Movie maker + Start > End frame
Hi there, is it possible to fusion the movie maker with the start end framer WF for make smooth transition between each segments instead of hard cut transition ? Have great day
Yes, with a different method of extending. Basically doing "extend a video" workflow for each segment.
Was thinking of making such. Will do ;-)
Also about to upload a workflow that is based on Prompt Relay, where you probably could also do that. But since that is 1 video (not multiple videos extended), it can only generate say about 30 seconds or so max.
But you can continue, and generate next 30 seconds of the song, and so on (and edit it together manually).
Will post when i got something ;-)
Yes, with a different method of extending. Basically doing "extend a video" workflow for each segment.
Was thinking of making such. Will do ;-)Also about to upload a workflow that is based on Prompt Relay, where you probably could also do that. But since that is 1 video (not multiple videos extended), it can only generate say about 30 seconds or so max.
But you can continue, and generate next 30 seconds of the song, and so on (and edit it together manually).Will post when i got something ;-)
That's amazing ! thanks again for work! Cannot wait :D
Hello for the movie maker workflow, is there a way to remove subtitles generated in the video. I saw some people manage to do it using the nag node. I added it but nothing happened. Maybe i've done something wrong. Does anyone know the right way to do it? thank you
is there a way to remove subtitles generated in the video. I saw some people manage to do it using the nag node. I added it but nothing happened.
Check the negative prompt if its subtitle there. But i think it should already be there.
Also make sure you are using version 1.1 of the distilled model (if you are using that) or version 1.1 of the distilled lora (if you are using dev model)
and lastly also version 1.1 of the upscaler model
Thank you! Yeah i'm using the GGUF distilled 1.1 version, I had to change the way the text is said in the prompt and I had to come up with a long list of negative keywords to make it work at the end. I actually did not even need the nag node as I disabled it and it worked. For anyone interested here is the negative prompt:
"NO TEXT, NO WORDS, NO LETTERS, NO SUBTITLES, NO CAPTIONS, BLANK SCREEN, NO OVERLAYS, NO DIALOGUE ON SCREEN, NO SPEECH BUBBLES, NO TYPOGRAPHY, NO NUMBERS, text, letters, numbers, captions, subtitles, closed captions, karaoke captions, dialogue captions, speech bubbles, written words, readable markings, typography, fonts, glyphs, UI text, lower-third text, floating text, visible dialogue, on-screen dialogue, printed dialogue, written quotes, poster text, sign text, machine labels, paper words, screen words, chart labels, chart numbers, logos, watermark, signature, copyright, blurry, oversaturated, pixelated, low resolution, grainy, distorted, noise, compression artifacts, jpeg artifacts, glitches, distorted sound, saturated sound, loud audio artifacts"
The NAG node is for when you have CFG 1 (that "disable" the negative prompt).
So while it might work without, using nag should at least in theory make the negative work better
But as long as it working all is good ;-)