Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

DaVinci Resolve Results #3

Open
camkerr opened this issue Dec 12, 2024 · 0 comments
Open

DaVinci Resolve Results #3

camkerr opened this issue Dec 12, 2024 · 0 comments

Comments

@camkerr
Copy link

camkerr commented Dec 12, 2024

Description

I have run the test suite through DaVinci Resolve. I have discovered some discrepancies with some of the time warps as well as marker placement on time warped clips. I have attached my time_warp_test.davinci_resolve.ocr_results.txt file with the full output.

Summary of Issues

  1. Whenever a time warp results in the clip running out of available frames for a given duration, Media Composer & Resolve do not display the same last available frame when they run out of frames and hold.

  2. None of the trim-to-fill clips have a Speed Change applied in the Resolve Inspector. Likely due to Trim-to-fill time warp effects appear as the wrong type #2

  3. None of the Backwards time warps are matching. There is also an issue called out in Reverse time warps have wrong time_scalar #1 It results in one of two outcomes:

    1. A frozen frame for the duration of the entire clip. Or...
    2. A backwards time warp is applied, but starting frame and rate do not match Media Composer.
  4. The last frame of clip Reverse 50% (frames 99 to 50, each 2x) reports media offline. When a render is attempted, the error Resolve gives is:

Screenshot 2024-12-10 at 9 41 11 AM

The timecode in that error is the timeline playhead at the location of the offline media. The clip is trying to play 00:59:59:23 which doesn't exist because test_pattern_media_1920x1080_24_h264.mov starts at 01:00:00:00. To get a render out of Resolve for the rest of the timeline, I had to trim the last frame off this clip to avoid this render error.

  1. The clip Marker is not visible on some clips: Freeze frame in middle, Freeze frame at end and all the Backwards time warps. The Marker is still attached to the clip itself, but it appears to always be attached to the very start of the clip. See below as I duplicate the clip to a new track and roll and extend the clip to the head, the marker becomes visible.

time_warp_marker_placement

Detailed Test Suite Report

Legend

  • ✅ Time Warp Matches
  • 🏷️ Marker Error
  • ❌ Time Warp Does Not Match

Clips From Test Suite

  • ✅ Full clip (no effects)
  • Segments:
    • ✅ Segment at start (no effects)
    • ✅ Segment in middle (no effects)
    • ✅ Segment at end (no effects)
  • Freeze frames:
    • ✅ Freeze frame at start
    • ✅ 🏷️ Freeze frame in middle
      • Error: Marker on retimed clip in incorrect location
    • ✅ 🏷️ Freeze frame at end
      • Error: Marker on retimed clip in incorrect location
  • Linear time warps by percentage
    • ✅ Identity 100% speed
    • ✅ Slow down to 99% speed (is 1 frame repeated?)
    • ✅ Slow down to 90% speed
    • ✅ Slow down to 50% speed
    • ✅ Slow down to 10% speed
    • ❌ Speed up to 101% speed (is 1 frame skipped?)
      • Error: Last frame of clip differs by 1 frame
    • ❌ Speed up to 110% speed
      • Error: Last 10 frames of clip are held and differ by 1 frame
    • ❌ Speed up to 2x (200%) speed
      • Error: The second half of the clip (last 60 frames) held and differ by 1 frame.
    • ❌ Speed up to 10x (1000%) speed
      • Error: The last 108 frames of the clip held and differ by 9 frames. Only the first 12 frames match the time warp.
  • Linear time warps fit-to-fill
    • ✅ Fit-to-fill 99 frames into 100
    • ✅ Fit-to-fill 90 frames into 100
    • ✅ Fit-to-fill 50 frames into 100
    • ✅ Fit-to-fill 33 frames into 100
    • ✅ Fit-to-fill 5 frames into 100
    • ✅ Fit-to-fill 100 frames into 99
    • ✅ Fit-to-fill 100 frames into 50
    • ✅ Fit-to-fill 100 frames into 33
    • ✅ Fit-to-fill 100 frames into 10
    • ✅ Fit-to-fill 100 frames into 9
  • Linear time warps trim-to-fill ("trim" is different from "fit")
    • ❌ Trim-to-fill 99 frames into 100 (BUG: TimeEffect should be LinearTimeWarp)
      • Error: No Resolve Speed Change Applied.
    • ❌ Trim-to-fill 90 frames into 100 (BUG: TimeEffect should be LinearTimeWarp)
      • Error: No Resolve Speed Change Applied.
    • ❌ Trim-to-fill 50 frames into 100 (BUG: TimeEffect should be LinearTimeWarp)
      • Error: No Resolve Speed Change Applied.
    • ❌ Trim-to-fill 33 frames into 100 (BUG: TimeEffect should be LinearTimeWarp)
      • Error: No Resolve Speed Change Applied.
    • ❌ Trim-to-fill 5 frames into 100 (BUG: TimeEffect should be LinearTimeWarp)
      • Error: No Resolve Speed Change Applied.
    • ❌ Trim-to-fill 100 frames into 99 (BUG: TimeEffect should be LinearTimeWarp)
      • Error: No Resolve Speed Change Applied.
    • ❌ Trim-to-fill 100 frames into 50 (BUG: TimeEffect should be LinearTimeWarp)
      • Error: No Resolve Speed Change Applied.
    • ❌ Trim-to-fill 100 frames into 33 (BUG: TimeEffect should be LinearTimeWarp)
      • Error: No Resolve Speed Change Applied.
    • ❌ Trim-to-fill 100 frames into 10 (BUG: TimeEffect should be LinearTimeWarp)
      • Error: No Resolve Speed Change Applied.
    • ❌ Trim-to-fill 100 frames into 9 (BUG: TimeEffect should be LinearTimeWarp)
      • Error: No Resolve Speed Change Applied.
  • Backwards time warps
    • ❌ Reverse 100% (frames 99 to 0)
      • Error: Appears to be a bug with the "Duration" calculation in the Speed Change section of the Inspector. A value of just 00:00:00:01 resulting in no visual time warp and just a held frame for the duration of the shot.
    • ❌ Reverse 50% (frames 99 to 50, each 2x) (BUG: -0.51 should be -0.50)
      • Error: A reverse speed change is applied, however starting frame is is not correct. Also a slight difference which frames get duplicated.
    • ❌ Reverse 200% (frames 99 to 0, on 2s) (BUG: -1.0 should be -2.0)
      • Error: Appears to be a bug with the "Duration" calculation in the Speed Change section of the Inspector. A value of just 00:00:00:01 resulting in no visual time warp and just a held frame for the duration of the shot.
    • ❌ Reverse 30% (BUG: -0.31 should be -0.30)
      • Error: A reverse speed change is applied, however starting frame is is not correct. Also a slight difference which frames get duplicated.
    • ❌ Reverse 120% (BUG: -1.0 should be -1.2)
      • Error: Appears to be a bug with the "Duration" calculation in the Speed Change section of the Inspector. A value of just 00:00:00:01 resulting in no visual time warp and just a held frame for the duration of the shot.

Optional


Environment

Operating System: Windows 11 23H2
Host Application: DaVinci Resolve Studio 19.1 Build 12

Reproduction Steps

  1. Create new Resolve Project.
  2. File > Import > Timeline

resolve_import_timeline_menu

  1. Navigate to and select time_warp_test_suite.otioz

resolve_load_otio_popup

  1. Go to the Deliver page and add an H.264 to the Render Queue for this timeline.

  2. Render

During this render is where I'm flagged to the Media Offline error mentioned above. The very last frame of the 2nd backwards timewarp clip gives a Media Offline and a Render Job 1 Failed pop up. To complete the render of the rest of the timeline, I just trimmed the last frame from that clip and re-rendered. I called this file time_warp_test.davinci_resolve.mov

  1. Use this QuickTime with the ocr_frame_counter.sh from the test suite to generate time_warp_test.davinci_resolve.ocr_results.txt

% ./ocr_frame_counter.sh time_warp_test.davinci_resolve.mov > time_warp_test.davinci_resolve.ocr_results.txt

Log output if appropriate

Full output of time_warp_test.davinci_resolve.ocr_results.txt is attached.

Sample data and other attachements

time_warp_test.davinci_resolve.ocr_results.txt

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant