I noticed a pattern while decoding Dolby Stereo tracks. Many of them seem to "lean left". Aka their left channel seems to often be a few decibel louder than the right one. The result is that in the decoding process, the left channel ends up slightly louder than the right one as well. But more worryingly, this also might affect the correct decoding since the balance between left and right affects what comes through the center channel.

So something that is supposed to come purely through the center channel might end up coming from the center AND left channel, depending on how strong the "lean" is.

What do you guys think the reason for this might be?

I just did a quick google and stumbled over this: https://soundforums.net/community/thread...uder.4646/

Could this be another story like the 16kHz noise coming from monitors during mixing that sound engineers just couldn't hear because they are old farts? Did they maybe do too much DJing with headphones on one ear, resulting in, erm, non-symmetric degradation of their hearing apparatus, thus leading them to wrongly balance Dolby Stereo tracks?

By the way, this applies to both 35mm optical audio and LD tracks at times. With the optical, I tended towards believing that the capture device wasn't properly calibrated but that doesn't explain the LD cases since they are bit-perfect PCM (or should be). Very strange.

How do you decode Dolby Surround?

By the way, try to swith left and right cables from LD; if it's the same, it's not in the stereo track, but in your capture card/software/etc.

@spoRv You should know, with an SDU4.

I'm not doing captures myself. But tracks I've been given had this pattern, both from prints (from different scanners too!) and from LDs which shouldn't have it since it's PCM and bit-perfect.

(2020-11-13, 03:11 AM)TomArrow Wrote: @spoRv You should know, with an SDU4.

I'm not doing captures myself. But tracks I've been given had this pattern, both from prints (from different scanners too!) and from LDs which shouldn't have it since it's PCM and bit-perfect.

I know, but maybe others don't!

I guess you feed it with you PC output; try to swap L/R inputs in SDU4; if it's still the same, it's SDU4 fault; if it's the contrary, there is something in your PC; if it's still the same after done both, the fault should be in the PC/LD/capture card/etc of the person who sent you the files.

You misunderstand. The difference is already visible in the source waveform. The decoding just confirms it.

Yeah, I've seen this as well, and I did wonder what the hell was going on.

Even more bizzarely than the left side simply being a little stronger than the right in terms of balance from start to finish, you may also have seen me talking about how in one particular instance (the widescreen RESERVOIR DOGS LaserDisc), the volume of the right channel abruptly dropped partway through a track, and remained significantly lower for the rest of the film. I initially thought it might be a problem with the capture since we knew that the original capture wasn't bit perfect, but I then got my own copy, did several bit perfect captures of it myself, and they have the exact same issue so it's definitely like that on the LaserDisc itself. The fullscreen version, which is listed as Pan & Scan but is actually open matte, doesn't have that issue.

Something's definitely happening to cause this across multiple tracks, and I would love to know what it might be! Re. the comment about devices not being calibrated for capture, I suppose the thing to bear in mind is that the audio on the LaserDiscs has to have been captured from analogue source(s) to get onto the LaserDisc, so the problem either exists at that stage or before it even got onto the optical / magnetic master in the first place; bit perfect captures will just reproduce any inaccuracy from source to disc.

Oh yeah for sure, but since it's also on the optical audio of film prints (old movies too), idk what to think. And why is it so consistent and always the left channel and not vice versa? Very strange. Think maybe they had some digitization hardware that had a bug that caused this on a widespread scale across different studios etc? Or is it limited to particular distributors/companies working on it?

Is it maybe some kind of inside joke? Since everyone knows Hollywood is ruled by the devil and the devil stuff is spiritually considered the "left hand path", maybe that's some weird wink to that?

(2020-11-13, 04:21 AM)TomArrow Wrote: Oh yeah for sure, but since it's also on the optical audio of film prints (old movies too), idk what to think. And why is it so consistent and always the left channel and not vice versa? Very strange. Think maybe they had some digitization hardware that had a bug that caused this on a widespread scale across different studios etc? Or is it limited to particular distributors/companies working on it?

Is it maybe some kind of inside joke? Since everyone knows Hollywood is ruled by the devil and the devil stuff is spiritually considered the "left hand path", maybe that's some weird wink to that?

Hahaha who knows... but it's as interesting to me as it is irritating!

I hardly think it's fair to generalise audio engineers as 'old farts' because of the CRT hum, after all CRTs were the predominant display back then so there would always have been a background hum.

As for the left thing as long as the phase is correct the centre channel will be played back correctly, it sounds like you're describing mixing decisions. It's the same with surround content, a lot of the time it's more prevalent in one channel, it's usually down to what's happening on screen

(2020-11-13, 10:30 AM)zoidberg Wrote: I hardly think it's fair to generalise audio engineers as 'old farts' because of the CRT hum, after all CRTs were the predominant display back then so there would always have been a background hum.

As for the left thing as long as the phase is correct the centre channel will be played back correctly, it sounds like you're describing mixing decisions. It's the same with surround content, a lot of the time it's more prevalent in one channel, it's usually down to what's happening on screen

Oh I meant no insult, that fate awaits us all. High frequency perception dwindles with age, you can find diagrams to that effect via a simple google search. But you might be right, even someone with young ears might not have been able to avoid that interference or not noticed it since the display was running during recording. Fair point.

Well I'm not sure about it being a mixing decision. It might be, however on a spectral analysis you can also see that in these cases the noise floor in the left channel is a bit higher, which I think supports the idea that it might not have been intended unless I'm making a thinking mistake. And it's not just single sounds, it's centered stuff like dialogue too and large segments of the audio, in fact the entire audio waveform looks bigger on the left channel in these cases. The lean towards the left channel means that the decoder has more bleed of the center into the left channel than otherwise. You can try it yourself by duplicating a mono source, decoding it and then just making the left channel a bit louder and decoding.