Activity
Mon
Wed
Fri
Sun
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
Oct
What is this?
Less
More
Muse Collective
648
Free
3 contributions to Muse Collective
9d •
First Look At Muse MiniMax V2!
Ok so I wanted to give you guys a quick look at the MiniMax Director V2 node. It will shortly be available for download. Please note that the V2 will only be available in the new 'Advanced Subscription' tier of the community. All new advanced workflows and nodes will all be available to those that subscribe. There is not a charge per node or advanced workflow. Your advanced tier subscription gives you access to any existing or future advanced nodes or workflows. There will of course still be plenty of free workflows and nodes that anyone in the community can use, however the advanced ones take a lot longer to create which is why I am creating this second tier. You can watch below or on YouTube. https://www.youtube.com/watch?v=zQ6SFUuW7f8 Thanks for your support!
1 like • 9d
OK. I've watched the video. Where do I sign up? Will it be $12 per monthly or $144 billed yearly?
27d •
Minimax H3 Director second pass workflow issue
Minimax H3 Director v1.4 second pass workflow issue Hi Andy - I'm using the prompt override box for test prompts (Prompt override from socket is checked) - without the second pass, the preview and final output all work ok. I'm setting 18 steps total; 14 steps for initial pass, then 4 for second refine stage. If I check Enable Two Stage Sampling, it crashes after generating the 14 steps first pass (no video output). I'm guessing it should do 4 more passes, then move to the latent upscaler? But the run crashes out; seems to be on the temporal chunks checkbox. Doesn't matter whether the temporal chunking is checked or not, same crash. Suggest the second refinement pass and the upscale be separated? (I might have it wrong, thought refinement passes were using the same first-pass model? Or is it trying to use the upscaler model to do the refinement and upscale at the same time?) [INFO] got prompt [INFO] [MuseMinimaxDirector] mode=Reference (Omni) — up to 7 images, 3 videos, 3 audio, 1376x768, seed=1111429277178777, seed_hunt=False, candidates=1, 1 chunk(s): ~15.0s (total 15.0s) [INFO] [MuseMinimaxDirector] chunk 1/1, seed=1111429277178777, length=362 frames, video_carry=False, audio_carry=False, hybrid=False [INFO] Requested to load MiniMaxH3VideoVAE [INFO] Model MiniMaxH3VideoVAE prepared for dynamic VRAM loading. 4965MB Staged. 0 patches attached. Force pre-loaded 128 weights: 348 KB. [INFO] Model MiniMaxH3VideoVAE prepared for dynamic VRAM loading. 4965MB Staged. 0 patches attached. Force pre-loaded 128 weights: 348 KB. [INFO] Model MiniMaxH3VideoVAE prepared for dynamic VRAM loading. 4965MB Staged. 0 patches attached. Force pre-loaded 128 weights: 348 KB. [INFO] Model MiniMaxH3VideoVAE prepared for dynamic VRAM loading. 4965MB Staged. 0 patches attached. Force pre-loaded 128 weights: 348 KB. [INFO] Model MiniMaxH3VideoVAE prepared for dynamic VRAM loading. 4965MB Staged. 0 patches attached. Force pre-loaded 128 weights: 348 KB. [INFO] Requested to load MiniMaxH3AudioVAE
1 like • 11d
Hi Rick, did you ever get this licked? Mine has always done the same thing. I have just been running with "Two Stage Sampling" unchecked and it runs OK
21d •
Variable Length Chunks?
In a workflow like Minimax director 1.4 would it be possible to make variable length chunks? Or do they all need to be fixed length?
1 like • 20d
@Rick Burlow I think that's the point of the director node. You can break down a longer clip into 15 second (max) chunks and each chunk can be broken down into multiple cuts. It would be the same as manually generating clips using a reference image and then generating the next clip using the last frame of the previous clip as the first frame of the next clip etc. it just automates it.
0 likes • 19d
Hi Rick, I am not sure, but my theory is that there might be a slight rounding error in the calculation for the cut lengths. The reason I say that is in a previous attempt with 15 second chunks some of the cuts added up to 14.9 seconds but the chunk said it was 15.0 seconds. Things were fine for the 1st two chunks but then the third chunk stopped singing and then started singing again on the 5th chunk. I changed the chunk duration and then things got jumpy and the singing stopped at different places. Andy is looking at it now. That's why I inquired about the possibility of variable length chunks. I could tweak each chunk slightly to bring the video back into sync with the sound track. Now it seems I have a heterodyne effect where it comes and goes in a beat frequency-like fashion. But not knowing how exactly, the inner workings of the node function, its just my theory.
1-3 of 3
@david-duley-6866
Engineer, Musician and all around polymath. Trying to learn AI video and music making to let out what's inside
Active 55m ago
Joined Sep 17, 2026