-
Luminance BUG in FCP 5.1.1 using Apple 10 bit uncmpressed NTSC codec
-
[Countdown] “Please post your rantings about Rules on White on another post. I’d like to stay focused on the bug, please.”
The bug is that you are trying to clamp a 100% graphic which is not accepted broadcast workflow. You’re setting up a white on white effect and you’re saying the filter doesn’t work and there’s a bug.
I’m saying your workflow is incorrect and it’s not a real world test. The workflow is create graphics with no more than 90% white. You apply the Color Corrector Filter to your clips to balance and correct levels. You apply the Broadcast Safe Filter as a fail-safe final level check.
You do not use the Broadcast Safe Filter by itself. It works best with the Color Corrector.
That’s not a rant, that’s a real world workflow that works.
Walter Biscardi, Jr.
https://www.biscardicreative.com
HD Editorial & Animation for Food Network’s “Good Eats”
HD Editorial for “Assignment Earth”“I reject your reality and substitute my own!” – Adam Savage, Mythbusters
-
Didn’t get that result here even with the Color Corrector filters applied as you said. Lowering highlights from 255 to 235 brought white levels generally down to 92. (not the effect I’m looking for anyway). The box however peaks at 104 with significant edging at 96. In other words white box on white background still produces a box with visible luminance difference edges. This is with Broadcast Safe at Extremely Conservative, sliders all the way to the left (ultra right wing setting).
-
Respectfully Walter, I have no control over the videography that comes in higher than 85 or 90 white levels. Expecting source footage and graphics to come in always below 85 or 90 is NOT a real world workflow. Broadcast Safe doesn’t work after applying the 3-way–there are still transient spikes that aren’t clipped. I get .7volts peak to peak in my signal and I’m going to use all of it. You can use .6v and feel fine, but this box should handle .7volts. That’s NTSC and that’s what I’m editing. Final Cut 4.5 can handle NTSC and I expect this newer version to handle a full, legal NTSC signal as well without creating artifacts around the edges.
-
and another thing. I’m not trying to clamp a 100% white. I’m just trying to move a white box over a white background.
-
[Countdown] “Respectfully Walter, I have no control over the videography that comes in higher than 85 or 90 white levels. Expecting source footage and graphics to come in always below 85 or 90 is NOT a real world workflow.”
Video from the field, no. There’s no way to control that. Have you watched an episode of Good Eats on the Food Network? The kitchen is full of stainless steel, the utensils and pots are stainless, there’s white all over the place and the stove is gas (fire) so I’m dealing with lots of hot video in that show. But I’ve delivered 22 episodes via the 3 Way Color Corrector, Levels and Broadcast Safe Filter in FCP with zero rejections. The last 8 were delivered via 5.1.1. The levels are checked in part by computer so if one single frame of any of this was over 100%, the show be rejected. If the Broadcast Safe Filter was not working correctly in 5.1.1 it would be costing me a lot of money in re-mastering.
[Countdown] “Expecting source footage and graphics to come in always below 85 or 90 is NOT a real world workflow.”
Graphics, absolutely I expect those to be created with 85 – 90% whites. If they’re not, they’re sent back to the artist to be fixed or I simply drop the brightness by 10% on everything in Photoshop. That’s something we DO have control over and it’s something that’s easily fixed before it gets to FCP.
CC and apply Broadcast safe to some regular video and let me know if you’re still getting the spikes. The 100% white on white test really doesn’t tell me anything.
Walter Biscardi, Jr.
https://www.biscardicreative.com
HD Editorial & Animation for Food Network’s “Good Eats”
HD Editorial for “Assignment Earth”“I reject your reality and substitute my own!” – Adam Savage, Mythbusters
-
Countdown,
Setting aside all this brouhaha about where levels should be…. I can confirm that we’ve had the EXACT same problem with the broadcast safe filter. Even stranger, we’ve found that the editor made an adjustment to “safe”, thought he was good to go, and then discovered that the “render” reverted back to the “hot” levels. In other words, the correction was made, but the results were luminence around 108 or 110. Double clicking on the offending clip showed the levels in the correct place. It was very frustrating, and resulted in one of our shows getting kicked back.
Yes we have outboard scopes, and yes we know what we’re doing. I’ve never had a show THAT far out of specs on so many clips. So… while I agree that you SHOULD be able to go up to 100, it really is much safer to tread the waters around 90 to avoid these kind of unpleasant experiences.
Mark
-
You do realize that if you want to change the levels you set the broadcast safe filter to custom, no?
-
There have been bugs in 10 bit renders in Final Cut for quite a while. Not sure why they can’t seem to get it working completely, and the bugs aren’t always the same from version to version. It’s pretty much a crap shoot from one version to the next. And maybe I’m pushing the envelope or breaking the Golden Rule, but I never worry about peaks at 100%. I’ve sent national spots to all the major networks and never had a problem with 100% white. I do use 90-95% white for white in graphics because they reproduce more cleanly on consumer grade monitors than if I use 100%.
Joe Murray
-
[David Roth Weiss] ”
In fact, I’m surprised that Boris 3D titler defaults to 100% white. Many titling apps, TitleMotion for example, default to 85 or 90% in order to keep neophites from sending out 100% white titles by mistake.”Wierd, mine defaults to 235, which is an RGB value so it gets contracted to 92% in a YCrCb colorspace.
The default 235 of Title 3D value is confusing because it’s a RGB value and it corresponds to the white value in the YCrCb colorspace.
RGB 255 = YCrCb 235, so an RGB graphic file should always be thoricaly luma safe (but can be chroma unsafe when in the extreme of R, G or B).
-
A. Here Walter is a test that should get beyond the White level diversion and focus on the edge anomalies:
Try this simple non-peak white test:
1. Create matte color white 85
2. Place that white on V1 and also on top of it on V2
3. Select V2 white clip and create a motion effect with scale at “50” to start and “80” at end of clip. (creating a zooming white box over a white background)
4. Render the effect.
5. Look at waveform monitor (internal FCP, or external). Notice the faint white dots and waveforms at about 95, not all at 85 as it should be.
6. Notice the peak transient white artifacts on the edges of the box in your NTSC broadcast monitor as the box moves. (kinda looks like a thin glass bevel effect on the box)
7. Confirm: Try 1-6 above on FCP 4.5. No problems. White boxes moving over white backgrounds look like, well, white.B. As I believe I said in my first post, the Broadcast Safe filter appears to clip the whites on regular video clips to safe levels until rendered, then FCP produces peak whites well above 100, nearly 110 as Mark reports in his post.
Reply to this Discussion! Login or Sign Up