"Official" Audyssey thread Part II - Page 89 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 44205Likes
Reply
 
Thread Tools
post #2641 of 7059 Old 03-22-2017, 03:19 PM
AVS Forum Special Member
 
garygarrison's Avatar
 
Join Date: Aug 2013
Location: The Milky Way
Posts: 1,299
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 867 Post(s)
Liked: 1216
Quote:
Originally Posted by mthomas47 View Post
... I have deliberately tried to avoid anything controversial in that new DEQ section, but as always, comments are welcome. I typically tweak these things a good bit over time, anyway.
Wow, Mike, you got 23 likes for the revised "Guide to Subwoofer Levels, Audyssey Setup, and DynamicEQ" post number 1296 (as of 3:10 pm Wednesday, Pacific Time), including many I have never heard of. It's good to know that so many people are reading along, and know a quality post when they see one!

Last edited by garygarrison; 03-22-2017 at 03:24 PM.
garygarrison is offline  
Sponsored Links
Advertisement
 
post #2642 of 7059 Old 03-22-2017, 03:42 PM
AVS ***** Member
 
mthomas47's Avatar
 
Join Date: Nov 2007
Posts: 8,271
Mentioned: 347 Post(s)
Tagged: 0 Thread(s)
Quoted: 5485 Post(s)
Liked: 10471
Quote:
Originally Posted by garygarrison View Post
Wow, Mike, you got 23 likes for the revised "Guide to Subwoofer Levels, Audyssey Setup, and DynamicEQ" post number 1296 (as of 3:10 pm Wednesday, Pacific Time), including many I have never heard of. It's good to know that so many people are reading along, and know a quality post when they see one!
Thank you very much for the compliment Gary! I don't know about the quality part, but I am pretty sure that there are a lot of people silently reading along, who aren't posting on the thread.

GUIDE TO SUBWOOFER CALIBRATION AND BASS PREFERENCES

* The Guide linked above is a comprehensive guide to Audio & HT systems, including:
Speaker placements & Room treatments; HT calibration & Room EQ; Room gain; Bass
Preferences; Subwoofer Buyer's Guide: Sealed/ported; ID subs; Subwoofer placement.
mthomas47 is online now  
post #2643 of 7059 Old 03-22-2017, 03:53 PM
AVS Forum Special Member
 
garygarrison's Avatar
 
Join Date: Aug 2013
Location: The Milky Way
Posts: 1,299
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 867 Post(s)
Liked: 1216
Quote:
Originally Posted by gurkey View Post
... a momentary dynamic compensation element within DEQ is counter productive, because it cures a "phenomenon" which in reality doesn't exist ... and would falsify the impression of the "real thing" ...
Quote:
Originally Posted by mthomas47 View Post
^^^

Just for the record, I also would eliminate the moment-to-moment aspect of DEQ's operation, as well ... And, I gave my specific reasons for that in my belief that it contributes to more abrupt transitions between loud and soft passages. I simply suggested that if such a feature were to be retained, it would have worked better with a listen-ahead component, instead of trying to make changes in real time, as the listener is actually hearing them. I think that it actually over-complicates the algorithm a bit to try to make individual channel adjustments, and to do that on the fly.
I agree that the moment-to-moment "compensation" DEQ provides is an attempt to cure something that doesn't exist. I've never understood why they included it in the first place. At reference volume the soft passages should have the same proportion of bass and treble they would have in real life, and at lower than reference volume, with DEQ engaged, wouldn't the simple compensation that is tied to the MV setting suffice?

Perhaps the possible over-complication of the algorithm Mike refers to, and asking Audyssey to do too much, etc. is responsible for the loss of clarity and transparency I hear with DEQ on.
mthomas47 likes this.
garygarrison is offline  
Sponsored Links
Advertisement
 
post #2644 of 7059 Old 03-22-2017, 04:04 PM
AVS ***** Member
 
mthomas47's Avatar
 
Join Date: Nov 2007
Posts: 8,271
Mentioned: 347 Post(s)
Tagged: 0 Thread(s)
Quoted: 5485 Post(s)
Liked: 10471
Quote:
Originally Posted by garygarrison View Post
I agree that the moment-to-moment "compensation" DEQ provides is an attempt to cure something that doesn't exist. I've never understood why they included it in the first place. At reference volume the soft passages should have the same proportion of bass and treble they would have in real life, and at lower than reference volume, with DEQ engaged, wouldn't the simple compensation that is tied to the MV setting suffice?

Perhaps the possible over-complication of the algorithm Mike refers to, and asking Audyssey to do too much, etc. is responsible for the loss of clarity and transparency I hear with DEQ on.
I have attributed some of the loss of clarity to the nature of the bass curve that DEQ uses, although I can't really think of a way to confirm that intuition. Another factor, for movies at least, could certainly be the bass boost in the center channel. I tried a variety of ways to circumvent that slight loss of clarity, including raising the crossover, but that was sort of an inescapable component of DEQ for me.

Have you experienced the transition problem between loud and soft passages to which I referred? To be fair, I only noticed that when I experimented without DEQ for a while. Without was just smoother to me, and I could hear the difference more clearly when I subsequently experimented with DEQ again. We all hear slightly different things, in slightly different ways. And, then we use slightly different language to try to describe what we are hearing, so who knows whether we actually hear the same things?
mogorf and garygarrison like this.

GUIDE TO SUBWOOFER CALIBRATION AND BASS PREFERENCES

* The Guide linked above is a comprehensive guide to Audio & HT systems, including:
Speaker placements & Room treatments; HT calibration & Room EQ; Room gain; Bass
Preferences; Subwoofer Buyer's Guide: Sealed/ported; ID subs; Subwoofer placement.
mthomas47 is online now  
post #2645 of 7059 Old 03-22-2017, 04:19 PM
AVS Forum Special Member
 
mogorf's Avatar
 
Join Date: May 2009
Location: Budapest, Hungary
Posts: 5,798
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 1252 Post(s)
Liked: 951
Quote:
Originally Posted by garygarrison View Post
I agree that the moment-to-moment "compensation" DEQ provides is an attempt to cure something that doesn't exist. I've never understood why they included it in the first place. At reference volume the soft passages should have the same proportion of bass and treble they would have in real life, and at lower than reference volume, with DEQ engaged, wouldn't the simple compensation that is tied to the MV setting suffice?

Perhaps the possible over-complication of the algorithm Mike refers to, and asking Audyssey to do too much, etc. is responsible for the loss of clarity and transparency I hear with DEQ on.
I'll try my wings again, so please bear with me.

Any time we turn down MV from 0 dB ref., DEQ will kick-in with a fixed compensation curve, e.g. at -30 dB MV there will be a curve that compensates the incoming signal that is 0 dB (electric) or let's just call it loud. But then in the passage there will be a signal that is playing at -30 dB (electric) and that signal will fall on another steeper curve resulting in a thinner sound (too much bass and treble cut). In order to tackle this existing phenomenon, the second-tier compensation will put it back to the same curve needed at -30dB MV setting. This is what DEQ does moment-by-moment and I've always regarded this as a valid point.

Let's try to leave a bit of time to digest this second feature of DEQ which unfortunately can not be turned off separately. Probably the only way to test this would be to setup a system with another AVR (Yamaha, Pioneer, etc.) that does not have this feature on-board and do a very careful A-B testing. Won't gonna happen soon, eh?
mthomas47 and garygarrison like this.
mogorf is online now  
post #2646 of 7059 Old 03-22-2017, 05:09 PM
AVS ***** Member
 
mthomas47's Avatar
 
Join Date: Nov 2007
Posts: 8,271
Mentioned: 347 Post(s)
Tagged: 0 Thread(s)
Quoted: 5485 Post(s)
Liked: 10471
Quote:
Originally Posted by mogorf View Post
I'll try my wings again, so please bear with me.

Any time we turn down MV from 0 dB ref., DEQ will kick-in with a fixed compensation curve, e.g. at -30 dB MV there will be a curve that compensates the incoming signal that is 0 dB (electric) or let's just call it loud. But then in the passage there will be a signal that is playing at -30 dB (electric) and that signal will fall on another steeper curve resulting in a thinner sound (too much bass and treble cut). In order to tackle this existing phenomenon, the second-tier compensation will put it back to the same curve needed at -30dB MV setting. This is what DEQ does moment-by-moment and I've always regarded this as a valid point.

Let's try to leave a bit of time to digest this second feature of DEQ which unfortunately can not be turned off separately. Probably the only way to test this would be to setup a system with another AVR (Yamaha, Pioneer, etc.) that does not have this feature on-board and do a very careful A-B testing. Won't gonna happen soon, eh?
I'm fairly ambivalent on this one, Feri. I do think that I understand the reasoning behind what they were attempting, and if I apply that reasoning only to 5.1 movies, where different things may be happening in different channels, at the same time, I think I follow it. Of course, whether it actually accomplishes exactly what they intended, in the way they intended it, even if the reasoning behind it is valid, is an entirely separate question.

FWIW, I find discussions of DEQ's application to music, and particularly to two-channel music, much less fruitful. That one is sort of like discussing how many angels can dance on the head of a pin, to me. Although DEQ can certainly be used with any listening material, we are well beyond the theory that was the whole basis for the development of DEQ, at that point. So, we are forced to rely solely on our personal impressions of how well it works, with little recourse to the way that theory matches application. Again, that is because DEQ was specifically designed to compensate for the attenuation of (primarily) low frequencies, that occurs at below Reference volumes, in 5.1 movies.

I hope that makes sense to someone else. I think it is a pretty critical point in discussions of DEQ.

Regards,
Mike
garygarrison likes this.

GUIDE TO SUBWOOFER CALIBRATION AND BASS PREFERENCES

* The Guide linked above is a comprehensive guide to Audio & HT systems, including:
Speaker placements & Room treatments; HT calibration & Room EQ; Room gain; Bass
Preferences; Subwoofer Buyer's Guide: Sealed/ported; ID subs; Subwoofer placement.
mthomas47 is online now  
post #2647 of 7059 Old 03-23-2017, 12:17 AM
AVS Forum Special Member
 
garygarrison's Avatar
 
Join Date: Aug 2013
Location: The Milky Way
Posts: 1,299
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 867 Post(s)
Liked: 1216
Quote:
Originally Posted by mthomas47 View Post

Have you experienced the transition problem between loud and soft passages to which I referred?
I value really fast, clean transient response and I noticed somewhat poorer transient attacks with DEQ ... I guess that could be due to a slight delay in the compensation changing between soft and loud passages. Also, micro dynamics and inner voices of the orchestra sound a little blurred with DEQ on, as opposed to off. I've long despised devices like automatic gain controls, so maybe something similar (but more subtle) is happening with the moment by moment changes in bass and treble compensation imposed by DEQ.
mthomas47 likes this.
garygarrison is offline  
post #2648 of 7059 Old 03-23-2017, 12:56 AM
AVS Forum Special Member
 
gurkey's Avatar
 
Join Date: Aug 2007
Posts: 3,141
Mentioned: 2 Post(s)
Tagged: 0 Thread(s)
Quoted: 289 Post(s)
Liked: 181
@mogorf

My understanding would be, that the DEQ algorithm shouldn't do anything which counteracts a "natural" behavior. It's fine with me, if an adjustable compensations is applied, when the (static) listening level is not at reference level as a mark, everyone is referring to, but in real life dynamics in a performance will change every minute without any momentary compensations added to it.
It just is what it is... And that is, what the original artist had in mind in his arrangement, especially in "classical" music.
mthomas47 and garygarrison like this.
gurkey is offline  
post #2649 of 7059 Old 03-23-2017, 03:16 AM
AVS Forum Special Member
 
David Aiken's Avatar
 
Join Date: Mar 2002
Location: Brisbane, Qld, Australia
Posts: 1,872
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 817 Post(s)
Liked: 1471
Quote:
Originally Posted by mogorf View Post
Hi David,

I think we all have a good understanding on how our ears work with regards to different frequencies at different sound pressure levels. That's all fine in a concert hall, everything sound "as is", no debate!

But once we get home and turn on our amps things start to change. We have something know as "The Volume Control" and that's what makes a huge difference when compared to a live performance.

Let's take a look at how Roger Dressler (in the Audyssey thread part I back in 2012) ) explains the "two-tier" operation sheme of DEQ:



"Let's use the example of a train recording (I guess it could also be a marching band!). The top line in each of these diagrams is the train's spectrum as it passes close by. The lower line is the final loudness as the train travels far away. The overall SPL reduces 40 dB during the train recording.






Blue = Playback at Ref volume. When the train is close, we hear all frequencies equally loud. When it is far, the bass is naturally weaker (just like real life), but it is audible.





Red = MV at -30. When the train is close, the bass is weak. When it is far, the bass is much weaker, and becomes inaudible below 60 Hz.





Green = Fixed Loudness Compensation applied. When the train is close, we hear all frequencies equally loud. When it is far, the bass is weak below 100 Hz, and inaudible below 30 Hz.



Grey = Dynamic EQ applied. When the train is close, we hear all frequencies equally loud. When it is far, the bass is weaker, but it is audible. The arrows show where the lower curve had additional bass boost applied in response to the quiet program level. The amount of compensation needed has changed as the train went from loud to soft. Fixed compensation cannot do that.



Blue/Gray = Before /After with Dynamic EQ"

Let's talk!
I've been taking a look at this, and at Mike's post, again and I wanted to revisit Feri's post, not because of anything he said but because of some things that Roger Dressler says in the post Feri quoted.

For a start, we really can't work with these graphs. What they show us is what level a sound at one frequency has to be in order to sound equally as loud as a sound at 1000 Hz. We're interested in something else entirely. We aren't interested in how much higher the SPL of a sound at 31.5 Hz, or some other frequency, has to be in order to sound equal in volume to a sound at 1000 Hz. Most of the time the sound at frequencies other than 1000 Hz does not sound equally loud to us as sounds at 1000 Hz. What we're interested in is not ensuring that 2 sounds at different frequencies perceived as equal in loudness maintain that equality as Master Volume level changes during playback. What we're interest in is ensuring that the perceived loudness balance between sounds at different frequencies is preserved as Master Volume level changes, and the amount of compensation required to ensure that varies depending on the difference in actual level of the sound at a given frequency to the level of a standard 1000 Hz tone at Reference Level.

Look at Dressler's caption to the first graph which says "When the train is close, we hear all frequencies equally loud. When it is far, the bass is naturally weaker (just like real life), but it is audible." Now that is simply wrong. If you did a frequency analysis of the sound of the train when it was close and the sound at 1000 Hz was 90 phons in level we would most definitely hear all frequencies equally loud. A plot of the level of the sound at each frequency would definitely not follow the equal loudness contour for 90 phons in Dressler's plots and because of that some frequencies would be heard as louder than others. So it is with the vast majority of sounds. There are incredibly few sounds in real life in which each frequency in the sound is perceived at equal in loudness to every other frequency. In fact I doubt such a sound occurs naturally but we could certainly create such a sound digitally. I have no idea what such a sound would sound like but I am willing to bet that it does not sound like a train.

Yet Dressler starts his example with graphs showing equal loudness contours and the indefensible assumption that in the sound of the train when it is near we hear each frequency as having equal loudness and then proceeds to demonstrate what is required to maintain that perception of equal loudness.

Unfortunately if we do not perceive each frequency in the sound as being equal in loudness the amount of compensation required to preserve the balance between the levels at different frequencies which we perceive is going to be different to what is required to preserve the perception of equal loudness as we adjust master volume, and there will be times when sound at a given frequency is not going to be audible and we don't want to make it audible, but if the sound is dropping in level as in the case of the sound of a train receding into the distance, we will want to preserve the point in the decay of the sound where a frequency becomes inaudible, and maintain the perceived shifts in tonal character of the sound as it decays while the train recedes into the distance.

Dressler's example is a simple, unique case, example but it's also a very unrealistic example if we consider the sounds confronting us in real life and on soundtracks, and statements like "When the train is close, we hear all frequencies equally loud. When it is far, the bass is naturally weaker (just like real life), but it is audible." in the captions to his graphs are misleading and don't help us to assess and understand what actually is going on in the special case he illustrates.

Now if you want to start to show what we need with DEQ you could start with 2 different sets of graphs. The first one could be of white noise at different SPLs and show our perception of the level of different frequencies in that sound as we reduce the playback level of the white noise. The second one would be similar but substitute pink noise for white noise because pink noise has a different frequency plot to white noise. If you repeat Dressler's series of 4 graphs, plotting the process of deriving a DEQ profile to preserve the perceived loudness of both noises at differing playback levels you would see that the amount of DEQ required to preserve the perceived character of the white noise will be slightly different to that required to preserve the perceived balance of the pink noise. Then try some real life sounds like, say, frequency plots of the actual sound of a close and of a distant train and repeat the series of graphs and you will find that the DEQ required to preserve the tonal character of the train sound in both cases is going to be different, and the DEQ required for the train sounds is going to be different to that required for white and pink noise, and different also to a noise which follows the equal loudness contours of Dressler's example. The differences are going to be small but they are going to be there, and some frequencies are going to drop into inaudibility which does not occur in Dressler's example.

Comment for Mike in particular: you said it took you ages to write the original version of your DEQ post. I figure it has probably taken me as long to write this and I can't guarantee that in a day's time I won't want to tear this up and start again, or just delete this post and give up on starting over. The requirements for calculating the amount of compensation required on a moment to moment basis got more and more complex to me as I drafted and redrafted. I tried to summarise them, I rewrote that part several times, and I finally left it on the cutting room floor. In the end I've simply focussed on showing up the issues in Dressler's presentation as I see them. He is right if we start from a sound in which each frequency appears equally loud but, as I said, that's a special case and it isn't going to occur in real life and the result for sounds in real life is going to be slightly different for each sound. You said you thought the calculations involved for the moment to moment calculation were pushing the capabilities of the software to deliver in real time. I came to the conclusion that they probably exceed the capacity of the processors in our AVRs. It could be done but I think what is being done is much simpler and what is being done is the generation of a compensation profile based solely on the master volume setting and some kind of adjustment determined from a very limited frequency/level analysis, perhaps in terms of octave bands or samples taken at a small number of frequencies. The computational task if you analysed levels at each frequency between 20 Hz and 20 kHz in real time is probably beyond the processors in our AVRs.
mthomas47 and garygarrison like this.
David Aiken is offline  
post #2650 of 7059 Old 03-23-2017, 08:02 AM
AVS ***** Member
 
mthomas47's Avatar
 
Join Date: Nov 2007
Posts: 8,271
Mentioned: 347 Post(s)
Tagged: 0 Thread(s)
Quoted: 5485 Post(s)
Liked: 10471
Important clarification for DEQ discussions

In my opinion, there have been several excellent posts, regarding DEQ, on the current page. There is Gary's post on microdynamics in orchestral compositions; an equally good post by Gurkey on natural sound, especially for classical music; and David's really excellent analysis of Roger's train example. All three posts share two things in common. They are all valid, in my opinion. And, they are all talking about listening activities that have nothing at all to do with DEQ's original design intent. This is a very important point!

These thread discussions of DEQ seem to always follow a common pattern. Someone mentions a dissatisfaction with some aspect of DEQ, and then someone (typically an engineer) like Chris, when he was active on the thread, or Feri in recent years, patiently explains the theory behind DEQ. That is a perfectly logical approach. If we understand the design, we will be better able to compare the design to its practical application. So, Design >/< = Actual (or Practical) Application. But, there is actually a component missing in that construction. And, that is the original design intent. If we are going to discuss the relationship between theory and practical application, then we need an intermediate piece to add to our logical construction. It should be: Design >/< = Intended Application >/< = Actual Application.

Perhaps an analogy will help to make this essential point a little clearer. Several carpenters and electricians are at a job site on a lunch break, and they are discussing the ideal flat head screwdriver and how it should be designed. They all agree that it should have a long shaft for leverage and torque, and a fairly broad head, with a thinly tapering edge. It should be extremely strong, without being brittle. Then a painter walks up and agrees that a long shaft, and a broad flat head are valuable, because after using the screw driver to pry open a paint can, he often uses it to stir the paint. And, the long shaft and broad head help with that. The others agree, and then go back to talking about the nature of the tool's design, for its intended function.

The effective use of a flat head screwdriver, as a paint stirrer, is a happy coincidence that has nothing whatsoever to do with its original design intent. DEQ is just a tool, more complex perhaps than a flat head screwdriver, but still designed for a specific purpose. That purpose was not to make two-channel orchestral music sound more natural and convincing, or to accurately replicate the sound of a train passing. It was designed to maintain the acoustic balance of the inherently artificial sounds of a 5.1 movie soundtrack (with 6 separate channels, mixed according to the tastes and whims of a director and film mixer) at below Reference listening levels. If it also happens to work effectively for something else, that is a happy coincidence which has nothing whatsoever to do with its original design intent. Just like the screwdriver.

This is not a matter of being pedantic. It is a fundamental issue of the logical disconnect that occurs for all of us when we start to discuss the theory of a surround boost, or moment-by-moment changes within individual channels, or the bass loss that may occur, due to the Equal Loudness Contours, and then try to apply those same concepts to something besides 5.1 movies. Those are all specific artifacts of the completely artificial sounds of a 5.1 movie, played in a home theater with multiple surround channels. And, they have nothing at all to do with the "natural" sound of a train, or of a symphony orchestra, or of a two-channel recording of a symphony orchestra, even if it is expanded into a 5-channel track, via a surround processor such as PLII.

So, we can talk about DEQ in any context we like: TV game shows, sporting events, two-channel music, whatever. And we will all have personal opinions as to how well the flat head screwdriver works as a pry bar, or as a paint stirrer. Or how well DEQ works for orchestral music, or to capture the natural sound of a train. But, it is only in the context of 5.1 movies, and Reference volumes, that we will be able to achieve any real alignment between design theory and practical application. In my opinion, that is why the logical disconnects in these discussions, going back quite a few years, keep occurring.

Again, that is because the application to 5.1 movies was the specific design intent behind the creation of DEQ in the first place. And, everything from its addition of 2.2db of bass boost for every -5 MV below Reference, to its moment-by-moment adjustments of individual channels, to its surround boost, were predicated upon that specific design intent. We may like those design attributes, or not, for their intended purpose, or within any other context that we choose. But, if we want our discussions to all be on the same page (as they have not been in the past), we really need to make the connection between the original design intent, and its actual, or practical, application for 5.1 movies. Anything beyond that will be a happy coincidence, in my opinion.

Regards,
Mike

GUIDE TO SUBWOOFER CALIBRATION AND BASS PREFERENCES

* The Guide linked above is a comprehensive guide to Audio & HT systems, including:
Speaker placements & Room treatments; HT calibration & Room EQ; Room gain; Bass
Preferences; Subwoofer Buyer's Guide: Sealed/ported; ID subs; Subwoofer placement.

Last edited by mthomas47; 03-23-2017 at 10:54 AM.
mthomas47 is online now  
post #2651 of 7059 Old 03-23-2017, 02:06 PM
AVS Forum Special Member
 
mogorf's Avatar
 
Join Date: May 2009
Location: Budapest, Hungary
Posts: 5,798
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 1252 Post(s)
Liked: 951
Quote:
Originally Posted by mthomas47 View Post
In my opinion, there have been several excellent posts, regarding DEQ, on the current page. There is Gary's post on microdynamics in orchestral compositions; an equally good post by Gurkey on natural sound, especially for classical music; and David's really excellent analysis of Roger's train example. All three posts share two things in common. They are all valid, in my opinion. And, they are all talking about listening activities that have nothing at all to do with DEQ's original design intent. This is a very important point!

These thread discussions of DEQ seem to always follow a common pattern. Someone mentions a dissatisfaction with some aspect of DEQ, and then someone (typically an engineer) like Chris, when he was active on the thread, or Feri in recent years, patiently explains the theory behind DEQ. That is a perfectly logical approach. If we understand the design, we will be better able to compare the design to its practical application. So, Design >/< = Actual (or Practical) Application. But, there is actually a component missing in that construction. And, that is the original design intent. If we are going to discuss the relationship between theory and practical application, then we need an intermediate piece to add to our logical construction. It should be: Design >/< = Intended Application >/< = Actual Application.

Regards,
Mike
Mike, I think this is one of the best and coolest post I've ever read on this forum. All my appreciations and credits to out to you. It is confidence building for the individuals and team spirit building for this little community we have here on the Audyssey thread making it out to be a kinda "family place" where particular people congregate. Kudos to Ya.

As a wrap up of my humble contributions to the DEQ topic here is something I'd like to share with everyone hopefully still interested in some technical stuff as a back up of all the talk we've been doing here in the last couple of day. (Especially for those who like graphs! )

Back in 2013 we had here another long discussion on "How DEQ works" and member "urwi" showed these two graphs he measured on the pre-out of his AVR with regards to the two-tier operation of DEQ in real life:

"Dynamic adjustments based on signal level are necessary if the equal loudness curves behave non-monotonically. Audyssey obviously believes this is the case although the ISO curves do behave monotonically. Here are Dynamic EQ equal loudness curves measured from an AVR preamp output.



The following graph shows the dynamic behavior of compensation curves for three different input levels with MV at -30dB."



I sincerely hope by now we have a much better understanding on the pure technical side of what DEQ does in the signal chain that will surely help everyone decide on their own subjective preferences when next time the HT system is turned on. Good luck guys, with the true hope I didn't rob anyone's precious time with my continuous efforts to shed some light on something that was also new for me when I started to dig into all the technical details of DEQ. I just tried to pass on some of my savvy to all wishing to be upgraded with more than enough details.

Finally I'd like to offer something as a bonus to our discussions which is a sound file of a train coming and going that represents what Roger Dressler tried to demo with his graphs. Please play it through your system. Let's have some fun.

P.s.: Mike et al, watch out, this is a 5.1 multi-channel recording!!

https://www.dropbox.com/s/cqmlezqyrm...Train.wav?dl=0

Last edited by mogorf; 03-23-2017 at 02:14 PM.
mogorf is online now  
post #2652 of 7059 Old 03-23-2017, 02:17 PM
AVS Forum Special Member
 
David Aiken's Avatar
 
Join Date: Mar 2002
Location: Brisbane, Qld, Australia
Posts: 1,872
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 817 Post(s)
Liked: 1471
Quote:
Originally Posted by mthomas47 View Post
In my opinion, there have been several excellent posts, regarding DEQ, on the current page…
A bit of a disclaimer on my part plus a bit of relevant personal history and a comment on music, real sounds, and their relevance to DEQ:

I have 2 systems, one with Audyssey and one without. The one with Audyssey gets used for movie/TV viewing only, the other system is my music system so DEQ never goes near it. Audyssey never goes near it.

My comments on DEQ relate to my experience of it with movies and TV audio sources because I haven't used it for music or other types or recordings, just soundtracks. Trains occasionally figure in those activities, actually more often than they do in my activities away from the screen. I started thinking and commenting on it again because someone, either Mike or Feri, made a comment about how it worked which differed from my understanding of how it worked and the more I thought about the way the comment suggested it worked, which apparently actually is the way it works so my original beliefs were wrong, the more I started having problems with the concept behind it.

Coincidentally with that I came across a podcast about some aspects of room correction and started experimenting with some different settings for Audyssey including turning DEQ off instead of having it on which had been my usual practice. I like some of the changes I'm hearing with DEQ off and that fed into my concerns about how DEQ worked.

Mike's point about the purpose of DEQ being to correct something in the playback of 5.1 soundtracks is very relevant, and yes, soundtracks are different to music and real sounds, but soundtracks do include music and real sounds and when they do the soundtrack usually can't deviate too far from reality in it's treatment of those things if they are going to work successfully in the soundtrack. If a movie has a train coming closer and then receding, then the sound of the train better sound reasonably like a train's sound and the way in which the sound gets louder as the train gets closer and then gets softer as the train gets further away had all better remind us of the way a real train sounds under those circumstances or the illusion gets broken. You can't substitute the sound of a car and have the sound get softer as the train gets closer and louder as it moves away, for example, and expect the soundtrack to work well with the movie. Music and real sounds get used in soundtracks and we expect them to behave the way music and real sounds behave in our normal experience. They can be modified to accentuate or diminish something for a specific purpose but if that is to be done the movie itself has to provide a context for the change in the behaviour of the sound or it simply doesn't work for us.

So while agreeing with Mike's comments about the purpose that DEQ was intended to serve and his statement that it was not developed for use with music or real sounds, I do think it has to be able of working with music and real sounds in a way which does not prevent music and real sounds from sounding the way we expect them to sound and behave when they occur in soundtracks and the soundtrack requires that of those sounds.

That is an entirely different thing to expecting DEQ to work "seamlessly" with music or other recordings which are not mastered in the same way as a soundtrack and which do not have a known Reference Level which can be used as a data point for calculation of the corrections.
David Aiken is offline  
post #2653 of 7059 Old 03-23-2017, 02:36 PM
The Ultimate chest punch!
 
Adamg (Ret-Navy)'s Avatar
 
Join Date: Jul 2009
Location: SW Florida
Posts: 1,725
Mentioned: 19 Post(s)
Tagged: 0 Thread(s)
Quoted: 947 Post(s)
Liked: 2087
Mike, David, Feri and et al,

Count me as one of the Silent readers. Very informative stuff, and served in chewable bite size portions to boot. I may not have a complete understanding of this, but I certainly know more than I did before following this recent DEQ conversation/Dissertation/Oration.

My humble gratitude to each of you for graciously sharing your knowledge and time with all of us. What DEQ does is amazing. I am personally a Fan. I use it on a daily basis. In my meager mind, DEQ allows me to experience the sound stage as it was intended even at reduced MV. I think it is Brilliant work and am thankful for its existence.

Knowing how it works and what it does specifically just makes it that much more Magical to me. Keep up the splendid work Gentlemen. I am certain there are many more like me about. Reading, absorbing, pondering and reading some more. Yet remaining silent as it would be hard for many of us to hold a candle to the current Subject Matter Experts.

I was going to say "Grey Beards" but we know where that will take us......

AdamG, ...out!
The government cannot give to anybody,
anything that the government does not first,
take from somebody else.
Adamg (Ret-Navy) is offline  
post #2654 of 7059 Old 03-23-2017, 03:20 PM
AVS Forum Special Member
 
David Aiken's Avatar
 
Join Date: Mar 2002
Location: Brisbane, Qld, Australia
Posts: 1,872
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 817 Post(s)
Liked: 1471
Quote:
Originally Posted by mogorf View Post
Mike, I think this is one of the best and coolest post I've ever read on this forum. All my appreciations and credits to out to you. It is confidence building for the individuals and team spirit building for this little community we have here on the Audyssey thread making it out to be a kinda "family place" where particular people congregate. Kudos to Ya.

As a wrap up of my humble contributions to the DEQ topic here is something I'd like to share with everyone hopefully still interested in some technical stuff as a back up of all the talk we've been doing here in the last couple of day. (Especially for those who like graphs! )

Back in 2013 we had here another long discussion on "How DEQ works" and member "urwi" showed these two graphs he measured on the pre-out of his AVR with regards to the two-tier operation of DEQ in real life:

"Dynamic adjustments based on signal level are necessary if the equal loudness curves behave non-monotonically. Audyssey obviously believes this is the case although the ISO curves do behave monotonically. Here are Dynamic EQ equal loudness curves measured from an AVR preamp output.


Finally I'd like to offer something as a bonus to our discussions which is a sound file of a train coming and going that represents what Roger Dressler tried to demo with his graphs. Please play it through your system. Let's have some fun.

P.s.: Mike et al, watch out, this is a 5.1 multi-channel recording!!

https://www.dropbox.com/s/cqmlezqyrm...Train.wav?dl=0
Feri,

First, a +1 on your first paragraph.

I got a lot out of both your post with the Dressler quote and this one. That doesn't necessarily mean that I agreed with everything in them, I obviously had problems with Dressler, but I did get a lot out of them because they both have made me think about what's going on in ways I would not have done without them, and that has also made me question my views and understanding which have shifted a bit too. We don't have to agree with something to find it useful and sometimes the things we don't agree with can be very useful to us.

I'm not certain that I understand urwi's comment that Audyssey believes that the equal loudness curves behave non-monotonically while the ISO curves do behave monotonically. I think I know what is being said but the only way I can think of testing in order to derive an equal loudness curve would actually be a monotonic test. Our hearing may not behave like that but I can't think of a way to test whether our perception of loudness of sounds containing multiple frequencies is driven by the SPL of the total sound or whether we can distinguish the level of the different frequencies within the sound and our perception of the loudness of each frequency within the sound depends on the level of that frequency.

The problem I see with that hypothesis is that without being able to demonstrate that it is correct, and without being able to derive non-monotonic equal loudness curves, anything used as a basis for determining the amount of DEQ to apply is going to be a guess based on someone's unverified view of how non-monotonic equal loudness contours work. Actually the problem is worse than that. If the non-monotonic theory is correct, then our perception of the loudness of different frequencies within a complex sound may actually depend on the relationships between the loudness of all of the frequencies within the complex sound and since those relationships are going to be different for every different sound we would be faced with an essentially infinite number of different non-monotonic curves so utilising a non-monotonic model as the basis for DEQ could prove problematic. It would probably mean picking the result from one particular complex sound as the basis for the calculation and the accuracy of the calculation then depends on how representative of all complex sounds your chosen complex sound was.

I guess my reaction to the idea of using a non-monotonic model would be:

1- I'd rather stick with an established and verified monotonic model than an unverified non-monotonic model. Even if the non-monotonic model is verified, it's going to be a hell of a lot more complex than a monotonic model and using it is going to require a hell of a lot more computational power for a moment to moment calculation and computational power is the thing that is likely to be in short supply in our AVRs. Both Mike and I have expressed doubts about the computational capability of AVRs for this kind of task and there are a lot of practical advantages to simplicity when it comes to implementation, even if the more complex model yields more accurate results. When it comes to whether or not using a superior model works better than using an inferior but still reasonably good model, the result often depends on which model is best implemented. A superior quality implemenation of the inferior model may in practice yield better results than an average quality implementation of the superior model. Cut corners in the implementation and you can easily fail to achieve the theoretical advantages promised by your model. Most AVRs are not brilliant examples of no holds barred, highest quality implementation of everything. This is a practical point which acknowledges the advantages of basing things on superior theory and data but also acknowledges that lots of practical and cost benefit issues can get in the way of both utilising and getting the most advantage from the superior theory and data when it comes to including it in a product.

2- I'd rather use perceived loudness contours rather than equal loudness contours. By that I mean a contour that is derived not from playing a reference tone at a fixed level and having the subject adjust the level of a tone at a different frequency until it sounds equally loud to them as the reference tone but a contour derived by playing the target frequency at a fixed level and having the subject adjust the level of the reference tone until they think that is equally loud as the target reference. The curves we would get from that would have an inverse contour to the equal loudness curves and would show how loud we think a sound at a given frequency is rather than how loud a sound at a given frequency needs to be to equal the loudness of the sound at the reference frequency. I don't know if someone has ever done the work to derive such curves but it would be interesting to see them if it has been done, and I wonder whether they would be an exact inverse of the equal loudness curves. I think not.

Apart from that I really appreciated seeing the graphs of the actual behaviour of DEQ. Do you happen to know what the test signal was for those graphs? Was it a frequency sweep or something like a white noise source?

Thanks for your input and keep it coming, and that goes for everyone in this discussion as well.
David Aiken is offline  
post #2655 of 7059 Old 03-23-2017, 03:24 PM
AVS Forum Special Member
 
David Aiken's Avatar
 
Join Date: Mar 2002
Location: Brisbane, Qld, Australia
Posts: 1,872
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 817 Post(s)
Liked: 1471
Quote:
Originally Posted by Adamg (Ret-Navy) View Post

I was going to say "Grey Beards" but we know where that will take us......
Keep saying it and I may decide to grow a beard back to prove you right. Of course if Samson lost his strength when Delilah cut his hair, I may lose my intellect if I grow a beard. There is a risk involved here

:-)

I think I can safely speak for everyone and say we genuinely appreciate your appreciation.
mogorf and Adamg (Ret-Navy) like this.
David Aiken is offline  
post #2656 of 7059 Old 03-23-2017, 03:39 PM
AVS Forum Special Member
 
mogorf's Avatar
 
Join Date: May 2009
Location: Budapest, Hungary
Posts: 5,798
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 1252 Post(s)
Liked: 951
Quote:
Originally Posted by David Aiken View Post
Keep saying it and I may decide to grow a beard back to prove you right. Of course if Samson lost his strength when Delilah cut his hair, I may lose my intellect if I grow a beard. There is a risk involved here

:-)

I think I can safely speak for everyone and say we genuinely appreciate your appreciation.
And I always love to see silent readers posting. Take care!
Adamg (Ret-Navy) likes this.
mogorf is online now  
post #2657 of 7059 Old 03-23-2017, 03:40 PM
AVS Forum Special Member
 
David Aiken's Avatar
 
Join Date: Mar 2002
Location: Brisbane, Qld, Australia
Posts: 1,872
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 817 Post(s)
Liked: 1471
Feri,

Addendum to my previous post:

I just remembered something from the podcast I mentioned earlier, and I think that a non-monotonic theory of loudness perception is almost certainly true. That podcast included a comment that you could have 2 sounds which measured with exactly equal SPLs but if the first of those sounds had a flat frequency response and the second had elevated bass and a depressed midrange relative, the second would be perceived as being less loud than the first because it was lower in level in the midrange where our ears are more sensitive.

So yes, the non-monotonic theory does make sense. I still have no idea how you would ever come up with a set of equal loudness, or perceived loudness curves for that model.
David Aiken is offline  
post #2658 of 7059 Old 03-23-2017, 03:41 PM
AVS ***** Member
 
mthomas47's Avatar
 
Join Date: Nov 2007
Posts: 8,271
Mentioned: 347 Post(s)
Tagged: 0 Thread(s)
Quoted: 5485 Post(s)
Liked: 10471
^^^^^

Since there are several posts coming all at once, I hope no one will mind if I sort of bundle (or bungle ) my response. First, thanks everyone, for your nice comments. And, Feri, thank you especially! That is high praise.

There is more that I might have tried to say in my last post, but I wanted to keep it simple for clarity and emphasis. But, David has raised some good points about realistic sounds, and music in movies, and this is a great time to try to address them in the context of DEQ. (And, Adam, I am really glad that it is working as intended for you.)

Very few sounds in movies are at all realistic. Think about voices for a moment. Do our voices really sound like that when we are outdoors, for instance? Or are our voices thin sounding, and lacking in timbre, when we are outside? Gary, with his experience in film recording will be able to address this issue better than I can, but I will give a few examples. Are footsteps really as loud as they sound in a movie? Or doors closing? For anyone who has ever been in a fight, or studied martial arts, or even watched MMA, does the sound of a fist striking really make a sharp "Crack!" sound, or is more of a meaty thud? And, do kicks really make any special sound at all, as they always do in movies?

As a former shooter, and one with a military background (there Adam, I said it) I can assure everyone that gunshots, particularly outdoors, don't sound at all like that. The sound is very sharp and quick, not drawn out for dramatic effect, as it is in movies. And, explosions don't sound anything like the way they do in movies. They aren't nearly as prolonged, and they are much, much louder. Burst your eardrums loud, at close range. None of that is intended as a critique. Movies are drama, and sounds are created for dramatic effect, and for a simulation of reality, not for reality itself. And, we willingly suspend disbelief, as we do for the story, and for the visual effects, as part of the entertainment process.

We also become acclimated to "movie sounds" over time and separate them in our minds a bit from real sounds. For instance, anyone who has watched fight scenes, from Western movies made in the 1930's and early 40's, knows that the visual and audio effects of those scenes were much more realistic than the ones that we have grown up seeing. They actually fought just as most people really would. The problem is, that on the screen, they didn't look or sound realistic. The punches happened too fast visually, and there was little sound when they landed (which they often did, in the early movies) or when they appeared to land. They just didn't have any dramatic appeal.

John Wayne and Yakima Canut are credited with revolutionizing movie fight scenes. Long, slow punches, with lots of windup. Loud "Cracks!" when the punches landed. That's just one example, from many that we could think of. But, after a while, we got used to movie and TV sounds, as opposed to real ones and incorporated them into our mental database of how things are "supposed" to sound in a movie. David made a similar point in his analysis of a real train, versus a movie train, which can sound any way the film mixer and director want it to.

But, what about music? Surely, if DEQ were designed to work for music in a 5.1 movie, it would have to work in the same way for a two-channel recording, wouldn't it? Well, yes and no. Others can probably think of other examples, but this is one of my favorites to illustrate the point. Anyone who has seen "Meet Joe Black" will probably remember the climactic scene of the birthday party at the end of the movie. There is a hauntingly beautiful song playing, and there are fireworks going off overhead. Many fireworks, which continue for quite a while. The director and film mixer (probably in collaboration with the composer) do a tremendous job of weaving multiple sounds together, from different channels, at different volumes. At times, the music swells, and the fireworks recede, while two of the characters dance or kiss. (One male/one female.) At other times the music is in the background, and there is dialogue, and sounds from the party. At other times, the fireworks dominate the soundtrack, but the music is always playing, throughout all of that. It's a brilliant job of scoring, in my opinion.

But, if you were to listen to the main theme from the soundtrack, on a standalone basis, you wouldn't hear any of the background party sounds, or dialogue, or fireworks. And, there would be no particular need for DEQ, because there would be no special acoustic balance to maintain, and especially not from individual channels. There wouldn't be a need for a bass boost, so that the fireworks would sound realistic, or for a surround boost, and there wouldn't be a Reference level to compensate for in any way.

In other words, movie music is just part of the overall dramatic presentation in a movie, and is mixed especially with that in mind. If DEQ has any validity at all, it is to maintain dynamic equilibrium for 5.1 movie soundtracks (including the on-screen music) at below Reference volumes. Take the the 5.1 movie, out of the music equation, and the situation changes dramatically.

Nothing I have said reflects my personal opinion on whether or not DEQ actually works exactly as intended for 5.1 movies. It just continues the theme of trying to keep us on the same page, as we discuss how well it works, in the context of its actual design intent.

Reflects,
Mike

GUIDE TO SUBWOOFER CALIBRATION AND BASS PREFERENCES

* The Guide linked above is a comprehensive guide to Audio & HT systems, including:
Speaker placements & Room treatments; HT calibration & Room EQ; Room gain; Bass
Preferences; Subwoofer Buyer's Guide: Sealed/ported; ID subs; Subwoofer placement.
mthomas47 is online now  
post #2659 of 7059 Old 03-23-2017, 04:15 PM
AVS Forum Special Member
 
mogorf's Avatar
 
Join Date: May 2009
Location: Budapest, Hungary
Posts: 5,798
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 1252 Post(s)
Liked: 951
Quote:
Originally Posted by David Aiken View Post
Apart from that I really appreciated seeing the graphs of the actual behaviour of DEQ. Do you happen to know what the test signal was for those graphs? Was it a frequency sweep or something like a white noise source?
Member "urwi" was using REW (Room EQ Wizard) to measure his AVR. REW is available from Hometheatershack.com. Free to download. REW has its own frequency sweeps available on-board for testing purposes.
mogorf is online now  
post #2660 of 7059 Old 03-23-2017, 04:17 PM
AVS Forum Special Member
 
David Aiken's Avatar
 
Join Date: Mar 2002
Location: Brisbane, Qld, Australia
Posts: 1,872
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 817 Post(s)
Liked: 1471
[quote=garygarrison;51731953]
Quote:
Originally Posted by mogorf View Post

I'm trying to digest this second feature, and I'm not in a hurry, but, sooner or later I'll probably need some more digestive enzymes.
I forget who said "If you think you're in possession of all of the facts, you haven't been fully informed." I'm currently mainlining on digestive enzymes. The facts keep coming and my digestive enzymes keep falling behind.
mogorf and Adamg (Ret-Navy) like this.
David Aiken is offline  
post #2661 of 7059 Old 03-23-2017, 04:22 PM
AVS Forum Special Member
 
David Aiken's Avatar
 
Join Date: Mar 2002
Location: Brisbane, Qld, Australia
Posts: 1,872
Mentioned: 7 Post(s)
Tagged: 0 Thread(s)
Quoted: 817 Post(s)
Liked: 1471
Quote:
Originally Posted by mogorf View Post
Member "urwi" was using REW (Room EQ Wizard) to measure his AVR. REW is available from Hometheatershack.com. Free to download. REW has its own frequency sweeps available on-board for testing purposes.
Thanks, Feri. Last time I looked REW was Windows only and didn't have a Mac version but that's not an issue. Knowing that it was a frequency sweep is what I was after.

It would be interesting to see if there was any variation in the results if white noise and pink noise were used as test material rather than frequency sweeps but I don't think that information would be of any use to us.
mogorf likes this.
David Aiken is offline  
post #2662 of 7059 Old 03-23-2017, 04:54 PM
AVS ***** Member
 
mthomas47's Avatar
 
Join Date: Nov 2007
Posts: 8,271
Mentioned: 347 Post(s)
Tagged: 0 Thread(s)
Quoted: 5485 Post(s)
Liked: 10471
Quote:
Originally Posted by mogorf View Post
Member "urwi" was using REW (Room EQ Wizard) to measure his AVR. REW is available from Hometheatershack.com. Free to download. REW has its own frequency sweeps available on-board for testing purposes.
Feri,

That was a very cool train recording, which perfectly illustrates the Doppler effect. I certainly wouldn't want to live too close to train tracks. But, as a kid, I remember lying in bed thinking how comforting the sound of a train, and train whistle, in the distance was. I guess I still think so.

Regards,
Mike
Adamg (Ret-Navy) and Matt2026 like this.

GUIDE TO SUBWOOFER CALIBRATION AND BASS PREFERENCES

* The Guide linked above is a comprehensive guide to Audio & HT systems, including:
Speaker placements & Room treatments; HT calibration & Room EQ; Room gain; Bass
Preferences; Subwoofer Buyer's Guide: Sealed/ported; ID subs; Subwoofer placement.
mthomas47 is online now  
post #2663 of 7059 Old 03-24-2017, 02:08 AM
AVS Forum Special Member
 
garygarrison's Avatar
 
Join Date: Aug 2013
Location: The Milky Way
Posts: 1,299
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 867 Post(s)
Liked: 1216
First, I'd like to say to everyone that the "likes" are not working on the forum -- or, perhaps on my computer -- at the moment, so, even though I liked all of the last several posts, you wouldn't be able to tell by looking in the "like" rectangle.

Quote:
Originally Posted by mthomas47 View Post
... That purpose was not to make two-channel orchestral music sound more natural and convincing, or to accurately replicate the sound of a train passing. It was designed to maintain the acoustic balance of the inherently artificial sounds of a 5.1 movie soundtrack (with 6 separate channels, mixed according to the tastes and whims of a director and film mixer) at below Reference listening levels ...

... the completely artificial sounds of a 5.1 movie, played in a home theater with multiple surround channels. And, they have nothing at all to do with the "natural" sound of a train, or of a symphony orchestra, or of a two-channel recording of a symphony orchestra, even if it is expanded into a 5-channel track, via a surround processor such as PLII.
Quote:
Originally Posted by David Aiken View Post
... Mike's point about the purpose of DEQ being to correct something in the playback of 5.1 soundtracks is very relevant, and yes, soundtracks are different to music and real sounds, but soundtracks do include music and real sounds and when they do the soundtrack usually can't deviate too far from reality in it's treatment of those things if they are going to work successfully in the soundtrack ....
Some 5.1 movies on Blu-ray have lengthy orchestral passages. Examples of these 5.1 (or greater) movies are 2001: a Space Odyssey, Amadeus, Fantasia 2000, etc. Other movies may have overtures/preludes or lengthy opening or ending credits with fairly long stretches of uninterrupted music. With all of these, DEQ, on my system, in my room, seems to slightly blunt transients and blur the microdynamics and inner voices of the orchestra I so love. I have recently tried a lower MV setting than my usual, for audiences with sensitive ears, so I have tried DEQ once again. It laid an egg every time. Leaving DEQ off, boosting the subwoofer level and using the bass tone control do not cause these problems, at least for me. Even well into a movie, when the music may be turned up and down to make way for dialog or effects, I'd hate to see the texture and detail of a Newman (any of them), a Herrmann, a Rozsa, or a Horner obscured by DEQ. Tonight we ran The 33. It must have been Horner's last score. The music was often interrupted by sounds of the mine shaft collapsing (not a spoiler, in this case), raised voices, etc., yet the musical detail came through with bass boost, sub boost, and no DEQ.

The same would apply to SACDs, usually 5.1 or 5.0.
mthomas47 likes this.

Last edited by garygarrison; 03-24-2017 at 02:12 AM.
garygarrison is offline  
post #2664 of 7059 Old 03-24-2017, 04:43 AM
Member
 
anothermib's Avatar
 
Join Date: Mar 2017
Posts: 190
Mentioned: 3 Post(s)
Tagged: 0 Thread(s)
Quoted: 171 Post(s)
Liked: 58
Quote:
Originally Posted by mthomas47 View Post

...

Again, that is because the application to 5.1 movies was the specific design intent behind the creation of DEQ in the first place. And, everything from its addition of 2.2db of bass boost for every -5 MV below Reference, to its moment-by-moment adjustments of individual channels, to its surround boost, were predicated upon that specific design intent. We may like those design attributes, or not, for their intended purpose, or within any other context that we choose. But, if we want our discussions to all be on the same page (as they have not been in the past), we really need to make the connection between the original design intent, and its actual, or practical, application for 5.1 movies. Anything beyond that will be a happy coincidence, in my opinion.

Regards,
Mike

All, this is a highly interesting discussion on DEQ. Very valid points being made from different perspectives and I am learning a lot on the inner workings of that tool. I would like to add my thoughts on the discrepancy between original intent and today's perception of DEQ.

As far as I can tell, from a high level the current situation may be described as follows:

- Given the software and DSP capabilities of modern AVRs the user is (I think rightfully) expecting some sort of automatic loudness compensation in addition to the ability of manually adjusting the tone settings or SW trim for each level of MV.

- In many AVRs DEQ is the only loudness software option provided. It is turned on by default with RLO=0 and does a couple of things (that may or may not be desired by the user) in addition to the basic loudness compensation. While this may be very good for 5.1 movies - many people feel that this sounds really poor for music and less than optimal for a lot of TV content (I am wondering if that is the underlying issue of people claiming e.g. that Denons generally have unnatural, thickened bass).

- The result can be improved substantially by adjusting RLO on a per content basis and tweaking channel trims. Some are happy with the result and some really are not.

While it probably would be unfair to blame Audyssey or the DEQ software for this situation I believe it is understandable why people feel that overall this is really not ideal. DEQ may have delivered on the original objectives. However, I wonder if it would be worth formulating a new set of objectives for such a software that will lead to more people being happy with the results. Ideally this should help both - the novice that wants a reasonable sound just after completing the Audyssey measurement and the pro that is happy to understand and tweak the parameters of what his system does until he has what he wants.

I am wondering if most of the basic capabilities aren't in DEQ already. It may just require first taking a step back, looking at it from a slightly different angle and (re)defining the usage scenarios and objectives.

What do you think?
mthomas47 likes this.
anothermib is offline  
post #2665 of 7059 Old 03-24-2017, 04:49 AM
Member
 
Join Date: Nov 2016
Posts: 71
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 28 Post(s)
Liked: 19
Does anyone know roughly how many db per db below reference DEQ applies to low frequencies? Would be nice to know so I can experiment with a manual bass boost...
CommanderROR is offline  
post #2666 of 7059 Old 03-24-2017, 05:07 AM
Member
 
Join Date: Nov 2016
Posts: 71
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 28 Post(s)
Liked: 19
Ignore my question. I found the reply above. 2.2db per 5db. That's pretty heavy handed. Means my -30db standard listening really drives up the bass...
CommanderROR is offline  
post #2667 of 7059 Old 03-24-2017, 05:25 AM
AVS Forum Special Member
 
mogorf's Avatar
 
Join Date: May 2009
Location: Budapest, Hungary
Posts: 5,798
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 1252 Post(s)
Liked: 951
Quote:
Originally Posted by CommanderROR View Post
Ignore my question. I found the reply above. 2.2db per 5db. That's pretty heavy handed. Means my -30db standard listening really drives up the bass...
Actually it varies with frequency and playback level, so your question can not be answered with 1 figure only.
mogorf is online now  
post #2668 of 7059 Old 03-24-2017, 06:36 AM
Member
 
Join Date: Nov 2016
Posts: 71
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 28 Post(s)
Liked: 19
Ok. I just tried boosting bass with Tone Control to it's max. of 6db, but that doesn't even come close to the effect of Dyn.EQ. I'm so angry that Denon/Audyssey won't allow us to turn off the stupid volume boost to the surround channels with Dyn EQ! I mostly like what it does to the sound in general, but the Surrounds and Top Rear get blown conpletely out of proportion for me. Even if I lower them by 6db (which then can cause problems on sources like PS4 where I have to deactivate Dyn EQ anyway and also unbalances things at volume closer to Reference) it still seems like they are boosted and the whole surround feeling collapses towards the back. This has been a complaint for ages, but Denon and Audyssey just don't see it as an issue...
CommanderROR is offline  
post #2669 of 7059 Old 03-24-2017, 07:15 AM
AVS ***** Member
 
mthomas47's Avatar
 
Join Date: Nov 2007
Posts: 8,271
Mentioned: 347 Post(s)
Tagged: 0 Thread(s)
Quoted: 5485 Post(s)
Liked: 10471
Quote:
Originally Posted by CommanderROR View Post
Ok. I just tried boosting bass with Tone Control to it's max. of 6db, but that doesn't even come close to the effect of Dyn.EQ. I'm so angry that Denon/Audyssey won't allow us to turn off the stupid volume boost to the surround channels with Dyn EQ! I mostly like what it does to the sound in general, but the Surrounds and Top Rear get blown conpletely out of proportion for me. Even if I lower them by 6db (which then can cause problems on sources like PS4 where I have to deactivate Dyn EQ anyway and also unbalances things at volume closer to Reference) it still seems like they are boosted and the whole surround feeling collapses towards the back. This has been a complaint for ages, but Denon and Audyssey just don't see it as an issue...
Hi,

I understand your frustration, but FWIW, Audyssey is a proprietary software, leased to Denon/Marantz, so any changes in the basic Audyssey/DEQ software would have to come from them. If you have a sub in your system, that is what you should be boosting first. The tone controls only affect your front speakers, and have a limited, although helpful, effect. But, if you have a sub, and are bass-managing your speakers with a crossover, using the tone control will only boost your front speakers in the mid to upper-bass range. All the low bass boost will still have to come from the sub, which isn't affected by the tone controls.

The subwoofer guide, linked in my signature, will help you to understand all of this better, so that you can get the bass you want, either with, or without DEQ.

Regards,
Mike

GUIDE TO SUBWOOFER CALIBRATION AND BASS PREFERENCES

* The Guide linked above is a comprehensive guide to Audio & HT systems, including:
Speaker placements & Room treatments; HT calibration & Room EQ; Room gain; Bass
Preferences; Subwoofer Buyer's Guide: Sealed/ported; ID subs; Subwoofer placement.
mthomas47 is online now  
post #2670 of 7059 Old 03-24-2017, 07:26 AM
Member
 
Join Date: Nov 2016
Posts: 71
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 28 Post(s)
Liked: 19
Thanks. I'll play around more with it over time. I contacted Audyssey a while ago and asked about configuring Dyn EQ via the upcoming Audyssey app, and they said that such a feature would have to be implemented by Denon. I guess it's the usual game of "Sombody elses Problem"
CommanderROR is offline  
Sponsored Links
Advertisement
 
Reply Receivers, Amps, and Processors

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off