Knowing Your Scopes: When to Trust Your Eye and the Most Useful Scopes of All
Depending on the acquisition system, waveform monitors and vectorscopes can guide quality control of your imagery from shooting all the way down the pipeline to grading, mastering, and compression for delivery. Scopes can seem a little intimidating and esoteric to the new user, but getting the basics down can really help in owning your image. Recently, Alexis Van Hurkman over at ProVideo Coalition has answered some key questions about scopes: find out which ones he considers the most indispensable below, plus when it may be a good idea to trust your own pair of eyes in making adjustments -- even when your scopes are reading 'A-Okay.'
Scopes Don't Lie, But They Can Get Tricked: When Your Eye's the Bottom Line
Alexis kicks off his new Q/A column over at PVC by confronting a rare and dreaded conundrum, with respect to grading: what if your scopes and your eye are telling you two different things?
What do you do when you have two shots that appear to be matching on your video scopes but it still isn't matching according to your eye?
I find this doesn't come up too terribly much, or perhaps it's just that I don't really make a note of it anymore, but I'm fond of saying you should keep one eye on your scopes, and another eye on your monitor. My short answer would be, if ever the two disagree, go with your monitor.
Alexis emphasizes that this preference weighs heavily on your proportionate experience with the monitor. If your monitor's tonal and gamma responses (compared to corresponding scope readings) are familiar to you and you're comfortable and confident in its representation of key characteristics, going by the monitor is a relatively safe bet. On the other hand, if you have any reason to not fully trust the monitor, you'll have to play it closer to the scopes.
Getting back to my original point, the overall goal of shot matching is to achieve a convincingly plausible match among the various shots in a scene, and that requires you to evaluate the characteristics of each image, overall. When people watch a clip, they're looking at two things–they're focused on the object of the scene, such as a person talking, and they're also taking in the overall atmosphere of the image, whether it's bluish or orangeish, high contrast or low.
Qualities such as the more diffuse, overall color context -- the 'warmth' of a scene -- can appear more skewed one way or the other in comparison to a preceding shot, depending on the contents of each. Interestingly, this is due to some of the same principles at work in the law of simultaneous contrast -- in its most extreme case, the law dictates that two complementary colors appear their most bold and saturated when juxtaposed directly adjacent to one another. As such, the balance of ambient details can become weighted differently in perception across multiple shots, even when their scope values seem consistent throughout those same shots. Alexis points out that due to this (among other things, we recommend you read his full post), the image on the left appears warmer than the cutaway to its right, which appears cooler:
This is a bit of a befuddling place to find oneself during grading, especially when comparing the two shots' RGB parades -- be sure to read Alexis's full post at PVC to check those out, as they'll yield the best comparison. As indicated above, the solution may simply be to 'go by eye,' and cheat one of the shots a little closer to the other than the scopes alone might dictate. After all, the final viewer will be judging by eye, and by eye alone!
What's 'The Scope of All Scopes'?
Alexis says that if he were marooned somewhere remote and had to pick just one scope to use, it would be the very flexible RGB parade -- but that the vectorscope would be his second choice. Vectorscopes do what no other scope can, he says, which is show you your overall color balance, which color(s) that balance most leans towards, and how saturated those dominant colors are all at once. Check his post for greater detail.
He's also devised (and open-sourced) a brand-new graticule for the vectorscope, which is the two-dimensional overlay that guides color calibration and context, pictured above. Color graders will have to chime in with their own thoughts, but to me, this graticule looks a lot more intuitive than the traditional one (which again, can be seen in the source post). The guidelines leading directly to the primary and secondary colors seems a lot more helpful than the concentric rings of other vectorscope models -- but again, that just might be me.
What are the scopes you most rely on in post (or on set), and why? When have you found your scopes to be 'tricked' by an edit, and what ways have you found yourselves correcting for perceptive mismatches?