You are very welcome, Massta.
When you say that clips go out of sync over time, I am curious:
- How much time? After 1 hour? 2 hours? Can you do a test and report back? Do you have the computer doing any other tasks that could be taxing the CPU/graphics? Are sources coming from the same BOOT drive? Is the external drive(s) – if used – a RAID of minimum 2-bay drives, or a single-platter/single-drive?
- Because I am rolling movie clips from multiple camera angles - at times for the entire movie (90min~140min, supplied by the movie studio and the actual film's editor), I can see if anything slips by even 1 single frame. Amazingly, without feeding Resolume Arena SMPTE timecode as a reference, everything is in lock and perfect (what we in Film/TV legacy call "Frame Accurate," which means it NEVER slips. That is a credit to Resolume's programmers! I have left a 90min movie with 4 camera angles on LOOP playback and walked away overnight, and came back the next morning 7 hours later and Resolume had not slipped 1 frame.
I bring this up because in my case, I don't touch Arena once I hit >Play. And all my clips are the same length. In your case, for DJ/VJ work, clips could all be different lengths (RT/TRT: Run Time, Total Running Time) and further, different frame rates, resolution, aspect ratios, which Resolume will 'conform' to whatever your OUTPUT standard is. And that could make it seem like things drift out of "Frame Accurate" sync.
Finally, Sir Zoltan has an amazing sidecar utility called "CLICKABLE TIMELINE" –see his other response here on this very page in response to you to find his web link). This has worked for us when a visiting producer, director or film teacher yells out to me, "Stop right there and show that scene again!" – I use this utility to click on ONE timeline in a side-window, and all Resolume Layers sync to that, so I am not trying to manually type-in long digits on each layer's TRANSPORT and clicking PLAY again. Using Zoltan's utility, ONE timeline is a "Master Control" of all CLIP LAYERS in a Column – and it works flawlessly (BRAVO, Zoltan!
- You may want to search my other, detailed explanation here in Forums of the difference between DROP-FRAME TIME CODE (DF or DFTC, which matches "Real" CLOCK TIME); and NDF (Non-DropFrame), which is 3.6 seconds longer per hour. This is mainly a US/Canada/Mexico and Japan/Asia issue, because DFTC had to be invented in the 1950s/60s by EECO for the TV Broadcasting and first videotape editing computers to make time code match the clock we time to.
If you are generating your own clips to send to Resolume, it will help if you are consistent and ensure all your timecode is IDENTICAL in Frame Rate and Timecode type, if you need to troubleshoot sync drift. Editing software allows you to freely mix NDF and DF and frame rates, obviously – BUT YOUR FINAL, SEQUENCE TIMELINE's timecode must be just ONE type/frame rate. In our case, even if we get NDF, we convert ALL video clips sent to Resolume to DF: Drop Frame, 29.97fps, so we match real clock time as a reference. In DF, the frame numbers :00 and :01 do not exist every minute, except at the top of every minute where it does appear. This does not mean your video has frames literally "dropped" – it just means we don't assign the frame numbers :00 and :01 for those 2 frames, it jumps from frame HH:MM:SS;29 to HH:MM:SS;02 – the "labeling" of the frame is affected (H: hours M:Minutes S:Seconds). Editing software allows mixed time code types by doing the calculation between DF/NDF and frame rates from different countries by doing the math for you automatically.
massta wrote: ↑Sun May 24, 2020 14:11
Thanks for your input and rendering settings. Yes, I've requested the long fade to black at the end of clips. It is probably the only solution. I've tried adding empty cells between clips but don't think its helping.
Another issue, columns do get out of sync over time and I find myself needing to trigger the first column ever so often.
I'm going to be using Resolume for another permanent installation but really wish these two issues were remedied.