logo Sign In

Williarob

User Group
Members
Join date
9-Apr-2007
Last activity
16-Jun-2025
Posts
915
Web Site
http://www.thestarwarstrilogy.com

Post History

Post
#1209316
Topic
<strong>4K77</strong> - Released
Time

Octorox said:

Williarob said:

The DNR version is coming along quite nicely I think:

Wow, those are a marked improvement. Can anyone explain exactly how the DNR process is being handled? It definitely doesn’t have the “waxy” look one might associate with DNR.

The levels of grain and noise in Star Wars vary scene to scene and shot to shot. So I split the entire film into shots and I’m applying Neat Video’s Denoise filter to each shot, one at a time, and adjusting the settings. Often, for example when characters are talking, a scene is split into three camera angles - over Luke’s shoulder, over Ben’s shoulder, wide shot and the edit cycles back and forth between the angles. Unless there are SFX in some of these shots, the same settings can usually be used for all of the shots from the same angle, so the neatvideo settings can be resused without having to set everything up again, which makes it go faster. In this fashion I was able to denoise all of reel 1 on Sunday.

Besides Denoising, I am also sharpening the image a little - removing the sharp grain makes the image look softer.

Finally, I regrain with nice clean (real - not emulated) film grain.

Post
#1208543
Topic
<strong>4K77</strong> - Released
Time

towne32 said:

pittrek said:

I’m just watching the 1080p version from myspleen right now and I absolutely love what I’ve seen so far.Great work guys!
But … is it really a non-DNVR version? I have noticed one weird shot - when the stormtroopers attack Tantive IV and are rushing through the whole in the door - it really looks like every trooper leaves a trail of himself while he moves, like that shot went through an extensive temporal denoising.

screenshot?

Sounds like a decoding error to me. Even going frame by frame I’m not seeing it. In any case, yes, I am 100% certain there was no temporal smoothing anywhere in this version. You’ll be able to see the difference (work in progress):

http://www.framecompare.com/image-compare/screenshotcomparison/FBEMNNNU

http://www.framecompare.com/image-compare/screenshotcomparison/KL6LNNNX

Post
#1208361
Topic
<strong>4K77</strong> - Released
Time

Yeah, all home video grades were rejected because most of them were regraded with neutral colors, and to fit within the color and levels limitations of NTSC, PAL and SECAM tube TV systems. If you have access to two unfaded, original Technicolor prints, why base your grade on an old laserdisc? I’m not saying such as grade is a bad idea, actually I think it would be awesome to have some nice home video colors at this resolution - that’s nostalgia right there, because kids of the '80s like me wore out those tapes and laserdiscs watching the films over and over so this is how I remember the film looking. But the goal of the project was to preserve the actual colors of the print.

Post
#1208265
Topic
<strong>4K77</strong> - Released
Time

ZigZig said:

To be able to compare 4k77 with other editions of Star Wars, I took the frame 5032 (GOUT-synced) of 6 different versions: 4k77 (2160p), SilverScreen Edition 1.6 (1080p), BluRay (1080p), Despecialized Edition 2.7 (720p), GOUT DVD (720p) and 16mm (480p)

I adjusted the proportions of each version so that they fit and the comparison is visually effective.

The result is available here : http://www.framecompare.com/screenshotcomparison/FBMFNNNU

In my opinion, the accuracy of the 4k77 version is similar to Blu-Ray and 4k is not really sharper than 2k (4k does not help much given the grain on the film).

This new 4k77 version has very natural and less aggressive colors than the BR or the DeEd 2.7, but the SiverScreen Edition(TN1) is still my favorite because it is less cropped and keeps the full width of the film (there is 11% more image than on the 4k77 version).

(I also compared the different versions laserdisc, VHS, VCD and 8mm, but I can only upload 6 images on framecompare).

Percentage wise, cropping is about the same as for the LPP - it was cropped to 2.35:1 which is less than the cropping at the cinema in '77 with was 2.39.1. The top and bottom and the junk on the right would have been removed by the projectionist too. It is also worth pointing out that the LPP used for the SSE had much more picture information on all sides than the tech prints:

http://www.framecompare.com/image-compare/screenshotcomparison/KL6WNNNX

Post
#1208255
Topic
<strong>4K77</strong> - Released
Time

GZK8000 said:

I have a few questions: what was the goal regarding the color grading? In the 4K77 blog, there are videos of both Technicolor prints, (the one Lucasfilm and Mike Verta has used and the cleaner one with worse colors), and the latter one was the main source for 4K77. So the idea was to mimic the look of a typical/ideal Technicolor print, or one of the prints used for 4K77? In which context (when screened in a cinema or dark room with 70s bulbs, in a bright living room)? I am curious about the blacks here (specially in many space shots), was it a limitation of the print used or what? I’m not criticising 4K77, the restoration looks amazing and for me it’s now THE version of Star Wars to watch, I’d like to know the technical details if it can be shared.

A single correction was made for each reel. In most cases it involved nothing more than white balancing the image using the optical track for the white point and then adjusting the contrast so that the brightest point on the reel is right at the top of the scopes and the blackest right at the bottom. With a single adjustment like that, you can’t make space or the end credits completely black or you will crush the blacks in other parts of the reel.

Colors and levels could be greatly improved with a shot by shot grade, but I wanted to preserve the original colors and levels as much as possible for this version. So the colors quite accurately represent the digital scan of the print, which isn’t necessarily the same as when projected in a dark room with a 70s bulb, but nor is it anybody’s idealized imagination of what it should look like.

Post
#1207180
Topic
James Bond - The Living Daylights - 35mm (Released)
Time

TomArrow said:

Awesome. I like these raw looking scans. It’s basically like the Grindhouse of the ROTJ and ESB scans, correct?

Is it available for sharing? I am not familiar with the etiquette here, pardon my ignorance.

Yes, it’s basically just a Grindhouse version - though apart from having plenty of dirt and dust at the reel changes, this one is very clean.

Send me a PM if you would like to check it out.

Post
#1206830
Topic
The Original Trilogy restored from 35mm prints (a WIP)
Time

dahmage said:

That speeder shot isn’t cool though guys…

But the Greedo shot was SE’d by adding some additional content at the end, which is just as easily deleted from the bluray. The rest of the scene is untouched. So I’d vote for “Look sir Droids” scene, or the Sandcrawler coming over the dune. Or the Speeder shot, but I think one of the other shots I mentioned might achieve more impressive results.

Post
#1206574
Topic
Movie Preservation and Home Media: An Opinion
Time

theMaestro said:

The Star Wars Blu-rays are a different case though, aren’t they? Because they seem to be the same version that the 2004 DVDs were based on (but with some further modifications). And so, if deadlines were what contributed to their botched colors, then it would be the deadline for getting the 2004 DVDs out in time for the holidays and to promote the last prequel movie. But what’s baffling is that the Blu-rays came out 7 years later…and still had the same color issues! You would think that they’d at least try to make some kind of effort to improve the colors. But nope, perhaps they were okay with all the characters looking magenta.

That’s true, it is odd that they didn’t fix it in 2011.

Post
#1206558
Topic
Movie Preservation and Home Media: An Opinion
Time

As far as modern films go I really don’t see any reason why the Bluray should look any different to the DCP sent to theaters, though if I had to guess, I would say that it is because the studios are probably handling the bluray releases entirely unsupervised by the directors. I wouldn’t be surprised to find out that the studios just take the DCP, apply a blanket yellow, teal, or teal and orange LUT to it, then apply some standard Edge enhancement and DNR and call it done. That way it looks like all the other modern blurays. The director probably hates it as much as we do when and if (s)he gets a look at it, but by then it’s too late, and they are busy with their next project anyway.

For older films I think it is a little more complicated. In the 1980s they would often just transfer an IP print to tape, with the levels and colors adjusted to work within the limitations of PAL/NTSC/SECAM/etc tube TVs. In the late 90’s and early '00s DVDs suddenly started selling like hot cakes and it was worth the investment by the studios to remaster as many of the old films as they could. Sometimes this involved A GOUT style rush job, where they just took the digital laserdisc master and slapped it on a DVD, while others took more care.

With the advent of HD video and bluray, digital scans from the negatives became much more common as a way to achieve the best possible picture quality - upscaling the DVDs just wasn’t good enough. However, by going back to the negatives the original color timing is lost completely and has to be recreated. Color timing digitally, rather than chemically, makes it much cheaper and easier to experiment with new “looks”. I like to imagine that the colorist for the new bluray is sat in a hi tech room with his digital workstation and a projector, screening a theatrical print of the film on a huge screen, allowing the colorist to try to match each scene as closely as possible… But I don’t think that happened very often. Perhaps they were handed the old DVD, or nothing at all. Perhaps sometimes the film’s director would supervise - but of course that doesn’t necessarily mean that the colors wouldn’t be “improved” in the remaster, made perhaps many years after the original film was released.

There is also probably some concern that the grainy, film like image will look out of place among all of the modern, blurays, shot and processed digitally from start to finish. Hence the need to scrub them clean, and sharpen them to hell.

Finally, as somebody who has tried grading scans of film prints, and encoding them, it is very easy for me to see how the addition of a fixed budget and a fixed deadline can lead to less than optimal results. Sometimes the encoder changes the colors, or the levels from your carefully created master, or you forget to change one of the settings, and a re-encode is necessary. But if the adverts have already been broadcast that this film will be available to buy on Bluray on this date - Best Buy already has the cardboard stands to display them, and the master tapes need to be sent to the factory for authoring, then your time is up… Now of course with professionals this isn’t supposed to happen, but I’m sure it does. How else do you explain the Star Wars Blurays? When the final encode was done, somebody must have piped up and said, “oh shit, these are as magenta as fuck!”, but they had to be in stores by October, so that’s what we have… Since then they have been repackaged a number of times, ample opportunities to fix the colors, but that would cut into the profits - and we bought them anyway, right?

Post
#1184084
Topic
team negative1 - star wars 1977 - 35mm theatrical version (Released)
Time

I’m pretty sure the LPP used for the SSE is a duplicate, so it’s 4(?) generations away from the negative.

Even if it isn’t a dupe, it’s still 3 generations away so it really shouldn’t come as a surprise that Bluray based sources have more detail and less grain than a scan of a release print. But like santakrooz says, release prints are what people saw in the cinema and Cinema prints were not struck from the original negative - contact printing was a destructive process, and if they created hundreds of prints from the negative it would have been destroyed. So they would create an interpositive, and use that to create a throw away negative to use for striking prints. When it wears out, they would simply create another. This allowed them to create hundreds of prints without damaging the negative. Sometimes a few prints would be struck directly from the negative for critics screenings, and special events, but these did not end up at your local cinema.

SSE = Original negative -> Interpositive -> internegative -> Positive print -> duplicate -> HD print scanner.
4K77 = Original negative -> color separation matrices -> Positive print -> 4K print scanner
Bluray = Original negative -> HD print scanner.

(According to Videography, the negatives were scanned on a Cintel C-Reality telecine, at 1920x1080 resolution, in 4:4:4 RGB, recorded to Sony SR tape in 2004)
Source: https://web.archive.org/web/20150219050324/http://secrethistoryofstarwars.com:80/savingstarwars.html

It would be really nice if Disney/Lucasfilm could dig up one of those original IPs and create a new 4K scan - it would be head and shoulders above anything else we have access to.

Post
#1181026
Topic
The Original Trilogy restored from 35mm prints (a WIP)
Time

RU.08 said:

RUN_LUKE_RUN said:

I honestly didn’t want to sound rude. If i did I’m sorry. I had a honest question and thank you for taking the time to post your comments to clarify. You have sacrificed so much so people can enjoy your work and i thank you. if i ever save up enough money for my home server i would be happy to lend my processing power to your projects. But for now cryptocurrency miners will continue to jack-up the prices.

I don’t think anyone was accusing you of that. To me it just sounded a little uninformed, not rude. I think the thing is there is a disparity of information between people who are involved (like poita), people who follow closely, and people who are new or don’t follow these projects so closely. In any case posting your enquiry in the forum rather than by private topic I think is helpful as it provides an opportunity for others who are thinking the same thing to get an answer. 😃

Exactly. Though typically, any post that begins with a phrase like “I don’t mean to sound rude but…” is usually followed by something that could be interpreted as rude - even the person writing the post can see that, hence the preface. In this case, I think you asked a perfectly legitimate question that I’m sure others were wondering about also, and I don’t think it was rude to ask. Those of us who have been on this board for more than a decade often forget that not everybody is as informed as we are, and lesser beings than Poita have been know to pounce on posts like this one with venomous wrath - which is totally unfair, but also human nature. If you had spent years of your life and thousands of dollars working on these projects, and then somebody new comes along and asks “why aren’t you doing this or that?”, it’s very easy to take offense and overreact when posting a response. I know I’ve been guilty of that myself.

So my response was really just to express my admiration for how Poita handled your question, in the hopes that others (myself included) might learn from it. If we want the next generation of fans to continue to care for and about the original trilogy, we need to continue to politely answer their questions, even if the answers seem obvious to us or have already been posted many times before.

Post
#1179889
Topic
The Original Trilogy restored from 35mm prints (a WIP)
Time

Poita I’d like to take a moment to thank you, not only for all you do for the community but also for continuing to take the high road in situations like this - it’s unfortunate that a lot of forum threads ultimately flame out because somebody says something rude, and others respond in kind. It’s all too easy to write a biting retort. Your thoughtful and reasoned reply is the kind of response that stops that nonsense in its tracks, restores perspective and allows discussions to continue. This is not the first time, I’ve seen you bite your tongue, and refrain from cutting somebody down to size, and I’m sure it won’t be the last. I applaud your patience, and wish the internet was filled with more people like you.

Post
#1172170
Topic
Dealing with People Selling Fan Projects
Time

J0E said:

dylanv1025 said:

LordZerome1080 said:

Found another one. https://m.ebay.com/itm/Star-Wars-Original-Trilogy-Despecialized-Edition-3-Blu-Ray-Set-SEALED-BRAND-NEW/162905430089?hash=item25edeb9849:g:FvoAAOSwIFtZ~TN1

Wtf, how did he get it “sealed”?

I work at a Party City and we sell giant rolls of cello wrap. It isn’t hard to wrap things in it, you just have to be good at wrapping and gluing. What would be REALLY impressive is if they were shrink wrapped.

Not really, I used to work at a place that sold software and we had a cheap, off the shelf, shrink wrap machine. It was basically just a bunch of cellophane bags, open at one end that we’d put the boxes in, and then use the attached “hair dryer” like device to shrink it and seal it. Took just a few seconds. I’m sure anyone can buy these replacement bags and use a household hairdryer to do the same thing.

Post
#1170291
Topic
Color matching and prediction: color correction tool v1.3 released!
Time

DrDre said:

lansing said:

DrDre said:

lansing said:

DrDre said:

lansing said:

Hi I ran into problem on color matching 2 color palettes, every color matched except that one on row 3 column 2, it always turn to pink on the result.

http://www.framecompare.com/image-compare/screenshotcomparison/0E1NNN8U

This is a discrete color matching problem. The color matching algorithm assumes images with a continuous color distributions with smooth color gradients, so it’s not perfectly suited for this sort of problem.

However, the explanation is in row 5 column 1, which in the source image has almost the same color as row 3 column 2. That color has to become a bright pink, and so row 3 column 2 also becomes pink.

So does that mean the problem can be resolved if I can sort the order of the color in the source image into one long smooth gradient before passing to the program?

It might be…

I isolated that two color and do the testing on the order, but the program didn’t run.

source:
https://i.imgur.com/58gYXRb.png

ref:
https://i.imgur.com/KmO21IJ.png

The algo matches color distributions. With two colors there is not much of a distribution, so I don’t think that will work.

Perhaps this explains why the program wouldn’t run for me when I tried to match the end credits of Star Wars.