You have absolutely no idea what you’re talking about.
You know in Star Wars where the actors have really red faces… That is what i am talking about.
So, how do you plan on playing back this finished color correction? Do you have a theater quality projector? If not, your aim is without merit. You will have to convert and then convert it back to RGB to do anything with it so what is the point in converting it in the first place? If you start with an RGB source (GOUT, DVD, and BR) and you are going to watch it on an RGB device (Any color television ever made) with an RGB file type (as DVD and BR require), I really don’t see the point of your exercise. You are trying to recreate the wheel and there is no reason to. Calibrate your monitor first. Then stick to RGB (what your source and end viewing will be in). That is the colors our eyes see after all. Unless you have calibrated professional equipment designed for theatrical projection, there is no reason to change the color space. If it starts out RGB and ends up RGB and gets to your eyes in RGB, then … it is an exercise in uselessness. Especially without a calibrated monitor. Yours is obviously way off.
I can not seem to get it through to you but I will try again.
Anything RGB will display the YUV or Ypryb colorspace correctly.
The filter to do the hue shift though can not be done from the roots of Red Green and Blue because the math is off that is why it creates the weird / odd error. Womble did have right sort of hue shift but the filter itself was poor.
What ever filter is in VLC is the right filter to use for hue shifting the Gout which Lightworks does not have only strict RGB root shifting.
Any hue shift does not require calibration because it is relevant only to how the footage or file is. There is no added or color taken away only shifted and the aim is to shift it so that it is not off base i.e. shifted away from where it would be normally. Correcting for the shift that it is plagued by.
you can watch it on any RGB device no problem.
I can’t say it any more times I am talking about a video filter and how it handles and trying to find away atound the problem, even you said yourself you can’t use it… I say you can if you use the right one.
Anymore of this sillyness, If you still can not understand that the filter acts differently in VLC than root shifting from RGB then I just at a loss of explaining this to you. This conversation also has nothing to do with film prints. It is a practical based thread on why color hue shifting is not a bad thing if done a differenh way.
So not only are you not calibrating your monitor before making color adjustments but you aren’t even encoding properly? You’re playing it back and recording it with FRAPS? I feel like we’re being trolled now.
this would only be for a reference. But I can’t seem to find any software that is not using RGB filters.
So it would only be for a color match sample. The purpose is to create a reference version outside of RGB filters. For which if you like is a way around the problem of an RGB hue filter (wrong math)
I am trying to find a way to not use the RGB hue filter within a NLE.
The reason I suggest Fraps capture as it will capture from a window view and you also can have the filter control visible amongst any other windows needed.
But it is the only way I can see of doing this to create a reference. It is Hackish and it is unauthordox like I said. There is possibly a better way in better software may do this but it is unknown to me.
Ok in some brief spare time I tested Womble Dvd Wizard and although the filer is yprbr colorspace it is just not a good piece of software as the filter was correct but just was innacurate and errors occured compared with the VLC media player filter.
So to offer up how I am considering the Solution now is as follows.
Load Gout in VLC.
Set hue shift in VLC media player to desired hue and capture to AVI in Fraps. this would entail having a large number of AVI clips to piece together in Lightworks and unauthordox but I can’t see anh other way of doing this. For any Color change only the color wheel should be used as not strictly RGB aside from master brightness / contrast / Gamma and saturation. Sort of creating a rule book at the moment about it.
unless for instance you can set the colorspace hue filter on programs like adobe after effects or similar. This I am unaware of capabilities of this software.
But I would vouch that once an assembled version is shifted where needed this is then color matched to a better source with Dr Dre tool or another color match filter.
This is as far as I got, as I am now not able to study any further at this time but shame about womble though.
So, one issue I have with what you are saying, Ronster, is that you really haven’t researched between the generation of the new interpositive the DE/Faces/Gout transfer was made from and how it ended up on the DVD/HD/Blu-ray digital print.
Here is what happened. Between when that interpostive was struck and 1993, the original negative of Star Wars faded terribly. Now, if you watch all the OOT home video widescreen versions, you will see some interesting differences in the color that could be the result of early fading, but overall they are very similar. For the Special Edition in 1997, they were working with a negative that had to be retimed. IN addition, they replaced a number of effects shots, added some effects, added some new shots. When you look at color timing, it is the process where each shot is timed when generating the interpositive. From there the prints are struck. So what you are seeing in both the GOUT and SE home video and broadcast is the telecine of a correctly colortimed interpositive. When it was scanned in 2003, it was not color timed during the scan, but was digitally color timed post scan. This resulted in all the original color timing to be lost. Not only that, but different batches of film faded at different rates so some shots are really badly faded while others aren’t. Some of the film had gone to Tunesia and back. Some to London and back, and some had never left California. This was taken into account in the color timing for the SE, but it seems not for the 2003 scan (which was done at Lucasfilm). Reportedly both the SE and the 2003 scan were timed to the Technicolor print Lucas has, but the results of both are questionable. It doesn’t appear that TESB or ROTJ suffered from the same fading as Star Wars.
To find what the colors should be we do have the telecines of the pre SE version of the film. But we are not just reliant on those telecines to find the original colors. Star Wars was one of the last films to have Technicolor prints struck. We not only have scans of Technicolor prints, but of other prints. When you put the Japanese Special Collection, Special Widescreen Edition, Definitive Edition/Faces/GOUT, foreign widescreen editions, the 16 mm prints, the 35 mm prints, and compare them to reference photos, photos of props, models, costumes, sets (or the occasions when people have been able to actually see some of those such as the costume exhibit that has been going around the country), we can arrive at a very good estimation of what the original colors were. We can see how well or bad the 2003 scan/color correction/cleanup was done on a shot by shot basis. Some are easy to recover. Some are not. Some are the result of the digital noise reduction, which removed not only actual dirt, but smoke as well.
So constantly saying the effects shots are the problem is not accurate. A lot of the effects shots have more accurate color than the no effects shots. A big reason was that the original negative of the elements were not as overused as the fully edited original negative. Some effects shots were never touched and are the same now as in 1977. Some were replaced in 1996 and some in 2003. I’ve noticed a mixed bag as to the quality of the special effects. For instance you have all these shots of Vader or an imperial pilot in the Tie fighter cockpit set that were filmed under similar circumstances and yet the red lights surrounding them come out different. A lot of those are original and not SE.
Now the inserted SE footage has been noted, even in the 1997 telecines, to be off. A bit too red and not quite matching the surrounding footage. I never noticed how bad the 1997 version was in the theater, but I think it is fair to say that the telecine is pretty accurate. We should soon have a good scan of a 35 mm print of the 97 SE to really see how it was in theaters.
So it isn’t the effects not fitting or the new footage not fitting, it is that there are so many sources and the SE just added to the number. The blu-ray really has to be cleaned up one shot at a time. It is a lot of work. The only reason I tried to do a global correction was to find the best way to fix the major issues and have a basis to take those fixes further in the necessary scenes. Some need more contrast, some less, some more color, some less, some need to be brighter and some darker. About the only thing I haven’t needed to mess with is the darks. When I compare it to the film scans, it is comparable. But the hue is not something that needs to be changed. I tried that and the results were not good enough. I wasn’t happy until I scrapped that and focused on balancing the colors. What you are seeing is that the yellows are not well represented in the blu-ray and pushing the hue one direction changes the reds to yellows and looks like it fixes the problem, but it doesn’t. It just creates other problems. You have not found the magic fix. There is no magic fix. You have to fix nearly everything. Especially in reels 1, 2, 3 and 6.
Well I think it was a bit confusing why the filter was not working as intended.
Are you using a HSV filter in the correct colorspace yprbr?
no it’s all interesting stuff and to be honest I like the way the old transfers look for obvious reasons, But I also think they can look better.
there are 4 combinations.
1.Hue Shift
Change Color
Hue Shift and Change Color (When you find one of these it is very cool in fact the most fun shots to do really)
Do Nothing at all it is correct.
Essentially I have no time at all to do a project perhaps until December. But I am happy a big hurdle has been overcome.
In terms of the 97 laserdisc Special edition I probably only want the Special edition Special effects Shots Only from Battle of Yavin, well everything after the turret / tie fighter attack. The begining as Dre said looks good in places but generally this is a step back from the GOUT in a way.
So yeah it’s on the back burner but I have no time to do anything at this moment.
basically Laserschwerts special edition trailer is the same color for special effects in 97 broadcast version although pan & scanned.
If the print special effects shots does not resemble exactly laserschwerts trailer then we have something unique.
Does the special edition Laserdisc resemble 97 broadcast version?
Look in the Archive…
Yes, they are from the same source.
Ok So basically I want to re-create 97 (broadcast?) version without crap horrible stuff at mos eisley but in yprbr colorspace.
Did anyone ever capture 97 pal laserdisc?
basically Laserschwerts special edition trailer is the same color for special effects in 97 broadcast version although pan & scanned.
If the print special effects shots does not resemble exactly laserschwerts trailer then we have something unique.
Does the special edition Laserdisc resemble 97 broadcast version?
Look in the Archive…
no I agree the film itself restored was combined with the special edition shots… But that does not mean the fx shots were designed against the print itself. As the restored print was probably was not available to design the fx shots to match to.
It seems to match the laserdiscs more so or broadcast telecines in my opinion and that colorspace.
This explains the discrepency i feel.
OK but does George Lucus have a calibrated monitor?
I imagine they designed them to match the gout master.
You imagine far too much.
Well I don’t imagine so far as the 97 broadcast version is concerned they do match up.
So this does mean it is not how i remember it in 2004 and also it feels like what I have been trying to nail down what has happened… Do we have mixed colorspaces within the same film as of 2004. Look at the difference of the green laser color and stuff weird stuff going on similat to color error of using filters in wrong colorspace that well this thread proves you need to be in the right colorspace to use filters effectively on footage so as you get a true result or intended result.
Is this part of what has happened post 2004?
OK but does George Lucus have a calibrated monitor?
definitely not… He has a calibrated laser gun.
I don’t think the special effects were designed to govwith restored 1977 footage.
I imagine they designed them to match the gout master. As that was the Master prior to any other version.
Here is how I feel about it.
Theatrical version preserved good job. I have nothing to add to this I have no desire to interfere with what ever you lot are up to.
I would however like to have a version that is very analogue looking similar to what i grew up with on TV. It’s not to hard to understand I have said this multiple times and i would like to try and recapture that. The 97 TV broadcast version I think is very good at the end of the film, I can not speak for the start really.
Do the laserdisc or VCD of special edition resemble the broadcast version? The thing is and the common complaint is that the special edition special effects shots do not match the film at all.
But if you watch the 97 broadcast version they actually do match up pretty well in my opinion. This is because the shots were not designed to go with the new Master in 2004 but another master and the broadcast version seems to me a good match. But yeah I hate the mos eisley parts and stuff horrible no matter what.
I’m going to try one last time. Telecines, and raw scans are not reliable color references. They capture what’s on an interpositive or print, but the colors are going to be specific to the apparatus and settings used for the scanning, and in case of a telecine to the person doing the telecine. They do not represent what´s on a print, unless the scan or telecine is corrected to roughly match a projected print. So, the telecine´s generally do not represent what was put on the o-neg, or what was seen in theatres. It may look pleasing to the eye, but that’s another matter.
Unfortunately the fact of the matter is…
If you hue shift the Gout in the correct colorspace what you get is not “revised color timing” But “fixed color timing” and it is practically equal and identical to the Special edition pre 2004 i.e. 1997 Special edition released to theaters.
How do you know this? Did you compare it to a projected 1997 SE print?
This is the point. George Lucus sat there and ordered the hue shifting fixed for this release , but now we are back to square one again post 2004.
So sorry for liking older versions. After seeing 1997 broadcast version of the special edition. It very much makes sense. But come 2004 until present that is like a mish-mash version.
It is theatrical print coupled with the special edition.
The Special edition though was not based on the theatrical film print but a telecine or the Gout version master. Or atleast the broadcast 97 version looks like a good version of gout master.
The 1997 SE broadcast version of the special edition for ANH is one of the worst telecines of any made for the OT. The color balance shifts constantly even within a shot, and large parts of the early parts of the film have a marked blue cast to them.
ok I only really was comparing the detail in the special effects shots initially and this is good although it is pan & scan.
But yeah I noticed how different it was and I thought this looked like a modified Gout master paired with the special edition effects. I honestly did not spend much time looking at it. But I think there is something right and something match up between the special edition effects and this broadcast version. It may be bad in your eyes but it may just have some problems.
It does seem that the general consensus is TV broadcast = bad looking not desireable and always wrong.
the only thing lost here is the master that the special edition effects made by ILM were to go with.
Now we have special edition effects paired with a new master that they were not designed to go with.
I don’t think the way it is now match at all rather than how it was designed to be originally in the Special Edition.
I did notice laserschwertz trailer for the normal film footage though looks nothing like the 97 broadcast version. The special effects shots however are accurate.
97 broadcast version looks like Gout Master and updated special effects.
Unfortunately the fact of the matter is…
If you hue shift the Gout in the correct colorspace what you get is not “revised color timing” But “fixed color timing” and it is practically equal and identical to the Special edition pre 2004 i.e. 1997 Special edition released to theaters.
This is the point. George Lucus sat there and ordered the hue shifting fixed for this release , but now we are back to square one again post 2004.
So sorry for liking older versions. After seeing 1997 broadcast version of the special edition. It very much makes sense. But come 2004 until present that is like a mish-mash version.
It is theatrical print coupled with the special edition.
The Special edition though was not based on the theatrical film print but a telecine or the Gout version master. Or atleast the broadcast 97 version looks like a good version of gout master.
if the actual special edition print looks different to broadcast version then you have a yet another version of special edition which does not use Gout master.
Ronster said: at the end of the day, I like that analogue look. Sorry if you don’t like that analogue look but i do. Sort of warm and fuzzy imperfect but it yeah I like it.
PLEASE stop this. You clearly have no idea what you’re doing, and you’re NOT restoring the film to its original look.
There are numerous people in this thread pointing out errors in everything you do. Maybe it’s not us, but you.Can’t a mod lock this thread please?
well I just caught a quick gimpse of the special edition broadcast and it is very true to laserschwerts trailer.
It is amazing how mugh ddtail has been lost after “The Restoration” What i am saying is that the special edition end part especially mostly looked way better than it does now. I am not all that fussed by it but it is like very different now also.
It is like as soon as a pure digital version was created the film has never been the same.
I would like to make a very analogue version that has that look and feel. It is trying to find what would be the best way to do that. But yes back to phosphor also something warm fuzzy and crisp.
What is the quality of special edition vcd like?
So… the correct color space is not RGB? Why? Do you know how colors are recorded on film? Are you aware that both color film and the old 3 strip technicolor process filter by Red, Green, and Blue light. Technicolor used actual filters, but color film uses chemicals sensitive to red, green, and blue light. So why wouldn’t RBG work when that is what is on the film? CYM is just the mirror image of RBG. The technicolor process filtered by RBG and then printed in CYM. It’s not like this is a new process.
And its not like what you are doing is actually fixing it. It is making it worse, just in a different way.
I don’t understand why you are angry…
at the end of the day, I like that analogue look. Sorry if you don’t like that analogue look but i do. Sort of warm and fuzzy imperfect but it yeah I like it. I have not even done anything, Jusg looking at “how” not this is exactly anything at the moment.
Do uou like the old special edition laserdiscs?
guys, the whole movie was meant to be played backwards. it all makes sense. the death star isn’t a weapon, it has a planet creating laser. and Luke was so bad at his job, he was left behind on a desert planet.
Don’t forget the hidden message from the devil as darth vader is finally bbq’s over a spit roast.
What would you say is the best special edition laserdisc version?
you need to use video filters in the right colorspace that is not RGB but ypbpr.
this is pretty much the moral of this story.
And that is about all i have to say about it.
First you say it is lovely, then you say there is a problem. Make up your mind. You could be doing some fun things, but you do need to take advice. That is what this site is for. Handle it right and the people around here will listen to you. Keep up the way you are an no one will.
I have no idea what tools you are using or what color shift you keep talking about. And it isn’t shift so much as balance. The GOUT, and I think all the other pre-dvd editions, were done from interpositives. If you aren’t familiar with them, that is an orange tinted positive print from which an internegative (an orange tinted negative print) is made which is used to make the distribution prints. How well a telecine removed that orange tint varies a lot. We were used to those methods as most movies on TV and home video were that. But to find the true colors, you have to take it a step further than the GOUT or JSC. They are nearly as bad as the DVD/BR, but in a different way. All those garbage mattes and other flaws that you don’t see in the DVD/BR or the film scans (and we have three of those - old and faded, Technicolor, and a low fade print). So I would say we have a pretty good idea what the original colors were (on the prints seen in most theaters there would be some variation depending on the quality of the duplication). When Pleasantville came out with all its B&W sections, they went to a lot of effort to make sure that each copy of the film that went out had a similar bias in the colors on each reel. So even if what you remember was more yellow and if your memory is accurate, it may not be the correct colors coming off the negative. We have a lot of copies of the film that we can back track how they were done and come up with a very good guess of the colors. What you are coming up with does not look even remotely accurate and looks as bad in a different way from the blu-ray. You are shifting the colors too much when 90% of what you need it to fix the blanance. Turn the yellow/blue channel more to the yellow and the cyan/red channel more to the cyan. Stop messing with the tint. The most you should have to nudge that setting is 1-2%. And for select scenes only. The setting you are calling color shift is one that you have to be very careful with because the moment you go too far, the results look horrible. Other settings can be done with a heavier hand, but that one you must be light with or the results look terrible.
The actual Color Hue Saturation and Value filter either needs to be in the correct colorspace or done direcg to mpeg is what I am finding at least not in RGB colorspace for Gout.
you can shift it as much as it has drifted away by “off base” or “X” to bring it back by that amount to same value as it is relative and requires no calibration.
In RGB vérsion of Filter it does indeed cause a problem for which you might try to compensate for the color error like I was doing.
But doing the shift direct to the mpeg does actually get really nice results. I am happy that this at least works correctly.
It could rapidly speed up a color correction project I think. As in the Gout looks not right on anything I watch it on in a lot of parts but in some it is fine.
Just so i have to carry on repeating myself the base or a very correct part of the GOUT transfer would be R2 and 3p0 on the sandcrawler interior hold. This I consider Base or no shift at all.
It’s not my fault the software i was using is not handling the colorspace correctly and does nog handle mpeg formats anymore.
I don’t have a lot of time on my hands to do a project at this time you are right.
But it is important to figure out how to go about doing something to get the best result. What I wanted to do was simple but the software lacked the capability.
I use professional hardware most of the time that is very good. I don’t claim to be some video editing genius but I am both identifying a problem and finding a solution of how to handle none of which anyone has any input in apart from calibrate your monitor guff when it comes down to actually going to the mpeg direct and using filters that will work in the right colorspace. Which my software did not… That was why I was using image program because I knew something was not right.
Things have clicked in to place now. A road forward has been at least found for myself and identified of a possible way forward on conforming the hues of the gout in the correct working colorspace.
I thought I did quite well to identify the problem and give a manual solution and a possible editing solution for reference that (should) work.
Unfortunately it took a while to get there. I knew the problem just trouble finding the solution and how things were going wrong.
Lightworks used to support progressive scanned Vob file but after mandatory update it is now gone.
sorry just finish work… Using Lightworks NLE but it no longer supports mpeg as of new version but still the result is different.
VLC is doing exactly what i want put this in NLE or even convert file it acts different the control is different and creates issues. Going to try womble.
Edit:
Effects might be in different colorspace is possibly the issue.
perhaps YPbPr or YCbCr is color space needed for effect to work correctly? Not in RGB anyway.
Go and see what it is meant to look like 😃 It looks lovely…
This is why you NEED to calibrate your monitor. If you think that looks lovely then your monitor is not showing you that gif is WAY too shifted towards green. It looks terrible.
Yep that is what happens in a NLE. And that was exported from. Really it needs to be done in Womble probably will then miss out the mpeg decoding issues.
The result is not true unless you apply it directly to the mpeg.
It crash and corrupt the file NLE file again and again. So you have to treat the raw mpeg after proper de-coding not in NLE for true result as it screws up the color once converted or Ripped. I think Womble or do it like I have in a player and capture the result.
Gout Hue Shifting Correction Reveals the true look of the Gout. It’s actually a doddle to fix this film using the GOUT. This was Loaded in VLC Media PLayer > Tools > Effects and Filters > Image Adjust.
This Yielded Better results than Loading it in to an non linear editor for MPEG decoding issues. Possibly womble DVD editor is probably the best thing to use to fix this and export for reference (free 30 day trial)
Perhaps still a bit pink on Tarkin.
The amount of Shift Changed
I think this is a fairly conclusive experiment and I now feel like I can see what the film is supposed to look like. I think it looks really really nice like this and it’s so simple and easy to get it where it needs to be.
To me it’s perfect like this… So to create a reference from a hue shifted GOUT would be the way to go, this could then be easily color matched to say the Blu-Ray and fix that pile of crap.
This is the best solution I feel to what is happening and the confusion over what the film should look like. Some sequences have shots that require different amounts of Shift but eventually after going through it piece by piece you will get to a conclusive answer on the Normal film footage.
The Special Effects as I said requires different Handling.
The reason I suggest using Womble is because it will handle the MPEG without having to decipher it as when I put this in NLE the results were slightly different and not as good. You xould also just use VLC for snapshots Like I have done and Color match.
This won’t give good reference for absolutely everything but I reckon about 80% some parts require no shift at all like R2 and 3P0 on the sandcrawler for instance. It only needs to shift when you see red face and so on.
I would probably trust the brightness and levels of the JSC to get a slightly brighter and less dark but overall I think I have found a good soloution.
Go and see what it is meant to look like 😃 It looks lovely…
The Gout has unfaded Highlights and the Door flash should look like this, Although I think there is a bit of a question over the light on the right hand side of the door is either Blue or Purple? I think it is Blue and we have an issue there. Not sure about the deep Blue that suddenly appear either before the explosion. Also note Soldier with purple shirt. It is all obvious evidence of Hue Shifting and the whole lot is swinging about like a windsock in a hurricane here.
The color is right but just shifting Hues so it needs to be conformed so that it maintains the correct Hue or the same Hue.
While most of these look good, I think you should avoid phrases like “the true look” of the GOUT unless you use some sort of reliable reference to define what such a definition entails. Any other method involves a great level of subjectivity and personal preference.
Ronster Remembers…Your Prints might be a bit different but this is how it used to look or very close with what I grew up with I am sure of it. The Hues are shifting that is about it. Like I said won’t get same result in NLE has to be done direct to mpeg.
No added color just hue shift. Be delicate it’s all there under our nose all this time 😃
Gout Hue Shifting Correction Reveals the true look of the Gout. It’s actually a doddle to fix this film using the GOUT. This was Loaded in VLC Media PLayer > Tools > Effects and Filters > Image Adjust.
This Yielded Better results than Loading it in to an non linear editor for MPEG decoding issues. Possibly womble DVD editor is probably the best thing to use to fix this and export for reference (free 30 day trial)
Perhaps still a bit pink on Tarkin.
The amount of Shift Changed
I think this is a fairly conclusive experiment and I now feel like I can see what the film is supposed to look like. I think it looks really really nice like this and it’s so simple and easy to get it where it needs to be.
To me it’s perfect like this… So to create a reference from a hue shifted GOUT would be the way to go, this could then be easily color matched to say the Blu-Ray and fix that pile of crap.
This is the best solution I feel to what is happening and the confusion over what the film should look like. Some sequences have shots that require different amounts of Shift but eventually after going through it piece by piece you will get to a conclusive answer on the Normal film footage.
The Special Effects as I said requires different Handling.
The reason I suggest using Womble is because it will handle the MPEG without having to decipher it as when I put this in NLE the results were slightly different and not as good. You xould also just use VLC for snapshots Like I have done and Color match.
This won’t give good reference for absolutely everything but I reckon about 80% some parts require no shift at all like R2 and 3P0 on the sandcrawler for instance. It only needs to shift when you see red face and so on.
I would probably trust the brightness and levels of the JSC to get a slightly brighter and less dark but overall I think I have found a good soloution.
Go and see what it is meant to look like 😃 It looks lovely…
The Gout has unfaded Highlights and the Door flash should look like this, Although I think there is a bit of a question over the light on the right hand side of the door is either Blue or Purple? I think it is Blue and we have an issue there. Not sure about the deep Blue that suddenly appear either before the explosion. Also note Soldier with purple shirt. It is all obvious evidence of Hue Shifting and the whole lot is swinging about like a windsock in a hurricane here.
The color is right but just shifting Hues so it needs to be conformed so that it maintains the correct Hue or the same Hue.