Multichannel Audio (Moved from MQA)

[pinnahertz]

Huh. Sounds like it was mixed in an uncalibrated room. This isn't the first time I've heard of this. There was a rather prominent movie trilogy a few years back that reportedly was mixed in a room with miscalibrated LFE resulting in a whole lotta LFE that was blowing subs in calibrated rooms.

Yeah, we don't need to pay an expensive tech to calibrated our rooms. Right.

 

[bigshot]

I'm wondering if someone doing the mastering had their info wrong. Historically, LFE channels in theatrical spec are attenuated 10dB to allow for extra bandwidth and boosted 10dB on playback. (This doesn't apply to home theater, where bass management is involved.) I think I've read that rears are attenuated 3dB for theatrical too. If the engineer patched the cut backwards and attenuated the rears 10 and the LFE 3, that would make total sense mathematically. About a 10dB boost on the rears and a 3dB boost on the LFE on playback would put it back into spec. Of course they shouldn't have been applying attenuation at all for a blu-ray because it's intended for home use, not theatrically.

I've only checked the DTS track so far. I plan to go back and figure out if the Dolby one is the same. The weird levels might be because they were mixing for a theatrical 7.1.4 Atmos setup, and forgot to readjust the calibration when they output 5.1 for the blu-ray. That's just a guess though. I have no idea how they wouldn't realize that they were 10 dB off unless they thought they were calibrated to correct spec and something downstream would correct it.

 

[pinnahertz]

You might want to check your own sys cal., just to make sure. Mark Waldrep just posted this (and he'd have picked up something as off as low LFE and low S):
http://www.realhd-audio.com/?p=6006

Not a complete review, that's coming. But I would have expected his initial impression to pick up on something like you've observed. Rather, he said,
"It’s so amazing to hear the background vocals, audience noise, sound effects, and discrete instrumentation coming from all around you."

Might be too early to point fingers just yet.

 

[bigshot]

My system is totally in spec. I play multichannel stuff several times a week and it all sounds fine. This album is the exception. I think people think that it's supposed to have a forward soundstage. They think the rear channel is just fill, but it isn't just fill if you boost it. All they have to do is to make the simple level change I've noted and the whole thing will mesh properly. Suggest it on that forum if you are a member there and have them try it. There will be armchair experts who poo poo it without even trying it, but the ones who do try it will instantly hear the difference.

One caveat... I've only listened to the DTS track, not the Dolby. It could be that there is a difference between the two.

 

[pinnahertz]

My system is totally in spec. I play multichannel stuff several times a week and it all sounds fine. This album is the exception.

I understand that. But when was the last time you calibrated?

I think people think that it's supposed to have a forward soundstage. They think the rear channel is just fill, but it isn't just fill if you boost it. All they have to do is to make the simple level change I've noted and the whole thing will mesh properly. Suggest it on that forum if you are a member there and have them try it.

That didn't come from a forum. It was his blog/email.

There will be armchair experts who poo poo it without even trying it, but the ones who do try it will instantly hear the difference.

I would suggest you check out who Mark Waldrep is. Just google him.

One caveat... I've only listened to the DTS track, not the Dolby. It could be that there is a difference between the two.

Yes, could be. How hard would it be for you to try the Dolby? Though I expect most players and systems default to DTS...

As I said, it may be early to point any fingers here. If this is due to a system calibration error, it seems less likely that someone producing a commercial product would ignore calibration of his monitor system, but it has happened. It could also be an artistic choice, as you say. My pointing to Mark's post was meant to show that someone else seemed to feel the presentation was immersive, and not stage-forward. If the universal opinion is that the mix is stage-forward and not immersive, then we have an artistic choice or miss-cal during production. And we may never know which. But given Giles work on "Love", it seems less likely it's an artistic choice (to be stage-forward and not immersive).

 

[bigshot]

I calibrated my system last about six months ago. I do fine tuning all the time though. It isn't likely a calibration issue if I play a wide range of other multichannel music and movies almost every day with no problem. This is the only one that needs this kind of boost. 7dB isn't subtle at all- it's almost double the perceived volume. I accidentally started playing a TV show without taking the rears back down to 0 again, and it sounded horrible. People often make excuses for mixes, thinking that it's "the intent of the engineer in the studio" when it isn't. Some mixers do have poor judgement, but they don't generally have poor judgement by 7dB consistently across a whole album. My suspicion is that there is a different calibration for theatrical multichannel than there is for home theater, and they accidentally mixed up the settings when they were mastering the blu-ray.

If you have the album, try playing it with the boosts I suggested. You'll instantly hear what I'm talking about. It's like the difference between a classical style "low bed of ambience in the rear" and a Steven Wilson style immersive mix. When the sound crosses from back to front or vice versa, it's more noticeable, because at 0dB, it doesn't mesh up in the middle of the room, it sounds like it's forward, then it dips out, then it reappears in the rears. With the boost, it's an even handoff from front to rear. That doesn't sound like an artistic choice. Bluray mastering can be tricky. I've had copies that were corrected and replaced by studios in the past.

Like I said, I haven't checked the Dolby version, the theatrical mix, nor the lossy surround mix on the DVD. One or more of those versions might be correct, and perhaps that's what other people who don't notice a problem have been listening to. But I'm sure anyone who listens to the DTS mix on the blu-ray and tries the boosts will see right away what I'm talking about.

 

[pinnahertz]

I calibrated my system last about six months ago. I do fine tuning all the time though.

That would be reason enough to take a few minutes and verify it. I'm not saying you're wrong, just suggesting you might consider eliminating a possible source of error.

It isn't likely a calibration issue if I play a wide range of other multichannel music and movies almost every day with no problem.

If someone were to mount a 5.1 project as significant as SP, the first thing they would do is verify their mix environment was dead-on. At least, one would hope so. Using other mixes as a reference were the specific surround level is subject to artistic is not a reliable means of verification. There are several places where level adjustments can occur.

Personally, I've recalibrated several "dead-on" systems only to find they weren't, even though the owner insisted they were. One was a professional THX certified studio that produced sound for film. It wasn't a matter of a couple of dB off either, it wasn't even close. And they'd been using it that way. If I recall, surround level was down, sub EQ was off, and center was hot. And because of the type of projects they worked on, this likely was never recognized as an issue.

This is the only one that needs this kind of boost. 7dB isn't subtle at all- it's almost double the perceived volume.

I agree that tends to support your hypothesis. Wouldn't it be nice to validate the hypothesis?

Like I said, I haven't checked the Dolby version, the theatrical mix, nor the lossy surround mix on the DVD.

Let us know when you do.

 

[bigshot]

Wouldn't it be nice to validate the hypothesis?

No, not really. If everything else sounds fine and this is the only disc that sounds incorrect, I'm happy to go with deductive reasoning on this one. I don't want to drag out all the stuff and recalibrate unless it's affecting more than just one disc. I'd rather spend that time listening to music and watching movies. If more discs turn up wonky, the first thing I'll do is recalibrate.

 

[gregorio]

I don't think you understand what I'm saying. Maybe I'm not saying it clearly. The probably mixed one track on the album at a time and the article says they did had multiple bounce downs on a single song... four tracks down to one in the first bounce down. Then three more tracks added. Then those four tracks were bounced down to the final mix. That is 4 + 3 tracks. In the new mix they are going back to the original tapes before the bounce downs and assembling them in a multitrack digital mix. 7 elements will give them more flexibility than 4, and digital mixing provides many more options than 60s technology.

I do understand what you are saying but you're not understanding what I'm saying! When they bounced down to another tape, the two machines were not sync'ed. The 4 tracks bounced down to one track of the second machine would not be sync, there would be a timing discrepancy. That's not a problem as the next 3 added tracks would be performed relative to the one track bounced track. Same with the next generation bounce down, not a problem because the added tracks are performed using the bounced track as reference/guide. However, as with adding noise, with each generation you're compounding timing errors. Again, no one will notice this unless you try to combine material on the last tape with something from a previous tape and then you'll have phase problems. There are digital processors which can correct this type of error (VocAlign for example) but it's not ideal and we still have the problem of only 7 elements (and the equivalent of at least 10 channels with Atmos) and, it still depends on what those 7 elements are.

Historically, LFE channels in theatrical spec are attenuated 10dB to allow for extra bandwidth and boosted 10dB on playback. (This doesn't apply to home theater, where bass management is involved.) I think I've read that rears are attenuated 3dB for theatrical too.

You seem to have a misconception about surround calibration, which is not surprising as it's a bit of a mess! Theatrical surround systems are calibrated to: LFE +10dB in-band gain relative to each of the mains (LCR) and the surrounds to -3dB. For Home theatre (bluray, dvd, HDTV) the calibration is the same as theatrical except that the surrounds are not -3dB, IE. All speakers calibrated the same and the LFE to +10dB in-band gain. So your statement is incorrect, the +10dB of in-band LFE gain DOES apply to bass managed systems but only to the LFE channel, not the entire sub output! Theatrical mixes are typically re-versioned (remixed) for consumer distribution. Music should in theory be more or less the same as home theatre (although there are some differences), except for SACD, where the LFE does not require the +10dB in-band gain on playback (IE. All channels the same including the LFE). BTW, accurately calibrating a surround sound system manually is not possible with a simple SPL meter, you need the right tools: A full range measurement mic + RTA software + full range AND band limited pink noise.

Your quoted problem is a pretty serious error on someone's part. 7dB out is significant and there's no reason for it but severe incompetence somewhere along the line. The relatively straight forward positioning you've described is in line with what I would have expected.

G

 

[bigshot]

Someone over at Home Theater Forum tried the fix and agreed with me that it improved the mix a lot. Interestingly enough, when he played the bonus tracks (Abbey Lane and Strawberry Fields) they sounded better at default settings. That makes sense because those mixes were done for the 1+ bluray a year or so ago and there was nothing wrong with that one. I think there definitely is something wrong with the Sgt Pepper track. Tonight I'm going to check the Dolby and DVD versions and I'm going to try to do the correction subtractively, by reducing the mains rather than boosting the rears. I think that may work better.

 

[bigshot]

When they bounced down to another tape, the two machines were not sync'ed. The 4 tracks bounced down to one track of the second machine would not be sync, there would be a timing discrepancy. That's not a problem as the next 3 added tracks would be performed relative to the one track bounced track. Same with the next generation bounce down, not a problem because the added tracks are performed using the bounced track as reference/guide. However, as with adding noise, with each generation you're compounding timing errors. Again, no one will notice this unless you try to combine material on the last tape with something from a previous tape and then you'll have phase problems.

I understand that, but Giles Martin has already dealt with far worse problems on the 1+ blu-ray. That disc included television performances of the Beatles where the backing track was a tape played back on set and the vocals were performed live with audience screams in the background- all merged onto a single mono track. Giles Martin was able to significantly reduce the level of the playback tape and replace it with a phase corrected instrumental track off the 4 track master, while still leaving the live vocals and audience sounds. That's even worse than trying to line up separate 4 tracks, because some of these TV performances were shot on video then kinescoped. You can't get more opportunity for timing error than that. He was able to take all the available elements and merge and conform them perfectly into a very good sounding 5.1 mix. I'm sure it took a long time and a lot of hard work, but with the Beatles, money is no object. The stuff being done here is cutting edge technology. It's ironic that something as simple as a calibration error could mess it all up!

I have a blu-ray that is similar, except with film, not sound. They discovered a deteriorating hand colored print of Melies's landmark film A Trip To The Moon. It only existed in B&W before that. They years carefully peeling the layers of film off the reel and scanning every fragment. They ended up with hundreds of thousands of bits, which they reassembled using computer analysis. After this was done, there were gaps in frames and sections of the film that were missing altogether. They patched the holes in the film by compositing it with a completely different B&W print. Then they contrast graded and digitally colorized the B&W elements to precisely match the hand tinting. I would never have believed it was possible if I hadn't seen it myself. The documentary on the restoration of the film is ten times longer than the film they restored.

 

[pinnahertz]

BTW, accurately calibrating a surround sound system manually is not possible with a simple SPL meter, you need the right tools: A full range measurement mic + RTA software + full range AND band limited pink noise.

True, but all of that is pointless unless there is also a means to accurately deal with measured response anomalies. Usually, in a home system, the EQ tools are either far too basic or nonexistent. So what you get down to is using the AVRs' built-in auto-cal system, then verifying. If you have no means to manually tweak EQ (most auto-cal systems have little means of manual tweaking), then an SPL meter and band-limited pink noise will get you close enough. The good auto-cal systems get things pretty darn close.

 

[gregorio]

True, but all of that is pointless unless there is also a means to accurately deal with measured response anomalies.

Agreed but I was talking just in basic terms of achieving a vaguely accurate LFE balance (getting the +10dB).

G

 

[pinnahertz]

Agreed but I was talking just in basic terms of achieving a vaguely accurate LFE balance (getting the +10dB).

G

Yes. What I said applies to getting vaguely accurate LFE in a home. Vaguely...that's the trick. Most home theaters/5.1 systems have a single sub in a little room where it looks good (not sounds good). Hence you have to be lenient with your use of "vaguely". My point was that if you had all the right measurement gear you could get the LFE level close, but still couldn't cal because the EQ tools aren't available. An SPL meter with the band-limited noise found on any test DVD will get the gain right, and frankly, eliminate the bulk of a 7dB error. Only if you actually use the SPL meter, though.

And AVRs don't know to kill the +10LFE for SACD, hopefully a disc player does.

 

[pinnahertz]

Tonight I'm going to check the Dolby and DVD versions and I'm going to try to do the correction subtractively, by reducing the mains rather than boosting the rears. I think that may work better.

Why would that work better? The critical part is the balance between all channels, how you do it is unimportant beyond having headroom issues. How do you know what you're starting with (yes, it's another shameless "why not recalibrate" point).

Sorry, I'm a very analytical person. When I hear of a problem the first thing I "need" is all the information. Not getting that here.