[Kevin Cannon]

I often switch between using project-per-reel and session-per-reel. I’ve had at least one 70-80 minute RED project where the complexity of the project bogged everything down until the project had to be duplicated, half the reels deleted and all the media from those reels removed from the master session.

I could be totally wrong on this, but I am under the impression that the RED clips are more RAM intensive because of the fact that they consist of multiple 2GB files that Resolve (and other systems) display as single clips but are adding unseen weight to the media pool. At least it’s more taxing on the system in some way…

Certainly with 12GB of RAM or less you are more likely to run into issues.

KC

Prehistoric Digital
PhD Grading Suite

[Kevin Cannon]

Well, the good news is that (according to Rohit in another thread) the Resolve dongle that manages the license is interchangeable between mac and windows… so the same purchase basically allows free and instant cross-grading if you need to switch.

If you are using graphics cards from MacVidCards and David Pirinelli (which many people on this forum use with no issues), he says those are able to be moved to a windows system no problem.

The Decklink HD Extreme can also be moved into a windows system, probably the other video interface solutions as well.

Any grading monitor you might get will work the same, as well as most or all storage solutions with PCI card interfaces…

So you could easily invest in an entire Resolve grading set-up where the only piece that is OS dependent is the Mac Pro itself. Is there other gear or software in the set-up that you’re considering that is?

Cheers,

KC

Prehistoric Digital
PhD Grading Suite

[Kevin Cannon]

I got the LP2475w so it would match the bevel of the Dreamcolor… so whatever that’s worth. The stand is about identical in terms of height and adjustments, which I always thought was pretty good. I think the cheaper HP looks like it has about the same bevel with a different stand. The backlight adjustment is good but can only go so far – I still find it a bit bright vs. projection, so I’m adding a glare reducing filter to cut some light.

KC

Prehistoric Digital
PhD Grading Suite

[Kevin Cannon]

Well, I’d pay special attention to the fact that the 2140 does not have a 10-bit panel, and does not have the same color fidelity engine that the 2461w does… While the Dreamcolor and the 2461w have native color gamuts that exceed Rec. 709 and approach DCI P3, the 2140 only approaches Rec.709… a “99% match” according to Bram Desmet over on the FSI forum.

Although the Dreamcolor requires that you calibrate it more often, it is something you can do yourself and quickly…

I would definitely agree that the Dreamcolor as a pain to work around and could cost more in total… the question is more about the color accuracy of the 2140w, and Flanders is not shy about stating that the 2461w is their monitor for color-critical applications.

Cheers,

KC

Prehistoric Digital
PhD Grading Suite

[Kevin Cannon]

Hi Ryan,

My experience with the Dreamcolor has been that, calibrated frequently and going through the effort to give it a signal that it can process (progressive, not PsF) you can get a very accurate result over HDMI. A few months ago I compared a Dreamcolor, a Flanders, and a Barco digital cinema projector all receiving the same signal from the DaVinci, and I was still quite impressed with the accuracy of the Dreamcolor.

While I would definitely recommend FSI’s grading monitors (especially the 2461w) they don’t necessarily recommend the 2140 for color grading… it’s possible that the Dreamcolor (and the calibration kit) would give you a better solution at a similar price point…

KC

Prehistoric Digital
PhD Grading Suite

[Kevin Cannon]

I have a 12-core 2.93GHz Mac Pro with 24 gig of RAM and do not use a Red Rocket card, and I get very reliable playback on 1/2 Res good on 4K and 4.5K footage in 1080p or 2K, but a few caveats: I use Ultrascopes, but if you have to use DaVinci’s scopes, they will put more demands on the CPU. Also, I believe that the bezier curves when you create a power curve require CPU computation, and so power curves can slow you down.

And I never was able to get 1/2 res good with Epic 5K footage, and 1/4 res good not doing the trick for me. I’m currently planning on borrowing a rocket card or pre-rendering to DPX when I deal with Epic, at least until I can see how a 16-core Mac Pro would perform…

KC

Prehistoric Digital
PhD Grading Suite

[Kevin Cannon]

Another consideration is that the way the director and DP have worked with the camera can lead you in the right direction. If you have a DP that has a color-critical Rec 709 or p3 monitor on set, in a tent or reasonable lighting conditions, they might be basing many of their lighting decisions – contrast ratios, highlights, fill for shadow details – on the exact RED settings or LUT that they had set in camera and monitored on set. In such a case, creating your own transform to replace the ones they selected might only take you further away from what they’ve already achieved, or take you longer to arrive at the same result.

Of course, you can also be working with a team that shoots using a light meter more than a monitor, perhaps envisioning a characteristic curve for the image that isn’t a typical camera setting. Those might require you to eschew the default settings and create your own curves to bring RedLogFilm or Log-C into line with the vision.

KC

Prehistoric Digital
PhD Grading Suite

[Kevin Cannon]

Yeah, being able to select multiple volumes, or having an “add folders and subfolders based on XML” would be helpful. In the past I’ve done this: Load the XML and select “add media” and point it to your first volume. Once it loads and gives you the log (which inevitably tells you that a number of clips can’t be found) delete that new timeline while leaving the media in the media pool. Then load the XML again, again selecting “add media”, but this time point to your second volume. Do it another time for each additional volume, and you should end up with a media pool full of all the clips you need. Then you can load the XML without adding media and it should connect….

KC

Prehistoric Digital
PhD Grading Suite

[Kevin Cannon]

I think that you have both the proxies and R3Ds in the master timeline and that Resolve is keeping the clips distinct from each other even though they share identical timecode and reel numbers. When you reconform from folders, you point your timeline to different clips in the master session which link up correctly, but those R3D clips on the master session don’t share grades with the proxies.

I’m not sure how you could create a relationship between those clips in the master session- i think it’s intentionally designed to keep that from happening so that projects with conflicts will keep distinct grades on the conflicting files in the master session.

KC

Prehistoric Digital
PhD Grading Suite

[Kevin Cannon]

Try going into Format > Input > Source blanking OFF

I think it’s in input…

KC

Prehistoric Digital
PhD Grading Suite