What one real 1440p Music Motion render tells us—and what it does not.
This report uses a completed production job recorded by the Music Motion render service. It publishes the actual renderer path, frame count and timing instead of turning a Task Manager percentage into a speed claim.
1440p output · 1,379 completed frames
Frame-production rate written by the GPU v3.2 completion log.
From the job-start timestamp to the completed MP4 timestamp.
About 7.9 seconds of work per second of source in this one project.
Verified render path
What the log proves
- The frame renderer ran through the GPU v3.2 path.
- ANGLE used the D3D11 graphics path on an NVIDIA GeForce RTX 3090.
- The H.264 stream used NVIDIA hardware encoding through D3D11 / Media Foundation.
- The completed MP4 contained the full 1,379-frame sequence.
Why 20% GPU does not mean 80% speed is unused
Different engines
Windows displays 3D, copy, decode and encode as separate engines. A low encoder graph does not measure shader work.
A serial pipeline
Asset decoding, scene preparation, browser-to-GPU synchronization, frame completion and muxing can wait on one another.
The safest next test
Compare the same source and effect at 1080p and 1440p, repeat three times, then test one and two concurrent jobs.
Limits of this first observation
- This is a production observation, not a controlled multi-effect laboratory benchmark.
- One render cannot predict every effect, source-video codec, resolution or project length.
- Wall time includes preparation and muxing as well as frame rendering.
- The result must not be interpreted as a guaranteed render time for customers.
Observation captured 25 July 2026. The report will be updated when a controlled multi-run matrix is available.
Judge the visual result before you judge the progress graph.
Open the live demo, choose an effect and test how it responds to your own track. Rendering performance varies by project; the editor lets you validate the look first.
Try the live demo