Zyrother said:
Looks awesome! But, can we get a couple screenshot comparisons of V11 and V15? Would help to determine how much of an improvement it is.
Thanks!
Here's a comparison between SRV11 and SRV15 (I've reduced the sharpening somewhat):
yotsuya said:
DrDre said:
yotsuya said:
I think I may be closer. watching different un-color corrected versions is educational, as is watching other films. I watched Lawrence of Arabia and am partway through Zeffirelli's Hamlet. Accurate fleshtones are elusive, especially in something with as varied color timings as ANH. Still, I think I keep getting closer to my goal. I still have some tweaking to do, but I think I am even nearer than the last draft. This is definitely nearer to the scans/images of the technicolor, though that is not my goal, it is an awesome reference. I also just about have all my sources converted into files that I can bring in to Vegas so I can compare them. Before I finalize anything, I want to compare the JSC print, the GOUT print, the 97 SE EU broadcast, and the Lowry scan of the negative. I want to see if I can figure out what scenes may have had additional adjustments.
In any case, I have reduced the saturation and adjusted the colors, and changed some settings. Anything that causes any loss of image must be avoided and I had used something that had washed out some details. I think I am getting close.
In my opinion that last one is too desaturated. I've discussed the basis for my regrade in my own thread, but here's how I would regrade this shot:
What you have it indeed much better for that shot. If I was color correcting each scene, I would have something more along those lines, but I am trying to find a color correction plan that covers the entire film. When I up the saturation to make this shot look good, 50 others look bad. This is one of the key culprints-
I didn't realize that you wanted a single setting for the whole film. That is indeed a big challenge.
archimedes said:
Here's the prologue of FOTR EE. It has no sound, as I'm currently lacking a program to edit dts or ac3 files.
https://mega.nz/#!XZJlQAoJ!jMiHPa0JsLqjm_LAx_uNh-YFgH4T3zWnPvCEvCDV5xA
It's not a final file and still work in progress. So far I've spent 50 hours creating this file. I've used a single color space model with stabilization parameter 500.
Shots which are using blending are really challenging to convert. There is also additional noise introduced by this algorithm. Maybe a stabilization parameter of 1000 can fix this.
It's looking really good!
yotsuya said:
I think I may be closer. watching different un-color corrected versions is educational, as is watching other films. I watched Lawrence of Arabia and am partway through Zeffirelli's Hamlet. Accurate fleshtones are elusive, especially in something with as varied color timings as ANH. Still, I think I keep getting closer to my goal. I still have some tweaking to do, but I think I am even nearer than the last draft. This is definitely nearer to the scans/images of the technicolor, though that is not my goal, it is an awesome reference. I also just about have all my sources converted into files that I can bring in to Vegas so I can compare them. Before I finalize anything, I want to compare the JSC print, the GOUT print, the 97 SE EU broadcast, and the Lowry scan of the negative. I want to see if I can figure out what scenes may have had additional adjustments.
In any case, I have reduced the saturation and adjusted the colors, and changed some settings. Anything that causes any loss of image must be avoided and I had used something that had washed out some details. I think I am getting close.
In my opinion that last one is too desaturated. I've discussed the basis for my regrade in my own thread, but here's how I would regrade this shot:
The soldier shot has always intrigued me, as the color grading for this shot always felt off to me in any home video release. His face was always too red, and although the man was obviously ginger in his younger years, this is never reflected correctly in the color gradings I've seen so far. I think the automated correction of Team Negative1's preview turns out to be a good basis for some manual adjustments, and for what in my opinion is the way this shot should be graded:
Team Negative1 original:
Team Negative1 regraded:
Using this result to for a regrade of the bluray leads to...
Bluray:
Bluray regraded:
It's been a while, but here's SRV15, which is a slight improvement over SRV11, and will be used for the release version:
http://screenshotcomparison.com/comparison/146454
archimedes said:
Thank you, DrDre, this helps.
Another suggestion: Transfering a whole film shot by shot is very time consuming. Let's say you build a color model based on the first frame of the shot and transfer the rest of the shot using this model. For the next shot, you build a new color model, and so on. Right now, you have to sit in front of your computer, even though it's a very automitc process. So it would be good if you could specify the shots, the reference frame and the model type in a text file. The program would read this file and perform the transfers automatically.
Example: The input frames are called "inputX.bmp" where X is a number, the reference frames "refX.bmp"
One line for a shot could look something like this:
100,110,m,500,100,400
input100.bmp and ref110.bmp are used to build a multi color space model with stabilization parameter 500. After this is done, input100.bmp to input400bmp are transfered. The idea is to write several of these lines in a single textfile. The program can work through them without somebody having to be in front of the computer.
I'm actually working on a script for this, so v1.4 should have an automated shot by shot correction feature, although you would still have to make sure the cropping for the test and reference frames are the same.
dark_jedi said:
towne32 said:
Images. Though I believe now you can export a LUT and apply that to whatever segment of video you want in Premiere, etc.
Thanks, so nothing for Avisynth or maybe Adobe After Effects?
For the current version (v1.2) you can export an image sequence with Virtualdub, correct them, and import the image sequence again with Virtualdub, and save it as a video.
For the upcoming v1.3 you should be able to export a LUT that can be used in Adobe After Effects.
archimedes said:
Great tool! I'm currently using it to color transfer Fellowship of the Ring EE. It does a good job. However, I've run into following problem and have no idea what causes it:
Input
Reference
Result
I'm using a single color space model and have tried the stabilization parameters 0, 1, 500 and 1000. I get similar results for all of them.
In some cases the result can be very sensitive to the cropping, and it may help to select part of the frame to construct a model. This is what I get if I select the left part of the frame in the reference, and then the exact same part in the test frame:
Reference:
Test:
Test matched to reference:
yotsuya said:
Well, what I am aiming for is a global color correction for the Bluray and for whatever GOUT source I end up with. I want them to match overall. The things I am aiming for are skin tone, C-3PO, R2-D2, Darth Vader, the Blockade Runner interiors, the Death Star interiors. Leia is my typical target for skintone. Carrie Fisher has pale skin with some blush and when you have the tone right, her skin should look peachish and you should be able to see the blush without it being overly obvious. Considering the movie, in all four scans that I am referencing of it, is not properly timed from scene scene, you have to decide which scenes should be matched. I'm going for the majority. So I have gotten rid of the overly red flesh tones, but in the occasional shot the people are still a bit too red and in a few shots (mostly Tarkin) they are too pale.
My plan is to create my own despecialized version (hoping one of the newer resolution enhancers works well enough and easy enough) and individually color correct the scenes that I replace to match that scene. I have already done it to two scenes. C-3PO on the sand dunes in front of the skeleton has a green splotch under his arms that Harmy got rid of, but I had to match his color correction to what I am doing. Then I recolored the added Biggs scene because it was glaringly off. and I'm doing this on different layers so I have the entire film corrected and then what I change on top of that I can turn on or off as I like.
It is all a set of compromises. I want the films as they were shown, which at this stage is pretty much a guess. Only an original unfaded interpositive would give us the real colors as they were. The technicolor is the closest we can get, but it seems to come out too green unless you go in and color correct ever scene separately. Not a level of work I want to go to. I want to fix it so I can live with it.
I see, but in the case of the Tantive IV soldier there's only so much you can do manually. The skin still has an unnatural tone in your regrade, and there is a greenish hue to his helmet. This is impossible to get rid off mannualy, without introducing other color issues in the rest of the frame. So, it's not a matter of getting close to the prints, but getting close to your superior regrading of the laserdisc, which does not have the problems the bluray has for this shot. I'm saying you can get those good looking colors for the bluray, but not manually.
Here's a direct comparison between your two regrades, the first your latest regrade of the bluray, the second my attempt to match the bluray to your earlier regrade of the laserdisc:
yotsuya said:
I had a chance to watch the full version of my 3rd draft this morning and the results look good. A few tweaks left I think, but not much that most people would notice. Here are the same shots as above, but this time from the my color corrected bluray.
I think it's pretty much impossible to manually match the bluray regrading to your other sources. Your grading based on the laserdisc looked much better for this frame, which is oddy green shifted on the bluray.
I would advise you to use the color matching tool for such cases. In this case I matched the bluray to your laserdisc regrading:
Here's a little test I did on matching a frame for The Dark Knight bluray to 35 mm:
35 mm:
Bluray:
Bluray matched to 35 mm:
Here are three more screenshot comparisons:
Flexicon9 said:
Other than possibly a very slight touch of yellow, it strikes me that each shot is nearly identical in color to its counterpart... right? Maybe my eyes are playing tricks on me here. Are they as close as they seem?
The colors are actually quite different, although the difference is difficult to see if you can't compare them directly. Here are a few screenshot comparisons:
Yeah, that's very slow. It takes about 10 min to build a model for a 1080p test, and reference frame on my 2.6 GHz laptop, which has 6GB memory.
That 16 mm frame is a real challenge. The colors are very different, and it's pretty noisy. Perhaps using a temporal denoiser will reduce the noise artifacts. Also, colors can probably be matched more closely if the 16 mm frame, and 35 mm frame are cropped in the same way.
Here they are for the latest update of the regrade (bluray top, correction bottom):
I think there's still something a bit off. It's looking too much like an old print. I think this is better.
35 mm:
Bluray:
WOWOW:
Regrade:
35 mm:
Bluray:
WOWOW:
Regrade:
35 mm:
Bluray:
WOWOW:
Regrade:
Here are a couple of frame comparisons for the video sample that will be posted soon of the latest regrade (bluray top, correction bottom):
Here's how the latest regrade compares to the bluray, and the WOWOW for the other two 35 mm frames, adjusted to match the bluray's saturation and contrast.
35 mm:
Bluray:
WOWOW:
Regrade:
35 mm:
Bluray:
WOWOW:
Regrade:
There were still some things that bugged me about the regrade, among them the fact that I did not use an actual 35 mm reference. So, I decided to go back to the 35 mm references, more specifically this reference, which has a good white balance, and consequently does not seem to suffer from color shift:
However, it is too saturated, so I came up with a solution. By matching the gray scale image of the 35 mm frame to the gray scale image of the bluray, I can adjust the saturation and contrast of the 35 mm frame to match the saturation and contrast of the bluray, without affecting the colors. After this process it looks like this:
I then matched the bluray to this reference, and got the following result.
Bluray:
WOWOW:
Bluray matched to reference:
TheUltimate said:
COULD SOMEONE PLEASE SEND ME A F***ING DOWNLOAD LINK?
Glad to hear you're home, poita! Take it easy, as they say...
Colek said:
DrDre said:
Colek said:
Hello,
I would want to really thank you for this great software you made here, it will help me a lot in my future projects.
As I presume, I could use it to color correct deleted scenes to match the movie from other source, is that right (the color prediction correction)?
I would need to extract whole movie into .jpg and color correct them frame-by-frame using this, yes? For best results I propably should use new color correction model for each scene?
I tried it quickly with this movie I have been wanting to color correct and work on for a very long time now, but considering the colors it was really tough work for me. Your tool will really help me a lot :)
Source (horrible red tint)
image removed
HDTV Broadcast (IMO best colors of this movie out there, better than DVDs)
image removed
Color corrected
image removed
As you can see it producted a lot of artifacts on that spot there. What should I do about it? Try to match both frames even more to each other?Thank you very much again :)
The best thing to do, is to do the color correction shot by shot, although sometimes it works very well for multiple scenes or even a whole movie if the differences in color are very consistent.
The artifacts seem to be caused by some compression artifacts in the source. Increasing the stabilization parameter might reduce these, so try a value of a 1000, and then reduce the parameter if possible.
Thanks for your reply, I will try to set stabilization parameter to 1000 when I am at my workstation PC :)
You think I should firstly go with all the filters for the movie (denoisers, Super Resolution etc. using AviSynth for that) or color correction first?
Color correction first could be useful, because you can remove any remaining artifacts with filters afterwards.
Colek said:
Hello,
I would want to really thank you for this great software you made here, it will help me a lot in my future projects.
As I presume, I could use it to color correct deleted scenes to match the movie from other source, is that right (the color prediction correction)?
I would need to extract whole movie into .jpg and color correct them frame-by-frame using this, yes? For best results I propably should use new color correction model for each scene?
I tried it quickly with this movie I have been wanting to color correct and work on for a very long time now, but considering the colors it was really tough work for me. Your tool will really help me a lot :)
Source (horrible red tint)
image removed
HDTV Broadcast (IMO best colors of this movie out there, better than DVDs)
image removed
Color corrected
image removed
As you can see it producted a lot of artifacts on that spot there. What should I do about it? Try to match both frames even more to each other?Thank you very much again :)
The best thing to do, is to do the color correction shot by shot, although sometimes it works very well for multiple scenes or even a whole movie if the differences in color are very consistent.
The artifacts seem to be caused by some compression artifacts in the source. Increasing the stabilization parameter might reduce these, so try a value of a 1000, and then reduce the parameter if possible.
pittrek said:
This automated method looks interesting, I love the colours.
Could it be used for fixing faded 8mm film, e.g. something like this ?
Image removed
In principle yes, but it depends somewhat on, whether the film meets the underlying assumptions of the color correction model. One of those assumptions is that over the run of a reel/film the fading is similar across frames, and another is that over the run of a reel/film, each of the colors magenta, yellow, and cyan, are equally probable to occur. Consequently, the model needs enough shifted frames, to predict an accurate correction.
