@ChenHaoYan-z1c Thank you for your comment. I'm glad the videos and workflows were able to help you a little. Am I remembering correctly that you wanted to get a 4090? I can't keep up with that, with my 4080 and 16GB VRAM I'm already reaching the limits with Flux, especially if I want to record a video in parallel. How is it working with the workflows with a 4090, do you ever run into memory problems? Best regards

@ Enhancing with Flux is more time-consuming, while using other models with the 4090 is effortless, with outputs occurring in seconds. However, the Flux workflow normalizes the processing, making it more efficient despite the longer time. /cry

Wow, what an honor! I'm absolutely thrilled to receive such a wonderful comment from the person I consider the best SDXL model creator out there. Thank you so much for your kind words-they mean the world to me! ❤️

@@murphylanga dixit the best noodles maker, aha !!

@greathawken7579 Ok, that's a clear statement, thanks for that.

@muhm00d Thank you, I'll keep my fingers crossed for you. Soon, if everything goes well this evening, we'll be posting a video explaining a simple workflow that you can do an incredible amount with. We recommend the SDXL version of it, which runs much faster and works even on a weaker computer.

@abdunabiaxmedov 1. Install Git Download Git from the official website. Run the installer and follow the installation steps. Make sure to: Select the option to add Git to your PATH during the installation process! This option is usually labeled as "Git from the command line and also from 3rd-party software." 2. Verify Git Installation After installation, verify that Git is working: Open a new Command Prompt or PowerShell window. Type: cmd copy the code from the next line: git --version Steps 3 and 4 are usually not necessary You should see the Git version displayed. If not, Git is not properly installed or not added to your PATH. 3. Add Git to PATH (if necessary) If Git is installed but still not recognized, you need to manually add it to your system's PATH: Find the Git installation directory (e.g., C:\Program Files\Git\bin). Open the Start Menu, search for "Environment Variables," and select "Edit the system environment variables." In the System Properties window, click on Environment Variables. Under "System Variables," find the variable named Path, select it, and click Edit. Add the Git bin folder path (e.g., C:\Program Files\Git\bin) and the cmd folder path (e.g., C:\Program Files\Git\cmd) as new entries. Click OK to save changes. Restart your Command Prompt or PowerShell, then try: cmd copy the code from the next line: git --version 4. Retry Installing ComfyUI Manager Once Git is recognized, run the ComfyUI Manager installation command again. Make sure you are in the correct directory for the installation.

@guojunWGJ For the portable standalone version of ComfyUI on Windows, you typically do not need to manually install or configure CUDA, as long as you already have the correct NVIDIA GPU drivers installed. Here's a summary: CUDA Drivers: You only need the latest NVIDIA drivers, which include CUDA support. If these are already installed, you're good to go. No separate CUDA installation: The standalone version of ComfyUI bundles everything necessary for GPU acceleration, provided your drivers are compatible. Automatic GPU detection: ComfyUI automatically detects and uses a compatible NVIDIA GPU without additional configuration. 💡 Are CUDA drivers required? Yes, you need NVIDIA GPU drivers that include CUDA support. However, if your drivers are already installed and up-to-date, no further action is necessary. Summary: As long as your NVIDIA GPU drivers are installed and updated, there’s no need to manually install or configure CUDA for the portable standalone version of ComfyUI.

@@murphylanga Very good, you explained it very professionally. Thank you

Hi @fesanza, Wow, your comment absolutely made my day! 🥹✨ Hearing such kind and thoughtful words about my work means the world to me. I'm genuinely thrilled that the workflow has been helpful for your projects and that you’ve found value in it-it’s exactly what I hoped to achieve. It’s incredibly motivating to know that my efforts can spark creativity and open up new possibilities for talented users like you. Your support and encouragement inspire me to keep improving and sharing with this amazing community. Thank you so much for taking the time to share your thoughts; it’s truly appreciated more than I can express. Wishing you endless success and creativity in all your future projects! 🚀🎨 Warm regards, Murphy ....and greetings from the Ziggys 😉💕

"Thank you so much for your loyalty and enthusiasm! It means a lot coming from one of our most dedicated fans. Stay tuned for more exciting content! If you agree, I should really get started on the splitting and organizing of these workflows-they've become quite unwieldy in size. My former apprentice will be thrilled to see that her voice, which she lent me to train a speech AI, is well-received; this was just the first shot. Do you think a video explaining how this works would interest viewers? I'm also considering showing how to easily train Loras with Comfy or how to transfer one's own expressions onto a virtual character using Comfy, just like in my recent post where the angel sings."

@@murphylanga you're asking the wrong person, as I find all AI interesting😉 These days, I like the generate an image, Then change things I don't like, or want to add. Then upscale. Then swop faces for friends and family. My wife now has a full collection that she uses for her WhatsApp profile. 😁I think she likes them, because they make you look younger. That's why I like multiple stages to a workflow. Keep up the good work and have a Happy New Year!

You may be interested in this collection of workflows. You have to join the Patreon 😒 But this is an amazing workflow for creating face Loras. kzbin.info/www/bejne/a5LJY4htf7WYeNUsi=XX6iQFdW-lrFrZEM

yes, that looks very interesting. i've seen it before, but i haven't joined patreon (yet). i installed part of it under pinocchio, which also runs ollama for me. f5-tts is available there as well, but i installed it completely independently because i want to create languages other than english and chinese-especially german. 😊

@@murphylanga I used to use pinocchio. But started having trouble with nodes failing after a while in comfyui,. I just use the portable version now. The patreon isn't expensive, and his workflows are interesting once you get used to everything being separate 😉

Thanks a lot! I'm glad you think it's cool. 🧛 Stay tuned for more awesome content coming soon!

Thank you very much, I have adapted some things to the current Comfy, the new video and the workflows will go online today or tomorrow...

@VietnamShorts That's fine, you can also go through several steps with image to image and use a very low denoise 0.1 - 0.3 value and Lora, and use only the style as a prompt. If you use one of my newer workflows, you can also integrate a Lora to optimize with the Ultimate SD Upscaler. In addition, there is also the option in these to use a Scribblel or Lineart Controlnet, then a slightly higher denoise value can be used. Finding the right denoise is something you have to find out by approximating. Try to get the created image closer to the desired style several times with low denoise values ​​and Controlnet. Here is the latest video with links to a workflow package, if your computer is too weak for Flux, simply do not switch it on in the Ultimate workflow, or deleting the groups should also work. kzbin.info/www/bejne/kH3EeGx_iKyfbM0

@@murphylanga So there's no good answer for that. if I apply lower noise i won't see the style :) I did several experiments every time when you aply a style it distort the character. It is ok when you use some non existing character but it is not very nice when you use real pictures with real people. Just try to use your own face, you will see how quickly it will become someone's face you don't know ;)

@@VietnamShorts It's not really possible in one step, but it works quite well in several. First change the style (possibly with Controlnet and image to prompt) + Lora and possibly IP adapter only style, then face swap (can be done in one step). Then 2-3 steps with low denoise values ​​to completely adapt only the face to the style. In the newer workflows, only directly with SD upscale/downscale and e.g. the Remarci upscaler

@murphylanga I will check that. Anyway Lora is here so that means whole process become very time consuming. I have my own dreambooth cktp - it works quite ok, but it consumes at least few hours of processing to get a final result. Or maybe you meant just a styling lora🙃 i guess yes you did😏

Helllo @BetatestBanned, because that's only a small part of what you can do with this workflow. If you watch the new video, you can see how you can put together a workflow yourself from the individual modules and then leave out all the things you don't need. But I almost always use the very large workflow, which means you hardly lose any time because no resources are needed by "switching off" individual tools/groups, but if I want to switch something else on, I can do that easily without switching to another workflow.

Hello @Dreddoom, thank you very much, comments like yours keep encouraging me to publish videos. I hope you're well again. Do you know SeaArt? They asked me if I wanted to publish workflows with them. 🤣The old man, that's actually me 🤭But I tried to optimize the voice a bit before the "training" and get my Franconian mumbling out. Unfortunately, I don't have a decent microphone for recording yet. I waited for Prime Day to buy a webcam and a microphone, but there's no offer for the ones I chose. I still have to convince at least one of my former trainees to do the voice(s) for Ziggy. I have a cool workflow, well "almost just copied" to adapt lip movements to spoken words (videos). I created some things with it, but I can't upload them for legal reasons, but I find them really funny.

Sorry I don't know sea art. Sounds like your work is getting noticed. At the end of the day, as long as you enjoy what you're doing, that's all that matters. I was watching matt Wolfe today. Can't remember how you spell his name. His channel reports on AI news. He was saying that KZbin was going to automatically dub your vids in different languages. Don't know when 😊 my illness is still lingering 😢 but feel much better thanks. Keep up the good work.

kzbin.info/www/bejne/pWSlmJ9si8-gbNksi=etGdR1iXvv70GM_4 matt wolfe

No problem, you can cancel your subscription if you like, I don't make any money here and I'm just doing this for fun. The voices are all generated locally and it doesn't always work the way I want it to.

In my last picture on CivitAI the latest Ultimate workflow is embedded, I hope I can upload it officially and create the video for it by the weekend

Thank you for your comment

🤭

Many thanks

Unfortunately, I don't quite understand what you mean. It would be possible, for example, to use multiple ControlNets one after the other, theoretically with different images. You could also use segmentation to automate the processing of only certain areas with different prompts and use loops for this. You can also do this manually using inpaint. I published an inpaint workflow a long time ago, and that should work. If you have a little patience, I want to build a special workflow for inpainting of all kinds by the end of the year at the latest. I'm only doing this on the side when I find some time.

@lom1910 Update your ComfyUI all nodes should be found For the reactor nodes you have to manually install something (Insideface) or delete these groups. This is explained in more detail in the Face swap video. Or start with the simple things (see latest video)

I find such comments presumptuous! I provide this workflow for free and there are people who really enjoy working with it. It's about modules that can be easily swapped or removed. You can do very simple image to image with it, and I could build a single node to do that. But you can also do things with it that are probably beyond your imagination....

@canaldetestes4517 Hi, these are workflows for the open source project Comfy UI. Information on how to install it locally can be found on my channel or elsewhere on the Internet. Sebastian Kamph recently published a video on installation for "newbies". My workflows are intended to make it easier to create high-quality images. I create these workflows primarily for myself. But I publish some of them for free to anyone who is interested. The latest package on CivitAI. Or on OpenartAI You can also try out a number of my workflows directly there. I don't know how to do that, and you probably have to sign up there too.

@MikevomMars Hello Mike, thank you for your comment. You don't seem to have seen the last video yet. You can put the workflow together from the individual modules (groups) or simply remove it from the large, individual groups. And everything should still work. The advantage of the huge one for me is that I can do different things step by step and with different models like SDXL, Kolors and Flux. And keeping a big one up to date is much easier for me than 1000 small ones. Because that's at least how much is "hidden" in the one big one.

@RamonGuthrie Hallo, this is manually filled, if I understood your question correctly. I use the RGThree Nodes for this. These custom nodes have very good things integrated.

Hi, I'm using ComfyUI in the standard embedded Python variant. Latest version and in the settings the tabs and beta menu are enabled.

@oliviertorres8001 Hello and thank you for your comment, what you are planning should work, if your graphics card is powerful enough I would use Flux + Controlnet and/or image to image. Or start with SCXL. The Controlnets for Flux are currently only in the large Ultinate workflow. Good luck

Thank you, have fun testing. Have a closer look at the multi workflows and play with the image to image. I created a painting of a candle on parchment and then mixed it with the Mad Max prompt and then just tweaked the denoise. Totally awesome 🥳

​@@murphylanga sounds cool, I'll have a play

Thanks for your comment, I'm glad you like it!

You're welcome

@salomahal7287 Hi, thanks for your message. I just looked into this, with my new workflow that isn't released yet and also with the V2.0 that is on CivitAi. I can't recreate it. Did you remember that there are 3 Save nodes that you can turn on. The blue one is always just for the start image. The green and brown ones are for the final image. The final images are saved in a different directory than the start image. You can freely define the directories and, if you want, enter the same one twice. I think I'll upload the current workflow at the end of the week, but first I want to install a Lora trigger for the latent SDXL. I just noticed when checking that I had forgotten that. ;-). There are also 15 small workflows with a different memory module. Not as convenient but much faster. You can add it pretty easily. The workflows are finished and the video instructions are in progress. But I can only continue a little after work.

@@murphylanga Ah yea I was using the 1.0 version the one in the 2.0 works nicely thx

@@murphylanga btw dunno if its already somewhere in there but the detailer daemon node is really nice and an ez implementation just for the futur mb.

Thank you for this info. I tested these nodes about 2 weeks ago and also implemented one of these nodes, but haven't published the workflow yet. I think on Friday there will be the new video and several small and the new big workflow with this node.

@@WhySoBroke Hello, I'm currently building a collection of simple workflows that I will publish soon. Flux upsale can then be easily assembled.

Next, I plan to make a small starter module to which you can connect the individual modules in order to build your own small, compact special tool. What do you think, does that make sense? Best regards and thank you for your comment

​@@murphylangathat would be appreciated

Sounds good to me. Not everyone likes very complicated workflows, and there's alot to be said for different workflows or modules for different tasks.

ok, then I will do that soon, I'm currently playing with text to speech F5TTS cool stuff :-)

@colinchambers48 OK, that wasn't the case yesterday. But today I also had some strange behavior from Comfy. Here are the 4 custom nodes that you can look for manually: It is probably this ComfyUI-styles-all github.com/aegis72/comfyui-styles-all SDXL Prompt Styler github.com/twri/sdxl_prompt_styler SDXL Prompt Styler (customized version by wolfden) github.com/wolfden/ComfyUi_PromptStylers JPS Custom Nodes for ComfyUI github.com/JPS-GER/ComfyUI_JPS-Nodes But the workflow should work without it, then only this styler should not work

Were you able to install ComfyUI on your device? It works a little differently than on Windows. But I don't know much about that. The workflow shouldn't have any restrictions if ComfyUI is working. Many error messages are very general. If you see "red" nodes, then not all custom nodes are installed. I haven't tested it at the moment, but it should all work automatically. I'll test it with a freshly installed ComfyUI and publish a list of all custom nodes. Then you can install them manually. The Flux variants and the Kolors variant require more memory. But they aren't switched on after loading, so it shouldn't make a difference. I don't know how that works on a Mac. But if you don't change the workflow for now, it should work. The next reason could be that you don't have all the models that are required. You have to check the "Loader". Maybe select it again.

@@murphylangaThere is no CUDA on MAC. Nevertheless the GGUF Models work fine. You should check if torch uses MPS support.

Hello. I'm sorry, but I don't have a Mac and I have absolutely no idea what you can and can't do with a Mac. It's true that CUDA only works with Nvidia graphics cards, which aren't available in Macs. I don't have anything else to say on this topic.

Thanks for the great comment, that motivates me to publish more in this direction again.🥰

@@murphylanga 👍😁Was an Pleasure help You,Buddy.👍😁👍😁👍😁🇪🇨💙🤟👽

@ВячеславШамрай-й4э 538 / 5.000 You're right, I have a simple solution for you. Right-click on the sampler node and select "reload node" quite far down. In August, some things were changed at Comfy, but unfortunately that meant that some things didn't work 100% smoothly. I didn't completely rebuild the workflow, I just changed an older one of mine. The reload node can sometimes lead to connections being lost. But that shouldn't happen here. Thanks for the info, and have fun with the workflow.

@@murphylanga Thanks for the prompt response, everything works now

Thanks for the tip, I'll take a look at it.

You'll get used to it pretty quickly I think. What I noticed is that I miss the Lora trigger near the SDXL Loras or/and Flux Loras. And I'll have to look into saving for something better. If you have something that you find inconvenient. Don't be shy, just write to me

​@murphylanga thanks. The only thing I've stated to add is big labels above the different groups with the label node. I find it easier to navigate when zoomed out. I always forget which bookmark is which 😅

Name and the bookmark number, why not

Sorry, das kann ich leider nicht mehr machen. Ich habe nur noch das fertige Videofile mit Sprache und Ton als eine Spur. Ich hoffe, in den späteren Videos ist es besser. Damals hatte ich überhaupt noch keine Ahnung, wie man so ein Video für KZbin erstellt. Danke für deinen Kommentar.

@ChenHaoYan-z1c , "Oh, thank you, you flatter me! In the past few weeks, I haven't even turned on my computer, even though I bought a new one. It only has a 4080, but that's still a huge upgrade for me. Currently, I'm expanding the workflow with Kolors and other features. One thing I know for sure, 96 GB of RAM (I was able to install 2x32 in addition to my existing RAM) makes a massive difference, but 64 GB would be just as good. However, 32 GB is definitely not enough, especially if you want to have fun with Flux and Supir, where more RAM really helps. The 4090 was too expensive for me, but I wish you a lot of fun with it. Thankfully, the weather has cooled down, allowing me to use my computer longer with a RAM upgrade."

@@murphylanga Congratulations on your upgrade! Upgrading to 64 GB is a wise decision, especially with the ever-growing demands of AI content. This enhancement has notably increased the efficiency of using Supir within the entire workflow, enabling it to process tasks in just minutes. The improved performance not only saves valuable time but also boosts productivity significantly. Upgrading the computer was certainly the right choice, enhancing the overall experience and making the workflow smoother and more effective. I am truly appreciative of this improvement, as it has made a remarkable difference.

Well, not really, once you understand the principle, it's easy. But with the heat we have here, it's heading in that direction 🥵 I haven't used the computer for a while and discovered that the last version contained a few errors. Nothing dramatic, but annoying nonetheless. The styler groups were not taken into account as planned. Once the custom nodes and the required models are all installed, it's very easy to use. I'll upload a corrected version later.

civitai.com/models/641359?modelVersionId=795110