or Connect
AVS › AVS Forum › Audio › Receivers, Amps, and Processors › "Official" Audyssey thread (FAQ in post #51779)
New Posts  All Forums:Forum Nav:

"Official" Audyssey thread (FAQ in post #51779) - Page 2030

post #60871 of 70896
Quote:
Originally Posted by Wryker View Post

Quote:
Originally Posted by AustinJerry View Post

This is not intended to be a flame, but any time you have a question, simply enter it into your browser search.  For example, the search for "What is reference level" produced many hits, including this appropriate one from the Audyssey site:

https://audyssey.zendesk.com/entries/73283-Dynamic-EQ-and-Reference-Level
From what I've read the past few weeks there's been many a-discussion regarding Audyssey's leveling and the f-3 etc etc so I wanted to pose my question here to see if everyone holds the same definition of "reference level". Thanks!

ps - if listening at "0db" on the AVR is reference I'll have a hard time ever watching it at that level since with my new AVR and speakers I have a hard time getting higher than -5db since it is very very loud (and I'm one who likes it loud)!

 

Most people find reference level too loud in a domestic environment. I listen loud too, often at -5dB or so, and I know I couldn’t really take any more for very long - and this is in a system/room that has substantial headroom so it is clean all the way up to 0dB.  It is interesting that AVS-er 'FilmMixer' - a professional film mixer doesn't listen at reference level at home as he finds it "too loud".

post #60872 of 70896
I've never understood full the reason for that. What is it about the home vs the theater that makes reference too loud? Room size? More reflected sound? Being closer to the speakers?
post #60873 of 70896
Quote:
Originally Posted by JHAz View Post

I am pretty sure that, for my ears anyway, what a musician's intent was when playing a specific passage does not change how overall volume changes affect my perception of the tonal balance. I'd assume Audyssey assumes that every dynamic change in every movie or piece of music is quite intentional. But it's not trying to make a piano played pianissimo sound like it was played forte. It's trying to make the piano played pianissimo then played back 10 dB below that sound, from a spectral balance perspective, like it would sound if I had my volume turned up so it was actually pianissimo.
 

 

I understand what you are saying - but I still don't really 'get it' in one sense. If I go a live gig and the pianist takes a nice long solo and at one point in that solo he plays so pianissimo that I, from my seat some way back from the stage, can no longer hear those softest notes at all - what does DEQ do?  Does it boost them so that they remain audible (if so that is distortion) or does it let them fall below the threshold of hearing (if so DEQ doesn't seem to be working as advertised)?
post #60874 of 70896
Quote:
Originally Posted by batpig View Post

I've never understood full the reason for that. What is it about the home vs the theater that makes reference too loud? Room size? More reflected sound? Being closer to the speakers?

Subjectively, I always find my own system to be much louder than the movie theater long before I get to 0 on the master volume control. I'm wondering if many movie theaters actually play at reference levels. I suspect many of them don't.

Anyone ever take an SPL meter to the movies with them? biggrin.gif
post #60875 of 70896
Quote:
Originally Posted by JHAz View Post

And it is literally impossible within any particular channel to turn up (or change the EQ of) the quiet piano while leaving the loud trumpet alone, if you're DEQ. In any given channel, the mix of stuff that is there is just the mix of stuff that is there, and at any given microsecond, the soundwave being played back through tat channel is only at one location, regardless of how many instruments are represented by the wave. Luckily for DEQ, we hear on a relatively slow, RMS basis so it can look at the totality of the sound moment by moment and make corrections versus the levels that would be present at the (assumed/RLO determined) "reference" level.

 

You are the musician and I respect your opinion greatly for that reason. But am I understanding you correctly in this paragraph?  I agree that DEQ cannot, at the same time, turn up (or change the EQ of) the quiet piano while at the same time leaving the loud trumpet alone. So this in itself means that Feri's notion of 'two-tier' equalisation is wrong, doesn’t it?  Forgive me if I have misunderstood you - I have house guests (who have just gone to bed) and I have consumed the best part of half a bottle of Lagavulin after dinner ;)

post #60876 of 70896
Quote:
Originally Posted by urwi View Post

Quote:
Originally Posted by kbarnes701 View Post

This is the part that nobody has ever explained satisfactorily. How is DEQ supposed to know which parts are deliberately played more softly by the musician?  I can see that maybe it is possible, but never seen an explanation of it, even from the most ardent DEQ enthusiasts. 

Dynamic EQ doesn't need "to know which parts are deliberately played more softly by the musician". It is simply monitoring the overall input level and modifies the applied loudness correction curve accordingly.

 

And when the same passage has a very soft piano and a very loud horn, at the same time (as in JHAZ's example), then what does it do?

post #60877 of 70896
Quote:
Originally Posted by pepar View Post

Quote:
Originally Posted by sdurani View Post

Has to be, at least at those frequencies that need to be boosted in order to keep those sounds from falling into inaudibility.

Compression is needed to compensate for the "expansion" caused by our non-linear hearing. So the net effect is zero. smile.gif

 

In an old-style 'loudness control' that is what happens (at a given level). In a dynamic version of that (dynamic equalisation) the amount of EQ will be changed 'on the fly'. But DEQ, so those who claim to know how it works say, does more than that and actually 'listens ahead' to what loud or soft sounds are coming along and then alters the EQ accordingly. I am not sure this is what actually happens, and those who claim to know are unable or unwilling to demonstrate it. However it works, it seems to be good for movies and next to useless for music. Maybe that is all anyone really needs to know (per your airplane wing example).

post #60878 of 70896
Quote:
Originally Posted by mogorf View Post

Quote:
Originally Posted by kbarnes701 View Post

Too technical for the FAQ - the explanation that is there already is enough IMO - so long as people have the basic understanding of what it does, I don't think they need to know how it works (and indeed AFAIK nobody really knows how it works anyway as it is proprietary information). All they really need to know is a) leave it turned on for movies and b) leave it turned off for music.

I see,...And batpig was talking about graphs for those who wish to learn! tongue.gif

Yes, but unfortunately, while Roger's graphs are interesting in what they are trying to explain, they are not attempting to explain how Audyssey's DEQ works. Did you think that is what they were showing?

post #60879 of 70896
Quote:
Originally Posted by batpig View Post

I've never understood full the reason for that. What is it about the home vs the theater that makes reference too loud? Room size? More reflected sound? Being closer to the speakers?

 

I have always assumed it is those sort of things - nearfield listening, the way small rooms impinge differently on the sounds in them (hence the reason Pro has different curves based on room size), far more chance of reflections bouncing the sound around and making it 'appear' to be louder (well it IS louder but you know what I mean - it seems louder than it actually is) and so on. I have noticed since I heavily treated my room that I listen at -5dB typically these days as opposed to the -9dB I typically listened at in the untreated room. I know that I have removed energy from the room with the treatments and that is a partial explanation of that, but -5dB 'sounds less loud' than it used to as well. 

post #60880 of 70896
Quote:
Originally Posted by kbarnes701 View Post

In an old-style 'loudness control' that is what happens (at a given level). In a dynamic version of that (dynamic equalisation) the amount of EQ will be changed 'on the fly'. But DEQ, so those who claim to know how it works say, does more than that and actually 'listens ahead' to what loud or soft sounds are coming along and then alters the EQ accordingly. I am not sure this is what actually happens, and those who claim to know are unable or unwilling to demonstrate it. However it works, it seems to be good for movies and next to useless for music. Maybe that is all anyone really needs to know (per your airplane wing example).

Post 30325 and post 30329 by Chris K. are deep down in this thread, but worth to read. smile.gif

Meantime Keith, the 'two-tier' equalisation is not Feri's notion, but Audyssey's!!wink.gif
post #60881 of 70896
Quote:
Originally Posted by mogorf View Post

Quote:
Originally Posted by kbarnes701 View Post

In an old-style 'loudness control' that is what happens (at a given level). In a dynamic version of that (dynamic equalisation) the amount of EQ will be changed 'on the fly'. But DEQ, so those who claim to know how it works say, does more than that and actually 'listens ahead' to what loud or soft sounds are coming along and then alters the EQ accordingly. I am not sure this is what actually happens, and those who claim to know are unable or unwilling to demonstrate it. However it works, it seems to be good for movies and next to useless for music. Maybe that is all anyone really needs to know (per your airplane wing example).

Post 30325 and post 30329 by Chris K. are deep down in this thread, but worth to read. smile.gif
 

 

 

I have said this before, Feri, and I will say it again - you are a marketing man's dream - and I say that as someone who has worked his entire life in marketing's evil twin - advertising :)

 

Giving me the extract from the 'brochure' (or in this case, from Chris K's typing fingers) doesn't help much in getting to understand how something really works.

 

I am intrigued that you have been privy to what is proprietary, and therefore secret, information. But as it seems you have - if you would answer the various questions posed in this thread recently, and posed long ago by your good friend Markus, but not actually really answered by your good friend, Chris, I would be hugely obliged.

 

 

Quote:

Meantime Keith, the 'two-tier' equalisation is not Feri's notion, but Audyssey's!!wink.gif
 

 

I didn't realise there was a difference.

 

PS. I thought you said you were "finished" with this and had "nothing to add"?  Ah yes - here it is: post #60808 


Edited by kbarnes701 - 3/19/13 at 4:59pm
post #60882 of 70896
Quote:
Originally Posted by pepar View Post

Compression is needed to...
As long as you understand it is compression.
post #60883 of 70896
Quote:
Originally Posted by sdurani View Post

As long as you understand it is compression.

Does characterizing it as compression make it more or less desirable? If neither, then why would it be necessary to understand that it is this or it is that?
post #60884 of 70896
Quote:
Originally Posted by kbarnes701 View Post

Yes, but unfortunately, while Roger's graphs are interesting in what they are trying to explain, they are not attempting to explain how Audyssey's DEQ works. Did you think that is what they were showing?
I thought so, from a "black box" perspective, anyway.
Quote:
Originally Posted by kbarnes701 View Post

And when the same passage has a very soft piano and a very loud horn, at the same time (as in JHAZ's example), then what does it do?
DEQ, like Dynamic Volume and Dolby Volume, use models of human hearing to inform the algorithm which sounds dominate the perception. If the horn is loud and tends to mask the piano, then the process will adapt based on the piano. When the horn ceases, the process will adapt to the piano. It of course takes a little time to do that, but since these processes are frequency selective, they can do a lot of changing below the radar of the casual listener. Listen carefully and you will hear the adaptation going on.
post #60885 of 70896
Quote:
Originally Posted by kbarnes701 View Post

However it works, it seems to be good for movies and next to useless for music.
How does it work for music in movies? Good or useless? If good, how come its able to work fine for music in movies?
post #60886 of 70896

Music in movies is mixed and recorded by movie sound techs and, imho, sounds nothing like a live (or well-recorded) orchestra.  It can sound spectacular, of course, and probably works as well with DEQ as does the rest of the soundtrack.

post #60887 of 70896
Hi,
Typically when I run Audyssey (EQ XT32) it sets most of my crossovers to 40Hz. After Audyssey does it's thing I change them to 60Hz on my mains and 80-100Hz on my surrounds. My question is that if I'm changing the crossover point after Audyssey runs, does that mean that Audyssey did not calibrate my sub in the 40-100Hz range. In other words I'm thinking that Audyssey calibrated my speakers for that range so when I give that range back to the sub does that screw up anything?
post #60888 of 70896
Your thinking is wrong smile.gif. Audyssey creates filters for the complete operating range of the speakers. It has no idea what the crossovers are and has nothing to do with the bass management. That's why we always try to disabuse people of the misconception that Audyssey sets the crossovers. It doesn't. All MultEQ does is measure and create speaker EQ filters, then passes the measurements to to the AVR which sets the crossovers.
post #60889 of 70896
Quote:
Your thinking is wrong

You sound like my wife smile.gif

Got it, thanks for the explanation.
post #60890 of 70896
Ouch man tongue.gif
post #60891 of 70896
Quote:
Originally Posted by pepar View Post

Does characterizing it as compression make it more or less desirable? If neither, then why would it be necessary to understand that it is this or it is that?
Because not everyone is willing to admit it involves compression (see batpig's post I was replying to).
post #60892 of 70896
Quote:
Originally Posted by mogorf View Post

Actually it was me who introduced the term "two-tier", but that's not so important.

Otherwise, I'm finished here, nothing to add to the subject. smile.gif

Yes, I borrowed that terminology from Feri. It fits well.

Harrison
post #60893 of 70896
Quote:
Originally Posted by kbarnes701 View Post

And when the same passage has a very soft piano and a very loud horn, at the same time (as in JHAZ's example), then what does it do?

Dynamic EQ knows nothing about single instruments and their loudness. All it sees is the overall signal level. When that level drops Dynamic EQ modifies the applied loudness equalization accordingly.
post #60894 of 70896
Quote:
Originally Posted by pepar View Post

Quote:
Originally Posted by sdurani View Post

As long as you understand it is compression.

Does characterizing it as compression make it more or less desirable? If neither, then why would it be necessary to understand that it is this or it is that?

 

I think compression is generally undesirable as it is another form of distortion. Sometimes it might be a necessary evil, or even the lesser of two evils, but it's still distortion.

 

For a long time people argued (me included IIRC) that Dynamic EQ, Audyssey-style, did not involve compression, but I think now that generally people realise that it probably does.

post #60895 of 70896
Quote:
Originally Posted by Roger Dressler View Post

Quote:
Originally Posted by kbarnes701 View Post

Yes, but unfortunately, while Roger's graphs are interesting in what they are trying to explain, they are not attempting to explain how Audyssey's DEQ works. Did you think that is what they were showing?
I thought so, from a "black box" perspective, anyway.

 

 

Well, your thoughts on it trump everyone else's thoughts on it, Roger! :)  I saw it more as explaining the difference between an 'old style' 'static' EQ (like a loudness control) and a more modern 'dynamic' form of EQ, but didn't realise that it was your implication that it represented the way DEQ works (or might work I probably should say, as, AIUI, nobody outside Audyssey really knows, despite some of the recent bluster in this thread). 

 

 

Quote:

Quote:
Originally Posted by kbarnes701 View Post

And when the same passage has a very soft piano and a very loud horn, at the same time (as in JHAZ's example), then what does it do?
DEQ, like Dynamic Volume and Dolby Volume, use models of human hearing to inform the algorithm which sounds dominate the perception. If the horn is loud and tends to mask the piano, then the process will adapt based on the piano. When the horn ceases, the process will adapt to the piano. It of course takes a little time to do that, but since these processes are frequency selective, they can do a lot of changing below the radar of the casual listener. Listen carefully and you will hear the adaptation going on.

 

 

Thanks for that. The simplest explanations always come from those whose understanding is deepest. I assume that the Dolby variants of Audyssey's DEQ work somewhat differently (as does THX LP) but the objective is the same?  Would that be a fair characterisation? Or do they in fact work in the same way regardless?  AIUI, Audyssey used real-life film mixers as their 'model of human hearing'. Others may have used Fletcher-Munson I guess, or some similar model. How important is that?

post #60896 of 70896
Quote:
Originally Posted by enthusiast8 View Post

Quote:
Originally Posted by kbarnes701 View Post

However it works, it seems to be good for movies and next to useless for music.
How does it work for music in movies? Good or useless? If good, how come its able to work fine for music in movies?

I suspect it is because we listen less critically to music in movies (because the on-screen action takes over a substantial part of our conscious (and maybe even subconscious) mind. When listening just to music we have no such distractions - I, for example, often listen to music with my eyes closed - that wouldn't be practical when watching a movie ;) Although there are some movies where it might be considered an advantage ;)

 

Have you ever watched the credits at the end of a movie and seen the music credits and thought to yourself "don't remember hearing that song..."This happens frequently to me (the first time I see the movie). I assume that the action or the story or some other visual element of the movie so overtook my mind that I didn't even consciously become aware of the music playing behind it. Probably on a subconscious level the music added to the creative content of the scene though. 

 

I do leave DEQ permanently ON in my system, which is used more or less exclusively for movies.

post #60897 of 70896
Quote:
Originally Posted by hclarkx View Post

Quote:
Originally Posted by mogorf View Post

Actually it was me who introduced the term "two-tier", but that's not so important.

Otherwise, I'm finished here, nothing to add to the subject. smile.gif

Yes, I borrowed that terminology from Feri. It fits well.

Harrison

 

I apologise for conflating the two of you :)

post #60898 of 70896
Quote:
Originally Posted by urwi View Post

Quote:
Originally Posted by kbarnes701 View Post

And when the same passage has a very soft piano and a very loud horn, at the same time (as in JHAZ's example), then what does it do?

Dynamic EQ knows nothing about single instruments and their loudness. All it sees is the overall signal level. When that level drops Dynamic EQ modifies the applied loudness equalization accordingly.

 

That's not how Roger characterises it in his post above. He says that the process adapts to the loud or soft instrument. Who's right? 

 

The main problem I personally have with DEQ is that the more I (think) I understand how it works, the less I actually like it ;)  Or like the idea of it perhaps is a better way of saying it. At my usual listening levels it doesn’t really do very much at all TBH, so I just leave it on anyway. I have experimented with turning it on and off and listening to the differences, but they are slight in my room at -5dB TBH. Of course, for those who listen at -20dB etc, it could well be an entirely different story.  The best way to listen to anything, if circumstances permit, is the way it was intended to be heard by the creators (movies closer to Reference than not, music at 'live' levels where possible). IMO of course, and fully accepting that not everyone has domestic circumstances that make this possible.


Edited by kbarnes701 - 3/20/13 at 3:04am
post #60899 of 70896
Quote:
Originally Posted by sdurani View Post

Has to be, at least at those frequencies that need to be boosted in order to keep those sounds from falling into inaudibility.

Yes, Dynamic EQ might be implemented as a multiband compressor.
Edited by urwi - 3/20/13 at 3:47am
post #60900 of 70896
Quote:
Originally Posted by Roger Dressler View Post


DEQ, like Dynamic Volume and Dolby Volume, use models of human hearing to inform the algorithm which sounds dominate the perception. If the horn is loud and tends to mask the piano, then the process will adapt based on the piano. When the horn ceases, the process will adapt to the piano. It of course takes a little time to do that, but since these processes are frequency selective, they can do a lot of changing below the radar of the casual listener. Listen carefully and you will hear the adaptation going on.

I'd guess that a LOT of work goes into crafting it so that the "adaptation" is NOT heard. I remember pumping in the early NR schemes ...

Jeff
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Receivers, Amps, and Processors
AVS › AVS Forum › Audio › Receivers, Amps, and Processors › "Official" Audyssey thread (FAQ in post #51779)