I’ve added a bit more yellow, and reduced the contrast somewhat:
Looks like a really nice compromise between your version and Neverar’s. Good stuff.
Well, one major point of correction is that there is no TN1 footage in Jedi 2.5. That’s a different scan, the whole of which you can watch in Harmy’s Jedi Grindhouse release.
I don’t know about those specific shots, but if you see the same exact motion in the Boba shot in a GOUT release, presumably it’s present in the negative. It could have been stabilized for the SE (I don’t know if the exact shots you’re talking about are present in the SE) I’m not familiar with the shot. If so, you might be the first to have noted that change. As described in the first post, this thread functions best if you provide a time stamp of what you’re discussing.
Gate weave in scans can be a pain to remove. Poita’s process seems incredibly stable. But for those of us using other less sophisticated techniques (I am pretty sure Harmy would have used some Adobe type software or else an avisynth script), the results can be quite variable, especially if there is a lot of motion on the screen. But it sounds like this might just be part of the original film.
I’ll stand corrected on that. But honestly wouldn’t have expected that to be the go-to source.
Well, the broadcast sources look pretty good (at least, in the days where 35mm sources were rare they did compared to GOUT). There just aren’t a huge number of changes that it is useful for. The melted bars in ANH and Anakin’s Eyebrows come to mind as other 97SE sourced materials.
Presumably Lucasfilm started cropping those cockpit shots in the later half of the film, for the 97 release. And then they finished the job in 2004?
Harmy, will the shots in the Millennium Falcon cockpit that are overcropped on the BD be replaced with 35mm scans instead of the 97SE in the 2.5 update, or re-done from scratch in 3.0?
The current v2.0 used upscaled GOUT footage to repair the cropped cockpit scenes, not 97SE footage. Harmy has said he’ll be fixing those edges with 35mm scans for v2.5.
Many of the cockpit zoom outs are done with 97SE upscales. source: https://plus.google.com/photos/109609428403596349302/album/6044645069501365185/6044645451698333346
edit: Looks like the earlier shots in the film were done with 97SE, the later ones with GOUT: https://plus.google.com/photos/109609428403596349302/album/6044645069501365185/6044645763911390130
Thanks. Most of that is of course Harmy’s doing. I’d say lucasfilm could do a fantastic job, or at least produce something that only needs a few more tweaks.
But I do think Jedi is in the best state by a long shot. With better rancor source material and a 1080p pipeline it’ll be near perfect.
SW still needs some GOUT upgrades, especially in the final battle. And some of the 35mm footage could be better than the SSE print (especially with the lower bitrate 1.0).
I think ESB probably looks a bit better or more passable than SW 2.7. But part of it is because the changes were less severe in the first place. And the remaining recomped SE shots are nothing offensive. Poita’s new scans are incredible, so 2.5 should be amazing.
Indeed it is. It can be seen through the whole trilogy in many different shots
Figured that was the case, though I never went looking for them. Pretty clear in that one, though! I do recall hearing about the efforts of removing all those reflections from the (shinier) 3PO of RoTS.
Totally off topic (I like both of your color corrections), but is that a boom mic (and presumably camera equipment) in the reflection on 3PO’s head? Never noticed before.
Yeah, thanks. I recently I noticed that too. Looks like that host went under. I’ll have to see if I have the images somewhere still. They should be on some HDD.
The first post is ALWAYS up to date 😉
Nice. It would get very confusing otherwise. 😃
Is this going to affect Adywan’s edit?
This time I say it with sincerity.
I feel like I’ve been dosed with something I didn’t pay to be dosed with reading these last few posts. wtf is going on? R word aside, was that actually a reply to that question? Are these just bots posting random messages from other threads?
Is this going to affect Adywan’s edit?
(sorry)
The blades look okay. But the contrast of the rest of the image makes it look way too washed out. Version 1 was better in this regard.
Also, it’s good to do color and contrast adjustments on just the picture area, with the black bars cropped out. Then, they’ll be added back when you render later. That way they’ll stay black throughout the film, instead of switching color and contrast. But in this case, leaving them there has helped depict what’s off about the contrast.
The file is a little smaller, but I couldn’t detect any difference in quality between screenshots, even zoomed in.
Ah, yes. That’s just due to the differences in overhead space in an m2ts vs mkv container. Perfectly normal. And perfectly frustrating if you’re converting in the other direction and think you’ve got something that fits on a disc…
Agreed, Harmy. It would’ve saved me multiple steps to get where I am now at least. As it is, I’m working from a new, movie-only ISO I made through tsmuxer, so up to that point nothing’s actually been reencoded. I tried MakeMKV, but I haven’t done a screenshot comparison yet. Looks good from what I can see though, with grain apparently untouched and a file size of 29.9 GB.
If your computer didn’t spend hours encoding anything, and it only remuxed it, you don’t need to worry about video quality. As long as all the audio is there, and the two videos (crawl/rest) are playing nicely, it sounds like you should be in good shape.
On that note, look at Neverar’s color thread to see what SW 3.0 might look like. Harmy will be using it as a base for that project, though he could implement changes on top of Neverar’s.
Where can I see the Neverar’s work ? 😃
http://originaltrilogy.com/topic/Neverars-1080p-Star-Wars-Color-Correction/id/16256
I forget if the first post is up to date, but it’s throughout that thread.
I don’t understand if the real color of the real movie is that we can see in the 2.5 or if is it more the 2.7 who have real colors ? Because in the 2.5 the picture is always yellow, but may be that is the natural color of the original movie (or a degradation of time on the pellicle). I have noticed the same thing with ROTJ 2.5.
I wonder if one day an official blu ray comes out, if the color would like the image of the 2.5 or more like the 2.7And when the 3.0 despecialized will come out, it will take the same color traitment has the 2.5 or the 2.7 ? (or neither of two ?)
The “real color” of the movie is a topic worthy of a million pages of discussion that will get us not much closer to a real answer.
2.5 was colored with the idea that the film references that were used had already been corrected to match the way the print looks. This turned out to be a misunderstanding, and the references images were far more raw than that. On top of that, technicolor prints have other issues and might not be considered perfect representations of the color, despite the fact that they have not faded.
Knowing this, I altered the color to make 2.7. I don’t claim that it is more correct, as references such as scanned print projects were used only loosely, and even those are graded by eye and feel. The goal was really just to undo a lot of the yellow/green effect that had been over-utilized based on the misunderstanding described above, as some of the shots stood out very strongly for me and I personally found those shots to be a little bit distracting, and not at all how I remembered the film (can of worms comment), but also to fit alongside ESB and Jedi despecialized, which had more natural tones. That said, plenty of people prefer Harmy’s 2.5, and it provides it with a rather pleasant ‘old film’ type of look.
The project eventually escalated to a shot by shot color adjustment project. I aimed for consistency, and sometimes achieved it. I tried to remove the yellow look, mostly to fit my own tastes and make it a bit more neutral. I did not counteract the very vibrant blues of the interior/exterior death star. I think it looks great as vibrant as it is, but it might technically be a bit much. And from the looks of a certain projects, it will probably be a bit more tame for 3.0.
On that note, look at Neverar’s color thread to see what SW 3.0 might look like. Harmy will be using it as a base for that project, though he could implement changes on top of Neverar’s.
No one knows the “real color of the real movie”, really. If an official Blu-ray comes out, it will be yet another interpretation of the real colors.
Yes, that’s the TLDR of it. I’m not actually too picky about color. As long as it isn’t as botched as the blu-rays, I will most likely be happy. Often in the threads containing people’s color experiments, I think the shots look great after the first adjustment or two. After that, there is usually another 5 pages on a given shot before moving on, but I think most is accomplished before those minute and never ending tweaks.
If a new blu-ray/UHD looked as good as the 1997 color, I would be happy.
Did anyone else feel like the dub was out of sync sometimes on 2.0? It could be from the original production, but it just feels inconsistent to me. Some scenes the audio sync seems spot on and other scenes it seems a frame or two off.
Maybe I just need more sleep though. I’ve seen this film probably a thousand times but I’m not used to scrutinizing it in HD.
I believe ESB was a frame off from GOUT sync (though someone may correct me about that). I never noticed an issue with sync, but then I would never be capable of noticing something being out of sync by one frame.
Yes, I still have a copy of v2.1… I remember the Ben Kenobi desert scenes were adjusted because they had been a little blown out to look like the IB Tech print from that theater screening and then Harmy sort of dialed that back to look less so. That’s around the time the whole '70s look was coming into the presentation…
[IMG]http://i64.tinypic.com/f1iogj.png[/IMG]
The colors have been adjusted quite a bit since this example was presented.
Here’s a gallery showing what was changed from 2.1 to 2.5. Not sure how to get the newfangled google photos to actually show a gallery, but you should be able to click through them or press the right arrow.
https://plus.google.com/photos/109609428403596349302/album/5932148501374386753/5932148505580432450
I thought there also used to be a 2.0 to 2.1 gallery, but I can’t seem to find it.
Edit: no, maybe I’m just thinking of 2.0 (i.e., remastered). Am I correct in remembering that 2.1 is essentially 2.5 without the few little fixes in that gallery? It was 2.0 that had some of the really bright desert shots (and of course much less 35mm source).
Here’s the gallery from 1.0 to 2.0 (again, you need to click left or right) https://plus.google.com/photos/109609428403596349302/album/5781873106061439201/5781874268260252418
I managed to make an MKV with the new 81 crawl, but it came out to a total of 20.3 GB. I used Handbrake, with video set to lossless with no crop, and the audio set to passthru (I used all 3 DTS-HD tracks from DeEd 2.7 for consistency). The starting file was 33.9 GB. It looks good from what I’ve checked, I can’t really detect a difference without going side by side, but I’m not sure why I got so much smaller a file.
JEDIT: I did a screenshot comparison and it looks like my MKV softened the grain a bit. I’m not sure what I did wrong though.
Well, re-encoding and reducing the bitrate will lose detail including grain. If it was actually lossless in handbrake, the resulting file would be enormous.
This is why people have tried to make MKVs without re-encoding, though clearly it gets messy trying to append the crawl cleanly.
I think that confusion is inevitable, whether he makes the change now or for 3.0. I guess waiting until 3.0 would give the 2.x releases a bit of symmetry, but that’s fairly trivial as long as the audio is made easily available.
There is not a version 2.7 for either V or VI. It is unlikely that there will be.
Speaking of versions, that reminds me of a few questions I’ve been meaning to ask about the DEED Project as a whole.
I remember that ROTJ 2.5 dropped on 2/7/16. But I was still pretty new to this site at the time, and hadn’t been around for the initial release of 1.0, or any of the versions of the other films for that matter.
Were the all iterations of the DEEDs prior to that point as follows?:
SW 1.0
SW 2.0
SW 2.5ESB 1.0
ESB 2.0ROTJ 1.0
And if so, out of curiosity about the past DEED project timeline, what were the dates Harmy initially posted each of those?
That’s pretty much it. RoTJ had a barely distributed temporary v2.0, but it was pretty much just for previewing and then became 2.5 by the time it was actually released.
I don’t know about the exact release dates (and it would be just as much work for me to dig through the threads as you 😉 ). It’s possible that looking up the releases on myspleen for the NFO/readme would be informative. But they’re doing their usual thing of showing an error because they are over their limit of 5-10 people on at a time, no doubt. 😛
There is not a version 2.7 for either V or VI. It is unlikely that there will be.
Further info from Hidalgo’s Twitter:
- Someone asked him if George made 4K versions of 5, 6 & 1. His reply: “@Sora96 @RealCGallant @druzod1012 No one’s said anything about those, so I’m not gonna be one to do it.”
- Hidalgo either doesn’t know about a 4K saga / OT / OOT set or isn’t saying
Well, at first I thought it was going to be the former, with “No one’s said anything about those”, but “I’m not gonna be the one to do it” clearly sounds more like he’s not going to be the source of a leak on that one. It doesn’t mean anything exciting in terms of it being pre or post SE. Just that he’s commenting on IV because it was already out there by Edwards (and the Reliance reel, presumably). Perhaps some progress was made on V and VI as well, or they were at least commissioned?
Definitely an improvement.
JEDIT: I don’t see why ripping an MKV would be problematic, it just takes a program that can connect multiple files together like what I did. A little tinkering would show which file is which.
It’s possible that some of the previous MKV upload attempts had a long pause between the crawl and the second (main) video sequence.
![http://i64.tinypic.com/f1iogj.png[/IMG]](http://i64.tinypic.com/f1iogj.png%5B/IMG%5D){kind=link}