upon.
Well, when stated so broadly I would agree. That's why I specified low frequencies in small spaces. This is a minimum phase phenomenon.
Quote:
Even when the room is small, you have problems using signal processing to control sound quality at middle and high audio frequencies.
True, if a bit off topic.
Quote:
There is one kind of room that breaks that rule - an anechoic chamber.
Well, it would certainly be free of nulls... but wouldn't it also be free of peaks? I don't think it breaks that rule.
Quote:
Exactly! +1!

Well said!
Thanks! I just thought a little counterbalance to the constant poo-pooing of acoustic treatments that has occurred on avs recently was in order. As much as one person may claim another's financial interest begets his promoting the need for acoustic treatments, so is his own company built around products promising to work magic without needing them. A little underhanded IMO to accuse one company of doing what your own does.
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 › Are audio companies all involved in a huge conspiracy?