I'm usually really good about getting skin tones to fall exactly on the skin tone line indicator on the vectorscope in Resolve. However, I have a client who routinely tells me that when I do this, the skin tones are too red. Is this just their personal preference, or should I not be going by the skin tone indicator at all?
Can You Trust the Skin Tone Indicator in Resolve?
- Thread starter Aaron Kessler
- Start date
Not everyone's skin tone falls exactly on the line. It's more of a guide than a definitive rule. People can have more olive tones (yellow) or red tones in their skin.
Maybe the skin your correcting falls more on the yellow side of the line or maybe that's just your client's preference. Could also be a display issue! Hard to say without more context.
Maybe the skin your correcting falls more on the yellow side of the line or maybe that's just your client's preference. Could also be a display issue! Hard to say without more context.
Yeah, I've talked to them about calibrating the monitor where they review color. I don't think it's happened yet, though. My monitor is calibrated with Spyder Datacolor. I do suspect, though, that most people prefer more golden skin tones. Also, a computer whiz friend of mine has pointed out that older monitors tend to skew more red.
If it's possible, the best solution is to get your client in your suite and have them see what you're seeing. If it's not possible then you could try a service like Louper or Evercast. It's a little bit better but you still need to know that they're watching on whatever display could be giving them issues. You'll just know that they're getting a bit of a cleaner feed. Another alternative is to have them review on different displays and devices. iPhones and iPads are decent, not wholly accurate but there's less room for user error (they can only control brightness).
iPhone/iPad would probably be the best option. I have both and have been VERY pleased with how close they match my calibrated monitor.
A lot of post houses send out iPad pros for remote reviewing. Not sure what their exact pipeline is for that type of review but it is a trusted method. There might be some documentation out there on how to setup an iPad review.
I observe that clients fall into two camps. They are either a War on Magenta or a War on Green fighter. I have to align to their tastes, but especially with skin tone there is a preference that is subjective. You can reason by comparing to common references that both yourself and the client have agreed look good.
The "Skin-Tone Line" is not really the skintone line; there's a long history as to why that line is there and what it was really used for. Read this free guide:
"Using Waveform Monitors as Artistic Tools for Color Grading" by Steve Hullfish
It will tell you everything you possibly need about using scopes for color. I've been doing this for more than 40 years, and even _I_ learned a few things from this article.
What you can use the I&Q line is to get a reference for where the client wants to see skintone, and then use that to match to throughout the project. Me personally, I'd be wary of whatever they're using as a monitor. Without calibration... it's all guesswork.
"Using Waveform Monitors as Artistic Tools for Color Grading" by Steve Hullfish
It will tell you everything you possibly need about using scopes for color. I've been doing this for more than 40 years, and even _I_ learned a few things from this article.
What you can use the I&Q line is to get a reference for where the client wants to see skintone, and then use that to match to throughout the project. Me personally, I'd be wary of whatever they're using as a monitor. Without calibration... it's all guesswork.
Skin tonalities cover a wider range then this line, do not have to be on the line to look right and can be on the line while looking off for a bunch of reasons.
Last edited:
Add it to the list of terminology that has been co-opted over the years. Kind of like "color correction" and "color grading" being two different things (don't know how that happened).
One reason to love OmniScope - agree with client on site where the skin tone 'feels right', create a target there and get going ;•]
My thoughts: I think this obsession with quantifying any colour, let alone 'skintone' in this way introduces the risk of ignoring how relative our perception of colour is - especially related to the surrounding colours. E.g. skintone in front of a neutral background will look less warm than the same skintone in front of a blueish background. You may think you've nailed it but perceptually it won't match at all. Plenty of optical illusions published on the internet to illustrate this phenomenon (simultaneous contrast).
Equally - I rarely find it a good look to have all the people on screen have the same skintone.
Equally - I rarely find it a good look to have all the people on screen have the same skintone.
On this we're in 100% agreement. I think some people now believe that "color correction" is what we always called in post a "dailies grade," and then what some people call "color grading" is (to me) just "final color" for delivery. It could very well be a U.K. vs. U.S. thing. "One-Light" and "Best Light" are also useful names for dailies grades.
It's actually called "skin tone line" in Resolve's scopes which is grossly misleading. Like Raj says, this is the "I" vector. Along with the 90 degrees out of phase "Q" vector these two were used for chroma quadrature modulation for NTSC broadcast. They have been obsolete for some time now. The fact that skin hue sort of falls along this vector is probably a coincidence, I bet you that Soylent Green had zero shots where skin color would have fallen along this line. It's counterproductive to chase it. What's more important is skin hue continuity, wherever that hue may fall. Even then there will be motivations for the hue to change from shot to shot, i.e. character facing a campfire turns around to face the moon.
