As these are StudioCanal releases, it's worth noting that people at SC recut soundtracks on their own to make sure it's in sync with the visuals, they sometimes slowdown material or accelerate it at some points. All this is done without any approval from any of the people who did the films.

(2020-06-16, 08:05 PM)Stamper Wrote: As these are StudioCanal releases, it's worth noting that people at SC recut soundtracks on their own to make sure it's in sync with the visuals, they sometimes slowdown material or accelerate it at some points. All this is done without any approval from any of the people who did the films.

Near Dark (and your comments about SC meddling while I was working on it) taught me to fear their logo... that jingle is like Pavlov's bell to me now except instead of making me hungry it strikes terror into my very soul.

I'm reasonably sure I can do this, it's just time consuming and I'm having to loop 16 kHz interference hum in some brief "silent" sections to resync which may or may not prove transparent (we'll see I suppose). At least retains the character of this mix, monitor noise and all, for those who might prefer it to the French BD version for any reason.

WOW discussions are far advanced for my knowledge, but I think I can help in a couple of points:

1- Even though I don't have Teamblu's release, back in the day, a member of the team informed me the video was sourced from the Japanese BD.

2- As for the Brazilian R4 Colubmia-TriStar, it's indeed a NSTC disc. Video releases like VHS and DVD were NTSC over here, while broadcast used PAL-M.
It was a very limited press and was very hard to find even when it was released. Even though I was an avid collector and used to scout for movies on video and music stores every week, I don't think I've ever seen it in store.

Sony released other Tri-Star Carolco films on DVD before the rights went to Universal/Studio Canal, like Universal Soldier and Basic Insticnt. I owned UNISOL, it was a flipper with P&S and WS on opposite sides and a DD 2.0 soundtrack. I'm not sure I've seen the BI disc around either.

But the good news, is I just found this T2 Columbia-Tri-Star DVD on an auction site. It's expensive, but if it's worth, I think I can get it.

(2020-06-16, 03:06 PM)pipefan413 Wrote: Well, I don't have the original 2008 Japanese BD but as far as I can tell that seems to be what TeamBlu based their T2 on. I'd much rather be looking at the original Japanese BD if I could! I'm just about to check if they did anything to jonno's sync of the 1997 US DVD 5.1; afaik he synced to the UK Skynet Edition rather than the JP disc I think TeamBlu's video is from, though they might already be exactly frame matched. So far, it does actually look like they've at least hacked off the start with a negative delay (looks like -817ms, a little over 20 video frames) but I dunno yet if there's any further tinkering.

They definitely used the Japanese master. That said, I wasn't involved in any of TB's editing/mastering decisions, so there may have been some further adjustment done from the original video that I wasn't aware of.

The conversation started because I'd already done one sync of the 1997 CDS 5.1 to the UK Skynet for my own purposes, and they requested a do-over to fit their own project. The job was based on DD 2.0 reference track they supplied - I was basically matching to a track they had in hand, *presumably* the stereo track straight off the JP disc. I was extremely careful with the matching (even logging the frames I added and removed to achieve the sync) and the result was certainly a good one for the time, though I'm happy to see the resolution of that mix has been superseded in the time since (384kb/s will only get you so far!)

I still have most of the files - let me know if any of it is useful to your further detective work.

(2020-06-16, 10:27 PM)jonno Wrote:

(2020-06-16, 03:06 PM)pipefan413 Wrote: Well, I don't have the original 2008 Japanese BD but as far as I can tell that seems to be what TeamBlu based their T2 on. I'd much rather be looking at the original Japanese BD if I could! I'm just about to check if they did anything to jonno's sync of the 1997 US DVD 5.1; afaik he synced to the UK Skynet Edition rather than the JP disc I think TeamBlu's video is from, though they might already be exactly frame matched. So far, it does actually look like they've at least hacked off the start with a negative delay (looks like -817ms, a little over 20 video frames) but I dunno yet if there's any further tinkering.

They definitely used the Japanese master. That said, I wasn't involved in any of TB's editing/mastering decisions, so there may have been some further adjustment done from the original video that I wasn't aware of.

The conversation started because I'd already done one sync of the 1997 CDS 5.1 to the UK Skynet for my own purposes, and they requested a do-over to fit their own project. The job was based on DD 2.0 reference track they supplied - I was basically matching to a track they had in hand, *presumably* the stereo track straight off the JP disc. I was extremely careful with the matching (even logging the frames I added and removed to achieve the sync) and the result was certainly a good one for the time, though I'm happy to see the resolution of that mix has been superseded in the time since (384kb/s will only get you so far!)

I still have most of the files - let me know if any of it is useful to your further detective work.

Thank you for chiming in, I actually almost messaged you for input then decided I probably didn't need to bother you (and besides I ended up figuring out a way to make it work anyway) so I appreciate hearing from you after all! (I did message you a while back elsewhere about the UK cinema intros, unless there's another jonno on the loose...)

One thing I'm wondering is whether your file posted in the top of this thread was the one made in the first place for your Skynet Edition or the one you did for TeamBlu, because it is different from the TeamBlu one: the offset is different and the bike-cut-to-Sarah-going-aff-her-nut moment is significantly changed in a way that presumably would have to have been edited separately and then encoded back into the AC-3 segment/frame because it doesn't exist in the source track. Was that you or did somebody else take what you did and then do something else to it?

It's interesting to hear that your sync target was the 2.0; that probably explains why it significantly differs from what I'm now doing because I'm using the centre channels of the 5.1. I feel like any reference is probably flawed now though because all these modern releases are using the remixes (which are not exactly synced to theatrical, evidently) and the older releases that did have theatrical audio seem to be missing frames in some places. The 2.0 original mix would be nice to refer to but this wasn't really a Dolby Stereo / Ultra Stereo film anyway, given it went for CDS; if they sent you the 2.0 they used in the release then I'm reasonably confident it's from the Japanese disc, it's definitely the remix in any case.

It's actually kinda nice to hear that our processes sound very similar: I also keep frankly ridiculously detailed notes and am doing precisely what you said and noting the precise frames I'm removing and adding. I unfortunately don't have Womble though so I'm basically just going through the waveforms in Audacity to create sort of a "sketch" version (.wav just to structure it) then will go through and actually make the edits to the DTS encode, most likely by splitting the file at precisely calculated points, rejoining the parts in the new structure, then loading it back into Audacity again to make sure the result is exactly sample-matched to the edited .wav. I imagine this would be significantly easier if I had software that could just knock out and insert frames as I wanted!

UPDATE: currently scratching my head over a bit not long after the 1h mark at which either the JP DVD (theatrical mix) loses about 10 DTS frames' worth of sound or the 2015 US BD (remix) gains it. Don't know if this is an example of the remix or indeed StudioCanal "fixing" an issue with the original mix and if so am wary about copying that but this is >100 ms so possibly noticeable and if it is indeed wrong in the JP DVD I'd be best to make sure it's made right. The trouble is not actually knowing whether the sync on the 2015 BD is correct in the first place! Like Stamper says, it's StudioCanal we're talking about here. I've seen what they're like.

It might help to compare to the original un-synced version of the AC-3 theatrical mix to see if it's much the same as the DTS one and if so, I may not "fix" the sync to match the 2015 BD here. It may also change my editing decisions earlier in the sync. Will test things out and see how it goes.

I think the differences you notice, are because this is a seamless branching point.
T2 master from the Extreme edition was mastered in several portions to allow for the branching.
The portions have probably sightly overlapping audio to allow to be used on the TC version or the SE.

If they are put back together using MakeMKV (automatic playlist), or MKV toolnix (one portion added after another), the audio samples may vary. The JP Bluray had separated versions on two discs, but you can't tell how they processed the versions initially from the mov files they were probably provided.

The motobike moment, especially, as motorbike audio that goes longer on the SE over the next shot.

IMHO you need to only use audio coming from the initial discs that had no such branching mastering.

1997 DVD
JP DVD
Columbia Tri Star DVD
Laserdiscs
Initial SC Blu-ray with the theatrical audio on one disc.

Ditch all the rest. Or you will add confusion to your project (plus the rest is available commercially on all versions).

Regarding some audio on the remix not being in the same sync as the theatrical, that is because Gary Rydstrom probably fixed stuff when he remixed (or stuff moved on the timeline and they didn't notice).

(2020-06-17, 11:07 AM)Stamper Wrote: I think the differences you notice, are because this is a seamless branching point.
T2 master from the Extreme edition was mastered in several portions to allow for the branching.
The portions have probably sightly overlapping audio to allow to be used on the TC version or the SE.

If they are put back together using MakeMKV (automatic playlist), or MKV toolnix (one portion added after another), the audio samples may vary. The JP Bluray had separated versions on two discs, but you can't tell how they processed the versions initially from the mov files they were probably provided.

The motobike moment, especially, as motorbike audio that goes longer on the SE over the next shot.

IMHO you need to only use audio coming from the initial discs that had no such branching mastering.

1997 DVD
JP DVD
Columbia Tri Star DVD
Laserdiscs
Initial SC Blu-ray with the theatrical audio on one disc.

Ditch all the rest. Or you will add confusion to your project (plus the rest is available commercially on all versions).

Regarding some audio on the remix not being in the same sync as the theatrical, that is because Gary Rydstrom probably fixed stuff when he remixed (or stuff moved on the timeline and they didn't notice).

Yeah that's exactly what I was wondering, I was about to say "Yeah like I said..." and then realised I didn't actually remember to type it in the earlier post, hahah. This was partially why I started off by using the JP BD as ref (separate discs) but when I saw how different it was (and considering that the 2015 video was what I was going to be watching anyway) I tried that instead. But the branching was something I was immediately wary of, and I've already got someone who's kindly offered to demux the branched 2015 audio using different software to see if it does anything different. What I will say, to be fair, is that I demuxed with eac3to which discards identical audio frames (the overlaps, like you said).

I'm not sure if maybe there has been some confusion here (maybe I didn't make this clear) but I'm only syncing one audio track (which is "theatrical"), but the problem is the reference. All the HD sources I have, which I believe we've established are all drawn from the same master (I don't yet have the 2017 restoration on disc, but I saw it in the cinema in 3D and was kinda frustrated by it) use exclusively remix audio so that isn't really that reliable as a sync reference. Like you said, and like I've been repeatedly wondering as I do this: things may either have been fixed during the remaster, or they may have been "fixed", or they may have been accidentally done wrong, and I really have no reliable way to know which of those any given difference from the '03 JP DVD is caused by.

The only semi-reliably reference I can think to use is probably that 2008 FR BD that I can't find a copy of, but even that isn't necessarily 100% accurate because it's SC. Remember the audio I'm syncing is from the 2008 JP DVD; if I just assume that the DVD is an exact (video) frame match with the 2015 BD etc. as I thought, then presumably the right approach would be to simply calculate the right delay to add on the start and then leave it alone otherwise, ignoring all the (occasionally severe) sync drifts over the course of the file compared to the remix (although I'm of the belief that you're correct about at least some of the being due to branching, or at the very least mastering to allow the branching). Alternatively, if I get my mitts on the FR BD, I can obviously just work to align the 2008 JP DVD audio to that, then there will effectively be a 1509 kbps lossy (and somewhat less processed) and lossless (and slightly more processed/cleaned up) version of the same mix with the same sync to use with the majority of the Blu-rays, except for the 2017 one. And I could potentially repeat that process again if I was able to confirm with some certainty which (video) frames were then restored in 2017.

I figure it was worth giving it a bash with what I had, but without the actually 2008 JP BD (rather than the TeamBlu project drawn from it, which may or may not have modified sync in some way) and/or the 2008 French BD, I'm like..



Ideally, I think I'd be most sensible to just take the 2008 FR BD and run with that as "correct" for sync ref purposes. I think out of sheer curiosity I'm going to try simply applying a delay as is and watching the bits of the film with the '03 audio that are the most out of sync to see if it's actually watchable or ends up looking/sounding ridiculous; my only semi-reliable ref that doesn't have any branching is the TeamBlu (probably same as '08 JP BD) so I might go back to aligning to that since that was what I was doing in the first place.

But what I should *really* be doing is finalising Snowpiercer, since I basically had that done several days ago before I got distracted with T2! So perhaps I'll go and do that...

You'll never get the 'CDS' track to 100% sync with the 5.1EX Rydstrom remix. The remix was built from the ground up with a lot of movement of tracks relative to one-another. So if the dialogue lines up, the background music will not, etc. The only sane way to approach it is to concentrate on one aspect (for example dialogue) or to sync using the picture as reference

(2020-06-17, 08:31 PM)zoidberg Wrote: You'll never get the 'CDS' track to 100% sync with the 5.1EX Rydstrom remix. The remix was built from the ground up with a lot of movement of tracks relative to one-another. So if the dialogue lines up, the background music will not, etc. The only sane way to approach it is to concentrate on one aspect (for example dialogue) or to sync using the picture as reference

Yeah, I know.

See, my original plan was to focus on video: IVTC the video that was originally presented with the audio source, go through it frame by frame and work out which 2002-sample chunks (a video frame) to remove or pad if there were any differences.

Reasons I didn't:

1. I've never done a manual IVTC and might make an arse of it (I did look into it though and am still considering it)
2. Disk space

So that led me to just looking at the audio instead. I knew it wouldn't be precise because I'm looking at remix vs CDS but I think I now want to go back to that original approach, initially just assuming that they are already frame matched (and just watching it to see how it runs) then checking it with IVTC etc. and adjusting that way. It might be fine just doing a delay but I'd never be able to let it lie unless I check it. But the 2008 FR BD awready has this mix presumably in sync, so that's also going to be useful.

EDIT: Just realised I didn't actually acknowledge one of the most important things you just said there. Re. focusing on only one element: that's precisely what I've been doing! I just isolated the centre channels and was syncing those, rather than trying to sync every single channel and element, since I knew that would be completely impossible.

I just got the Brazilian Columbia Tri-Star R4 NTSC DVD in an auction , so we can find out what it's DD 5.1 track is all about. It should be here in the next few days.

BTW, has anyone heard or have any information on the AC-3 LD [LD68952-2DD]? Is it based on the CDS mix as well?
Since AC-3 LD tracks are know to sound better than DVD ones, could it be a better source?

What about it's PCM track, is it based on the regular 35mm Dolby Surround track?