This helped me out massively! Thank you

Thanks mate but the volume is too low.

That's awesome, thanks! How does one light up those splines/ noodles like lasers, as shown on the Aegisflow Git page?

man, work on your AUDIO... not possible to listen

can lora be specified in the prompt? <lora_name:weight>

Thank you foor your lesson. Have a happy, healthy and prosperous 2025👍

خدا نکنه آدم گیر هندی جماعت بیوفته لامصب مثل انسان حرف بزن دیگه

Is it possible to upload an image from my machine and then remix that? Really trying to find something that can generate similar images to ones I upload.

If you have nodes with the same types of input connections, but you don't actually want the Anywhere blinking noodles to connect to it, you can drag a noodle from that input and leave it attached to a reroute mini node, with nothing else connected to it, like a dead end, but only "upstream" and that'll block Anywhere from sending data to it.

Thank you1 Very informative indeed. Do you mind if I ask how you have the queue buttons and CPU info on the right side please?

The cmd doesn't recognize the 'https' in the git hub link.

Thanks for the video, but the Canvas option isn't available in Capcut PC 4.6

Well I heard the audio just fine. It looks like the system will go through all the layers by default at clip skip -1, and the more layers it goes through, the more the system tries to refine your image according to your prompt. So by skipping a couple layers you are basically telling it to "almost" go through all the layers, i.e., skip layers starting with the last one in the sequence, then the second to last one, etc.

Is this possible for Flux? I miss it

The documentation on the GIThub for this node needs to show a couple examples of what exactly it's doing.

This site is kind of trash. I should just be able to upload an image and tweak it as I please yet this garbage site INSISTS I use it's models. No... I want to use my own. Kindly F OFF :)

Am very interested in this please can you be my boss

this was driving me up the wall, thank you so much!

Thanks, that video was what I needed. I was really trying to figure this out, and there you were with a short and easy video.

Nice! Thank you

How to preview more than one image when you're running a batch of e.g. 4 or 6 and wanna stop it if they all suck?

constructive feedback. you feel too close to your microphone.

what graphics card are you using?

How do we use concat conditioning with the efficient Loader?

What an incredibly lame video! You show nothing at all, just talking. This is a waste of time.

"KZbin police" 😂🤣 ...thanks for the great short tutorial, btw.👍

Does this work for Flux?

great video!!! Thank you for the proper syntax on the { .... I was wondering if you could do a video on the uses of " \ "....... example: \(black dog:1.2\) What does the \ do?!?! there isn't any good explanation's on the net and nothing on the comfyUI site.

thank you, it is what i am looking for

Excellent Technique

Great indeed. Wish it could connect properly to ComfyUI server..../alpha 0.23 stage now/

thank you

This is really cool.

cheers mate

Hello Joe. I am the coder of these nodes, so thank you for this video, I am glad to see it is used :). Initially I did this to simplify photorealistic prompts in ComfyUI following principles in PromptGeek's awesome pdf on this subject.

joe why are you only in my left ear

Looking at their pricing, I feel that there's room for a subscription between Basic and Professional for perhaps $20 and 60 minutes per month and without the 1 minute limit per video.

Does anybody know of open source (free) alternatives?

How do we do this with flux in comfyui?

i installed through the github repo instructions. and it gave me an error for the requirement line. then in other cases, i went to manager and tried to install through manager and it says import failed. i've never gotten crystools to work on PC which is baffling because i've gotten it to work on mac. and ive gotten more complicated to work on the pc. i would paste the github url issue here but youtube will just take it down. can you help me?

works!! your a legend!

This has been a good laugh sofar, inputing one liners and jokes. What a time to be aive

This was interesting. I paused to read the CLIP notes, and I think the concept is pretty fascinating. My understanding is that it's doing this magic by injecting itself into the image generation process to fine tune the original prompt using a referenced LORA and then inserting the processed image directly back into the generation workflow as a substitute, so the values of -1, -2 and so on reflect how many steps backwards from the final stage of the image process the standard generation gets before CLIP is triggered. Cool technique and effective!

Very clear and informative ! Thanks

DON'T USE HEDRA! It's a privacy TRAP!

It's not clear at all how the text file format should be, I tried to make a text file with 200 prompts and for some reason getting errors everytime but if I run the example it works fine. How do I fix my text file?????? positive:Futuristic survey ship, sleek design, solar panels, exploring Greenland fjord, summer season, melting icebergs, clear blue water, rocky green shores, wildflowers, distant snow-capped mountains, drone swarm mapping coastline, holographic displays on deck, scientists with high-tech equipment, midnight sun low on horizon, reflections on calm water, mist rising from fjord, 8K resolution, hyper-realistic, cinematic lighting, aerial perspective negative:text ----------------- positive:Futuristic Greenlandic city, vast plain atop towering cliffs, overlooking majestic fjord, gleaming skyscrapers with Nordic-inspired architecture, transparent dome covers sections of city, advanced wind turbines, vertical farms integrated into buildings, elevated mag-lev train system, hovering personal transport vehicles, holographic billboards, melting glaciers in distance, midnight sun low on horizon, vibrant aurora borealis in sky, blend of modern and traditional Inuit design elements, lush rooftop gardens, waterfall cascading from cliff to fjord, 8K resolution, hyper-realistic, dramatic lighting, wide-angle panoramic view negative:text ----------------- positive:Futuristic Greenlandic city, vast plain atop towering cliffs, overlooking majestic fjord, gleaming skyscrapers with Nordic-inspired architecture, transparent dome covers sections of city, advanced wind turbines, vertical farms integrated into buildings, elevated mag-lev train system, hovering personal transport vehicles, holographic billboards, melting glaciers in distance, midnight sun low on horizon, vibrant aurora borealis in sky, blend of modern and traditional Inuit design elements, lush rooftop gardens, waterfall cascading from cliff to fjord, 8K resolution, hyper-realistic, dramatic lighting, wide-angle panoramic view negative:text

Generating something costs energy and GPU time, they don't care if you like the outcome or not.

국수 같은 선이 아주 깔끔하게 정리됩니다. 휼륭한 강의입니다

I can't reproduce any of these connections. You should build the workflow from scratch while explaining things.

Thanks for the tour. Good to know the tools and follow how technology transforms and manifests into different solutions. Not suspecting it was generated output, your photographic face had me fooled. The illustration is a nice test, and can see a bit of softening of the painted texture. Good observation on the dread of commiting to an output that might contain the unexpected, or need to be refined with iterations to meet expectations. But it’s free for now as you say. I’m curious about results on 3-quarter perspectives for more extreme parallax. We will see-Sub’d