Thank you for this! I tried following other tutorials and kept running into strange errors, but since you broke it down step by step it now works! Much love
Thanks for the shout-out... I actually made my DD video input tutorial because I couldn't find anyone online doing it right.. Glad I could help out... My goal is to spread some of my creative knowledge... You should really look into Stable Diffusion if you haven't already... Next level AI... Just finally completed my tutorial for Video Input in SD... Keep up the good work my friend...
My dude. Thanks for sharing your knowledge! And yes very few people are making vids on video input. I’ve been messing around with SD a bit but I feel like its a bit inconsistent when it comes to coherency. I think it looks cool when each frame is random cause it looks like hand drawn images but then when it comes keeping it one style, sometimes it changes too much. I will def experiment more with it.
@@enigmatic_e there's tons of settings to tweak in SD to get you similar results... I feel DD is still unique in its own way but after I see a ton of DD projects i start to think thats DD... With Stable i keep falling deeper and deeper into rabbit holes and the results just impress me more and more as I see it surprising me again and again while I felt DD results kinda hit a wall... for me anyway but I completely understand why people still love it... and at the end of the day it is all preference.... I get endless people asking me why i dont warp or heavily interpolate my frames for example and as a "traditional" digital animator by trade I just always loved the more choppy lower frame rate look... Either way it is all so fascinating... I can only imagine where this tech will be in 5 years... Overall I see the popularity increasing dramatically but 90% of everyone is doing the same thing... I always enjoy seeing someone push the boundaries and use the tech as a tool to enhance their craft... Look forward to seeing what you create next... Have a great night...
@@enigmatic_e like Disco or VQgan or PyTTI... the more the use it the more you'll discover... But i personally feel the way it can imitate styles and artists in such a formidable way really let's you push the limits where all DD really did was give you mostly abstract results...
this was a really helpful video thankyou - i saw your post on Reddit and was blown away so i followed this and subd lol. Really great overview - i ended up creating a mess but i'll post a little video result soon to see what i belched out haha!
Great work man, i have an issue maybe you can help, its says OutOfMemoryError an tells me to "allocated memory try setting max_split_size_mb to avoid fragmentation" how i do that :(
I love your videos. I am hoping that there is a way to use an image i made as a prompt to guide a video of me. Basically use a 3d image of a werewolf to turn a video of me into a werewolf?
First of all, thank you for doing this. It really helped me to set up the workflow. That said, after all these years, I came to hate GitHub and Colab,...even though I can see how important these platforms are. It is their sluggish, error-prone, needlessly complicated interface that bothers me. Every time, something new will disrupt the workflow. This time, I see two identical fields for the video init address, represented by an If Else statement in the code. I just don't get it. The video won't load, and in the tutorial, you have only one field, not two. It's a pain in the butt to deal with stuff that otherwise is very easy to deal with in a software package. Still looking forward to see more creative uses of AI. Thanks again.
I totally agree. I hadn’t used DD for a few months and I recently decided to try it again and I encountered errors that I never had before. Now I have to research a for around. I guess that’s why I work more with Stable Diffusion lately.
@@enigmatic_e I see, well, that's interesting because I was just searching for Stable Diffusion. Best of luck to you and I'm looking forward to your content.
Thank you for this! I attempted to make this edit, but encountered an error in the final step stating 'No such file or directory.' Would you kindly provide me with guidance on resolving this issue?
i have the some issue too, and I found that in 3.SETTIi NG, I had set the quantity of SKIP STEPS higher than the STEP. After correcting it, everything started working. I hope this information helps you.
I personally dont like v5.61. It gives you more options with settings but for me, those extra features just made it harder to get good results for video input. Maybe I was just using it wrong but I always enjoyed working with 5.2.
Hey man! Thanks for this tutorial! I seem to have problem with transparency in the video - what codec did you use? I used Prores with alpha channel and it seems like DD can't work on it. Did you just use uncompressed?
Thanks man! I'm on pre-production on a music video and wanted to incorporate ai animation into it. These tuts are real helpful. Any tut for animating with midjourney?
Do we have to put the final clips together in After effects? Can we use other normal video editors like Windows video editor etc.? Like I dont have after effects.
Hello there, I have a question. Everytime I start the render, it runs a while and then unconnect me from runtime environment. After that I can't run it again because it is disconnected.
Hi, this is really cool and I believe AI generated videos will be very widely used by everyone in the near future. Do you know how to setup Disco Diffusion on M1 PRO/MAX MACBOOKS ?
i have this error ''/content An exception has occurred, use %tb to see the full traceback. SystemExit: ERROR: 0 flow files found. Please rerun the flow generation cell.''
Sorry haven’t used DD for while. I did make a trouble shooting video though have you watch it? Disco Diffusion Errors (Troubleshooting) ua-cam.com/video/eA5iTXtsXGA/v-deo.html
I was wrong about uploading it as transparent file. You should remove the background though. Once you put It into After Effects then you use the transparent file as a matte to remove the black background.
Рік тому
Thx a lot for sharing your knowledge. I need your help, I have this message "Error Count API failed" , what does it mean ? thx
Great tutorial. Thanks for sharing. I’m having an issue though, disco diffusion seems to only render 12 or 13 frames and then saves the video and shuts down the process. When I go to restart that frame, it says there’s no such file. Ever run into this? Looking for help! I think this stuff is so cool and I’d like to get to know it better.
Try running it with the most default settings as possible to trouble shoot it. I would try starting with a fresh new colab session and test the video with little changes and then make small changes until you find out what the issue is.
Hi, I was wondering how long step 4 diffuse should take? I have uploaded a 32s video of a resolution of 828x720 and its set to take like 30hrs. Is that normal?
Hey great video man! Would you be able to help me out? Every time i insert a video_initial_path which is the actual file, it will load but it wont star, mine doesnt say the name of the file once its done uploading, it will keep saying "insert text here". I am uploading the file just like you explained, what am i doing wrong?
GREAT tutorial, but i cannot upload the file in the first place.. i do have enough space on drive but after i finish uploading the file is not shown but there is a red circle in the upload section.. any help thanks
@@enigmatic_e after trying different settings and spending whole day today it’s working now.. thank you so much for quick response.. already subscribed 🙏
thanks for the work, but i saw you skipped the optical flow ? what's that and i can't generate without running that first, but then i get errors running it
@@enigmatic_e hmm I think i got the version confused lol, i had the v5.61 opened while looking at settings in v5.2, just so many versions out there 😮💨😵💫
Would you still recommend this method or is it outdated? Is stable diffusion better? if it is can you do an updated version where you can do something similar to this using google Collab stable diffusion.
This should still be useful, I’ve seen some creators make very cool stuff following this. I have gotten a few request to do a more in depth tut on SD in colab and I might try that.
Hi bro can ı ask something to you. I am using macbook air m1 16 gb. But this machine is not working well after effect. I want to try AI animation things. İf ı will buy macbook pro m1 pro 32 gb. is this device enough for that stuff
Hi ! When I render from AE transparent video in quicktime format and after I upload that vid to DD, it's still sees the black background , any ideas why?
@@enigmatic_e thank you for respond, my problem is that DD still can see black background and it's making animations on it. Perhaps I messed up the settings?
thanks man! have you tried the similar approach in A1111 SD? For me it outputs not on transparent layer, but SD adds a white background on my rotoscoped model. i can't seem to find a way around this! curious if you have?
Hey Viktor. I was wrong about the whole transparent bg when it comes to generating the AI art. However, rotoscoping is still good because you could alpha matte to remove bg after generating art.
@@enigmatic_e so you would re-rotoscope the finished art? argh that is less than ideal. for me it, a non computer scientist, it seems obvious to not add in a white background. would be so good if the transparent layer could be kept
@@enigmatic_e fully appreciate your work and sharing these video and knowledge. you where my entry point into the world months back. okay, done with the charming :D so it seems there isn't a way to have the AI render out with transparency alpha, but only work arounds like the copy of rotoscoping you are demonstrating in the other video. bummer!
Do you guys know why Disco is creating images even on the transparent background? I exported it as a Quicktime video with RGB + Alpha turned on and it still uses the background, do you know the way around it?
You don’t need it to be transparent. I know I suggested that but it’s because you might want to use the rotoscope layer as a matte to remove the background.
Curious if you'd be able to help with the following error: "FileNotFoundError: [WinError 2] The system cannot find the file specified". I uploaded the video, made sure the frame size was correct, and copy and pasted the path into the correct spot. I DID notice that mine did not say "/content" before the file name, but not sure why. Wondering if that has something to do with it? Please help!
Might be that your clip has a lot of movement. If it does it may also require you to generate every frame instead of every second frame if there is a lot of movement.
The subjects are actually standing still. And I should clarify, by distorted I just mean that it’s farther and farther away from the original frame. Frame zero it’s a slight texture on the subjects and by frame 13 they are becoming swirly blobs.
I actually have it on that. I was thinking that maybe it’s the ratio between the cgs and the init_scale? It’s just weird because it’s the same settings that you had on this video but yours seems to stay pretty consistent. Mine seems like it’s using the last frame output to work the next frame instead of referring to the original frame.
Hello, please Help me! What is the correct way to format the parameters to resume running? display_rate: 20 n_batches: 9 resume_run: run_to_resume: [0] resume_from_frame: [118] retain_overwritten_frames: ****MY Basic Settings: TimeToDisco the name of my last frame is - TimeToDisco(0)_0118 I have 313 total frames How do I resume work from frame 118? Help me, please!
@@enigmatic_e I can not run the same error as I threw off above. is it better to use another version? or where exactly to write the desired value. write please
Disco works and looks different. Disco diffusion gives more of a psychedelic look. You can technically use colab free but you have to be active on the browser or else it stops the session which sucks if you want to walk away from computer.
Hei mannn Really appreciate your work here. But somehow DD read my transparent background as black!!? 😱😱😱 how to export transparent background from DD?
Hey! Don't worry about it, DD will make it into jpeg image, which will make the background black anyways, so even a mp4 works great. But keep the transparent vid, you can alpha matte it to AI animation to remove the black background after!
Yea it doesn’t but you can use your original rotoscoped footage with alpha and use it as an alpha matte. Thats how rotoscoping can be useful and also doesn’t have the background to distract from subject. I went over that in pt 2
Remember people, never support ai arts with your money or praise as they already destroying what is great about humanity which is "creations" and quality craft with thoughts and life knowledge that brought us happiness. AI art doesn't have any clue or idea for our inspiration and greatness as much as a living human itself. What corporations and ai programmers have done is not okay at all. They don't care about the quality of experience and entertainment that brought people greatness, they only care about themselves collecting huge cash and doesn't care of the earth they living. It's getting worse and worse of how they can just insult us all with these ai arts techs that will make our entertainment media worse.
@@enigmatic_e nice. There is a mess up at the disco diffusion notebook at the moment it seems. Do you know anything about this? Its missing a file that you need for section 1.4. Many people are asking about it on the discord.
@@studio85amsterdam not sure what it is but I get a lot of people with this issue. I haven’t used DD for a while but I’ll get back into it to see if I can figure out the issue.
![АЛАУДИНОВ у Скабеевой: эти их ВСУ нас НЕ ДОГОНЯТ 😁 [Пародия]](http://i.ytimg.com/vi/an0anqU4WGQ/mqdefault.jpg)
Very cool of you to share your knowledge bro...taking time out of your life to provide other creatives with your knowledge is a noble thing
🙏🏽🙏🏽
I second this. Much appreciated 🙏
FIRST VIDEO WHEN I UNDERSTAND EVERY ENGLISH WORD! Thank you for your simple language.
🙏🏽
Hey man thanks for following through and posting a tutorial. I know me and a lot of other people appreciate it!
No problem!!! Thanks for checking it out
@@enigmatic_e FACTSSS
Thank you for this! I tried following other tutorials and kept running into strange errors, but since you broke it down step by step it now works! Much love
🙏🏽 glad to hear that! I’ve had many people tell me they’ve ran into errors so I’m glad to hear when people get it to work.
Thanks for the shout-out... I actually made my DD video input tutorial because I couldn't find anyone online doing it right.. Glad I could help out... My goal is to spread some of my creative knowledge... You should really look into Stable Diffusion if you haven't already... Next level AI... Just finally completed my tutorial for Video Input in SD... Keep up the good work my friend...
My dude. Thanks for sharing your knowledge! And yes very few people are making vids on video input. I’ve been messing around with SD a bit but I feel like its a bit inconsistent when it comes to coherency. I think it looks cool when each frame is random cause it looks like hand drawn images but then when it comes keeping it one style, sometimes it changes too much. I will def experiment more with it.
@@enigmatic_e there's tons of settings to tweak in SD to get you similar results... I feel DD is still unique in its own way but after I see a ton of DD projects i start to think thats DD... With Stable i keep falling deeper and deeper into rabbit holes and the results just impress me more and more as I see it surprising me again and again while I felt DD results kinda hit a wall... for me anyway but I completely understand why people still love it... and at the end of the day it is all preference.... I get endless people asking me why i dont warp or heavily interpolate my frames for example and as a "traditional" digital animator by trade I just always loved the more choppy lower frame rate look... Either way it is all so fascinating... I can only imagine where this tech will be in 5 years... Overall I see the popularity increasing dramatically but 90% of everyone is doing the same thing... I always enjoy seeing someone push the boundaries and use the tech as a tool to enhance their craft... Look forward to seeing what you create next... Have a great night...
@@iamYork_ ok you convinced me to dig deeper into SD. Thank you for the support man. I always look forward to your videos.
@@enigmatic_e like Disco or VQgan or PyTTI... the more the use it the more you'll discover... But i personally feel the way it can imitate styles and artists in such a formidable way really let's you push the limits where all DD really did was give you mostly abstract results...
@@iamYork_ True. Each tool has its different purpose. I guess it depends on what you’re trying to accomplish on each project.
Duuude thanks for making this. ive been studying a bunch and stilll need this! lol :)
This is the only video which helped me understand .. great job brother
this is going to be in my next todo video list
Nice!! Hope it helps!
Thnx e. The portal video is really innovating. Again thnx for sharing your knowledge. Keep it up.
Thanks to you ive been using this in my music video edits and couldn't be happier. Appreciate you boss.
I FOUND YOUR CHANNEL AND I SUBSCRIBED FAST!!! Thanks for this Brother!
super cool brother ... keep it up ...
Great video. Thanks!
it means alot that you are sharing your skills and energy for us 😄
Thank you for sharing this!
really great tutorial :) !! thank you for sharing such great information :)
this was a really helpful video thankyou - i saw your post on Reddit and was blown away so i followed this and subd lol. Really great overview - i ended up creating a mess but i'll post a little video result soon to see what i belched out haha!
Glad i could help! It does take a lot of trail and error to get results you‘re happy with.
will you post it or you changed your mind?
Great video!! Thank you
Hey Man ! I love your work
Thank you!
Great work man, i have an issue maybe you can help, its says OutOfMemoryError an tells me to "allocated memory try setting max_split_size_mb to avoid fragmentation" how i do that :(
Mine is not working, I've been following it step by step for hours and so far it hasn't worked, can anyone help me please?
Thanks to @MDMZ to share you in he's video, you both are great!
Love your tutorial. shot in the dark here, but I got as far as here '4. Diffuse! Do the Run!' and it failed. Come across this before? thanks loads!
i think it just fixed itself!
thanks bro
thanx man You´re great
I love your videos. I am hoping that there is a way to use an image i made as a prompt to guide a video of me. Basically use a 3d image of a werewolf to turn a video of me into a werewolf?
First of all, thank you for doing this. It really helped me to set up the workflow. That said, after all these years, I came to hate GitHub and Colab,...even though I can see how important these platforms are. It is their sluggish, error-prone, needlessly complicated interface that bothers me. Every time, something new will disrupt the workflow. This time, I see two identical fields for the video init address, represented by an If Else statement in the code. I just don't get it. The video won't load, and in the tutorial, you have only one field, not two. It's a pain in the butt to deal with stuff that otherwise is very easy to deal with in a software package. Still looking forward to see more creative uses of AI. Thanks again.
I totally agree. I hadn’t used DD for a few months and I recently decided to try it again and I encountered errors that I never had before. Now I have to research a for around. I guess that’s why I work more with Stable Diffusion lately.
@@enigmatic_e I see, well, that's interesting because I was just searching for Stable Diffusion. Best of luck to you and I'm looking forward to your content.
Thank you for this! I attempted to make this edit, but encountered an error in the final step stating 'No such file or directory.' Would you kindly provide me with guidance on resolving this issue?
i have the some issue too, and I found that in 3.SETTIi NG, I had set the quantity of SKIP STEPS higher than the STEP. After correcting it, everything started working. I hope this information helps you.
Is "Disco Diffusion v5.2 Warp" still the way to go, or has the team merged all features already into the latest Disco Diffusion v5.61?
I personally dont like v5.61. It gives you more options with settings but for me, those extra features just made it harder to get good results for video input. Maybe I was just using it wrong but I always enjoyed working with 5.2.
I get this error how to solve it
ModuleNotFoundError: No module named 'utils.utils'
Hey man! Thanks for this tutorial!
I seem to have problem with transparency in the video - what codec did you use? I used Prores with alpha channel and it seems like DD can't work on it. Did you just use uncompressed?
I use quicktime and GoPro CineForm RGB 12-Bit With Alpha At Maximum Bit Depth. I typical do that export from media encoder.
@@enigmatic_e gonna try this out. i thought it was not possible until i saw the video! niceeeeeee and thanks
Thanks man! I'm on pre-production on a music video and wanted to incorporate ai animation into it. These tuts are real helpful. Any tut for animating with midjourney?
Do we have to put the final clips together in After effects? Can we use other normal video editors like Windows video editor etc.? Like I dont have after effects.
Hello there, I have a question. Everytime I start the render, it runs a while and then unconnect me from runtime environment. After that I can't run it again because it is disconnected.
Is there a way to upload your own images instead of using prompts?
I believe there‘s a way. If you upload image and copy link to ini_image I think. I never really used it but I know its possible
Hi, this is really cool and I believe AI generated videos will be very widely used by everyone in the near future. Do you know how to setup Disco Diffusion on M1 PRO/MAX MACBOOKS ?
Yea, I think the same. And sorry I don‘t know how to set it up like that. I‘ve only used it in colab.
Someone Know why says Error Count Api Failed exactly below the title Disco Diffusion v5.2 - Warp by Alex Spirin? HELP PLEASE!
Hi, generating optical flow and concistency maps throws this error: OutOfMemoryError: CUDA out of memory. What can I do? Please help me, thank you :)
Mine keeps crashing. Do you think it’s because I got the free version?
i have this error
''/content
An exception has occurred, use %tb to see the full traceback.
SystemExit: ERROR: 0 flow files found.
Please rerun the flow generation cell.''
Stuck 1.4 define Midas Function
wow mee too, idk why.
Sorry haven’t used DD for while. I did make a trouble shooting video though have you watch it? Disco Diffusion Errors (Troubleshooting)
ua-cam.com/video/eA5iTXtsXGA/v-deo.html
太棒了
great tutorial! j.w what type of file u export and upload for transparent ?
I was wrong about uploading it as transparent file. You should remove the background though. Once you put It into After Effects then you use the transparent file as a matte to remove the black background.
Thx a lot for sharing your knowledge. I need your help, I have this message "Error Count API failed" , what does it mean ? thx
hello it stops at "Define optical flow functions for Video input animation mode only"
what can I do ?
Great tutorial. Thanks for sharing. I’m having an issue though, disco diffusion seems to only render 12 or 13 frames and then saves the video and shuts down the process. When I go to restart that frame, it says there’s no such file. Ever run into this? Looking for help! I think this stuff is so cool and I’d like to get to know it better.
Try running it with the most default settings as possible to trouble shoot it. I would try starting with a fresh new colab session and test the video with little changes and then make small changes until you find out what the issue is.
Hello everyone, does anyone know why I would get - Index Error: List index out of range? Thank you
Hi, I was wondering how long step 4 diffuse should take? I have uploaded a 32s video of a resolution of 828x720 and its set to take like 30hrs. Is that normal?
Mmm yea seems like a lot. Do you have the basic version of Colab?
@@enigmatic_e this was on the pro one, also how fast does compute units go cause it seems like an expensive way to maintain this sort of thing?
Hey great video man! Would you be able to help me out? Every time i insert a video_initial_path which is the actual file, it will load but it wont star, mine doesnt say the name of the file once its done uploading, it will keep saying "insert text here". I am uploading the file just like you explained, what am i doing wrong?
how to upload video to sample data, i keep failing when uploading?
thanks
GREAT tutorial, but i cannot upload the file in the first place.. i do have enough space on drive but after i finish uploading the file is not shown but there is a red circle in the upload section.. any help thanks
Did you create a copy of the colab session and saved it to your drive, and did you link your account to the session.?
@@enigmatic_e after trying different settings and spending whole day today it’s working now.. thank you so much for quick response.. already subscribed 🙏
@@LakpaSona What did you change to fix this ?
I got error on diffuse called "use%tb to see the full traceback
thanks for the work, but i saw you skipped the optical flow ? what's that and i can't generate without running that first, but then i get errors running it
Mm not sure why. Does your video dimensions match the dimensions in disco diffusion settings?
@@enigmatic_e hmm I think i got the version confused lol, i had the v5.61 opened while looking at settings in v5.2, just so many versions out there 😮💨😵💫
@@musyc1009 yea i’ve gotten confused like that before. All good!
excuse me, could it be a issue using an mp4 file instead than a mov? thank you
Yes it could be mp4.
As long as you have a plain background.
Error in ''Define optical flow functions for Video input animation mode only'' help me please
Have you checked my troubleshooting video? Disco Diffusion Errors (Troubleshooting)
ua-cam.com/video/eA5iTXtsXGA/v-deo.html
Would you still recommend this method or is it outdated? Is stable diffusion better? if it is can you do an updated version where you can do something similar to this using google Collab stable diffusion.
This should still be useful, I’ve seen some creators make very cool stuff following this. I have gotten a few request to do a more in depth tut on SD in colab and I might try that.
Hi bro can ı ask something to you. I am using macbook air m1 16 gb. But this machine is not working well after effect. I want to try AI animation things. İf ı will buy macbook pro m1 pro 32 gb. is this device enough for that stuff
Hi ! When I render from AE transparent video in quicktime format and after I upload that vid to DD, it's still sees the black background , any ideas why?
Yea it will always do that. You could use the footage without the background as a matte to remove the black bg after it’s been generated.
@@enigmatic_e thank you for respond, my problem is that DD still can see black background and it's making animations on it. Perhaps I messed up the settings?
thanks man!
have you tried the similar approach in A1111 SD? For me it outputs not on transparent layer, but SD adds a white background on my rotoscoped model. i can't seem to find a way around this! curious if you have?
Hey Viktor. I was wrong about the whole transparent bg when it comes to generating the AI art. However, rotoscoping is still good because you could alpha matte to remove bg after generating art.
@@enigmatic_e so you would re-rotoscope the finished art? argh that is less than ideal. for me it, a non computer scientist, it seems obvious to not add in a white background. would be so good if the transparent layer could be kept
@@Slothen no just once. i go over it on part two of this video ua-cam.com/video/SgNtc2c3Y5Y/v-deo.html
@@enigmatic_e fully appreciate your work and sharing these video and knowledge. you where my entry point into the world months back. okay, done with the charming :D so it seems there isn't a way to have the AI render out with transparency alpha, but only work arounds like the copy of rotoscoping you are demonstrating in the other video. bummer!
he still working on phone ? ty for vidéos
hey I am having issues at 2.0 is there a problem with the software being copyrighted now ?
Do you guys know why Disco is creating images even on the transparent background?
I exported it as a Quicktime video with RGB + Alpha turned on and it still uses the background, do you know the way around it?
You don’t need it to be transparent. I know I suggested that but it’s because you might want to use the rotoscope layer as a matte to remove the background.
@@enigmatic_e Thank you
Are generations with diffusion limited or UNLI?
Curious if you'd be able to help with the following error: "FileNotFoundError: [WinError 2] The system cannot find the file specified". I uploaded the video, made sure the frame size was correct, and copy and pasted the path into the correct spot. I DID notice that mine did not say "/content" before the file name, but not sure why. Wondering if that has something to do with it? Please help!
Hello, sorry for late reply, are you using the same version as I‘m using in video?
@@enigmatic_e No worries! I am. I tried a different version too, but couldn't get either to work unfortunately.
I put in your exact settings and for some reason each of my frames is more distorted than the last. Any idea why that might be?
Might be that your clip has a lot of movement. If it does it may also require you to generate every frame instead of every second frame if there is a lot of movement.
The subjects are actually standing still. And I should clarify, by distorted I just mean that it’s farther and farther away from the original frame. Frame zero it’s a slight texture on the subjects and by frame 13 they are becoming swirly blobs.
@@jadesleeper4713 ah ok. Try bumping up frames_skip_frames to 70% let me know if that works
I actually have it on that. I was thinking that maybe it’s the ratio between the cgs and the init_scale? It’s just weird because it’s the same settings that you had on this video but yours seems to stay pretty consistent. Mine seems like it’s using the last frame output to work the next frame instead of referring to the original frame.
@@jadesleeper4713 yea thats kinda weird. It could several things. What is the size of the video pixel wise?
Hello, please Help me! What is the correct way to format the parameters to resume running?
display_rate: 20
n_batches: 9
resume_run:
run_to_resume: [0]
resume_from_frame: [118]
retain_overwritten_frames:
****MY Basic Settings: TimeToDisco
the name of my last frame is - TimeToDisco(0)_0118
I have 313 total frames
How do I resume work from frame 118?
Help me, please!
put iteration number then put frame number and make sure the resume box is checked
@@enigmatic_e I can not run the same error as I threw off above. is it better to use another version?
or where exactly to write the desired value. write please
Thx for the tutorial bud but it doesn´t work for me I got this error: "ModuleNotFoundError: No module named 'utils"
Hey, at what point do you get that? At settings? Happens to me a lot, often times it's something minor you overlooked.
How fix the various error of disco diffusion? i follow your video step by step but there is an error like ( ERROR: 0 flow files found ! )
Can you give me more info. What is the video format? What is the dimensions? At what point do you get the error?
Hey I made a new video to help with troubleshooting. Let me know if you find that helpful.
Any reason to go for Disco over Stable Dissusion? I don't like to run a paid colab, as long as I can make use of 8gb RTX 3070
Disco works and looks different. Disco diffusion gives more of a psychedelic look. You can technically use colab free but you have to be active on the browser or else it stops the session which sucks if you want to walk away from computer.
@@enigmatic_e 10x!
my frames just go to drive unaltered :/
may have to do with settings. Might be it's not affecting the footage at all.
Hei mannn Really appreciate your work here. But somehow DD read my transparent background as black!!? 😱😱😱 how to export transparent background from DD?
Hey! Don't worry about it, DD will make it into jpeg image, which will make the background black anyways, so even a mp4 works great. But keep the transparent vid, you can alpha matte it to AI animation to remove the black background after!
@@enigmatic_e thankss for the answerr mann 🙏🔥
@@enigmatic_e Will do 🔥🔥🔥
cool, too bad DD dont save PNGs with alpha tho
Yea it doesn’t but you can use your original rotoscoped footage with alpha and use it as an alpha matte. Thats how rotoscoping can be useful and also doesn’t have the background to distract from subject. I went over that in pt 2
Dont work anymore :(
Gotta try warpfusion, the upgrade to this, I have tutorial on it, check my newer videos
Remember people, never support ai arts with your money or praise as they already destroying what is great about humanity which is "creations" and quality craft with thoughts and life knowledge that brought us happiness. AI art doesn't have any clue or idea for our inspiration and greatness as much as a living human itself.
What corporations and ai programmers have done is not okay at all. They don't care about the quality of experience and entertainment that brought people greatness, they only care about themselves collecting huge cash and doesn't care of the earth they living. It's getting worse and worse of how they can just insult us all with these ai arts techs that will make our entertainment media worse.
Dude your adress in shown 10:00
Thanks! It’s not my address, luckily, it’s a public location but I’ll try to see if I cut that part.
@@enigmatic_e oo good. Watching out for you 🙏🏼🙏🏼
@@studio85amsterdam i appreciate that. I totally missed it.
@@enigmatic_e nice. There is a mess up at the disco diffusion notebook at the moment it seems. Do you know anything about this? Its missing a file that you need for section 1.4. Many people are asking about it on the discord.
@@studio85amsterdam not sure what it is but I get a lot of people with this issue. I haven’t used DD for a while but I’ll get back into it to see if I can figure out the issue.
disguisting AI!
very "artistic"
There's only one H in 'height'
Error -
---------------------------------------------------------------------------
FileNotFoundError Traceback (most recent call last)
in
1 #@title ### 1.4 Define Midas functions
2
----> 3 from midas.dpt_depth import DPTDepthModel
4 from midas.midas_net import MidasNet
5 from midas.midas_net_custom import MidasNet_small
2 frames
/content/MiDaS/midas/backbones/next_vit.py in
6 from .utils import activations, forward_default, get_activation
7
----> 8 file = open("./externals/Next_ViT/classification/nextvit.py", "r")
9 source_code = file.read().replace(" utils", " externals.Next_ViT.classification.utils")
10 exec(source_code)
FileNotFoundError: [Errno 2] No such file or directory: './externals/Next_ViT/classification/nextvit.py'
SAME!!!!
Sorry I haven’t used DD for a while. I’ll get back into it to see how to solve these issues.
Update - i just had to delete the 1.4 cell :) anyways thanks
@@lentopronto and now everything is works?
@@denizersoy8496 yeah it's working fine for me