Can the model take animated sequence of depth-map as input (or even better, refeed modified depth-map as input while in the running process to continuously move the scene how we wish)?

the highresfix-script for the Efficient Nodes works beautifully, but it's def. not appropriate to use highres-fix to 4x an image without controlnet. It'll easily get between 1k and 2k for me, though.

"restore_cfg -1" - disable restoration. So why -1?

you have a natural talent for explaining and caring about your work, and you have a cute accent

How can I get those Reroute like nodes i.e. you have Get_BASE_IMAGE, as I understand it's some kind of route transfer without actually making Reroute 100times if node is far from things you need.

Your method of hi res fix is strange (4x upscale and 1.0 denoise will yield bad results). Latent hi res fix in general is bad. What's called non-latent hi res fix is much better. Use an image space upscale model at 2x, then run the upscaled image through a KSampler at 0.4-0.5 denoise. 0.4-0.5 is ideal as beyond 0.5 it begins to change composition. Doing hi res fix at 1.0 denoise would be the same as creating a new generation at a much higher resolution than the model was trained on, which will of course result in cloned bodies everywhere.

Love these quick tips videos. Straight to the point!

Mostly, people from comfyUI tutorial tries to describe it like some some form of art: canvas, paint, etc... Which is more confusing... It is much more sense to describe comfyUI like programing (you)... While me, I see comfyUI like flowchart in production... I think people are secretly have that production mentality thats why some people just love comfyUI. It just that they're unconscious to it and can't describe properly...

I find these kinds of summaries useful. Clearly he has verbal diarrhea at the very least (blatantly saying a number of things that are clearly untrue like he was a "secret agent") and approaching George Santos levels of ridiculousness. People can help bring something useful to the world (like Stable Diffusion) while at the same time being a train wreck and dumpster fire of a leader or in other areas of their life or things they do. We've seen this pattern over and over again with a bunch of these tech bros CEOs. Some just lie or scam in their core business (Elizabeth Holmes or Sam Bankman-Fried anyone) and some lie and exaggerate in less fatal ways. Just because they might do something useful or seem intelligent or charismatic doesn't mean we should just give them a pass when they lie and grift. The fanbois need to realize that a "visionary" with a "reality distortion field" might bring something useful to the world, but it doesn't mean their hero is right about everything (or even most things). You can still believe that some things they did were useful while also acknowledging that they're full of crap in other ways or that a lot of things are completely outside their areas of knowledge or expertise. A lot of these "rock star" tech ceos have verbal diarrhea, offering "information" on things they know nothing about. Some will pop off and say crazy things at any opportunity. I guess being brown nosed and told "yes you're always right" by everyone around them makes them believe it. Soon every whim they have or passing thought that goes through their mind is pure gold to them, so why not spout it? Especially when part of their fan base eats up every morsel with a spoon. Because they did something right it changed something, doesn't mean they shouldn't be called out when they're wrong, though.

but he still has teeth ?

1:20 I think you shouldn't put that info cuz it cause confusion... Its like Streisand effect... I understand you want make people understand but it become the opposite... It will just fine if you just point the direct download...

You’re tutorials are so on point. You actually get “under the hood” Cheers! 🍻

¿Y el link para entrar a Supir?

Hi, seems to be great work, combining a lot of skills. Do you think it's worth it to learn from this video or is it too old now? I learned those couple of months some comfyui nodes like IP adapter's and controlnets but not creating Videos still. Could you recommend me something to continue with?

Greetings, I have tried it with Matrix, Pinokio and the portable. Maybe they are compatibility problems but the "SaveVideo" nodes do not load. I don't know how many the same thing has happened to.

Is there such thing as buying a PC with ComfyUi already installed..?

i have error error occurred when executing supir_upscale

Doesn't work. Image Resize node has red outline almost immediately, no idea why: Error occurred when executing ImageResize+: 'bool' object has no attribute 'startswith'

Very useful information ! Thanks!

ahahahah the owl tutorial joke killed me

great job, but for render video we need quantum pc...

Hi Stephan ! Whare are you been ? 2 months we haven't seen you Dude.

I want you to know how much this is appreciated. Thank you.

I think I upscaled my brain after watching this. Thanks bro.

super bien fait mec merci beaucoup 💙

Thank you for explaining the whole process with every little detail. Loved that you got the example of someone that could be any of us and showed why it didn't work and how to fix it. Please continue like this!

Does this work for Animatediff too?

Thank you so much for your lessons, I made a short AI film using your lessons, I set up IPADAPTER and AnimateDiff and got an amazing result, I like it very much kzfaq.info/get/bejne/g6ydh9aY0LK8gnU.html

Great video, Stephan! I've been using ComfyUI for about a week now. Everything was working well for the first couple of days until this 'reconnecting' error started to appear whenever ComfyUI is loading a checkpoint. Initially, I thought it might be a memory issue, but it's not even loading the VRAM usage. Do you have any suggestions on how to fix this? Thanks in advance!

Thanks!! Great content!! 😊

ComfyUI is literally the least comfy ui I've ever seen. What a piece of hot garbage.

I am a little confused by one thing. The SUPIR Sampler Node seems to support only DPM++2M and EDM Samplers. Quite a few of the faster models are optimized for DPM+ + SDE, Euler A, or another sampler. On some models it makes a huge difference. Does this matter a lot for the SUPIR model? Should we spend time testing a model we want to use with SUPIR with DPM++2M and/or EDM ahead of time? SUPIR can be quite slow, so I'd like to make sure that the issue is not simply a question of model/sampler compatibility. Thanks

Someone told me that the portable version is using python_embedded and some nodes che be installed wrongly, while a venv installation should be preferred. Is that any good as advice?

Were do we import our image?

I'm impressed! Big thanks for your effort and your time to spend in this video. For me, it's a bit to fast and I have to stop and play it again... but its very helpfull!

The one thing I do not like about SUPIR is that it's not for commercial use.

It's strange, you have a French accent...

Hi Stephan, I have used your workflow for some time now and everything worked flawless so far, but since I've updated comfyui and the nodes today I get an error when it comes to image resizing: Error occurred when executing ImageResize+: 'bool' object has no attribute 'startswith' Any idea how to solve this?

Can you use supir for video upscaling? Would it stay consistent or not really?

cut and paste for xformers did not work for me. Copy paste should be: python.exe -m pip install -U xformers --no-dependencies This is what is shown in the video @1:20, but is not in the C/P text.

Amazing! I need more of there compare vids

how to run run_nvidia_gpu.bat?

how to show CPU, GPU, RAM and VRAM like your Comfyui?

The size of the source image is 2048 * 2048. How to divide it into tiles? The image generated after I divide it into tiles is also divided, not the entire image. What settings are wrong?

please link Civitai

why don't you show the final node where the image preview is shown. what is that node? jumping around aint helping. take it slow, and make it clear dude.

Great video. Thanks for the effort

Hello, This is this first workflow that really works (at least for me), thank you! I wrote you already I was looking for img2img workflow to get a realistic skin other than upscaling, this work perfect! I'm now trying with different Loras to achieve what I want, thank you again! Just a thing, the SUPIR Conditioner node doesn't have a positive prompt but just the negative, in the workflow I downloaded. EDIT: I realized now I have downloaded the next version you did :)

Thank you! This works well for just redoing an inpainting group without having to have a seed fixed for the original output!

Hello and thanks for the video! Do it exist a easier wotkflow, this one is huge for a newbie as I am :), just to make face skin more realistic and less plastic? Just an img2img to add detail and do a skin more realistic :)