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 1647

post #49381 of 70896
Quote:
Originally Posted by J_Palmer_Cass View Post

You won't get many people to agree on what "very high listening levels" means.

We would just need to agree if they are above or below "movie listening levels".
post #49382 of 70896
Quote:
Originally Posted by mogorf View Post

Keith, the practical solution with the same player here is to asign one HDMI input for DVD/BD and another one (coax/opt) for CDs, then adjust intellivolume (and save it) per input channel. It's not a global setting but can be set differently for each input.

Ah - brilliant. Thanks, Feri. I've never needed to use Intellivolume so I had no idea it was that flexible.
post #49383 of 70896
Quote:
Originally Posted by markus767 View Post

We would just need to agree if they are above or below "movie listening levels".

Markus, here are some excerpts from "Wikipedia: Loudness War"

1990s
In the early 1990s, CDs with louder music levels began to surface, and CD levels became more and more likely to bump up to the digital limit[note 1] resulting in recordings where the peaks on an average rock or beat-heavy pop CD hovered near 0 dB[note 2] but only occasionally reached it.[6][not in citation given]
The concept of making music releases "hotter" began to appeal to people within the industry, in part because of how noticeably louder releases had become and also in part because the industry believed that customers preferred louder sounding CDs, even though that notion might not have been true.[7] Engineers, musicians and labels each developed their own ideas of how CDs could be made louder.[citation needed] In 1994, the digital brickwall limiter with look-ahead (to pull down peak levels before they happened) was first mass-produced. While the increase in CD loudness was gradual throughout the 1990s, some opted to push the format to the limit, such as on Oasis's widely popular album (What's the Story) Morning Glory?, which averaged −8 dBFS on many of its tracks[6]a rare occurrence, especially in the year it was released (1995). In 1997, Iggy Pop assisted in the remix and remaster of the 1973 album Raw Power by his former band The Stooges, creating an album that, to this day, is arguably the loudest rock CD ever recorded. It has an average of −4 dBFS in places.[6]


And here's the link: http://en.wikipedia.org/wiki/Loudness_war
post #49384 of 70896
mogorf, thanks but I'm well aware of this. I was a combatant in that war for some time.
post #49385 of 70896
Quote:
Originally Posted by markus767 View Post

We would just need to agree if they are above or below "movie listening levels".


This is where I get lost a bit in that comparison. Are you talking about listening levels used to monitor sound in the music mixing studio, or dBFS levels that exist on the final recording?

The final recording levels on a CD/DVD can be measured, but the reference SPL level used to mix music is an unknown. For that matter, even DVD's are all over the place as far as average level is concerned even though there is a standard. That issue drives me crazy when I record a movie off of cable.
post #49386 of 70896
Quote:
Originally Posted by J_Palmer_Cass View Post

Are you talking about listening levels used to monitor sound in the music mixing studio

Yes, and "music or other program material that is mixed at very high listening levels" is the same thing, no?
post #49387 of 70896
Quote:
Originally Posted by markus767 View Post

Yes, and "music or other program material that is mixed at very high listening levels" is the same thing, no?


I don't know for sure what is meant by "mixed at very high listening levels". That is a very vague statement. What they say and what they mean can be two different things.

Is the front seat area of a live rock concert a "very high listening level", or is recorded music being played back in a home setting (no specific volume level) a "very high listening level"? Very subjective.

I did read the information in the link below (which you posted in), and the way that I read it Chris is talking about average levels on a rock recording. Yes, average levels recorded 10-15 dB higher than movie "reference level" is a high level recording relative to the average level that is used in a movie. Peak levels are the same, but dynamic range is reduced because the average level is louder. That is one (or a few) reason why you must lower the master volume level to playback a rock music CD as compared with your normal movie setting.

Here is a graphic representation of what I am talking about:




Chris mentioned the result of this dynamic range compression. "As you go from 0 (default) offset to 15 dB you are telling Dynamic EQ to compensate less. In other words you have to turn the volume down much more before it starts to make the adjustments."

Chris also noted the use of Intellivolume that was mentioned in a few postings above here. Per Chris "Intellivolume is just a fancy name for "input trim"."


http://ask.audyssey.com/entries/7328...eference-level
LL
post #49388 of 70896
Quote:
Originally Posted by J_Palmer_Cass View Post

I don't know for sure what is meant by "mixed at very high listening levels". That is a very vague statement. What they say and what they mean can be two different things.

Is the front seat area of a live rock concert a "very high listening level", or is recorded music being played back in a home setting (no specific volume level) a "very high listening level"? Very subjective.

I did read the information in the link below, and the way that I read it Chris is talking about average levels on a rock recording. Yes, average levels recorded 15 dB louder is a high level recording relative to the average level that is used in a movie.


http://ask.audyssey.com/entries/7328...eference-level

But the level on a recording is irrelevant for equal-loudness compensation. All that matters is the actual monitoring level during mixing/mastering.
post #49389 of 70896
Quote:
Originally Posted by markus767 View Post

But the level on a recording is irrelevant for equal-loudness compensation. All that matters is the actual monitoring level during mixing/mastering.


That seems to be contrary to what Chris is saying about the way that dynamic EQ works. He is talking about heavily compressed music (15 dB setting) not needing as much dynamic EQ as lightly compressed music (5 dB setting).


Per Chris "It's more complicated than just reference level... It also has to do with how much dynamic range compression is used in creating the content. So, I would just treat the numbers as just labels. The 15 dB offset is for highly compressed (in dynamic range) content that is mixed at high overall levels (e.g. hip hop music)."


Chris is talking about the difference between the average level (be it high or low) and the peak level in a recording (AKA dynamic range). Highly compressed music will pretty much still sound balanced if the master volume is reduced by 10 dB regardless of the original mixing volume level.
post #49390 of 70896
Quote:
Originally Posted by J_Palmer_Cass View Post

That seems to be contrary to what Chris is saying about the way that dynamic EQ works. He is talking about heavily compressed music (15 setting) not needing as much dynamic EQ as lightly compressed music (5 setting).

No, it does not contradict the way Dynamic EQ's Reference Level Offset works. It's not about compression either. The simple issue is that a recording will have a different spectral balance depending on the actual monitoring level when it was created. If the monitoring level is very low then the recording will contain more bass than a recording mixed at higher monitoring levels.

Again, "monitoring level". If THX and Audyssey use the term "listening levels" as a synonym for "monitoring level" then their explanation is backwards.
post #49391 of 70896
Quote:
Originally Posted by markus767 View Post

No, it does not contradict the way Dynamic EQ's Reference Level Offset works. It's not about compression either. The simple issue is that a recording will have a different spectral balance depending on the actual monitoring level when it was created. If the monitoring level is very low then the recording will contain more bass than a recording mixed at higher monitoring levels.

Again, "monitoring level". If THX and Audyssey use the term "listening levels" as a synonym for "monitoring level" then their explanation is backwards.


Music mixing / monitoring level is still undefined, so yes some CDs (music recordings) do have too little and too much bass. That is easy to notice just by listening to music, and dynamic EQ is not going to fix that issue. Same bass problem can be noted based on how flat the speaker system is at the LP, as well as the bandwidth of the subwoofer (at both ends of the chain -mixer and consumer).


Per Chris "It's more complicated than just reference level... It also has to do with how much dynamic range compression is used in creating the content. So, I would just treat the numbers as just labels. The 15 dB offset is for highly compressed (in dynamic range) content that is mixed at high overall levels (e.g. hip hop music)."
post #49392 of 70896
Quote:
Originally Posted by J_Palmer_Cass View Post

Music mixing / monitoring level is still undefined, so yes some CDs (music recordings) do have too little and too much bass. That is easy to notice just by listening to music, and dynamic EQ is not going to fix that issue.

It could fix it with the correct Reference Level Offset but that is not the topic.
post #49393 of 70896
Quote:
Originally Posted by markus767 View Post

...The simple issue is that a recording will have a different spectral balance depending on the actual monitoring level when it was created. If the monitoring level is very low then the recording will contain more bass than a recording mixed at higher monitoring levels.
Again, "monitoring level". If THX and Audyssey use the term "listening levels" as a synonym for "monitoring level" then their explanation is backwards.

Well if somebody's got it backwards I'm not putting any $ on it being both THX and Audyssey. Mostly they seem to get things right.

Shall we have another hundred or so speculative posts back and forth on this topic or instead would someone simply "Ask Audyssey"? Nah, too easy...
post #49394 of 70896
Quote:
Originally Posted by markus767 View Post

It could fix it with the correct Reference Level Offset but that is not the topic.


I doubt it. Too much depends on the room the music was mixed in!






http://seanolive.blogspot.com/2009/1...confusion.html
LL
post #49395 of 70896
Quote:
Originally Posted by SoundofMind View Post

Well if somebody's got it backwards I'm not putting any $ on it being both THX and Audyssey. Mostly they seem to get things right.

Shall we have another hundred or so speculative posts back and forth on this topic or instead would someone simply "Ask Audyssey"? Nah, too easy...

I did and it was linked a couple of times in the last "hundred or so speculative posts": http://ask.audyssey.com/entries/73283
post #49396 of 70896
Quote:
Originally Posted by J_Palmer_Cass View Post

I doubt it!

Then why is it that virtually ALL music recordings have bloated bass with Dynamic EQ engaged? And why does Reference Level Offset correct only in one direction and not both?
All of this has been discussed during the last few days, there's really no need to repeat all of it.
post #49397 of 70896
Quote:
Originally Posted by markus767 View Post

I did and it was linked a couple of times in the last "hundred or so speculative posts": http://ask.audyssey.com/entries/73283

That's a long discussion you linked to. What are you referring to exactly?
I stopped skimming when I got to this relevant xchange:

Q: "I'm confused on one point. I quote:
0 dB (Film Ref): This is the default setting...
15 dB: ... material that is mixed at very high listening levels ...
10 dB: ... content ... that is usually mixed at 10 dB below film reference.
So 10 dB is for content mixed at 10dB *below* film reference, while 15dB is for content that it is mixed at a "very high listening level", presumably *above* film reference. If I'm reading this right, perhaps they should be labeled "-10dB" and "+15dB". Is it possible something got reversed when that guide was typed up? ...
November 13, 2010 06:09 am.

A: Hi CF,
No, it's not a typo. It's more complicated than just reference level... It also has to do with how much dynamic range compression is used in creating the content. So, I would just treat the numbers as just labels. The 15 dB offset is for highly compressed (in dynamic range) content that is mixed at high overall levels (e.g. hip hop music)."
post #49398 of 70896
Quote:
Originally Posted by markus767 View Post

Then why is it that virtually ALL music recordings have bloated bass with Dynamic EQ engaged? And why does Reference Level Offset correct only in one direction and not both?
All of this has been discussed during the last few days, there's really no need to repeat all of it.



Has anyone done any preamp level sweeps of the main channels FR when signal levels change to see what DEQ does to cause this bass bloat? Or do you have to sweep the subwoofer preamp level output? Bass is more than subwoofer frequencies.

Like sweep with an input level of -10, -20, -30, -40 and -50 dB FS with DEQ setting of OFF, 0 dB, 5 dB, 10 dB and 15 dB? That would be 4 charts.

Repeat above sweep with different MV settings that would normally be used. Say 0 dB, -10 dB, and -20 dB?

I don't have DEQ, or I would do the tests myself!
post #49399 of 70896
Quote:
Originally Posted by J_Palmer_Cass View Post

Has anyone done any preamp level sweeps of the main channels FR when signal levels change to see what DEQ does to cause this bass bloat? Or do you have to sweep the subwoofer preamp level output? Bass is more than subwoofer frequencies.

Like sweep with an input level of -10, -20, -30, -40 and -50 dB FS with DEQ setting of OFF, 0 dB, 5 dB, 10 dB and 15 dB? That would be 4 charts.

Repeat above sweep with different MV settings that would normally be used. Say 0 dB, -10 dB, and -20 dB?

I don't have DEQ, or I would do the tests myself!

I posted this about Reference Level Offset before.
I think MV was at -20. 0dB in the graph is level of MultXT32 without Dynamic EQ, then the different lines are offsets to that.
So with Dynamic EQ and RLO = 0, it would boost about 9dB at 20Hz.
Red = RLO 0
Brown/yelllow = RLO 5
Blue = RLO 10
Green = RLO 15.
post #49400 of 70896
Quote:
Originally Posted by SoundofMind View Post

That's a long discussion you linked to. What are you referring to exactly?
I stopped skimming when I got to this relevant xchange:

Q: "I'm confused on one point. I quote:
0 dB (Film Ref): This is the default setting...
15 dB: ... material that is mixed at very high listening levels ...
10 dB: ... content ... that is usually mixed at 10 dB below film reference.
So 10 dB is for content mixed at 10dB *below* film reference, while 15dB is for content that it is mixed at a "very high listening level", presumably *above* film reference. If I'm reading this right, perhaps they should be labeled "-10dB" and "+15dB". Is it possible something got reversed when that guide was typed up? ...
November 13, 2010 06:09 am.

A: Hi CF,
No, it's not a typo. It's more complicated than just reference level... It also has to do with how much dynamic range compression is used in creating the content. So, I would just treat the numbers as just labels. The 15 dB offset is for highly compressed (in dynamic range) content that is mixed at high overall levels (e.g. hip hop music)."

Just search for "Markus" in the linked "Ask Audyssey" discussion.

FilmMixer did confirm what I was saying: "When they mixed it, they have their monitors at a level which is comfortable for them, which is probably around 65-75db SPL 0dbfs if you were to check it.... that is still plenty loud."
post #49401 of 70896
Quote:
Originally Posted by sdurani View Post

Just so you know, ASA is Spatializer-like processing that makes a pair of speakers sound like they're farther apart than they physically are. Similar to the surround-from-2-speakers processing built into some TVs. The closer together the rear speakers are, the more ASA processing is applied. This is why you have to dial in rear speaker spread during initial set-up. Of course, the reason you need ASA to make the rear speakers sound like they're apart is because THX recommends placing them right next to each other.

FWIW, I think that they have come off that recommendation. I got that from John Dahl about a year ago.

Jeff
post #49402 of 70896
Quote:
Originally Posted by markus767 View Post

Exactly, so I guess claiming music is mixed at higher levels is misleading at best.

And once again, it isn't..

Music engineers as a general rule monitor at higher SPL than we do....

How hot they print and what SPL they listen to are two completely separate things..... they print hotter than we do, and I've yet to go into a music mix and not have them listening it at an average level of 90+db.....
post #49403 of 70896
Quote:
Originally Posted by FilmMixer View Post

And once again, it isn't..

Music engineers as a general rule monitor at higher SPL than we do....

How hot they print and what SPL they listen to are two completely separate things..... they print hotter than we do, and I've yet to go into a music mix and not have them listening it at an average level of 90+db.....



Would that be an average of 90 plus dB in the:

SPL C scale?

SPL A scale?

Maximum levels would be even higher than that level based on the crest factor of the content in question!
post #49404 of 70896
Quote:
Originally Posted by markus767 View Post

Just search for "Markus" in the linked "Ask Audyssey" discussion.

FilmMixer did confirm what I was saying: "When they mixed it, they have their monitors at a level which is comfortable for them, which is probably around 65-75db SPL 0dbfs if you were to check it.... that is still plenty loud."

Once again, you are confusing the calibrated SPL of the monitor level to what is coming out of the speakers....

if you are printing at close to full scale, and the monitors are calibrated down even 20db, you are mixing at a much higher average level than films... the only thing turning down the monitors calibrated reference does is reduce dynamic range and the max SPL..

It is completely irrelevant to how loud the mixers are listening.... not sure why you think this is debatable....

As a general rule, music mixers listen/monitor at a higher average SPL than film mixers.... that is why the RLO is a negative number.

There is no contradiction in how it works, or the reasoning behind it.

If you agree that most music engineers would end up with a 0vu/-20dbfs SPL of 65-75 vs. our 85, you need a number to subtract from the AVR calibrated reference, which is "0db" MV= 85db Film Standard Reference.

Hence, you need to subtract 5, 10 or 15 db from the MV 0 reference point to offset the compensation of Dynamic EQ.
post #49405 of 70896
Quote:
Originally Posted by rickardl View Post

I posted this about Reference Level Offset before.
I think MV was at -20. 0dB in the graph is level of MultXT32 without Dynamic EQ, then the different lines are offsets to that.
So with Dynamic EQ and RLO = 0, it would boost about 9dB at 20Hz.
Red = RLO 0
Brown/yelllow = RLO 5
Blue = RLO 10
Green = RLO 15.

Very good, but with an RLO of 15 and a MV near -20 I don't see much boost that could cause bass bloat.

I was thinking more or less about the input level that you used for the sweep. Does a -10 dB FS sweep look like a -40 dB FS sweep with a RLO of -15 (or whatever)? Would a constant pink noise or PPN be better to use in that the time period of a sweep is very short. Just wondering how DEQ behaves with different input signal levels.
post #49406 of 70896
Quote:
Originally Posted by FilmMixer View Post

It is completely irrelevant to how loud the mixers are listening....

Why irrelevant? This is what equal-loudness contours are about. A voice will sound slightly different when monitored at different levels. The mixer will use EQ to compensate and make it sound "right" at the specific listening level. It can't be "right" at all monitoring levels. That's why Dynamic EQ is needed in the first place.
post #49407 of 70896
Quote:
Originally Posted by FilmMixer View Post


if you are printing at close to full scales, and the monitors are down, you are mixing at a much higher average level than films... the only thing turning down the monitors calibrated reference does is reduce dynamic range and the max SPL..

It is completely irrelevant to how loud the mixers are listening.... not sure why you think this is debatable....

As a general rule, music mixers mix at a higher average SPL than film mixers.... that is why the RLO is a negative number.


I have a general question on that. How does one define dynamic range? The way that I use the term is the difference between the average signal level and peak level is the "dynamic range" of the recording. I know that there is more range available in the recording medium, but it is not noticed when the average level is say 12 dB down from the peak level.
post #49408 of 70896
Quote:
Originally Posted by FilmMixer View Post

And once again, it isn't..

Music engineers as a general rule monitor at higher SPL than we do....

How hot they print and what SPL they listen to are two completely separate things..... they print hotter than we do, and I've yet to go into a music mix and not have them listening it at an average level of 90+db.....

So you're saying they monitor hot and record/print even hotter.

Yep, I can concur that they monitor hot alright (especially for pop and hip-hop). Don't know how it compares to a film mixing studio since I've never been on one of those, but I've been in a couple of music mixing studios.

Markus, I think you're getting a little hung up on semantics. "Very high listening levels" is a pretty ambiguous term. You're taking it to mean monitoring levels, but it could very easily mean listening during playback levels.

As FilmMixer and Roger have been trying to point out, because there are currently no music recording studio calibration standards, there is no way to know what their monitoring levels are and more importantly, how they relate to the recorded levels, so why obsess about it?

What matters appears to be that playback on a calibrated system results in much louder volumes at the same MV setting, requiring much less compensation by DEQ, thus the louder the average levels of the recording, the lower the MV will be set and the less compensation needed by DEQ hence, the RLO settings hierarchy being the way they are.



Max
post #49409 of 70896
Quote:
Originally Posted by FilmMixer View Post

If you agree that most music engineers would end up with a 0vu/-20dbfs SPL of 65-75 vs. our 85

In my mind that's LOWER listening levels not higher.

P.S. What Dynamic EQ's Reference Level Offset does is correct - don't know how often I've said this now.
post #49410 of 70896
Quote:
Originally Posted by J_Palmer_Cass View Post

I have a general question on that. How does one define dynamic range? The way that I use the term is the difference between the average signal level and peak level is the "dynamic range" of the recording. I know that there is more range available in the recording medium, but it is not noticed when the average level is say 12 dB down from the peak level.

Traditionally, dynamic range is the difference in db between the loudest and softest signal (not the average and loudest) that can be reproduced by any part of the signal chain, and not used to describe content..

In my experience, we use the term "compressed" to describe content that has such little dynamics in the average vs. peak, and not "reduced dynamic range."
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)