Williarob

NeverarGreat said:

Excellent news! I wonder about this Tech print though, since Mike said he had 2 Tech scans, one in rough shape and one that was quite clean. Are we sure that this isn’t the same as the clean one that he is using? The only print I know of that is in better shape than that is the one used at the Senator theater screening.

Also, perhaps the ending music on your videos a tad loud? It’s kind of startling.

No, we got the better of Mike Verta’s two prints. The other one is in even worse shape. Nor is this the Senator print. Mike Verta does not have this particular Technicolor print (or at least he didn’t at the time we scanned it, and with Legacy completed I see no reason for him to have acquired it since). At his request I did send him the end credits from it because he was looking for some missing frames there, I don’t know if he found any use for it though.

And this time I can be 100% sure of all this because I have contacted the owner of this Technicolor print myself and have also arranged to borrow and scan some 35mm James Bond trailers from him.

Two new videos were posted in the last 24 hours, one on some of the new techniques we’re using to clean the film:

and another about the 35mm sources:

Anyone who was wondering, but are they using THAT Technicolor scan will be pleased to see that we actually have a BETTER technicolor print to play with than the one we were experimenting with earlier in the year…

Dreamaster said:

joefavs said:

Maybe some light discretionary grain removal though? SSE had bits that almost looked like television snow. Honor the spirit without hewing so rigidly to the letter, is what I’m saying.

I agree with this… the Tatooine scenes… it’s not “film grain” we’re seeing there. Please consider at least doing the “glob removal” on those scenes.

OK, when I get there I’ll give you some samples and you can all argue over which way to go at that point…

Cool. No more DNR version then.

Behold the magic and the mystery:

pittrek said:

The colours look really nice.
However there are some problems in the DNRed video. Watch closely the white door and the white panels behind R2 - it almost looks like some “mist” going from the bottom to the top. It’s not in the “no DNR” version

I’m thinking I might abandon the DNR version. I’m just not liking it as much as I thought I would. Would anybody here be too upset? If anybody needs to use scenes for Despecialization, it will be easy enough for them to add the DNR to the shots they need to mix with the blu-ray…

For your viewing pleasure:

https://we.tl/hf7Rf86pTq

h.264 this time, presented in both 4k and 1080p, DNR and no DNR. The first clip was shots 18-40, this is shots 16-70 and I included Hairy Hen’s stereo mix. I also made a third cleanup pass.

The color is based on the latest LUT that Dre sent me (which was last weekend, so it’s obviously 30 or 40 generations old!)

Thoughts?

These look really good. I like the colors too.

https://ifdb.fanedit.org/james-bond-007-never-say-mcclory-again/

yoda-sama said:

Williarob said:

Let’s hope so! I just went back to the drive with the original ESB scan on it and it just died. One minute it was readable and I was checking out the scans, the next it’s making some mechanical grinding noises and, well that’s that. Can’t run any diagnostics on it because it no longer shows up as a connected drive (tried multiple ways). Probably the control arm is stuck. I doubt you’ll be surprised to learn that it was a cheap Seagate drive. Also on there was the original LPP scan, but I had backups of that. I didn’t have backups of the ESB files (though other people do). So, I guess I won’t be testing out Dre’s tools on that scan any time soon. It might not matter though. I don’t actually know which scan was used for the Grindhouse version so it might well have been this one.

One suggestion that can help a bit (though I don’t think it would work long enough to get that amount of data off) is to stick the drive in a freezer for a while. That’ll often bring it back to life for a short while, but generally only good for one shot (doesn’t work over and over). Generally best for generating a list of contents of the drive, so you know exactly what you’re missing. Again, probably won’t do much good in your case, but it is a surprisingly little known trick that’s worth repeating.

Freezing the drive sounds like a bad idea to me:

https://acsdata.com/hard-drive-freezer-trick/

I already tried switching the controller card on the back of the drive with another Seagate 3 TB drive, made around the same time, but that didn’t help, and I opened it up and confirmed that the platters are clear and the heads are not stuck, so short of spending hundreds and hundreds of $$ to recover data that other members of TN1 should still have somwehere I think that’s all I can do. Another Seagate Drive will RIP.

AllAboutThatSpace said:

canofhumdingers said:

Just discovered this today. Amazing. Haven’t read the thread yet so forgive me if it’s been said, but I really hope ESB and ROTJ are planned to get similar treatment too! There’s been so many outright amazing projects on the original film in recent years sometimes it feels like the sequels are almost forgotten.

Nevertheless, GREAT JOB!!! Can’t wait to see the finished project. I really need a raise so I can start upgrading my equipment to both get and watch all this new stuff! I still have to trek over to a friends house just to burn a single layer BD…

I could be wrong, but I think Poita is the next port of call for a good ESB restoration. He’s already got some terrific tests up on his own thread.

Let’s hope so! I just went back to the drive with the original ESB scan on it and it just died. One minute it was readable and I was checking out the scans, the next it’s making some mechanical grinding noises and, well that’s that. Can’t run any diagnostics on it because it no longer shows up as a connected drive (tried multiple ways). Probably the control arm is stuck. I doubt you’ll be surprised to learn that it was a cheap Seagate drive. Also on there was the original LPP scan, but I had backups of that. I didn’t have backups of the ESB files (though other people do). So, I guess I won’t be testing out Dre’s tools on that scan any time soon. It might not matter though. I don’t actually know which scan was used for the Grindhouse version so it might well have been this one.

Blackout said:

Williarob said:
Empire on the other hand will require considerably more work. Perhaps Poita or somebody has a better print, but I don’t have access to any great scans of Empire at the moment.

A color corrected grindhouse would suffice for awhile.

The problem with that print though is that the colors are all over the place. Even within a single shot the colors flare from green to red often with parts of the same frame being wildly different. The fact is that the Grindhouse ESB has already been color corrected. TN1 spent months correcting the colors to make it look as good as it does, but short of photoshopping every single frame individually it’s just not going to get much better.

What I will do though, is go back to -1’s original 35mm ESB - the reels that are completely faded to red and that we really haven’t even looked at since about 2012 - and see if Dre’s tools can do anything for them.

Cobra Kai said:

DrDre said:

Here’s an update on the color grading to remove the remaining green/yellow cast:

DrDre, everything looks pretty great so far, although I was wondering about this shot. Most of the stars here appear to be crushed out. Is that just a by-product of the degraining process? The star field is slightly more visible on the non-DNR video sample, although it is still not as visible as the Silver Screen Edition, just for comparison, so I was also curious if perhaps the contrast needs to be dialed back a bit.

I know this is still a work in progress, so again, just wondering…
Thanks!

This is not a by-product of the cleanup, the de-graining or the color correction. I checked the raw scan and this particular print just doesn’t have such bright stars in this shot. It may come down to which do you prefer - a sharper star destroyer with more detail or brighter stars. Unless I can combine the two prints together somehow, or separate the star field into a new layer and restore it separately… I’ll have to experiment.

Actually I was planning to work on Jedi in 4K before this project, but I’m still waiting on the scans. Figured I’d make a start on this while I wait. Jedi should be a relatively quick project because we have several good scans and much of it is very very clean. If those scans arrive before I finish this project, I’ll probably work on Jedi and knock that out first, then come back to this later.

Empire on the other hand will require considerably more work. Perhaps Poita or somebody has a better print, but I don’t have access to any great scans of Empire at the moment.

LexX said:

Also wondering about the audio treatment here. I really wish you can add the hairyhen mix into this mix.

Since it will be a GOUT Sync’d MKV, expect all of the audio tracks and subs from the Despecialized Edition to be included. Some of The Eastman reels are Stereo and some are Mono, but neither set is complete.

Aye, the sample is missing a frame here and there, which is odd because there aren’t any missing frames in these shots. When I put all the shots into a single composition, I lose some frames, probably a 23.976 vs 24fps frame rate setting problem. In any case, I have all the frames for this particular sequence so they will all make it into the final render.

Go for it.

Attention grain lovers. As promised, here is the same extended clip as before, but this time without the DNR:

https://we.tl/jUPOL363xc

Beautiful ^ Seeing it at this resolution (to fit on the OT page), makes me think that this is how the GOUT could have looked… Forget 4K, let’s just render it at 480p - it’ll be gorgeous!

Blackout said:

Shuggy said:

Williarob said:

For those of you wanting a closer look at the difference between the Denoised and non-denoised version:

http://screenshotcomparison.com/comparison/188230

I know some of you will hate the DNR, while others will prefer it, which is why I figured I should make two versions…

Jedit: By the way, the DNR is calibrated on a shot by shot basis, rather than just applying it to the whole film with one setting. As we all know, some scenes are much grainier than others!

I’m firmly in the anti DNR camp so extra thanks is in order for your being so considerate to we grain fetishists. 😃

^

Quick question, why does the Final Result crop off some of the image?
http://screenshotcomparison.com/comparison/188245

The cropping in the example is not final and crops too much from top and left and I didn’t bother to make sure it remained at 2.39:1, but the final version will need to be cropped to remove the tape at the top and bottom on scene changes, the left side needs to be cropped because you can clearly see the line where the Eastman print ends and the SSE repair layer begins, and the right side needs to be cropped because there is still damage there. If you compare it to the blu-ray frame you will see that the cropping on the right is the same and the bottom is actually less.

http://screenshotcomparison.com/comparison/188248

Most of the prints I have are cropped slightly differently, with the LPP seeming to have more picture information on all sides.

DrDre said:

Here’s the official 4K77 frame (work in progress at 4K):

http://screenshotcomparison.com/comparison/188242

Love it!

AllAboutThatSpace said:

Wow, brilliant news!
Why shots from the blu ray though? I thought all your print scans are basically better in detail and closer to the original. Regardless I’m very excited 😃
Are you thinking of this as a step up from the SSE? How much will it improve on the SSE when viewed at 1080p?

The Blu-ray at least gives the illusion of more detail. A handful of frames at the reel changes in SSE 1.5/6 actually came from the Blu-ray - there was just no other suitable source available (except upscaled GOUT / blu-ray sourced Despecialized). In many cases I was able to interpolate / extrapolate missing frames using PFClean’s Fix Frame effect but the end of Reels 3 and 5 used a few Blu-ray frames. I did my best to make them blend with the LPP and at full speed it’s barely noticeable, but slow it down and it is much more obvious…

I do think this is a big step up from the SSE. This is much shaper (the Eastman prints are perhaps even sharper than the Technicolor), more detailed, will have more consistant colors, and won’t have any of that light streaking that our scan of the LPP had. If I had to put a number on it, I’d say 75% better than the SSE at 1080p. If it was only 10% better it wouldn’t be worth the effort.

For those of you wanting a closer look at the difference between the Denoised and non-denoised version:

http://screenshotcomparison.com/comparison/188230

I know some of you will hate the DNR, while others will prefer it, which is why I figured I should make two versions…

Jedit: By the way, the DNR is calibrated on a shot by shot basis, rather than just applying it to the whole film with one setting. As we all know, some scenes are much grainier than others!

Blackout said:

Are there any plans to color correct the uncleaned scans and put out a grindhouse release at at least 1080p?

At this point, even just color correcting a whole reel is a huge undertaking. Assembling a full cut of the film from dozens of Hard drives and GOUT syncing it would also be a lot of work, requiring additional disk space that I just don’t have at the moment.

Let’s wait and see how Mike Verta’s pitch at Disney goes (if he ever gets in the building to deliver it). If he succeeds, then this whole project becomes unnecessary. At that point, a grindhouse version would be the way to go. In the meantime I can assemble the Grindhouse version in parallel to the cleaned up version which will be considerably easier.

Beber said:

That looks very promising. Will you render this at true 4K or will it be UHD? I’m asking this because if this project ends up looking that great as it starts, I will seriously consider renting a theater in Paris for a private screening, and a true 4K DCP would be better for projection purposes.

I’ll probably render it at full 4k, though it will probably look better at UHD resolution (and fantastic at 1080p). As I said, I don’t have a 4K monitor, so most of my work is done at 1080p. Whenever I zoom in to 100% I see a heck of a lot more dirt that could be cleaned, as well as flaws in the image caused by the cleanup process - more on that later.