Welcome to our community

Be a part of something great, join today!

  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

Correct Gamma - Discussion about good communication practice in a production

The interesting thing is to know what do experience colourists think is best practice for RED footage? This could be a white paper to help both post houses and clients alike...

To me there is no single "best" practice. Only the best strategies for the job at hand. Most of my grading for actual gigs happens via REDlogFilm, in which case I'll back Blair's thoughts. However, I've done broadcast work where REDgamma4 (and other gammas) were the starting point. It all depends on the job itself.


Just communicate...

I'll echo the importance of communication.

I feel somewhat fortunate to have been a colorist in the time frame where it was my primary function. There was an interesting moment in the early 2000s where we went from integrating digital correction in combination with lab print lights that was utterly terrifying. Digital Color manipulation at the time was limited more or less to RGB mult or add values and it was stressed hard not to touch the contrast of the image, which was difficult as one of the common issues at the time due to a variety of reasons was the existance of "milky blacks".

In the grading of features films at that time much of it was the wild west. Communicating and experimenting with different workflows and strategies became pretty much a weekly endeavor as things were developing so rapidly.

After I was a bit more "unchained" in regards to what I could do on the color grading front, it became far more interesting. Back at the studio I developed a few tools that helped everybody "know" and see exactly where things were at. Even still different films represented different approaches. For instance I've worked on projects that lasted 2.5 years and I've worked on projects that existed for a month. You get to explore different techniques in those situations.

I made handful of tools that were rather useful that worked well at the studio both for our internal needs and for the client's needs.

- Succinct Version-ing. Changes to color were stored in the database and I had the liberty to make notes to track what exactly was done. Pulling up previous corrections and grades was very quick suddenly.
- Key Frames. I made a special little program called "key" this would tie a unique frame or group of frames that represented an entire sequence. This was often the exact frame from the hero Match Clip that was either created at the Lab with the DP or by me, the DP, and the Lab via the laser recorder. In the earlier days we'd just get the Match Clips and had to start there, then it became a much more collaborative process as digital color work became more of a norm.
- Color Wedges. Running out script to generate color wedges that a DP or Director can look at on the light box and make selects from was very useful. Essentially generating a bunch of options and finding the "hero". This was later replaced by digital variations and approval.
- Color Quilts. I still do this. Essentially grabbing a single fame or up to 3-5 frames from each shot and laying them out. I used to film these out and stick them into binder sleeves for client approval and review on the light box. If something stood out it really was clear. Digital Quilts became a bit more interesting as I could use the SQL database to limit and expand our view of the color palette. Meaning pulling up individual scenes, shots, or the entire film.

One notable that happened before the big shift to digital cinema cameras were DSLRs landing on sets. Some DPs and myself included would shoot quick stills on set and generate looks "ahead of post" which was interesting. These days you don't really need to do that as it's fairly trivial to send a clip after shooting it.

Now that I'm more on the DP side of things I find that more and more both for my own sanity and post's sanity it's best to get the image right in camera. Saves time, saves money, saves headaches. I do get to shoot and do my own color work on a variety of productions now, which is very nice. Often I'm sending off the match clip or master look and people run from there.

The last job I shot only needed a one light which I did through REDcine-X Pro and it was a matter of just providing the post house the Look Preset. Nice and simple.

That's not every job though. Bigger jobs require more robust solutions and workflows.
 
I agree. If I were a client, I would only be asking myself one question: do I like what I'm seeing? How the professional in the chair is achieving it is not your concern.
That is the right answer. On the other hand, if the results are clipping, I'd ask the colorist to show me the raw file to see if it's actually clipped within the raw, or if the exposure or ISO can be changed to provide a little more range.

If there's very little time for a job, I see why people would grab RedColor3 / RedGamma3 (or 4) for a job just to blow through it quickly. But ideally, RedColor3 and RedLogFilm work for me. In some cases -- usually in terms of screw ups -- I wind up having to tweak the color temperature a little bit just to come up with a better starting point, but the rest is all manual. It's a bad day if it takes me more than 1 minute per shot.

I just finished my first Red 5K job, and was gratified at how much highlight detail we were able to save compared to their dailies, which had all been done in RedColor/RedGamma. There's lots and lots of range there if you know how to dig for it.
 
There's absolutely nothing wrong with working from Redgamma. I often prefer it in many situations. That being said. If the footage looks bad and it looks bad because the colorist would rather be working in log but is either ignorant to the tools or unaware that R3Ds can be changed then I would blame the colorist. I sometimes set my viewer LUT in the camera to linear. That doesn't mean I want the colorist to start from linear that just means I was checking my highlights. I also might have white balances all over the board since I don't assume that someone will actually use the metadata settings. I assume someone will nuke the RMDs and start from a consistent base.

Here is the test. You ask for something specific. "I don't like how the highlights are clipping could we ease out of them." If they still can't get it ask something like "Could I see it in log view? I feel like I had more highlights. Maybe I overexposed that a bit more than intended but it would help me understand what I shot a little better to see it in LOG." Assume it was your fault and be apologetic. If they at that point don't use the log footage because they don't know how-- it's the colorists fault and you need to either teach them about the capabilities of R3Ds or you need to find a new colorist. Some people are receptive to learning new things. Some people aren't. I had to walk someone through how to do a proper alpha unpremult-grade-premult workflow once in a session. They seemed to appreciate learning something new and the project benefited since it fixed a problem they were fighting. I don't trust people who resent learning/refuse to learn new things.

To create a car analogy. If you're in a Taxi and you request that they take a route you know... but they insist they know a faster way but then get lost: you have a bad Taxi driver. I've had those before, where I felt bad speaking up because I didn't want to be a bad backseat driver but then when they clearly miss an exit have to step in--but then they insist they know where they're going and turn down your offer of directions. At that point I've had to negotiate the rate at the end. "Look, you drove an extra 15 minutes and 3 miles past the correct exit. I'm not paying for that time and distance." I've seen this in post before too. I'm more than happy to let people do things whatever way they want, I'll usually offer up a suggestion for something I know will work well. But when they fuck it up and it turns out all wrong then it's on them. If however their way turns out great, fine by me the more ways the merrier. I had a job a couple years ago where an artist I was supervising was falling behind. I did a small demonstration on the task he had to help show him a better faster way to work but he refused to follow it and insisted his way was better. Well his way took 3x longer and ended up looking far worse than anyone elses' and after that shot was done we let him go.

There is backseat driving, but some people are just bad drivers and when you're along for the ride better to be annoying than dead in a ditch.

THen again there is always this. :D

 
Back
Top