Loudness Wars - Reducing (decoded) "Quad" effect from Stereo?

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

kfbkfb

2K Club - QQ Super Nova
QQ Supporter
Since 2002/2003
Joined
Feb 20, 2003
Messages
2,126
Location
Midwest USA
I have some remastered stereo popular music albums, I haven't tried the original and then the remaster thru my DPL surround sound system, I'm wondering if the volume compression reduces the (decoded - DynaQuad, QS etc.) "Quad" effect (FM radio audio processing does distort the phase difference between L and R)?


Kirk Bayne
 
I have some remastered stereo popular music albums, I haven't tried the original and then the remaster thru my DPL surround sound system, I'm wondering if the volume compression reduces the (decoded - DynaQuad, QS etc.) "Quad" effect (FM radio audio processing does distort the phase difference between L and R)?


Kirk Bayne

Kirk, this is an interesting question. I don't know the answer really, but it would not surprise me if it did have a detrimental effect. I have a remastered redbook CD of America and the SACD of the same album. The more carefully mastered SACD definitely exhibits more ambience than the redbook version when played through either the Surround Master or Logic7. I don't know if more compression on the redbook version is responsible for this difference.

I do know that ambience extraction systems will take whatever "crap" is in a recording and emphasize it in the surround/rear channels...surface noise, clicks and pops, etc. (For whatever reason, tape hiss does not bother me even if emphasized further in the rear channels.)

I also know that devices such as the DBX 3bx dynamic range expander suppress ambience extraction. I stopped using the 3bx when I bought a Lexicon CP-1 back in 1989. The 7.1 surround field from recordings processed through the CP-1 was not as impressive when the signal was first passed through the 3bx.

I also wonder about the impact on "Quad" effect of excessive de-noising processes.
 
I haven't played around with the upmix plugins. I thought they worked with phase tricks. ie isolating stereo vs mono content from a stereo pair. And then adding in targeting by frequency band. Multiband MS<->stereo kind of stuff if you will. (You can do this stuff 'manually' too.)

I don't think any of these work by dynamics profile? Or do they? Target different level content in the mix and separate/isolate based on that? This would get into actual DSP tricks vs just passive phase tricks. I thought I saw a "multiband" limiter that did the "multiband" as volume bands across the dynamic range. Never tried it...

Anyway...

If the upmix plugin was targeting dynamics, a volume war style limited master would be ruined for the process. If it's just the usual phase tricks, the volume war damage will still be what it is but the process should still work. There would probably be some extra artifacts but not complete failure.

And bla bla speculate...
Smash something down and turn it up 12 db with an L2 and try it! What happens?
 
I haven't played around with the upmix plugins. I thought they worked with phase tricks. ie isolating stereo vs mono content from a stereo pair. And then adding in targeting by frequency band. Multiband MS<->stereo kind of stuff if you will. (You can do this stuff 'manually' too.)

I don't think any of these work by dynamics profile? Or do they? Target different level content in the mix and separate/isolate based on that? This would get into actual DSP tricks vs just passive phase tricks. I thought I saw a "multiband" limiter that did the "multiband" as volume bands across the dynamic range. Never tried it...

Anyway...

If the upmix plugin was targeting dynamics, a volume war style limited master would be ruined for the process. If it's just the usual phase tricks, the volume war damage will still be what it is but the process should still work. There would probably be some extra artifacts but not complete failure.

And bla bla speculate...
Smash something down and turn it up 12 db with an L2 and try it! What happens?

The Pan/Expand function (I'm talking expand here) in AA 3 is not multi-band & it doesn't work by simple dynamics. It is a variable M-S set up and it works by doing zero enhancement towards Side when there is no level difference & max towards Side (with in user selected limits) as the level difference increases. I like this approach has it does enhancement for stereo to surround with out affecting the center front material like bass or vocals.

It is a pleasant side effect that transient dynamic range is increased when using this. The shot below shoes a Blue Man Group track enhanced at 250 Expansion and over all normalized to -.5dB:

1620144275506.png


It's easy to see that even tho the peaks are pretty high at -.5dB the average over all level is decreased while dynamic range over all is increased. I'm not suggesting that this be used as a dynamic range control, just that is good for doing S2S & it also does help the DR.

I'm sure Kirk's suspicion is right that heavy brick walling compression does hurt stereo imaging.
 
The Pan/Expand function (I'm talking expand here) in AA 3 is not multi-band & it doesn't work by simple dynamics. It is a variable M-S set up and it works by doing zero enhancement towards Side when there is no level difference & max towards Side (with in user selected limits) as the level difference increases.

Yeah, purely passive phase relationship tricks like this shouldn't be affected.

Limiting would need to get into distortion before it started to 'break' that. And of course, the volume war extremes are quite distorted!

I'd suggest that MS<->stereo based tricks are simple and screwing with dynamics are not, but that's another discussion! :)

You could try pummeling something yourself and then do an A/B before/after and get a sense of what you're missing when it comes up.
 
I was thinking about Matrix decoding only (Matrix decoders use the phase difference between L and R to help route the sounds to LF/RF and LB/RB).

https://www.thebdr.net/a-history-of-audio-processing-part-7-the-pioneers-explain-how-they-did-it/^^^
...the second most important recent advance in our FM processing is the ā€œMultipath Mitigatorā€ phase corrector, which eliminates interchannel phase shifts to minimize L-R energy...


I don't know if the record companies are using similar audio processing for their remasters (they probably are) as part of the Loudness Wars or if their processing changes the phase difference between L and R.


Kirk Bayne
 
How is it that stuff like Gooove Is King, which has what Iā€™d call mild brick walling can upmix so well? There must be more to it.

I would expect it would attributable to phase, not so much amplitude, between the stereo chs. You could have matching levels L/R but if the phase/polarity is 180 deg opposite in one ch, the baby will pop up right out of center back. Reaching into my AA 3 Bag O' Tricks again it has a Phase Analysis window that gives real time visualization of both phase & level relationships in a 2 ch pair:

1620150597061.png


I think the image is fairly self explanatory and in practice you would be playing the track & see how it goes in motion.

We all love good dynamic range but it is only one aspect of audio quality in reproduced music. The current obsession with DR sort of baffles me. I'm sure there's some wide DR recordings that might sound pretty awful because of high distortion or noise in the original. And some bands use that louder than loud compression as a trademark for their bands sound. I think most of the complaints about reduced DR legitimately comes from re-realese of material that had wider dynamic range. I've heard this myself in regard to Klaatu's albums going from LP to several "improved" CD issues. But I would never ignore music I liked from a group I liked just because it has limited DR.
 
We all love good dynamic range but it is only one aspect of audio quality in reproduced music. The current obsession with DR sort of baffles me. ...

I'll suggest it's a red herring. What people really don't like (more than the heavy compression) is the high end eq boosts in those volume war masters. DR readings can often follow the compression and the two often go hand in hand. Up to 4db limiting with the ancient L2 is transparent no matter who you are. Even 8db might be nearly impossible to hear in an A/B. It's the high end eq blast that makes for the stepped on sound. Especially this move: Limit so much it actually noticeably dulls the sound. (eg. more than 12 db). NOW hit it with high end boosts and make it scream!
 
I think most of the complaints about reduced DR legitimately comes from re-realese of material that had wider dynamic range.
Yeah, that's mostly the case for me. It's so effing predictable, not even funny. But, honestly. 99.999 percent (OK, maybe 60% lmao) of all new music that is "rock type" is brickwalled with horrific DR values.
 
Does it sound shrill?

Shrill is eq. Horrific limiting by itself sounds muffled.
 
Does it sound shrill?

Shrill is eq. Horrific limiting by itself sounds muffled.
When I think about newer music (newer bands honeststly) that has poor DR......yes. It sounds shrill, and it sounds muffled (mostly drum sounds) at the same time. Yes, they coexist. Is it on purpose? I dunno, but if it is, they should be slapped across the mouth for thinking that sounds good.
 
Haha. Shrill and muffled at the same time is a special level of inept for sure! Grabbing for the treble is a brute force way to force presence. It's probably as simple as that. It's either quick and dirty or inexperienced at the end of the day. Both I suppose. With today's tools and the expertise possible, even budget releases should sound excellent just matter of course. Some of that went with the volume war approach. It still seems weird that the "professionals" would do this crude of work sometimes. Intentional crude or cheap examples should be at a higher level than this stuff.

There's this secret trick record companies don't want you to know about something something something clickbait! It's called... a volume control!
 
I have done this kind of doctoring of recordings. Usually I do this so a classical piece can be heard well in a car on the highway.

Most of the tools I have do only what they are supposed to. I haven't found any sneaky changes not in the instructions.

I do have a tool that can be used to build a noise profile for noise removal. If the noise sample contains out-of-phase material, then that profile used to remove noise. can remove out-of-phase material.

I have another tool that allows me to vary the ratio of L+R to L-R content in a stereo recording. Turned fully one way, you get mono. Turned fully the other way, you get nothing but L-R (surround).

Note that remastering might also include remixing.
 
Back
Top