Please, watch the video demonstration in 720p.
The same issue like in version CS5.5
The same like in this thread about a very strange and not usual problem that occurs when I try to use almost all of the tools from Color Corrector, and the Scopes show an anormal change on the colours/tones of the video once applied one of the effects of the bin Color Correction (Three Way Color Corrector, Fast Color Corrector, Color Balance HLS, RGB Curves, etc...) but also changes a little bit even the distribution (if I may say so) of the pixels in the video.
This thing can be checked very easy zooming the Program Monitor at 100% for exemple, and enabling/desabling simultaneously Three Way Color Corrector or another tool, and you can see in the Preview how the pixels are moving around.
I hoped that this issue has been solved in the new version of Premiere CS6, especially that I sent a bug report.
Don't be impatient, I tell you right now that the bug still exist. I would like that when you'll test the trial of CS6 return in this thread and tell us if you noticed the same inconvenient.
Anyway, if I'm not wrong you also participate in the other thread from the link that I put here and contrary of what I, spreeni, shoternz... noticed, you told us that you didn't see any strange issue.
Is set to full, we tried everything, in other computers too, with MPE on and off, with Maximum Bit Depth and Maximum Render Quality on and off in the sequence settings. It was of no help, in any other NLE soft doesn't happen something like this.
Adobe, PLEASE, FIX THIS BUG before launching the new version of Premiere CS6, is not possible to make a DECENT and PROFESSIONAL color correction and grading in this conditions.
Of course we can upload a video, although allow me to tell you with all respect that we hardly believe you made the test and in the "All Scopes" you didn't see strange differences just when you apply (without change anything) any tool of Color Correction tab. As a matter of fact with any graphic or video material that you use, you'll can see this in "All Scopes", at least in every pc, we can't guarantee in a mac.
Thanks for you participation.
In our case the pixels shift and even more changes the tones of the video (particularly in the midtones) no matter if the MPE is on or off. Just like before the All Scopes show this clearly.
I uploaded 3 videos from 3 different cameras: Sony EX1, Panasonic AG-AC160 and Panasonic TM900. Here is from where you can download: http://www.filedropper.com/rawvideofiles
We also put on Youtube (watch at 720p) the test we made with this 3 RAW video file. This time we show you clearly pointing you with the finger where to look. At the 01:50 is so obvious how much the tone of the asphalt (street) changes and the pixels shift only when you apply Three Way Color Corrector without changing anything else.
Please Adobe, fix that, too many editors saw this bug!
There are some ads on the page. I got a "put in the letters box" and downloaded 67Meg. I went back after your post, and I can't tell the ads from the "real" download button. It is a single zip file.
Yes, this happens in all the 3 videos, actually happens in all videos and graphics that we have tested.
Imagine how difficult is to work with such an inconvenient, for example in the third video with the snow, we try to make a good/professional white balance and when using any tools of the Color Correction tab and add it on the event, we can see very easy how the tone of the asphalt/street changes from grey to slight green tone. So now we have two problems: the initial problem of the raw video and the new one made from this bug of Adobe Premiere. So what color correction are we doing now?!?
And this is an old issue, we were hoping that at least in the new Adobe Premiere CS6 the things would be different.
Thanks all of you for your participation.
This test can be done by everyone and can get the answear.
Yes, we did it. No change. The RGB Parade, Vectorscope and the other tools show a change of the tones and the brightness only by a simple adding of any tool from Color Correction.
In a previous comment you said that you couldn't download the 3 raw video files from the link that I left. I uploaded them again. So, have you finally downloaded them and made the test? As a matter of fact, you needn't really this 3 video files of mine, the test could be done with any other video or graphic. With the video/event on timeline and the eyes on the All Scopes is so very easy!
We didn't actually verify all the effects, especially the ones we were not interested in. But I can tell you that with almoust every effects from the Color Correction tab this bug happens. It does with Arithmetic and Blend from the Channel tab, Basic 3D and Bevel Alpha from the Perspective tab, too. The All Scopes shows it clearly.
So, we got to believe that it is not a matter of form but a matter of substance, and this bug/issue is related with the very structure in which the soft is made in.
Well, it does seem as if Steve is personally invested (to some degree, at least) in making sure the team addresses this issue. So that's a good thing.
For myself, I tend to prefer Colorista for my CC work. And with SpeedGrade included, that's another option. It may not be the work flow you want, but it will probably allow you to get the work done at the quality you desire until the issue gets resolved.
In our case both Colorista II and Magic Bullet Looks changes the tones and the brightness only by applying them. Verified in Adobe Premiere CS5.5.2 with one of the videos I uploaded.
That's why I sad that "is not a matter of form (effects of Premiere or other plugins) but a matter of substance, and this bug/issue is related with the very structure in which the soft is made in".
So nothing. The official trial is here, I tested it and there is no change, the pixels still shift when applying an effect, the tones and the brightness too, the same "All Scopes" shows it clearly.
I see that SteveHoeg didn't come back to tell us something.... can be repaired.... it can't be repaired.... in a future update or never....
I feel like adding my own 2 cents here....
I am not a professional color grader, and color grading has a limited application in most of my work. However, when I do any color correction, I am usually making subtle adjustments and NEED them to work well and so it does infuriate me a bit to have this issue brought to my attention. If I am wrangling not only the original color in the video along with PPro's accidental color adjustments as well, that is just exhausting to think about.
Again, admittedly I never knew about this or noticed it or anything like that, so you could argue that I don't have a foot to stand on in this discussion. However, the thing that is such an annoyance is that apparently at least one user already brought this to Adobe's attention (a user who seems quite capable with color work) and it went unresolved for whatever reason....deadlines, cutoffs, resources, whatever.
I'm quite pleased to see that Steve has acknowledged the problem and is working on a fix (confirming that it CAN be fixed, which is nice). But I'm still having a hard time wrapping my head around the idea that this even happens in the first place on the 6th generation of Premiere Pro software and the first version to practically be banging people's heads in with the whole "best professional NLE" ad blitz all over the web.
Anyway, succinctly put...."Adobe, WTF?" and also "thanks for fixing this Steve."
I had a look at this a long time ago and are certainly not denying it's existance it's real and it's there.
What I tested led me to believe that the shifts in colour were to me a very slight error with colour value mapping as if it was an incorrect LUT
The errors I got were between a 1% & .5% colour shift.
I found a full red field with 30% luminance gave no errors at all.
25% luminance gave a .5% rise in blue only
20% lum gave a drop of .5% in red
40% lum gave a 1% rise in red and a .5% rise in blue.
60% lum gave a 1% rise in blue only.
The variations are prevalent and different for Blue and green... you could go all day..
With my limited knowledge I thought, well you would be hard pressed to make .5% adjustments anyway.
and isn't the idea of colour correcting to change the colour to what you want anyway?
I too like everything pristine and in order and would like to see it fixed but hardly think it's a ball breaker.
So for you the fact that this is 5% more or less doesn't matter, that grey is grey just like in the video with the street/asphalt and when applying an effect of CC the grey takes a sort of a green tone, it doesn't matter for you because it's only a diference of 5%. That the luminance/colour/lum... may vary 5% when applying or not an effects it doesn't matter. But what kind of professionals are you?!? You know what you are saying? : "The problem is there, the "All Scope" shows it, but it doesn't bother us".
"and isn't the idea of colour correcting to change the colour to what you want anyway?" [Butch2oc]. No my friend, when I use CC to fix the snow's white, this is all what I want to do, not to fix also the tone of the street that was change once applied any tool of CC.
I'm glad that you at least don't deny the problem,or say that it doesn't exist, when so many of us saw it and when All Scope shows it clearly. Look at SteveHoeg, he said that in order to solve this problem, it takes some time .Maybe the next update will fix it. But what is even more important, is that Dave didn't deny the problem in his comentary and said that the problem will be fixed. That means he also saw the problem and realised that something is not wright. That't how a professional should be!!!
I have a strange feeling that many of you are using in the process of editing the technique of the "eye-scope", not the professional technique of the "vectorscope". For a real professional, that difference of 5% showed in "All Scope" is important. But with this kind of "professional" editors, I'm not surprised that Adobe will never fix this bug. Bug that in all of the other NLEs doesn't exist.
So nothing, we feel tired. Adobe, please, leave it the way it is. With such "professionals" you won't loose anything.
I'm sorry Hewlet that you feel so frustrated. My actual testing indicated .5 % not 5%.
I can assure you 2 things.
1) If it was 5% we would all be just as concerned
2) I have been in this industry for 30 Years and actually learned my colour grading skills from a person who was colour blind. I learnt how to grade by scopes only therefore I consider myself a proffesional and so do all our clients that keep coming back for grading.
Please don't take this the wrong way but this forum is for people to help each other out, give opinions and workarounds for what I think is a great piece of software.
We all hear your concerns but now I feel they are becoming a bit loud.
I assume that you are a pro and have external scopes that perform more vigorous and acurate testing... so, maybe you could focus on the issue and test like I did.
It sounds like you have tested all other NLEs
in all of the other NLEs doesn't exist
so I think you have solved your problem... go to a different NLE