Forum Replies Created

Page 6 of 107
  • Marc Wielage

    January 11, 2020 at 11:01 pm in reply to: Input color space issue

    It may not matter. If you’re monitoring in Rec709 and looking at a calibrated Rec709 display, and you;’re delivering to a Rec709 audience (like Vimeo), then I’d say it should be OK. Upload a short test and look at it for yourself and see if it looks OK.

    The input color space is just there for people who would rather work in the camera’s native color space. I generally don’t like to do that, but I can see the value of using a Color Space Transform node in the node tree, which can essentially do the same thing, shot-to-shot.

  • Marc Wielage

    January 11, 2020 at 3:15 am in reply to: Input color space issue

    It depends. What space was picked? What are you delivering in? What is the source material?

    I routinely do everything in DaVinci YRGB rec709 gamma 2.4, but I also deliver in Rec709 and used calibrated Rec709 displays. If you’re using a display-referred workflow, what you see is pretty much what you get.

  • Marc Wielage

    December 24, 2019 at 1:27 am in reply to: Low light limit

    I think your entire approach to correcting Red Camera files is not right. Have you done this before?

    Typically what I do is something along the following:

    1) set project Raw settings to Color Science IPP2, Color Space RedWideGamutRGB, Gamma Curve Log3G10

    2) set project Camera Metadata to use ISO, Exposure, and Color Temp.

    3) from then on, I just treat it as if it were film. In the first node, I use an Offset adjustment (essentially printer lights) to rebalance the picture and correct for any black level issues. After that, I use a Custom Curve to push the picture in what I call a “Faux Rec709” area with an S-curve. It helps to have a camera color chart for accuracy.

    4) In subsequent nodes, I adjust overall levels and color balance, and I loosely match them to the client’s dailies, brought in as a reference movie. If we decide the story is told better with a bigger change, I can always compare their original editorial “look” to the new look. Multiple Versions allow you to keep their original look or go with a radically new look.

    5) I generally correct in DaVinci YRGB and Gamma 2.4, on a Rec709 2.4 display. I use a calibrated Rec709 display and external scopes, but I match the latter to the internal scopes so there’s no discrepancy. Being an old-school guy, I tend to work in IRE values, 0 for black and 100 for white, but it’s just a system for numbering; you can go with 0 and 1023 if you prefer for 10-bit. This will automatically scale if you render out to an 8-bit medium for YouTube or something like it.

    I think going beyond this methodology introduces a level of complexity that’s going to confuse you. There are pros and cons to ACES, and that would be another potential way to go.

  • Marc Wielage

    December 24, 2019 at 1:26 am in reply to: Low light limit

    I think your entire approach to correcting Red Camera files is not right. Have you done this before?

    Typically what I do is something along the following:

    1) set project Raw settings to Color Science IPP2, Color Space RedWideGamutRGB, Gamma Curve Log3G10

    2) set project Camera Metadata to use ISO, Exposure, and Color Temp.

    3) from then on, I just treat it as if it were film. In the first node, I use an Offset adjustment (essentially printer lights) to rebalance the picture and correct for any black level issues. After that, I use a Custom Curve to push the picture in what I call a “Faux Rec709” area with an S-curve. It helps to have a camera color chart for accuracy.

    4) In subsequent nodes, I adjust overall levels and color balance, and I loosely match them to the client’s dailies, brought in as a reference movie. If we decide the story is told better with a bigger change, I can always compare their original editorial “look” to the new look. Multiple Versions allow you to keep their original look or go with a radically new look.

    5) I generally correct in DaVinci YRGB and Gamma 2.4, on a Rec709 2.4 display. I use external scopes, but I match them to the internal scopes so there’s no discrepancy. Being an old-school guy, I tend to work in IRE values, 0 for black and 100 for white, but it’s just a system for numbering; you can go with 0 and 1023 if you prefer for 10-bit. This will automatically scale if you render out to an 8-bit medium for YouTube or something like it.

    I think going beyond this methodology introduces a level of complexity that’s going to confuse you. There are pros and cons to ACES, and that would be another potential way to go.

  • Marc Wielage

    December 17, 2019 at 3:48 am in reply to: particles in adjustment clip

    I think because of the speed and processor load issue, you would be better off rendering the particles, then taking that rendered clip and compositing it over the other shots, rather than doing it entirely in one operation.

  • Marc Wielage

    December 15, 2019 at 9:36 am in reply to: p3 flanders cinema grading

    To me, the only time to color-correct for cinema is with a projector. You’ll never know how it’s going to translate until you see it in a DCI environment.

    What we have done is color-correct entirely in Rec709 on an OLED display, then move the whole session at the very end into a theater, and do a trim pass from start to finish on the projector. Generally, this can be done in 2 or 3 days. You wind up with a Rec709 version that works fine for broadcast, streaming, home video, and so on, and a DCI P3 version for theatrical/festival screening. No surprises either way.

  • Marc Wielage

    December 9, 2019 at 12:19 pm in reply to: Using De-Flicker Filter for NTSC/PAL Problem

    I can’t tell you if it’ll work or not without seeing the actual material. Try it and see. Often, confining the deflickering to one area will help. But the reality is that if there’s a lot of movement in the material (particularly camera movement), it can affect the results and create a lot of artifacts.

    I just had a case on a recent project last week that took me 20 minutes to figure out a combination of settings that eliminated 90% of the flicker but caused no artifacts. But it took me about 10 tries before I hit on the method. Keyframes helped. Confining the area (with Advanced Settings) helped. Comparing the various Deflicker modes helped. But my situation isn’t your situation. There are no rules — just hard work and experimentation.

  • Marc Wielage

    December 9, 2019 at 5:06 am in reply to: Using De-Flicker Filter for NTSC/PAL Problem

    These are tough problems to solve, but the Advanced Settings of Deflicker may help to a point. This is the kind of thing that you’re better off fixing there on set instead of in post: there are often prices to pay with processing like Deflicker.

  • Marc Wielage

    December 9, 2019 at 5:04 am in reply to: Output too high contrast on macbook pro

    Read these:

    Grading for Mixed Delivery: Cinema, Home, and Every Screen in Between
    https://blog.frame.io/2019/10/14/grading-mixed-delivery/

    and I think it covers the issues and the solutions very well. Understanding color management is also helpful:

    Color Management for Video Editors
    https://jonnyelwyn.co.uk/film-and-video-editing/colour-management-for-video-editors/

  • Marc Wielage

    December 5, 2019 at 11:26 pm in reply to: Render time increasing indefinitely.

    [Jamie Hutber] “I have just purchased Resolve Studio but and having some initial teething issues.”
    Which hardware (CPU, GPU, RAM, disk I/O speed)? What OS? What source material? What timeline resolution? What framerate? What format are you trying to render to? How does your drive rate with Blackmagic Speed Test?

Page 6 of 107

We use anonymous cookies to give you the best experience we can.
Our Privacy policy | GDPR Policy