TServo2049

Technically, this IS available on DVD/Blu-ray, but...

I found the deluxe laserdisc box set of Snow White and the Seven Dwarfs for 10 bucks at a Half Price Books. I'm not sure if it has any extras missing from the DVD/BD releases, but I do know that it has a lossless version of the original mono mix, which is only at 192kbps on the current DVD/BD. It's only in CX analog (with the stereo remix on the digital track), but it is the only uncompressed version of the original sound mix ever released.

I definitely think this track needs a preservation. I'll send it to anyone who can capture it.

dvdmike said:

 

First encode, not 100% happy yet 

http://screenshotcomparison.com/comparison/148709

http://screenshotcomparison.com/comparison/148710

 

Wow, that looks great so far. I still can't figure out why we get these historical-revisionist retimed versions.

Star Trek II is another - great picture quality, but the colors are way too cool. I'd love it if someone did a color-corrected version of that film based on the Director's Cut DVD...

msycamore said:

Also, member Gregaton mentioned earlier this year: http://originaltrilogy.com/forum/topic.cfm/Hopefully-the-last-70mm-vs-35mm-ESB-audio-differences-thread/post/573758/#TopicPost573758 Yoda's "Ehhhh!" was there at some point before the SE. I have an old VHS tape with a recorded 1980 tv review of the movie, and a clip from that scene shown during the review features the "Ehhhh!"

I wonder if this version is the sound mix that was on those clips sent to TV stations/etc. by Fox/Lucasfilm for promotional purposes. TV was still exclusively in mono, so would it be too much of a stretch that the clips sent out used a mono mix?

(I know I've seen a Siskel and Ebert episode from 1980, around the time ESB came out, which showed clips of the first film with the mono mix.)

Judging by the EoD clip, the '97SE ROTJ subtitles were actually in the same font as the original '83 version. It's not a pre-SE clip because in that doc, the non-SE clips are softer and have noticeable jaggies.

When Return of the Pug is finished and released, we'll finally be able to compare the two (or at least, the lines which show up in EoD).

msycamore said:

Darth Mallwalker said:

FWIW JSC has a video splice at that point, right after Gringo's last words.
There's two consecutive 3's in JSC's telecine pattern (in GOUT terminology 71933/71934).
Doesn't prove, but suggests JSC's main source could've had subs...

Good point, the different aspect ratio and framing/cropping in the Technidisc master in the Gringo scene I demonstrated in my thread, suggests the main source for that may have had subs as well.

But I thought they specifically tracked down the "first-generation IPs" because they were subtitle-free. Does that mean that they only used said IP as a source for the Greedo scene, and not the rest of the film? (If so, that could rule out my "Technidisc/GOUT source was used to make the foreign versions" theory.)

I think it would be a good idea to try to do a visual comparison of the scene as it appears in the JSC, Technidisc and GOUT. Perhaps the "first generation" source tracked down for the Definitive Collection was not used in the Technidisc transfer (which was done in 1992 for the VHS box set). It could have even been dropped in from the JSC master.

This could mean that every pre-SE home video transfer had the scene cut in from a different source via video editing. The 80s P&S video masters dropped in the version made for cropped Academy flat prints (the kind that used to be shown on airlines; at least one has turned up on eBay before). The subtitles are optical, and the cropping is identical in the time-compressed and uncompressed versions.

msycamore said:

msycamore said:

So we might perhaps start to refer to these versions as the Stereo and Mono versions.

Actually, on a second thought it's probably better to refer to this as an opening day vs. wider release variation, sort of like what happened with Empire that had modifications seen a month into its run when it was time for the 35mm versions. As Dolby Stereo prints have these modifications as well, it only gets confusing. ;)

The most likely scenario on the info we have is:

May '77 - Dolby Stereo prints - containing the rare credit roll and cloudy composite, possibly an early more unstable composite of the opening crawl.

June '77 - Wide release with a third Academy Mono mix for a large number of prints, with the above shortcomings corrected from then on.

That sounds plausible. Some comments:

-The IB Tech prints don't have the "shaky" crawl Mike V was talking about. For the 1981 re-release, Fox's UK division took the IB prints in their inventory and spliced on the Episode IV crawl in Eastmancolor. All of the known surviving IB prints have this alteration, so we don't know which opening crawl they originally had. (The fact that the film was released in the UK in December 1977, so it makes sense that it's the "revised version.")

-If the IP used for the 80s video masters was from an older source (see: lack of burn marks, lack of Greedo subs, "cloudy" composite), why does it have the revised credits? Were they spliced in at some point? (I'm guessing the cut point would have been at the final group shot...)

-What happened to the burn-mark free, subtitle-free IP with the "cloudy" takeoff composite? I can think of three possibilities: 1.) It was either too faded or too damaged to use again in '93; 2.) It was misplaced; or 3.) It was sent to Japan to make the JSC telecine, and was never returned.

(I'm not implying that it was stolen, simply that Fox Japan could have misplaced it or forgotten to send it back. That also doesn't mean it would still be there now; Fox/Lucasfilm could have recovered it during the element search for the SE restoration. And I myself admit that the Japan theory is the least likely one. I don't even know if the JSC transfer was done *in* Japan, I'm just guessing based on the fact that Fox had to use the hardsubbed Japanese masters for the American widescreen LDs.)

-Your theory may provide a clue to the origin of the Technidisc/GOUT source. It was assumed to be the "first-generation" IP, and was found in "a special vault having only opticals" after a "massive search" for a subtitle-free IP. If it's a chronologically earlier source, why did it have the cloudless composite, and why did have the burn marks that were missing from the IP used for the earlier transfers (which was ALSO subtitle-free)? And why was it found in a vault intended for storing opticals? I'm beginning to wonder if it was actually an element used to make the export versions...

-Does anybody here have the old, pre-THX French/German widescreen LDs? They seemingly used a different film source than the JSC master, because while they also have the Greedo scene without the English subs, that one scene is cut in from a clearly inferior print.  This suggests that their main source had the burned-in subs, and may not be the same as the JSC, in which case the takeoff composite may be different.

-russs15, could you maybe provide screenshots of this scene from the ITV broadcast? That version seems to be a film-chain from a flat, cropped Academy print (like the ones that used to show up on eBay from time to time). It has the EpIV crawl and lacks the Tantive burn marks, but it has the 1977 flyover instead of the recomposite. Perhaps the Yavin takeoff shot is different there too?

poita said:

None of them. If you compare them to a print, they are all off. The SC is in the correct spectrum but the saturation is way too low and the colour is way off in some areas. The THX laserdisc releases are closer to correct saturation, but the tint is off. The bootleg tapes are also not all that close. The PAL THX ones are closer than the NTSC ones.

The only reliable source is a print, anything else and you are only matching to the telecine colourist's choices of the day.

Agreed. That's also the view Mike Verta takes.

I've talked about how the PS78 bootleg transfer offers important clues to how the original '77 Eastmancolor prints looked*, but even so I understand that it is not a truly accurate representation.

Old video footage still intrigues me. We don't know how much of the color variations in the official releases was due to the telecine colorist, or how much was due to the timing of the film sources. However, when clips were shown in TV reviews and such, they were sourced from film that contained the original color timing. There's a clip of the TIE attack on the Falcon in a 1980 Siskel and Ebert episode, and it has a nice gold tinge. That's another clue, but no, it doesn't tell the whole story.

I am fully aware that you have to project an actual print under the correct conditions to see how the film actually looked. It's a pity that any surviving Eastman prints have faded.

*For example, the IB Technicolor prints have some unique color flaws, like some shots of the lightsaber duel having a blue tint. All of the bootlegs I've seen have a normal color balance in those shots, proof that even the IB prints don't provide a full picture of how the film looked in theaters in '77.

Erikstormtrooper said:

So, all these soundtrack revelations, aren't they the same as the differences on the Puggo Edition ESB (8 mm)?

No, they're not. For example, "Sensors are placed, I'm going back" is the same familiar take.

Anybody having trouble burning this? I tried twice, and though ImgBurn got through the writing part fine, when it tried to verify, it got stuck at 58% and gave me an "L-EC Uncorrectable Error."

When I re-insert each of my attempted burns into the drive, it detects the content as DVD Video, including the label "PSB", and when I play the burned discs' VOB files in VLC they seem to work just fine. And yet, when I try to verify either of them, there's a read error.

I've never had a bad burn on this drive that I can remember; is anybody else having a problem, or is it maybe something on my end?

hairy_hen is right, the 70mm mix is only very slightly different from the 35mm. The mix on this print is an entirely different beast, made *after* the 70mm 6-track and 35mm stereo mixes (just like the "Story of" LP and the Super 8 digest version). It's obvious that Lucas and Ben Burtt kept tinkering with the mixing throughout 1980, even after the 35mm version was out, and that they carried over some of their changes into the SE in '97.

As I've said before, some of the confusion about the content of the 70mm may stem from the fact that the 8mm, which shares some sound variations with the SE, used the 70mm cut as its picture element. However, the soundtrack was put together *after* the 35mm mix; for example, even though it has the 70mm edit of the Rebel fleet establishing shot, you can still hear some of the tracked "Yoda and the Force" music that was added when the scene was extended.

Anyway, back to the topic at hand: I've started watching PSB, and I haven't gotten as far as any of you, but I noticed another difference. When we see the snowspeeders taking off (after the scene of R2, "Imperial walkers on the north ridge", etc.), there is no music.

In all other mixes previously accounted for, the music starts during that scene of the speeders taking off. In this mix, the music doesn't cut in until the next shot of the walkers.

I had no idea that this was anything more than a mono fold-down of the standard 35mm stereo mix, until I got to that scene and thought "hey, isn't there supposed to be music here?" Then I switched to the stereo track and, sure enough, there was music there that's missing from the 16mm. I immediately pulled up this thread and saw that you all had already found other differences. Good work, everybody!

Here's another test. I used You_Too's corrected version of another one of -1's hi-res images from the colortiming thread.

First, the GOUT (and it's from the official released GOUT this time):

Now, here's my approximation of how that fits into -1's image:

Again, a bit more on the sides, a little more on the bottom, a little less on the top. And again, the aspect ratio of -1's image, when the proportions are corrected, is wider than the intended ratio.

Right-click on it and view it at full size, and notice the weird garbage on the left side; from the uncropped frames I've seen, the exposure level often "drops off" near the left edge of the frame. In this case, I wonder if the picture may be trailing off "sooner" (i.e., more to the right) than it does in the uncropped frames of other scenes that I've seen.

Here's the same image cropped to 2.39:1:

And here's what it looks like compared to the GOUT framing:

Yes, if the image is re-proportioned, then cropped to 2.35:1-2.40:1, it actually has a teeny tiny bit *less* on the sides than the GOUT.

I think that the cropping for the standard -1 release should be a bit more "open" than that. Here's a suggestion:

The red box is an aspect ratio of 2.35:1, the green areas are if the width were extended out to 2.40:1. This framing fits nicely inside the smaller of the two camera apertures in the film. Here's what this framing looks like on one of the shots with the larger aperture.

Unfortunately, any fixed cropping of the film has to lose that extra picture on the right side. If that picture area were included, the shots with the smaller frame size would have a black bar on the right edge.

I'm not sure, but the Death Star scenes I've been working on may be in the larger aperture, which could explain why they lose more on the sides.

For the main release, I'd probably go with something like the green box I drew, and then adjust the horizontal position so that the opening crawl is centered. (Again, I do think there should be a secondary "uncropped" version which shows as much of the frame as possible, so that we can see all that extra picture, even if it means that some shots have black area on the right.)

I guess I use V3 because it's anamorphic.

Maybe you're right, maybe I should use the official GOUT discs, because I have those too. I didn't know there was any cropping difference, however small.

Here's my first test with the high-res images: http://originaltrilogy.com/forum/topic.cfm/Cropping-the-Original-Trilogy-35mm-vs-dvd-gout/post/595901/#TopicPost595901

I think he's said that any version would have to crop at least a little bit of the picture for stabilization purposes?

I did another cropping estimate, this time against one of -1's high-resolution images from the color timing thread. First, here's what the GOUT looks like (or more specifically, DJ's V3):

Now here's the GOUT area traced into -1's high-res image of the same scene, scaled down to fit in this page, and with the 95% height adjustment to correct the proportions of the image (thanks You_Too for pointing that out in the colortiming thread). Right click "view image" to see it at full size.

As you can see, there's a decent amount more on the sides of -1's image, and a sliver more on the bottom, though the GOUT has a tiny bit more on the top (which is why the red box has no top edge - that area is not visible in -1's example image).

I think that the framing of -1's images from the colortiming thread is good. There's nothing wrong with it being cropped a little bit less than the official releases. Although, with the 95% height correction, it's obvious that -1's version is wider than 2.35:1, closer to about 2.50:1.

Here's -1's screenshot, with the proportions adjusted, and then cropped to 2.39:1.

And here's the same image with my GOUT framing estimate.

As far as the sides go, the cropping just about matches GOUT. I'd maybe open the picture up a little bit more on all four sides. At least, as much as possible without showing the black borders.

I do agree there should be a version that is as close to uncropped as possible, it's fun to see that extra image.

EDIT: none pointed out that DJ's V3 crops off a few pixels, so here's the same shot as seen in the official GOUT:

The GOUT is actually cropped a teeny tiny bit less than I've been assuming from the V3. It's a very, very, VERY small difference, but nevertheless, I'll take this into account and work from the official GOUT in my future comparisons.

negative1 said:

 

what you see is the original screencaps

from the english VOB version. i take it

you are using the same?

For GOUT reference, I am using my copy of DarkJedi's V3 DVD.

I'm going to do more examples of the GOUT framing as compared to your high-resolution images from the color timing thread. I'll link to them here once I've posted them in the cropping thread.

I'm not sure exactly what caused -1's initial overlays to be so off. Partt of it may be that some of those early test reel images looked distorted or keystoned:

Also, he didn't warp/scale the GOUT images properly. When I divide out the pixel dimensions on that supposedly heavily cropped shot of Obi-Wan, the GOUT image comes out to 2.76:1!

With a ratio like that, the GOUT overlay is clearly vertically squished. You can see that the images don't line up. It's even clearer in this shot of Han which was supposedly closely cropped:

None of this is intended as a slight to -1. He's doing amazing work, but this was earlier on in the project, and the bugs were still being worked out. Furthermore, my skepticism about these comparisons is what led me to do my own, which you can see in the "Cropping the original trilogy" thread.

Again, none of this is meant to detract from anything -1 has done. I'm just trying to more accurately determine how the GOUT was cropped, and consequently to help work out the best framing for his version.

Not sure, I'm not thinking in percentages. I'd think no more than 10%.

If I were to crop it, I'd at least make sure that the rounded corners on the right side are never visible. A 2.35:1 frame that falls inside those borders should work. And remember, there are two different camera apertures during the film. I'd make sure the cropping fits inside the smaller one.

One guideline I'd give you: Make sure that each line of text in the opening crawl is visible from end to end as soon as it scrolls up into frame.

Way back in page 1, -1's initial GOUT vs. raw cropping overlays seemed to suggest that the GOUT framing zoomed in as Han was chasing the stormtroopers down the hall:

Later, I contested this, saying that I didn't detect any framing change during the scene when I compared the GOUT against the Blu-ray.

Now that I have the red reel tests to reference, I can come up with even better GOUT cropping estimates for those scenes. Here's what I came up with for those two shots, using my technique of eyeballing where the edges of the GOUT image are in the uncropped frames.

Here are my estimates, by just comparing one frame to its GOUT equivalent:

The positions of the left and right edges don't match, but that's mainly because the combination of the motion blur of the scene, the DVNR smearing of the GOUT, and the compression of the test clips made it hard to find the left and right edges of the GOUT vs. this clip. But in general, it matches my comparisons from the reel 3 LPP images - the GOUT is framed toward the bottom and to the right.

Then I decided, why not make a more definite comparison using a more static shot, with more clearly visible "landmarks" on all of the edges? So here's a shot from earlier in the scene:

Again, it basically matches the "down and to the right" cropping of all my other GOUT estimates. Next, I decided to map this box over the images of Han, and here's what I came up with:

They fall in basically the same area as my initial estimates, further evidence that this scene is not as cropped as -1 thought, and that the cropping stays in one place.

Next, here's that shot of Ben walking down the hallway that -1 thought was extremely cropped in the GOUT:

I thought this seemed fishy - the proportions don't match (the GOUT overlay here has aan aspect ratio of 2.76:1!), the two images don't line up well, and I could not believe that much would be cropped at the bottom.

-1 just put up a test clip with this shot, so I decided to do another estimate. This one was even harder, because all four edges are dark, and the test clip is so crushed and compressed. Without any "landmarks" on the edges of the frame, I really had to eyeball it. Here's what I came up with:

As you can see, my estimate shows that it's no more cropped than any of the other shots I've tested. -1's initial comparison was just very inaccurate - no offense to him, of course. Those comparisons are what got me thinking about the GOUT framing in the first place.

Now, here's the same shot with the cropping estimate from that stormtrooper image mapped over it:

Again, it is very close to my first estimate, and it basically matches all my other estimates of the GOUT cropping area, framed toward the lower-right of the raw image.

And here's a shot from the uncropped, uncorrected test clip, with the approximate GOUT framing as a blue box:

Yes, the GOUT is cropped from the raw film frame, just like any video transfer. However, from what I'm seeing, it's not as severe or as inconsistent as it seemed from -1's initial comparisons.

The Blu-ray cropping is different than the GOUT. Some shots may have been reframed for the 97SE, or the 2004 version, but in general, it seems that the 2004/11 transfer is positioned a bit higher and more to the left than the GOUT.

EDIT: Here are the cropping estimates I promised: http://originaltrilogy.com/forum/topic.cfm/Cropping-the-Original-Trilogy-35mm-vs-dvd-gout/post/595678/#TopicPost595678

The more GOUT framing estimates I do, the more I see that the cropping is very consistent, and not as severe as your original comparisons imply.

negative1 said:

i'm not using the cropping diagrams.

 

this is a just a straight dump of the image

from the DVD VOB file, against the frame captures

we have.

 

I'm sorry, I wasn't clear. I didn't mean the diagrams you showed in that thread, I meant that your raw-to-GOUT comparisons are likely inaccurate. I just can't believe that isolated shots would be cropped that close. It doesn't match with my findings on how reel 3 was cropped.

My point was, I don't try to overlay images when I do my comparisons. I just try to find the edges of each transfer and draw them onto the raw frame. That way, I don't have to worry about the warping/distortion of the transfers. (As I said, I think the GOUT has a "curve" to the image, with the picture starting to get narrower on the sides.)

Do you have a test clip containing that scene with Ben walking down the hall?

I thought we determined that those cropping diagrams were not accurate due to the way the image is warped in the GOUT transfer not correctly lining up with the red reel transfer. (I think there's some kind of distortion in the GOUT transfer, where it's more squeezed on the edges than in the middle...)

I always based my comparisons solely on what was at the edges of the image, without doing any warping to match the two sources to each other.

Do you have a test clip of that bottom scene?

Looks good to me, except the images are also flipped horizontally - the soundtrack is supposed to be on the left side, not the right.

I'm certainly intrigued about Vertigo. I know that the restoration's remix was screwed up:

"One of the holy grails of moviedom, the restorers of Vertigo found it acceptable to record entire new footstep and incidental-noise tracks, and then mix them at relative volume much louder than they ever were in the original. Sure, it sounds good, but James Stewart didn't SCOOT SCOOT SCOOT every time he took a step, in fact often the Bernard Herrmann score was almost alone on the track for minutes at a time."

Pre-release specs for the upcoming Hitchcock Blu-ray set say Vertigo has a mono track, but there's always the possibility that it was a misprint, or that it's going to be another The Good, The Bad and the Ugly where it's just the 5.1 track folded down to mono. I'd get that Beta, just as "insurance."

And as far as the Paramount logo on Grease, that was another idea I already had. Just grab the logo from something like Star Trek: The Motion Picture and synch it up.

The 1997 laserdisc credits were video-generated. The credits on the 1997 digital broadcasts and the 2004/2011 DVD/Blu-ray releases are the ones that appeared on the SE theatrical prints. (The end credits were not changed for any of the films in 2004; instead, the new credits appeared as a supplement on the fourth bonus disc.)

I'm not sure why they altered the credits for ANH; was it to add those names, or just to make the credits easier to read?

I do know that at least the opening "fade in/fade out" sections of the end credits of ESB (and possibly ROTJ) were also video-generated on VHS/LD, because they forgot to add James Earl Jones to the SE credits the first time. (Consequently, Jones' credit was absent again in 2004.) The scrolling credits for Empire and Jedi were still the theatrical versions, at least as I remember it; I recall the ESB credit roll being almost illegible on the widescreen VHS.

The colors really remind me of the screenings of Close Encounters of the Third Kind and The Blues Brothers that I recently attended at the Paramount Theatre in Oakland, CA.

Both of them seemed to be modern LPP prints, and CE3K was the late-90s "Director's Cut" restoration, but they both had similar color timing to this. Lots of green and blue in the shadows/darker midtones, slightly off-white/pinkish whites, etc. (And like these images, they seemed to be a tad dark/dim, as if the projector's light source wasn't bright enough. No slight to -1, I know the brightness/contrast/gamma will be corrected later.)

My point is, with the higher resolution and improved compression, I can finally appreciate the quality of this print. I could imagine seeing an LPP print projected in a theater, looking like this.

I wasn't, I don't know anything about this technical stuff.

What I suggested was that someone could possibly do something like dark_jedi's Terminator mono project - that is, create a version that can be synched up with the official Blu-ray.

Just having a preservation of the original Dolby Stereo mix is enough for starters - as I said before, the '98 mix of Grease is a textbook example of both how different a 5.1 remix can sound from the original, and all the ways a remix can be screwed up.

OK, I looked up how to give SABnzbd the password, I downloaded the NFO first and got it from there. I've spent all morning *trying* to download this, first from MySpleen and now from Usenet, so I hope it works now.