[Joseph Owens]

I guess it doesn’t really help that this forum is for the most part almost moribund.

Used to be pretty active, the highest traffic on the Cow when Resolve was first introduced to the not-specifically-colorist community. Its also too bad that the general traffic elsewhere is about how to pick the cheapest hardware possible and either “why doesn’t this work the way I think it should?” or “Hey guys, this is my first post, what do you think of my fantastic grade? TIA.”

There are employment classifieds (for some time now) that specifically state that “Resolve colorists” need not apply.

Tough branding, and a pity because it used to be a big deal to be associated with daVinci. When it was davsys.

jPo, CSI

“I always pass on free advice — its never of any use to me” Oscar Wilde.

[Joseph Owens]

[Michael McCune] ” a different shift at different exposure levels.”

Film is about density. Which is a power/log transform.

I wouldn’t say the ProRes codec is a problem, per se… but it’s going to limit what you can do, to a certain extent.
If the option was there to pull the scan in a 444 codec, with a transform profile (ACES), I might have opted for that. In any event, working in the Academy color space with an appropriate IDT might give you some wiggle room.

If you are dealing with faded layers (remember color film is a gel sandwich) then maybe think about simulating a Technicolor 3-strip approach. Set up a splitter node and assign a grade path for R / G / B. Work the curves. Look for shading errors… sawtooth, parabola…

If you for some reason have a China Doll at the head, pay some attention to it. If your roll of film has been sitting in a can for a couple of decades, there is a chance that the outer wrapped footage is more faded than the inner, so anticipate that the image will not be entirely consistent throughout.

I used to deal with film that in some cases was more than 70 years old, and did not have the tools currently at our disposal. What I would have given…

jPo, CSI

“I always pass on free advice — its never of any use to me” Oscar Wilde.

[Joseph Owens]

Hey, Ken…

Its not the good old days of OMF anymore… sadly. I just worked through something like this with Johnny B. and eventually we had to resort to an XML or AAF out to Premiere (local circumstances in the studio where the project was based) and then exported a ProTools-compatible EDL from there. It still didn’t keep the original names, everything was “Untitled.” It was a short short but overall the schedule was so after midnight he just threw in the towel, and decided to hand-relink because at least the source codes were right.

I know this is not helpful in a positive sense. But I think 14 is still not really ready for Prime Time, and it is Fairlight – holding things back for the moment, but once it works and the hooks agree, then it will be business time.

jPo, CSI

“I always pass on free advice — its never of any use to me” Oscar Wilde.

[Joseph Owens]

[Glenn Sakatch] “Many have a node for the LUT, and node for the balance, a node for NR. If you use the same workflow and same order of nodes each time, you don’t really need to name them after a while, you just know what they are. If you are sharing work, then you may want to name them.
“

That’s basically how most for-pay colorists operate. Clients are continually asking to be shown what effect a particular action is having on an image and you have to be able to click things on-and-off, before-and-after. Sometimes there will be an issue where some artifact is being introduced, and being able to work through the node-tree is pretty much the only way to troubleshoot a process. Y’know, to figure out where those super-saturated red pixels are coming from.

And, of course, I was recently made aware of a situation where someone (prominent) was tasked as an alternate to re-do a grade, but even though the client apparently went away satisfied…. no, really, not so much. The feedback, by someone else in the client’s production totem pole, groused that there really weren’t enough nodes used to achieve the grade, and therefore not really a very impressive effort.

You are aware, of course that throwing a lot of switches and pulling levers several times is the sure sign of someone knowing absolutely what they’re doing.

jPo, CSI

“I always pass on free advice — its never of any use to me” Oscar Wilde.

[Joseph Owens]

[Duke Sweden] “Kind of odd that I don’t have enough VRAM
GeForce GTX 1050ti”
As if there weren’t enough mysteries. Although it is important, your onboard RAM is not that significant; it is the GPU VRAM statistic – in your case, 4GB – that is causing this warning. Stacking up several frames of 4096 on a single 1050Ti is going to exhaust that resource very quickly, and that is what is happening if you are adding a third-party OFX plugin.

Your render probably won’t crash, but the frame rate is obviously in the dumpster.

jPo, CSI

“I always pass on free advice — its never of any use to me” Oscar Wilde.

[Joseph Owens]

[Luke Sheldon] “if I want to output the image in Resolve to a dedicated screen, it should just be an HDTV and not a computer monitor because of the signal?

So if we get a really color accurate TV (rec.709 at 100%), it will do the job as both a color ref for me and a client monitor?”

Question 1. Not 100% correct, because many graphics displays claim gamut compatibility that is just an input specification. What it does with the image after that is a manufacturer fudge. DVI/HDMI and so on are just interfaces, and whether the display is valid or not depends on:

Question 2. If it is calibrated. This does not mean “covers [Colorspace ie. 709 or sRGB, or…]” but is doing so, as you say, accurately, which will take some deliberate color management on your part. You will be required to purchase a Black Magic device as an interface between Resolve and your judgement display. The User Interface is not a default color reference. It can be done… but it actually requires quite a bit more effort and technology to make it work than simply deploying a valid reference monitor, the way the application is intended to function.

jPo, CSI

“I always pass on free advice — its never of any use to me” Oscar Wilde.

[Joseph Owens]

[Michael West] ” it says I need the CUDA for OSX 10.10.5.”

macOS should be insensitive to CUDA version. Are you sure it is not objecting to a GPU/driver mis-match, because OSX is notoriously hostile to nVidia — especially with non-Apple qualified GPUs (shouldn’t be an issue with an iMac) where version and driver are constantly at war.

jPo, CSI

“I always pass on free advice — its never of any use to me” Oscar Wilde.

[Joseph Owens]

[ellen pollard] “e do have a back-up on the computer, a .drp file.”

I would try importing that drp… it will not open automatically. You have to navigate to it in the Resolve project browser and explicitly open it.

jPo, CSI

“I always pass on free advice — its never of any use to me” Oscar Wilde.

[Joseph Owens]

[David Mathis] “f it is posible to do in Resolve. “

Yes, it is, using keyframes. Review the Chapter covering the Edit Page, Variable-speed effects.

jPo, CSI

“I always pass on free advice — its never of any use to me” Oscar Wilde.

[Joseph Owens]

Haven’t encountered this for a very long time… used to be common with synthetic time codes. Never fully explained, but my impression is that it might have had something to do with a sub-frame mismatch, similar to the TC difference that would occur with 23.976 vs. 24.000 — will conform on a macro level, but technically a mis-match.

jPo, CSI

“I always pass on free advice — its never of any use to me” Oscar Wilde.