Home Grading Could We Split the Difference?

Could We Split the Difference?

by alexis

No matter who the client, or what the project, sooner or later you’re going to be asked “can you split the difference?” between your interpretation of what the client wanted, and what they discovered they really wanted once they saw what you were up to.

This will make you quietly, politely crazy, and is one of the reasons you need to cultivate a great reservoir of equanimity to do this job.

At the end of the day your client’s needs are more important then your mad skillz, so you’ll make the change, render out the project, and hopefully leave work on time to go knock back a beer or two during happy hour, recalling fondly how cool that program would have looked had they only let you off the chain. This is one reason why colorists still do music videos, despite the woefully poor budgets the majority of them have. Because most low budget music videos seem to want wall-to-wall insanity in the grading.

I was flipping through a fashion catalog (Free People, November) that featured some nice, gentle faded and flared film-ish treatments, and mulling over how I’d achieve those looks in different applications (as I am wont to do over my morning coffee). My wife Kaylynn is a photo stylist who works on these kinds of things, so she gets most of the relevant fashion magazines and catalogs, and we often compare notes on the changing styles of photography from season to season, which is a nice bit of casual research.

Then, of course, I dig into my grading project of the week, and inevitably they don’t want any of that; I’m told they want a nice clean grade, a little warm, with good contrast but no crushing, and FOR GOD’S SAKE DON’T CLIP THE SKIN TONES.

All of which is fine. Your average documentary is not wanting to look like a music video. Still, it makes me treasure all the more those projects that are looking for bolder color treatments. So when I get a project with a flashback or dream sequence, or for which the client is wanting signature looks for specific scenes or acts, and they let me go a little crazy, the pang I feel when I hear “could we split the difference” is just a little more pronounced.

Here’s a pretend example of what I’m talking about. This is an amalgam of different experiences; the example clip did not undergo this, I’m simply using it because it’s at hand (it’s available as one of the clips on the disc of my Color Correction Handbook), and I feel the need to point out that the particular client who brought that project to me was great to work with.

One of the first things I usually do is a simple, non-destructive and neutral grade of the image just to see what I’ve got to work with. In this instance, a very simple set of Lift/Gamma/Gain adjustments and a modest YRGB curves adjustment to compress the toe of the shadows yielded the following image:

At this point, the client tells me, “Yeah, I saw these great color treatments in the Free People catalog, and I really like that faded color with blue shadows, and a faded light-leak on the side. Could you do that? Let’s go crazy!”

And I say, “Heck yeah.” And proceed to start abusing the image, first using the YRGB curves to create nonlinear, per-channel color adjustments to the highlights and shadows to create a warm/turquise disparity, with high contrast specifically targeted to the tonality of the image to maintain a smooth falloff, and a blue lift via the blue channel’s YSFX slider (a DaVinci Resolve-specific adjustment).

Then, I use a Luma vs. Sat curve to create a gradual desaturation of the highlights, muting the colors of the skin tone.

Finally, I add a really, really soft window, and use it to limit another curves adjustment to create the light leak effect.

Then, I show the client the result. Predictably, after a period of silence, the client asks, “I’m not sure about the flaring. Could we split the difference?” The remainder of this narrative could go on and on, but to make a long story short, oftentimes situations like this have the following evolution.

My first take based on the reference imagery the client used:

Splitting the middle by fading/dissolving the curves adjustment literally by 50%, and losing the light flare:

What the client eventually signed off on:

The final solution ends up being slightly warmer midtones and very neutral shadows, easily accomplished by deleting all my other adjustments and making two simple color balance tweaks to Lift and Gamma. The reference image turned out to be a macguffin that served only to show the general direction of the correction. It was not, in fact, what the client wanted.

This happens all the time, and consequently I find I’m a bit skeptical when someone asks me to do something incredibly brash and bold. I don’t want to spend too much of the client’s time working up an elaborate grade when all they really want is something pretty simple. On the other hand, you want to take the client seriously, and if they really are looking for something bold, you don’t want to seem too meek, lest you’re thought of as a creative simpleton.

At the end of the day, I find it all boils down to getting to know your client as well as you can, and your first two hours are critical. Pay particular attention your client’s verbal and nonverbal cues as you create the initial, exploratory grades for a new piece. Chances are, you’ll know within three adjustments if you’re on the right track, and you can swiftly change course if you’re not.

And besides, if you create some super-cool look that the client doesn’t ultimately want, you can save it for some other job. That’s what the still store is for.

Download This Example

Added Nov 3rd, 2012—At reader’s request, I’ve uploaded a saved still and the grade I use as an example for you to download via this link. If you want to apply it to the clip in this example, this very clip comes with the media that accompanies my Color Correction Handbook. To import it into Resolve 9, download the file and uncompress it (it’s a .zip file), then open Resolve, Right-click in the Gallery, choose Import, then select the SplittingDifference_1.30.1.dpx file, and click Import. The still and its grade should import.


You may also like


Philippe Kiener November 12, 2012 - 10:22 am

Thanks for that article to comes at the exact right time for me. I just had 2 experiences exactly like you mentionned and I was wondering what was wrong with me.
As a starting colorist, I just had an experience with a client that wanted something different, a cinematic look and so on. So I sat with him and we made something with a bleach by pass look and a nice s Curve. Then I did all the clip with that and he says: “It’s way too much, I don’t like it, I prefer the flat image”. It was hard for him to put into words what was wrong.
I also do the editing prior to color grading, I shoot with a 5D with a flat or cinestyle profile and with that particular client, the clip was watched almost 100 times on Vimeo (with a password) with the flat profile before we worked on the grading. So I’m suspicious that for some clients they watch so many times their clip with a flat profile that they have a hard time to go beyond that and go for something more creative.

Alexis November 12, 2012 - 11:08 am

What you’re experiencing is not at all uncommon. Unfortunately, clients can sometimes get really attached to the look of ungraded dailies, and it makes it difficult for them to be open to other approaches. One more reason to produce simply graded dailies up front, when possible. Not that anyone wants to spend for that these days…

Ian Carruthers November 12, 2012 - 12:27 pm

Hey, Alexis! I’ve been a pro videographer for 4 years, but I’m just starting to drown – er, swim – in the waters of color correction. One of my first reference materials is your book, and I’m diggin’ it.

Thanks for the words on client relations too. The whole “quietly crazy” thing is something I’m working on. That is, I love the magic of color correction, and the learning curve right now is STEEP. So, I’m taking the crazy out on other things (usually pixellated storm troopers), and returning to color with a bit less crazy-steam.

I’m also noticing that, after what feels like (and sometimes is) hours of back and forth work on a single clip, I finally “clunk” into the right look. Often, that look is much simpler to achieve than I initially knew. Do you find this is an issue with new graders, in your experience?

Thanks again!

Alexis November 12, 2012 - 12:36 pm

I still find that’s often the case, especially when I’m tasked to create a signature look. Working in Resolve, my initial grade of exploration may well be a thicket of nodes, adjustments and readjustments, as I figure out just where we’re trying to go. Once the client and I are satisfied, I generally find a simpler and more elegant way of creating that look on subsequent shots. However, without that first mess of a grade, I never would have got there.

In terms of time taken, experience is the only thing that makes colorists faster. The more clips you grade, the more situations you encounter, the more you’ll build up an internal database of “solutions to common problems” that you can draw upon, shortcuts to various situations that will accelerate your day.

Keep on keeping on.

Goran November 18, 2012 - 7:06 am

Nice article and I am very interested in this grade using curves … maybe You could attach still that me and others can import in Resolve

Alexis November 19, 2012 - 11:17 am

That’s a fine idea. I’ve updated the article to include a download link for the still and grade. If you’ve got my Color Correction Handbook, then you can download this grade and apply it to the actual clip I reference. Have fun!

Chris K Jones November 25, 2012 - 5:26 am

Great post, it’s just how we think, and it applies to all human decision making. I.e. at which point when faced with three choices did I feel most confident and positive? Usually when I had just seen the first option. My sister is an (real) estate agent, when she is selling to a client she will always take them to the most suitable property first, then follow up with two less suitable. The client almost always goes for the first property – when they felt most positive. Choices offer context, feelings make our decisions for us.


Leave a Comment

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More