contrast and color change when clip stops
I use FCP6 with too little frequency so I haven't gotten used to a "problem" when editing on the canvas.
As I make a change using Color Correction 3-Way filter, it looks right as the clip is moving. But when It stops it changes to a "lighter" version which when exported, then converted to a QT .mov, the changes I made seem a distant memory. So is the end of the clip, the "lighter" version, what I'm actually going to get? Or is it what I see on the canvas as the clip is running?
Thanks so much, Ken
What is the codec of the footage you are working with?
FCP's interface...the Canvas and Viewer...aren't color accurate. In fact, they will lower resolution in order to play back the footage without dropping frames, so if you have high res footage on slower drives, or non-RT media...this will happen.
For proper monitoring in FCP Legacy (5, 6, 7) you need an IO device and external TV. FCX fixed this issue...
Little Frog Post
Read my blog, Little Frog in High Def
Thanks....in answer to your question, the imported codec is ProRes 422.
About the external device, GEEZ, I'm basically a photographer these many years and use two monitors...one for images, the other for tools, etc. They're properly calibrated ad nauseum. BUT, I have been shooting video for about two years, clips for stock and usually haven't had the issue of transferring what I see in FCP and the final QT clip I upload to agencies. This time I've got a close-up of hands and a white envelope so the color and contrast are important. More to the point, when you say that the viewer and canvas "aren't color accurate" that blows me away. I'm positive you know what you're saying however, I thought the equipment in use...computer+display needed the accuracy and that it was a "given" that FCP's algorithms are accurate and as good as the user (me) is in interpreting what's on the screen.
If you'd kindly clarify, I'd appreciate it. Thanks, Ken
[Ken Tannenbaum] "More to the point, when you say that the viewer and canvas "aren't color accurate" that blows me away. I'm positive you know what you're saying however, I thought the equipment in use...computer+display needed the accuracy and that it was a "given" that FCP's algorithms are accurate"
This debate got blown wide open about 4 years ago when Apple introduced "COLOR" as part of the Final Cut Suite (and then solved the problem by blowing away the software). Users began complaining very loudly that what they got back from the grade software round trip wasn't what they were expecting... well, a problem for those who were ignoring the recommended practice of using a Rec709-calibrated Video I/O card (AJA, Blackmagic, Matrox, etc.) and calibrated SDI display. Otherwise, the roundtrip between FCP and COLOR was subject to any number of colorspace, Quicktime, and ColorSync disparities that revealed a fundamental weakness in the FCP Viewer/Canvas interface. At best, they are just slightly better than a wild guess, and not really capable of reproducing the values that are present, even though the sRGB GUI display is supposed to be a relative of broadcast Rec709 -- it is just that, and *not* the same thing, sort of like a third cousin. Most graphics display devices, even the "calibrated" ones are just about good enough for print photography and camera-ready artwork, but CCIR-sourced video is actually on another planet from that, but Apple didn't really let that be a worry that they dealt with; up until they completely rewrote, and tried to reinvent, the video editing industry. Which didn't actually work out the way they guessed it might. But not for the first time.
"I always pass on free advice -- its never of any use to me" Oscar Wilde.
Beautifully said. Now, would you mind stopping by my place to help my wife remove the noose. She doesn't like heights.
It is interesting to me that although as stated, I don't use FCP much, when I first got involved a few years ago...when FCP6 was The One, I checked it out very carefully and typical of me, learned what I needed to learn to get the job done. Well, my memory does not include any reference to what you've laid out. Of course, what you know and what I NOW know took awhile to be revealed and understood when the shit hit the fan four years ago.
I'm not usually four years behind the times and have zero interest in changing the technologies I work with...so for the time being I'll do the following UNLESS you or another generous soul chimes in. My plan is to see a file through the process to a final QT movie. If it looks good, fine. If not I'll extrapolate a "backward" plan to change the color, etc. for the better and check it again as a "final" version. I see no other way around the carousel.
Again, thanks for your insight. Ken
[Ken Tannenbaum] "My plan is to see a file through the process to a final QT movie. If it looks good, fine. If not I'll extrapolate a "backward" plan to change the color, etc. for the better and check it again as a "final" version."
Its a plan.
What most people do is create a "transfer LUT" -- implemented as a "Look-Up-Table" which bridges the difference between colorspaces or presentation media. Some people use them simply as "look" generators. Whatever. Every color correction is a LUT, in a general sense.
There is a limited ability in Compressor to adjust the overall brightness, contrast, etc., as part of its image processing algorithm, so that would be another way of exporting different treatments for different release streams.
"I always pass on free advice -- its never of any use to me" Oscar Wilde.
Great avatar. Is it any wonder you're holding your head? All the stuff in there just itching to get out!
Me, I'm just a lowly photographer!
Just to revisit Shane's point about an I/O and indeed Joseph mentions this too. Even a calibrated computer monitor will not correctly display video.
This happens for all kinds or reasons, different colour spaces as mentioned, different scan rates and different native resolutions to mention just a few.
So the I/O bridges this gap and ensures correct monitoring of a video signal. Ideally this should be carried out an a calibrated video display. I reality we cant all afford or even need a fully calibrated broadcast monitor as we are not all making broadcast spec delivery to broadcasters. You can compromise with a decent plasma. (some of which can indeed be calibrated.) Or some of the "prosumer" monitors offer a usable solution. Sony and others have some smaller entry level monitors starting around £650 (UK) While perhaps not ideal for colour grading a broadcast show or feature you will get a much better idea of what is going on.
The advantage with these monitors is that depending on the model you buy they will likely have 1920x1080 panels that display HD natively. You will be able to more easily asses potential errors in interlacing as computer monitors can mangle those files. You often see posts here of people not being able to correctly work out if there are issues with interlacing/progressive and the answer is alway. "Look on a video monitor"
If you have a MAC that supports Thunderbolt then a really cost effective couple of monitoring solutions are the AJA TTap or Blackmagic Ultrastudio Mini Monitor.
If no Thunderbolt your options vary depending on your MAC but check out those companies as a starting point.
You are more likely to need an output solution rather than both in and out as I guess you capture from tapeless sources.
Anyway hope that helps in terms of why you might need one and what to look for.
8 Core MacPro, Kona 3, Tangent Wave, Mackie Universal Symphony 6.5 FCP7
i7 2.7 Gig MBP (non retina) 16Gigs Ram Blackmagic Monitor Mini Symphony 6.5 FCP7
I'm grateful for the insight and will look into it. Thanks, Ken