kijai

Flip a switch and stop your Videos from Degrading. 50 Clip Test No Degrading at the end. Geiru Context Suite I did the thing where I nuke the first post because I forgot to do something also the title sucked and I forgot to credit literally the most important contributor in the video. Skip to 8:11 to see the results. Also, if you guys are curious you can visit this repository to read about everything I tested to get to this point. roycho87/NerdNotes Lots of information in there. Maybe you guys can read it and come up with an even better solution. If you have some good info, please leave an issue on the Github and I'll compile it all together. Any experiments you've done and the results, what you've learned etc. All of it is helpul. Geiru Context Suite: degradation protection for long MiniMax H3 chains The main part of this that allowed this to work came from Ethanfel. He created the context noise method that allowed this to work so giant shout out to him. The Clown Context Suite is now the Geiru Context Suite. She's named after the clown, and the new name keeps it from getting mixed up with ClownShark's nodes. Your old workflows still load fine. The big addition is a degradation protection switch on Geiru Context that turns three nodes on together. Geiru Keystone locks a clean frame 0 from a source clip into every shot. Geiru Late Lock tacks that clip's capstone onto the back late in the steps. Geiru Context Noise, based on ethanfel's idea, gives the held frames part of the noise instead of handing them over perfectly clean. Turn it on for long stretches of one scene and off around cuts, since scene changes reset the drift on their own. Keystone can also load its source clip straight from your latent folder now, so when the scene changes you only change one number. Also new: a context noise switch on Geiru Hybrid Windows, and a ready-to-go example workflow, noisy_clown_example. I tested it on a 50-clip chain (about 5½ minutes): running on a beach, then a café, a walk through town and a garden. The plain chain was falling apart by clip 20, while the protected one still looked clean and barely drifted in the still scenes. Side by sides and all the numbers are in my notes. It's not perfect yet, since fully static shots can still slowly darken, but it's the best I've gotten so far! Repo: roycho87/geiru_context_suite Test notes: roycho87/NerdNotes Credits (none of this exists without them): - seitanism, for the masked latent context everything here is built on: seitanism/ComfyUI-H3-Motion-Context-MultiRef - NikoDemon80, for Motion Context, the Chain node and a ton of drift testing: NikoDemon80/ComfyUI-H3-Motion-Context - ethanfel (Lumifel), for the context noise idea: ethanfel/ComfyUI-MiniMaxH3-Context-Loop - drozbay, for H3 latent masking: drozbay/MaskVidExperiments - kijai, for H3 support and context windows in ComfyUI: github.com/kijai - JalenBrunson, for Hybrid Windows: Jalen-Brunson/ComfyUI-HybridWindows - TrentHunter82, for TrentNodes, whose degrade node made the frame 0 tests possible: TrentHunter82/TrentNodes

I need something new to work on. Skip to 8:11 to see the results. Geiru Context Suite: degradation protection for long MiniMax H3 chains The main part of this that allowed this to work came from Ethanfel. He created the context noise method that allowed this to work so giant shout out to him. The Clown Context Suite is now the Geiru Context Suite. She's named after the clown, and the new name keeps it from getting mixed up with ClownShark's nodes. Your old workflows still load fine. The big addition is a degradation protection switch on Geiru Context that turns three nodes on together. Geiru Keystone locks a clean frame 0 from a source clip into every shot. Geiru Late Lock tacks that clip's capstone onto the back late in the steps. Geiru Context Noise, based on ethanfel's idea, gives the held frames part of the noise instead of handing them over perfectly clean. Turn it on for long stretches of one scene and off around cuts, since scene changes reset the drift on their own. Keystone can also load its source clip straight from your latent folder now, so when the scene changes you only change one number. Also new: a context noise switch on Geiru Hybrid Windows, and a ready-to-go example workflow, noisy_clown_example. I tested it on a 50-clip chain (about 5½ minutes): running on a beach, then a café, a walk through town and a garden. The plain chain was falling apart by clip 20, while the protected one still looked clean and barely drifted in the still scenes. Side by sides and all the numbers are in my notes. It's not perfect yet, since fully static shots can still slowly darken, but it's the best I've gotten so far! Repo: roycho87/geiru_context_suite Test notes: roycho87/NerdNotes Credits (none of this exists without them): - seitanism, for the masked latent context everything here is built on: seitanism/ComfyUI-H3-Motion-Context-MultiRef - NikoDemon80, for Motion Context, the Chain node and a ton of drift testing: NikoDemon80/ComfyUI-H3-Motion-Context - ethanfel (Lumifel), for the context noise idea: ethanfel/ComfyUI-MiniMaxH3-Context-Loop - drozbay, for H3 latent masking: drozbay/MaskVidExperiments - kijai, for H3 support and context windows in ComfyUI: github.com/kijai - JalenBrunson, for Hybrid Windows: Jalen-Brunson/ComfyUI-HybridWindows - TrentHunter82, for TrentNodes, whose degrade node made the frame 0 tests possible: TrentHunter82/TrentNodes

Moss (turbo) without the blur for realism photography... I'm about to head out for the rest of evening so just wanted to quickly drop this tip here in the hope that others might tinker with this and come up with something better or trash it as you see fit. The setup is a regular Moss workflow you can get from Comfy UI but with the settings outlined below. I stumbled on this setup by accident but so far it's the best results I've gotten using turbo and without any upscaling. It also gives me great variety. Each of these loras by themselves seems to trash the diversity or gives too much blur, yet combined I get a good variety of people without noticeably degrading the overall quality. I've been struggling with Moss for a week, trying all sorts of tips that never got rid of the Moss "blur" with turbo loras. So this is for those of you who've been having issues like myself. For those who swear by 50 steps, yes your results are superior and we should all be ashamed of wanting to generate images faster :( Checkpoint: Solordz Moss Base Int8 (I'm on v10) civitai.com Loras: Add both, yes BOTH, of Lodestone's turbo Loras from here: huggingface.co The newer one at strength 1, the older one at strength 0.5 And add this lora at strength 0.5: civitai.red I'm running at 12 steps, 1.5 megapixels, Euler, Simple. Other than that it's a regular Moss workflow. For good measure, here's my negative prompt: worst quality, bad anatomy, watermark, jpeg artifact, bad face, bad eyes, beauty retouching, skin smoothing, waxy skin, plastic skin, bad environment, doll-like face, piercing, piercings, text, watermark, severe anatomical deformation, extra fingers, extra limbs, text, out of focus, depth of field, blurry And the positive prompt was simply: A photo of a man hiking through a snowy forest, walking through the snow. Note: the first image was slightly different as it wasn't focused on walking but included it for variety. Also, whilst it's not needed, I do use the "Moss Image Prompt Builder" node from KJ which you can get from here: kijai/ComfyUI-KJNodes It allows you to define regions the same way as Ideogram 4 and it's fantastic. Also, for those whom it concerns, this setup is very adult friendly.

Endless MiniMax H3 (with Endless LipSync) v1.6 (with less Motion Context degradation) Endless MiniMax H3 (with Endless LipSync) v1.6 When MiniMax H3 came out, I was badly missing the Endless Wan 2.2 I2V (SVI 2 Pro) features, so at first I look at other options, but most of them included an AIO node that could do everything, and I couldn't use most of my workflow, because they did everything inside that huge node. The only exception to this, was ComfyUI-H3-Motion-Context, which I could easily integrate into my setup. Its big shortcoming though, was that it just created single videos. No way to concatenate them without the lossy step of decoding and re-encoding in a video editor. So, I created a custom node that could do just that, and voila.. Endless MiniMax H3 (with Endless LipSync) v1.0 A simple workflow to create MiniMax H3 videos of unlimited duration, using ComfyUI-H3-Motion-Context and H3 Motion Context Clip Stitcher. I can easily create a 1:30 lip-synced video myself, with a RTX 3060 12GB at around 2 hours (with retries). It can use both FL2AV and Ref2AV, and can also create normal MiniMax H3 videos. The extra parts though, are the saving/loading of the latent from every generation we do, and the stitching of all (or some) of them, whenever we want a full video. Nothing visible at the connections, no indication that there were more than one video. Now, with much less degradation for multiple clips without a camera cut. We can try and re-try every generation, looking for the best one, and then proceed to the next. We can re-do any previous generation if we like too, but we will not be able to use the clips after it (previous clips are not affected), because they are in a way, "fused" with the replaced one. Version 1.6 - Recreated the Configuration panel and other subgraphs, to make them more compatible with the newer front-end versions (fixed the reset values after refresh bug) - Added an optional Clip Refiner optimization group Version 1.5 - Project name that creates a directory with all project's generations - Added Purge all clips group - Connected the 5th Ref Image that was disconnected - Simplified Audio 1 Forced Multi - Refiner that fixes (almost completely) the overexposed/over-contrasted clips from H3 Motion Context, on a long video without much movement. Controls The RED nodes Enable/Disable parts of the workflow. - Generation Mode (select only one) - T2V / I2V (FL2VA) - REF2VA - Optimizations (select as many as you want, but only 1 Attention and/or only 1 Cache) This panel controls - All the nodes that are inside the Optimizations/LoRA subgraph. - The H3 Motion Context Clip Refiner subgraph, that removes much the degradation that accumulates when stitching together latent clips without camera cuts (it's clearly visible after 4-5 clips, but is apparent from the 2nd-3rd clip). It's not needed if there are camera cuts in the clip, so you can turn it off to speed-up the generation. - Reference Items (select as many as you want) Special usage for the <Audio 1> Forced Multi, more for them later. - Setup (select as many as you want depending on the goal) - Generate starts a generation. You don't need this if you just stitching clips - Preview enables the main video preview that can show you where the generation is going before it finishes, so you can stop bad generations - Use previous clip, uses the last part of the previous generation to start the current one, continuing the video. Enable it if you want the current generation to be stitched with the previous video - Stitch clips, stitches all the clips (depending on the Stitcher's settings) from the h3_context folder (this is the default folder that the H3 Motion Context node uses to save the latents) - Stitch last only, stitches only the last video with the current - Save Video just saves the generated video - Purge all clips, deletes all clips from the current Project directory The GREEN nodes are various settings nodes that must be setup. Apart from the Prompt, Seed and LoRA nodes, the most important are these: - Configuration It contains all the settings for the video. The most important setting for the clip stitching, is the Save to clip Index number. This specifies the clip file's number, that the latent of the current generation will be saved to (overwriting any previous existing file). This number also tells us who is the previous clip file that we use to start our current video clip. - Setup multiple Forced Audio clips This is used if we use our own audio for the video, and we want to also stitch many clips together. More about it in the Usage section. Usage Most of the settings are self explanatory (like Optimizations or enabling Reference items). Here, I will just list the main goals of the workflow. The first thing to do, is to set the Project Name and the other video options at the Settings. This Project Name will be used to save your generated clips and videos in a folder of the same name, under the EndlessMMH3 directory. - Create a normal video - Select Generation Mode - Enable Generate, Preview and Save Video - Save to clip Index to 1 - You get a video and that's it. - Create a lip-synced video - Select Generation Mode - Enable Generate, Preview and Save Video - Enable <Audio 1> Forced Multi and load an audio file (never mind about the settings, except for the Start offset) - You get a video that is lip-synced with the provided audio - Create an Endless video - Create a normal video - Enable Use previous clip - Save to clip Index to 2 and generate - Save to clip Index to 3 and generate - Save to clip Index to 4 and generate - ... - You get many small videos, each one of them starts with the ending of the previous one. At a later time, you will use the Stitcher, to stitch them all together to one full video. - Create an Endless video, stitched - Enable Stitch clips With every generation, the Stitcher will stitch all the previous clips with the currently generated one. This way you always get the full video to check. - If you also enable Stitch last only, only the previous and the current videos are stitched together, so you can check the connection without waiting for the full video to be created. You can always stitch them all together at the end. - You get a full video every time, or just the last 2 videos connected, for previewing the connection. - Create an Endless lip-synced video, stitched - Create an Endless video - Enable <Audio 1> Forced Multi - At the Setup multiple Forced Audio clips panel there are some settings. - Start offset: At the 1st gen, you put here the initial offset that you want for the song (e.g. where the lyrics start). After every successful generation (when you advance the Save to clip Index number), you must copy here the value that is in the Copy to Next Start offset box. - Frame offset (ignore if 1st clip): Never mind at 1st generation. After every successful generation (when you advance the Save to clip Index number), you must copy here the value that is in the Copy to Next Frame offset box. - context_length must be the same value everywhere (here, at the Motion Context, and at the H3 Motion Context Clip Stitcher). It's the number of common frames the 2 video clips use to blend together. - You get a full lip-synced video every time, or just the last 2 videos connected, depending on the Stitch clips and Stitch last only settings. - Just stitch the clips together You just have to enable the Stitch clips and the Save Video All (or some of them depending on the settings), of the clips in the current h3_context folder, will be concatenated to a single full video. Notes: - All generated clips that need to be stitched, must have the same dimensions. - You can organize past generations inside EndlessMMH3 folder, using the Project Name to create the directories. - You can stitch together clips that are generated from either fl2va or ref2va. - In this workflow, I don't use the normal ref2va model in the Reference to Video node, but rather the fl2va with the ref_lora_layer20-49adaln LoRA that has better quality. You can check some LoRAs with different weights here, or totally bypass the LoRA and use the normal ref2va model. Models used: - minimax_h3_fl2va_pruned_w4a8_mixed - minimax_h3_ref2va_pruned_w4a8_mixed - qwen3vl_32b_heretic_minimax_h3_nvfp4 - minimax_h3_video_vae_int8_convrot - minimax_h3_audio_vae_fp32 - minimax_h3_fl2v_turbo_8step_v1.0_comfyui_resized_avg_rank_21_bf16 - minimax_h3_ref_lora_layer20-49adaln_proj_rank_256_bf16 Custom Nodes used: - rgthree-comfy - ComfyUI-KJNodes - ComfyUI-Easy-Use - ComfyUI-VideoHelperSuite - ComfyUI-H3-Motion-Context - ComfyUI-noEmbryo - ComfyUI-Spectrum-MiniMax-H3 (optional) Get the workflow at Civitai or in a gist..

How to Generate AI Videos (For Reference Only) May update this post to a proper guide if I find the time, for now, Google will be your best friend. I would highly recommend learning ComfyUI through Pixaroma's course - youtu.be (no, this is not paid nor sponsored) to help you understand the more nuanced image and video workflows you will be trying. 1\) You will want at least a 4060 or faster GPU for your own sanity. Install SageAttention first. From my experience, this is easier done on a portable or manual installation of ComfyUI. 2\) Choose between Wan2.2, LTX 2.3, or MiniMax H3. I only have experience genning NSFW with Wan2.2, the oldest and lowest quality of the three. This being said, MiniMax, though state of the art, was not trained on NSFW data and therefore only excels in SFW gens as of writing. I have not tried recent releases of LTX. 3\) Open an official template from Comfy-Org or download a workflow for the model of interest from a reputable source, e.g. Kijai - github.com/kijai or AI Search - youtu.be