Fair enough, as color management is a massive black hole of rabbit holes. But what I'm trying to do here is show you how to actually evaluate in a potentially successful, usable manner. And down below, I'll talk about the probable difference between Resolve and Premiere exports.
First, as noted above, there isn't any such thing as a consistent color view across devices. Period. Isn't even physically possible. So trying to see what Vimeo shows on one computer only shows you what Vimeo shows on that computer, and has no bearing nor relevancy to Vimeo, or any other playback app, on any other screen in existence.
Frustrating? Of course. But sady, true.
You do realize that no device actually sees or records color, right? All that sensors can record is brightness ... and by the use of filters over the sensing elements and a ton of math processing, the engineers totally manufacture what appears to be color to our eyes. Karl Soule has even frequently quoted me on that to get his audiences to begin to understand the journey of light to color.
And all manufacturing lines have tolerances ... we simply cannot make any two sensor elements identical in response, let alone, a whole panel of them. Which is why all sensors and LED panels are graded after assembly. The more even/accurate panels are sold at much higher prices for special uses, the less accurate ones for general prosumer gear.
And yes, that is a large part of why a Grade 1 Reference monitor costs a lot more than a normal monitor. And why "but that Reference monitor and this computer monitor for a quarter the price use the same panel" is also an irrelevant statement. That basic monitor starts with (comparatively) a lesser quality panel.
And all camera makers ... for each camera ... assemble a series of parts, come up with a 'firmware' to drive them someplace together, and sell that to us to use. So they combine parts from all over, that are only used in that one camera model as an assembly. Hence each camera is a bit to a lot different in response to other camera models.
But even with identical cameras, the responses will be somewhat to notably different depending up lenses settings and all.
The same is true of screens. Which is why any assumption that all those seeing your production on Vimeo will ever see what you saw on your machine is simply a fallacy of understanding. They simply can't.
But even on the same screen, take an iPad ... watch a dark scene in a show on a park bench at noon, and in a dark bedroom at night. It will look quite different due to simply the changes in the surrounding ambient light and reflections.
So chasing "accuracy" by looking at the image on a monitor or TV is simply a fruitless endeavor.
The only way to ensure that your media is given the best shot to be seen across devices ... is have it done on a properly calibrated tested system, and then ... let it go as is.
Because then, as I commented before, you kow that your program will look, in relative terms! ... in similar fashion to other professionally produced media on that screen.
That's all a pro colorist can do, and essentially, you are trying to outsmart the professionals at this. I know it seems sensible to try doing this, but ... it isn't.
Differences between Colorist media and Premiere media
I'm wondering if the colorist used the Rec.709-A export option from Premiere. That doesn't change the media, but it does put a different number in the file header's NCLC tagging. Which is a non-standard use of a number that has no setting by any standard. But the BlackMagic devs realized that the use of that number somehow makes the Mac ColorSync utility use gamma 2.4 for the display transform.
But as that is not a set standard use of that number, the Adobe devs do not use it. In testing that I've seen colorists do, on other non-Mac systems that 'extra' NCLC tag can cause other issues, so while it's sort of a cadged fix, it's a bit dodgy at best. I think that might be another bit of why Adobe doesn't use it, but they won't say of course.
So Premiere will not export with that number.
If you have part of the bit directly from Resolve, and part directly from Premiere, you would have two different views on a Mac without Reference modes!!!! ... if the colorist used the Rec.709-A setting at export from Resolve.
Final comments:
1) You can't match across screens let alone OS/Programs/devices.
2) Macs without Reference modes (set to HDTV) will display using a transform different from all other devices, including Macs with Reference modes set to HDTV. This is not "fixable".
3) Premiere's gamma compensation LUT changes the program file image data at export.
4) Resolve's Rec.709-A export option does not change the file image data, but it does change the file header NCLC tag.
5) So the two workarounds to the "gamma issue" on Macs without Reference modes set to HDTV are both ... partial ... and potentially problematic.
6) The majority of colorists simply grade to the standard and export stock Rec.709 as that is the safest way to get a view "similar to other pro produced media" on any screen out there. And is required for broadcast/streaming deliverables.
30
Replies
AdChoices
.png){kind=link}
.png){kind=link}
