hairy_hen

The 1997 mix of SW doesn’t sound anywhere near as good as the 1993 version. It has less dynamic range, lesser fidelity, more noise and distortion, a number of bad edits, a tinny EQ balance, and most of the time less low end. It’s still the best-sounding version of the SE, but the 1993 version is significantly better in all respects. By all rights the '97 should sound better, since it comes from an earlier generation source, but it doesn’t. I think the 4-track master tape it was copied from was not stored properly, whereas the 70mm printmaster used to make the '93 version must have been better preserved.

Oh man. Seriously, of all the worst luck . . .

Moth3r said:

What are the minor differences between the 35mm and the 70mm audio? I seem to remember something about a missing effect during the snowspeeder crash.

Am I correct in thinking that the list of audio differences in the wikipedia article is incorrect and actually represents differences between the theatrical and SE mix?

Yes, the wikipedia article on this is totally wrong. The changes it lists are just differences for the SE, and have nothing to do with the 70mm version. For a long time, since there was no recording of the 70 for anyone to check against, it was often assumed that it must have contained many of these SE differences; but once the in-theater recording surfaced, this rumor was proved to be incorrect. In actuality, it is very nearly identical to the 35mm version. In addition to the in-theater recording, morgands1 also provided a link to an article from 1981 where the mixing of The Empire Strikes Back is specifically discussed, in which the mixers say outright the two were very close, and that very few changes were made for the 35.

That article can be found here: http://www.in70mm.com/news/2015/mixing/index.htm

Most of the differences between the 70 and the 35 are so small as to be virtually inconsequential. Here is a complete list from what I remember:

1. The sound of Luke’s lightsaber deactivating as he leaves the Wampa cave is clearly audible in the 70, while in the 35 it is greatly reduced in level. This was probably changed because he is far from the camera and moving away.

2. When the walkers first open fire in the snow battle, the explosions are significantly louder in the 35.

3. One of the snowspeeder attacks on an Imperial walker uses a laser sound normally used for TIE fighters. In the 35 this was changed to the X-wing laser sound, which the speeders use everywhere else in the snow battle.

4. When 3PO says “R2, you take good care of Master Luke now, understand? And . . . do take good care of yourself,” the word ‘and’ is clearly audible. In the 35, the ‘and’ has been reduced significantly in level. (Not sure of the reasoning behind this one; perhaps it was to make him sound less hesitant in telling R2 to take care.)

5. In the transition from Dagobah to the Imperial fleet after the cave scene, a different TIE fighter sound effect is heard.

6. The end scene is edited completely differently. It does not contain tracked music from earlier in the film, as it does in the 35, due to the scene being shorter. Lando’s line “When we find Jabba the Hutt and that bounty hunter, we’ll contact you,” is absent. His lines “Luke, we’re ready for takeoff,” and “Princess, we’ll find Han. I promise,” are different takes.

That’s it. There are no other changes in content between the 70mm and 35mm mixes of Empire. The re-editing of the last scene, which was mainly done to accommodate the insertion of additional shots of the rebel ships, is by far the largest change, and the only one that is particularly significant. While it might be possible to recreate the other changes by editing in bits of other mixes, the end scene would be difficult. Perhaps not impossible, but there wouldn’t be any point unless the picture were also edited to match the difference in length; and there wouldn’t be any point in that unless the other visual differences could also be reproduced.

The 8mm version, contrary to what was stated earlier, is not at all the same as the 70mm. While it does use a number of the early visual composites created for the 70 which were replaced in the 35, the audio track is entirely separate, and its numerous differences in ADR are not present in the 70mm mix.

The snowspeeder crash sound being missing in the 1993 mix is an odd error that doesn’t appear in any other version. Both the 35mm and 70mm mixes do have it. My theory is that the four-track master omitted it mistakenly, and that it wasn’t corrected on that copy because only parts of the mix would have been re-recorded for later changes. Therefore, this early mistake ended up on home video even though it had already been fixed for the theatrical releases. That’s the danger of going back to earlier generation copies: better sound quality is obtained, but any later changes may not be present.

There seems to be some confusion about the audio tracks . . . CatBus explained it well, so I’ll reiterate what he said to make sure it’s clear.

The Empire Strikes Back had two versions of the film released in theaters: the 70mm cut, which visually was about 99% complete and had a six-channel mix; and the more familiar 35mm cut, which had a two-channel mix and some small refinements to some effects and editing. The 70mm mix and the 35mm mix are both largely derived from a common source, the original four-track master tape, which had a channel arrangement of Left, Right, Center, and Surround channels. The 70mm mix contained those four channels exactly as they were, plus two tracks devoted solely to bass content. It is a powerful track with a lot of dynamic range and plenty of low frequency support. The stereo mix contains the original four tracks folded down into two, and is intended to be expanded back into four channels during playback with Dolby Prologic or something similar. Its dynamics and bass are reduced appropriately to fit into the technical limitations of optical film audio. There are a few very minor differences between the two versions, but nothing really noteworthy. For the most part they sound exactly the same in content and mixing balance.

Since the 70mm mix was only ever used with the earlier cut of the film, and every subsequent release used the completed 35mm edit, technically I could not call my 5.1 version a reproduction of the 70mm audio because it would not match up with the timing of the 35mm edit, and I don’t really have a way to duplicate some of the small changes it included. Therefore, I elected to make a multichannel version, with the full dynamics and bass which that entails, that matched up with the 35mm edit we actually have available for viewing. The best source for this was the 1993 laserdisc, which was actually derived from that same original four-track master the theatrical mixes came from. I converted it up to five channels so that I could bundle it with a custom LFE track (the .1) for additional bass. The result is virtually identical to what the 1993 version sounds like when it is upmixed in a receiver, but with a stronger low end presence. So while it isn’t an exact match for what was heard in theaters, it is nonetheless very authentic in its design. The 35mm cut never had a six-track mix made for it, but if it had, it probably would have sounded quite similar to what I did.

For those worried that the number of channels isn’t high enough compared to the Bluray, or that it isn’t surround-y enough, don’t fret. Like CatBus said, it isn’t just a regular stereo mix that’s been spread out; there is plenty of real surround sound going on. Just listen to the launch of the probe droids in the very beginning and you’ll hear them passing into the rear channels as they leave the Star Destroyer. There’s plenty of other things of that nature throughout the movie, and you can hear all kinds of ambient sounds on Dagobah in the surrounds as well. The rear channels are also better balanced in the original mix compared to the special edition: in the SE, they couldn’t resist the temptation to make the surrounds really loud and flashy, which is distracting and calls too much attention to itself in a bad way. The original mix had more seamless panning from front to back and was really well done. Everything blends together nicely, while the SE never achieves that level of cohesion.

In short, you’ll get an excellent aural experience with the 5.1, as well as plenty of authenticity. I’ll add my standard reminder that it only really sounds the way it should on an actual 5.1 system, and that when viewing the films on a stereo system, the stereo mix will give a better result. The version of that I chose from schorman’s archive was the 1989 Special Widescreen Edition laserdisc, which seems to be the most dynamic version of the stereo mix ever put on home video. It too sounds excellent, and I highly recommend it. If you don’t want lots of bass, you can upmix that and it gives just as good a surround experience as the 5.1 does.

If you want the original stereo mix, just select track 2 and watch the film. I assembled that version especially for this release of Despecialized, and it was edited together from schorman’s laserdisc audio archive, so what you’re looking for is already included.

That is true: the old laserdiscs of the SW films are always missing frames at reel changes. Usually it’s around 5 to 7 frames, but I’ve seen it go as high as 30 in a few extreme cases. It seems to be the oldest home video versions that are the worst offenders in this regard.

When assembling the theatrical stereo mixes for the Despecialized Editions, I always opted to use small pieces of the 1993 laserdisc soundtracks to fill in the holes. The alternatives would be either to let it drop to silence and then come back in, or to try to create a loop out of the existing audio. I found neither of these to be acceptable options, particularly in the larger gaps. Looping the audio risks being highly noticeable, and dropping to silence breaks the sense of immersion, so substituting a small piece of a different soundtrack yielded a much more seamless result, provided care was taken to match the level and equalization as closely as possible.

On rare occasions I also noticed examples of missing or repeated frames in the middle of a reel, but these were always much smaller and harder to pick out than the reel changes. I expect that Colson’s explanation is the correct one; the interpositive and/or internegative may have had small variations in frame count from the negative itself, and these would then make their way to release prints, which would deviate from home video versions that were derived from earlier generation sources.

These colors look fantastic.

Within whatever small margin of error, I’d say the look of these scenes has been determined with great accuracy. They just look right, you know?

I believe Darth Mallwalker said that one extra frame was somehow added to ESB 2.0, making all audio tracks play one frame too early. It’s not easy to notice this while watching…

The 2-frame difference at a reel change during the Luke and Vader duel was a discrepancy between the PAL and NTSC versions of the GOUT. ESB 2.0 deleted these two frames from the video in order to match the NTSC audio tracks, so there shouldn’t be any additional mismatch there.

Analog audio doesn’t have a sample rate or a bit depth… it does have a noise floor and an usable frequency response range, but these are not hard limits and the distortion caused by moving beyond them is not as harsh or destructive as it is with digital audio.

When digitizing analog audio, you should record it as PCM with a bit depth of at least 16 at the absolute minimum; but using 24-bit is much better because the digital noise floor is well below anything a human could ever hear, and you have more breathing room in setting your recording levels without worrying about pushing the signal into clipping. With good quality converters, 48 kHz is quite acceptable as a choice of sample rate; with bad converters, a higher rate such as 96 kHz may be helpful since the converter artifacts will be pushed farther up into the ultrasonic range, where we can’t hear them. It can then be converted down to 48 with high quality software later on in order to fit with video standards.

If the bit depth is reduced to 16 on the final released copy to save space, dither must be used to preserve sound quality and avoid adding more digital distortion. Encoding lossy copies from a 24-bit PCM master may result in higher quality than encoding from a 16-bit version. Also be mindful that leaving at least 1 dB of headroom in the PCM is advisable so that clipping distortion is not added during a lossy encode, or by the conversion to analog during playback.

The more I see of Star Wars from original film sources, the more I think that this warm appearance is the only way that it looks even remotely correct. Applying a natural or cooler color scheme totally destroys this distinctive look to the Tatooine scenes, making them much more generic and uninteresting, even if they can be ‘prettier’ that way. The yellow sand and rocks, the nearly colorless skies, the grain and softer focus all combine to create the feeling of an empty wasteland, which is certainly what they were going for at the time. Later revisions to this look only serve to undermine this aspect of their appearance.

The appearance of trailer #3 most closely matches what that scene looked like in the print I saw. Since there was some fading, this would account for the moderate blues to have lessened to the point that it appeared neutral when I saw it. Other than the blue, the rest of the color was intact for this scene, more so than elsewhere on the print (some other scenes had become rather pink, but not this one).

Based on seeing those trailer shots, and the notion that the blueness fits with the cold exterior of the scene, I withdraw my theory that it was a mistake, and concur that it must have originally been an intentional part of the look. The red of later releases is simply revisionism after the fact.

My theory is that the blue cast, which is the opposite extreme from the red in other versions, is a mistake introduced at the interpositive stage. If I remember correctly, color timing on film is done with a series of red, green, and blue light filters. It is not unreasonable to suppose that on the interpositive this print was derived from, they meant to use a red filter in the bar scene, but messed up and accidentally used a blue filter instead, then didn’t have time to go back and reprint it. So any print made from this interpositive would show a blue cast to the scene; it is possible that other interpositives were made that did not have this blue cast.

Back in 2007, I went to a screening of a very old print of Raiders, and I remember very clearly that the bar scene did not have the red tint that the DVD and other recent home video releases have. Since it was an old print, and had significant fading throughout the film (the amount of which varied considerably between scenes), I unfortunately cannot say what the color actually should be. It definitely was not blue on the print, but the fading was such that the color blue was somewhat scarce throughout the movie, so who knows. If anything, I just remember it looking neutral.

There isn’t any content to make a 6.1 mix, since the original material has only a mono surround channel. In theaters this would be played by the entire array of surround speakers, producing a very wide sound filling the entire side and back areas of the room.

In order to approximate this in a home environment, with a typical 5.1 setup, I used the output from Prologic II in movie mode and then manually inverted the polarity of the right surround channel. Doing this prevents the mono surround effects common to both rear channels from forming a phantom center behind the listener, making them wider and more diffuse, as was the intention. Mono surround films typically aim to create a more ambient and less easily localized effect than modern formats, which can pan surround effects in any number of specific directions. (Inverting the right surround also has the side effect of making the upmix at least somewhat listenable if it is downmixed back to stereo again.)

The number of channels used in the delivery format is less important than its ability to reproduce the aesthetic used in initially creating the mix. Sending the mono surround effects to speakers located only in the back of the room would defeat the intention and would have little similarity to the way it was presented in theaters.

If I end up making another version at some point, I may use some different techniques than the existing version, in order to reduce front channel crosstalk into the surrounds and make the LFE more authentic, but the overall approach would still be similar. It is possible that a 4.1 mix might actually be more friendly to those with 7.1 systems, since it is possible to expand it during playback: according to Roger Dressler, playing a mono surround film with the receiver set to Prologic II in music mode will spread the surround effects equally among all four rear speakers. Note that this must be done in music mode, NOT movie mode, or else they will be sent only to the back, which is not at all what we want. This trick will not work with my current version, since the surround effects are decorrelated from each other, and I don’t know how that would sound since I only have a 5.1 system myself.

Here is, finally, some proof that the bar scene in Nepal did not have a red tint as it has on every recent home video release.

No, R2 tastes bad in both 35mm and 70mm.

Like I said earlier, the 70 is virtually identical to the 35 for most of the movie. Aside from a few small changes in sound effects, the only place it differs is the very last scene, which was re-edited for the 35 and has changes to both music and ADR.

It’s been ages since I saw it in the theater as a kid, so I can’t claim any kind of reliability; but when I think back on it, my memory of Beauty and the Beast was that it was often dark and gloomy, which would go along with the beast’s angry nature and Belle’s captivity. When I watched the film again on DVD several years ago, everything seemed very bright and colorful, which did not match my memory of the movie at all. I can’t recall specifics of either very well, but it definitely felt wrong to me as a viewer: the appearance of the movie seemed rather at odds with the tone of the story.

The above description of the 35mm film print being more dimly lit suggests that my admittedly vague recollections were on the right track. I’d be very interested to see what it is really supposed to look like.

My first thought on hearing the mono mix was that severe noise reduction had been applied. I have to concur with pittrek’s observations that a great deal of background detail has been largely removed along with the hiss, and that the music often suffers from digital distortion artifacts as a result of the noise reduction, giving it a watery, warbling sort of sound.

The good news is that from listening to the mp3 of the raw, unprocessed track, it seems like it has potential for a better sound quality to be achieved. Whether it could turn out better than the Belbucus track is hard to say; I’d categorize them as roughly similar in terms of fidelity based on what I’m hearing right now. If the original, uncompressed audio captures are still available (preferably at 24-bit if possible) and can be sent to me, I’d be willing to take a shot at cleaning them up.

By the way, I’ve watched the SSE v1.6 all the way through, and I was very impressed with the results that have been achieved. I wasn’t too keen on Star Wars in general for a while, owing at least partially to the weird things that happened about scans and teams and so forth, but now I’m starting to get back into it again. Being able to watch the whole movie looking like 35mm film, without a trace of the SE in sight, is very rewarding. Thank you guys for making that possible.

I think it is a holdover from the days when nobody knew what the 70mm mix of ESB sounded like, but this idea that it was similar to the SE, or to the 16mm, or the 8mm, or to any other odd variant is incorrect. Examination of the in-theater 70mm recording by morgands1 shows clearly that the only version the 70mm mix resembles is, in fact, the 35mm.

The 70mm version was made first, with the 35mm being completed shortly thereafter to go along with the visual edits. They are both extremely similar throughout the movie and show only minor differences from each other: a few small sound effects changed, a few level differences, a few lines of different ADR at the very end, etc.

It doesn’t make sense to think that any of those major differences in other versions would have appeared in that first mix, been completely deleted from the second, and then somehow reappeared in subsequent versions. Just like the changes to the mono mix of the first movie, all that extra stuff was added much later. The 70mm and 35mm share a common 4-track source in their ancestry, and are therefore very similar in any part that wasn’t re-recorded for the 35.

Well, it worked. I haven’t been interested in Star Wars for months now.

It sounds like the 35mm stereo mix comes from the Japanese laserdisc, the one with digital audio which was discovered a few years ago. If this is indeed the version I’m thinking of, then it has numerous dropouts throughout the movie and a pretty severe instance of tape distortion during the sunset scene.

For the Despecialized Edition, I put together a corrected version of this track, which eliminates every instance of dropout and distortion by replacing them with segments from Belbucus’ analog capture. I also adjusted the gain of the whole mix so that the dialog is at the same level as in the 1993 version, eliminating any perceived difference in average volume when switching between audio tracks. Looks like it’s too late to include it in this release, but most of the errors probably aren’t very noticeable anyway, so it’s not worth worrying about.

For the sake of clarification about my 5.1 version, it was created using the 1993 mix as a base, while removing all the extra sound effects that weren’t present theatrically and replacing them with segments taken from the 1985 mix, level-matched and EQ’d to blend as closely as possible. It was then upmixed to five channels with Prologic II, and combined with a custom LFE channel containing bass effects for each explosion, spaceship flyby, and other things of that nature. Each individual instance of bass was heavily edited (in level, dynamics, and timing) in order to blend into the rest of the soundtrack. Since the 1993 mix was mostly derived from the 70mm printmaster, the changes I made allow my 5.1 to be considered a close approximation of the original 70mm six-channel mix. Obviously it is not exact, since it does not have the true discrete channels of the original, but it does provide a very similar feel. The surround effects are monaural, the dynamic range is for the most part very similar, and the bass provides a solid foundation for the action which is missing in other versions.

I considered this to be worth trying to reconstruct, since Star Wars was the first movie ever to be made with a dedicated LFE channel, and the only other way to hear such a thing on home video is with the rubbishy special edition mixes. Since it is a reconstructed approximation and not completely a preservation as such, and since the real version was only present on 70mm prints, perhaps it does not really belong in a project that seeks to deliver the experience of watching the film on a 35mm print, whose content is 100% authentic. But if people want to include it in their own derived copies (probably with the branching removed to make authoring easier), I certainly won’t try to stop them. 😉

Anyway, enough of that: since I haven’t been around, I had no idea this project was ready for release. It looks to be quite impressive, and I’m certainly very interested in watching a version that is completely sourced from film and contains no traces of the SE whatsoever. Looking forward to seeing it!

It does, and it’s one reason why I’m not altogether thrilled with the idea of eliminating the color tints of DeEd 2.5 and making everything look neutral.

Mike V made this point back when he first introduced us to the idea a few years ago: in order to get back the original colors of the movie, you have to view it through a 70’s-era light source. If you don’t, you’re not seeing it the way the filmmakers or the general public would have seen it at the time. They made their color timing choices by viewing the film through projection equipment of that era, with its warmer and more yellow-tinted output. A neutral light source just doesn’t look right for this movie, because that whole effect is completely removed. The amount of yellow in the image will be significantly reduced compared to how the output would have appeared when the color timing was being done, so in order to see what they would have seen, it has to be put back in.

Obviously very few people are going to be able to project a film print with vintage light source, so incorporating it into the color timing of a release is the only way to actually ensure that it comes across the way it should to the end viewer. Now obviously this doesn’t directly affect what is being done with the algorithm this thread is about, because a neutral color scheme is exactly what it will most easily create. But it is something that needs to be kept in mind for any release that claims to desire color accuracy to the movie’s original appearance.

Yes, it’s ‘prollerby’ or nothing.

I think the blue weirdness is just how the scan came out, for whatever reason. In Mike Verta’s video “10 Minutes with an LPP”, he shows how his own scan of this same print can be corrected, and his version looked nothing that on the raw scan. It isn’t blue or dark at all, but just mildly green-shifted. So the print itself isn’t faded or weird, but the raw scans come out looking strange for whatever reason.

Shadows of the Empire works near-perfectly in Project 64. The problem is that the default settings are wrong. You have to make sure the Counter Factor is set to 2, not 1, and the VI Refresh Rate is at 1500, not 2200.

Just these two settings alone will fix a great many games that were broken in the more recent revisions. There are only a few that actually need them the other way. You also need to make sure you’re using Glide64 for your video plugin, because the default plugin is riddled with graphics glitches. Turn on ‘hardware frame buffer emulation’ and ‘software depth buffer rendering’ to enable many effects in various games that would otherwise be shown incorrectly or not at all.

Having said that, I’ve become increasingly dissatisfied with the state of Nintendo 64 emulation, since it still has many errors and shortcomings that get in the way, and don’t look to be fixed any time soon. Because of that, I’ll be using it less in the future, and spending more time on the actual console. The biggest problem with this is the joysticks are so prone to wearing out, so I’ve been looking into ways to repair or replace them. I got one of the ‘Gamecube-style’ sticks, which are very well made and won’t wear out, but since the sensitivity on those is completely screwed up and it’s really hard to aim properly in GoldenEye with it, I ordered a custom circuit board from a guy in Germany which is designed to fix the sensitivity and make it respond much more like a new original. I’ll have to learn to solder in order to install it, but it’s very much worth the effort.

On the subject of Zelda—if you’re going to play Ocarina of Time, for god’s sake, make sure you get a copy that is either version 1.0 or 1.1. Version 1.2 introduced the absolutely horrible revision to the music in the Fire Temple, which eliminated the awesome chanting and replaced it with some generic rubbish that just doesn’t fit. Listen to the original here: https://www.youtube.com/watch?v=VwGVb8Ivb9E

For our English friends, also known as bollocules.

Did you know that the alternatively-pronounced Testicules was the greatest of all the ancient Greek heroes? (Though he may have been the Roman version, going by the spelling.) He had the hugest, hairiest, shweatiest sack of all the sons of Zeus, and none could withstand his mighty swing.