Probably one of the best TD tutorials out there. I learned so many different techniques just from this one video alone. Thank you for sharing your work with us! This is also the closest thing to concert visuals I've seen a TD tutorial. Would love to see more projects and videos of similar complexity to help some of us learn more about making groundbreaking concert visuals!
On point as usual! Honestly, i was trying to find a way to do smooth recording with SD and there you are with a tutorial for exactly that. Thank you! :)
Really appreciate these videos! It's great to learn a lot about Touchdesigner, eventhough I've only used the program once before. In other words your great at doing tutorials, and I'd love to see more of these AI explorations!
Thanks a lot for all your videos. I watched you complete beginner course. And I have to admit, I'm still a little confused. Even though I'm not unfamiliar with nodes in Foundry Modo and Fusion (Davinci Resolve). Do you also have experience with Cavalry? I'm currently torn about which software to learn. TouchDesigner seems to be even more powerful. But also harder to learn. Or am I wrong? I could't find a comparison between Cavalry and TouchDesigner.
When you say you started incorporating SD into feedback loops, were there any particularly important problems to solve or "gotchas" about it? Like working with Realtime flag off, checking when things finish cooking etc? I've been using TD in feedback loops since before StreamDiffusion, using A1111 API for example (either by DotSimulate or Oleg Chomp), and I've repeatedly had issues synchronizing everything and having it ticking reliably if Diffusion happens somewhere in the middle of a complex network. The issue usually was that these operators execute in an async manner, meaning it's not following the usual TD dependency chain of operations, leading to unexpected behavior. It could be that none of that is even an issue with SD by DotSimulate. If so I'd need to investigate.
Hey, just a quick question, do you think it could work even with just a videofilein? Like if I want the crazyness of Streamdiffuion over a music videoclip, could I just input the video, set the fps I need and It should work right?
Wonderful. All the applause for your contribution. Really cool! When replacing with another music, in order to gain more influence, I need to play and make changes into the parameters right? Like your example uses snare, but could be another element from the music right? Say, the piano notes?
thanks so much ❤️ yea exactly. there are other beat detection components on my patreon or use any channel from the audio analysis in the pallete :) piano notes could come from something like my midi keyboard visualisation tutorial
@@elekktronaut Just checked the midi keyboard tutorial. I was thinking how a simpler method would be to use a 3rd party tool that isolates each instrument, then bring the piano track into TD and use it to drive the visuals?
Hey pal, love your channel, but sorry for off topic - did you put away Notch, and what do you think about future of TD and Notch? In theses, maybe? Just interested about your thoughts. And of course, thanks for another amazing video! Feel free to make a long comment or sort of podcast, I mean everything is changing so fast, especially real-time things and I think you're one of the guy who can tell something. Thanks again!
I've turned away from Notch, I want to completely focus on TD and now AI as well :) Idk about the future of Notch, it's definitely an amazing tool but I don't see a reason for me to get into it now TD is this powerful. You're very welcome!
Thanks for this :) how do you get absTime in the independent component to follow the independent timeline? for me that runs along with the main time....(ok I guess to use a timer/timeline chop instead, trying to figure out the best way with that) edit: yeah timeline chop set to seconds seems to be the way to go with that, will experiment :)
Quick question, in the beginning when I change the time to Run Independently - when I hit pause in the main project folder, it pauses. And then when I go into the "independent" node and look at the local node, it also seems to pause. I'm unable to get them to run separately but even after following your steps, I dont see the timeline running independently when my project is paused. Can you help?
Your tutorials helped me with Touchdesigner so much. I can now create textures and lots of cool stuff using your techniques. Bless you!
❤
Probably one of the best TD tutorials out there. I learned so many different techniques just from this one video alone. Thank you for sharing your work with us! This is also the closest thing to concert visuals I've seen a TD tutorial. Would love to see more projects and videos of similar complexity to help some of us learn more about making groundbreaking concert visuals!
On point as usual! Honestly, i was trying to find a way to do smooth recording with SD and there you are with a tutorial for exactly that. Thank you! :)
you're very welcome :)
Incredibly clear tutorial. Amazing. Thanks!
this is insane!! I've missed your tutorials man, wish this was available for Mac, already!
Really appreciate these videos! It's great to learn a lot about Touchdesigner, eventhough I've only used the program once before. In other words your great at doing tutorials, and I'd love to see more of these AI explorations!
❤️🙏
Danke für dieses tutorial aber ich bräuchte echt deine hilfe
Very cool tutorial!
Does the graffiti move to music or to the spoken voice?
Hey! super video as always!
Do you suggest having a gpu with more than 8gb for that or the 4070 could also manage this?
Thanks a lot for all your videos. I watched you complete beginner course. And I have to admit, I'm still a little confused. Even though I'm not unfamiliar with nodes in Foundry Modo and Fusion (Davinci Resolve).
Do you also have experience with Cavalry? I'm currently torn about which software to learn. TouchDesigner seems to be even more powerful. But also harder to learn. Or am I wrong? I could't find a comparison between Cavalry and TouchDesigner.
most of the times it's a good signal when something is harder to learn
@@thechrisdo You could be right. 🙄
When you say you started incorporating SD into feedback loops, were there any particularly important problems to solve or "gotchas" about it? Like working with Realtime flag off, checking when things finish cooking etc?
I've been using TD in feedback loops since before StreamDiffusion, using A1111 API for example (either by DotSimulate or Oleg Chomp), and I've repeatedly had issues synchronizing everything and having it ticking reliably if Diffusion happens somewhere in the middle of a complex network. The issue usually was that these operators execute in an async manner, meaning it's not following the usual TD dependency chain of operations, leading to unexpected behavior.
It could be that none of that is even an issue with SD by DotSimulate. If so I'd need to investigate.
Hey, just a quick question, do you think it could work even with just a videofilein? Like if I want the crazyness of Streamdiffuion over a music videoclip, could I just input the video, set the fps I need and It should work right?
dotsimulate
machineburners
lets fucking goooooooooooooooo
Wonderful. All the applause for your contribution. Really cool! When replacing with another music, in order to gain more influence, I need to play and make changes into the parameters right? Like your example uses snare, but could be another element from the music right? Say, the piano notes?
thanks so much ❤️ yea exactly. there are other beat detection components on my patreon or use any channel from the audio analysis in the pallete :) piano notes could come from something like my midi keyboard visualisation tutorial
@@elekktronaut Just checked the midi keyboard tutorial. I was thinking how a simpler method would be to use a 3rd party tool that isolates each instrument, then bring the piano track into TD and use it to drive the visuals?
Hey pal, love your channel, but sorry for off topic - did you put away Notch, and what do you think about future of TD and Notch? In theses, maybe? Just interested about your thoughts.
And of course, thanks for another amazing video!
Feel free to make a long comment or sort of podcast, I mean everything is changing so fast, especially real-time things and I think you're one of the guy who can tell something.
Thanks again!
I've turned away from Notch, I want to completely focus on TD and now AI as well :) Idk about the future of Notch, it's definitely an amazing tool but I don't see a reason for me to get into it now TD is this powerful. You're very welcome!
Thanks for this :) how do you get absTime in the independent component to follow the independent timeline? for me that runs along with the main time....(ok I guess to use a timer/timeline chop instead, trying to figure out the best way with that) edit: yeah timeline chop set to seconds seems to be the way to go with that, will experiment :)
me.time :)
@@elekktronaut ah that's even easier, thanks :)
Great share. Thank you. To bad it only works with TD 23.
Quick question, in the beginning when I change the time to Run Independently - when I hit pause in the main project folder, it pauses. And then when I go into the "independent" node and look at the local node, it also seems to pause. I'm unable to get them to run separately but even after following your steps, I dont see the timeline running independently when my project is paused. Can you help?
nevermind! I restarted TD and it worked lol
This is fucking crazy man
❤
damnit, how do I get this working on ubuntu?
Did you record this in a church? the reverb is nice though :)
new apartment that isn't finished setting up 🙃
🤍🤍🤍🤍