Automatic Infinite Zoom Using Midjourney - No Video Editing Required!
ฝัง
- เผยแพร่เมื่อ 1 ก.ค. 2023
- Here's a quick and simple to follow tutorial on how to automatically create videos with Midjourney's zoom feature introduced in version 5.2.
Google Collab: colab.research.google.com/dri...
Original Author Github: github.com/mwydmuch/ZoomVideo...
If you've been seeing a lot of videos popping up of people using Midjourney's zoom out feature, but don't have experience with video editing software to make your own, here's a simple guide on how to automatically generate videos using my Google Collab notebook.
No installation or video editing knowledge required!
FREE PDF Prompt Guides, Tutorials, etc:
taoprompts.gumroad.com/
Instagram: / taoprompts
#midjourneytutorial #midjourneyai #aiartcommunity #aiart #aivideocreation
Any reason why one of my images gets out of order in the sequence? Number 10 or a higher number jumps to the second slot. What could cause that? Edit: I'm guessing because of double digits. Or maybe I start numbering from 0 instead of 1? Another Edit: So I had to put double digits to keep the sequence. Starting from 01 to 20. I'm using 20 images. Thank you for putting up this super easy tutorial. Better than using video editing software like other youtube TUTS.
If you have more than 10 images you need to have 2 digits for each image at least, so put a zero in front of the 1 digit numbers. For example 01.png, 02.png, 03.png,...... 10.png. let me know if you have any other questions!
The names of the images are actually text, so name the images so the text-name ends up the same length across all images. For example: Image01, Image02 .... Image09, Image10, Image11. That way they will sort into the correct order. If you have a hundred images, then you'd need the numbers inside the text to be 3 digits long, by adding another zero in front.
hey i tried the way you said for some reason the 8th image dosnt merge properly @@nancykay3109
Wow!
woah, that's a super automated approach to creating these vids! thanks, extra helpful
You're welcome! It's always nice to automate
works perfect!! thank you!
Hi Tao,
It works great! Thank you, so much!
Awesome, glad to be of service!
Thank you so much man. I used it for my last music video.
That's awesome, it's cool to see people are still using this tool.
Thank you so much for this tip!!...I was looking for something like this, cheers! 👍
I'm glad you found this helpful, the zoom feature is great!
@@taoprompts ...yes it is, and this tool you shared is amazing!...the clips generated are seamless, no bumps nor any kind of interruption, just great, thanks again!
@@arturoarvizusart6127 You're more than welcome!
Awesome thanks been looking for this solution!
Glad this helped!
life saver you are!!
Great vid! Do you know by any chance if it is possible to use the zoom functionality in combination with an AI video generator, so a combination of moving images while zooming out?
Thanks! this script is only meant for images. You'd probably need a video editor to get the same effect for an ai video generator.
nice very helpful, tysm!!
Happy to help!
Great! Thank you very much!
Your welcome! I'm glad this helped.
Thansk a lot.
love this. thank you for the tutorial! For some strange reasons, one of the images out of say 12 images will show the edge on the video, apparently the program skipped the "blending" of that particular image. I have tried 2 sets of images in different durations and no. of images and both videos turn out to have one "busted" image. Any clue how to fix this? thanks!
Is it working for a smaller set of images, like 3 or 4? If so then it could be that Midjourney is not zooming out "correctly" and that particular image isn't centered perfectly.
Hallo, in "Install Prerequisites" i get the Error Message WARNING: The following packages were previously imported in this runtime:
[PIL]
You must restart the runtime in order to use newly installed versions. I pressed the Restart Button but it seems that nothing is happening. In "Run The Zoom Video Composer" than i get the Error Message "python3: can't open file '/content/zoom_video_composer.py': [Errno 2] No such file or directory" Any Idea? Thanks
Hey, refresh the page. And run the Install prerequisites again. This time ignore that Error message, don't press the restart button, and proceed through the notebook.
That should fix the problem. You don't need to worry about that error message.
Is there a way to connect this automation to take the pictures from a Drive folder? Bc the automation crashes in the uploading process. I'm using the paid version of Google Collab, and even that didn't solve it
You would have to duplicate this collab notebook and modify the code inside to access images from your own drive.
So much easier than after effects! 🎉
Sure is! Thanks for watching.
this really helps
Your welcome!
I keep getting this error when I try to run the "run the zoom video composer" section: "python3: can't open file '/content/zoom_video_composer.py': [Errno 2] No such file or directory". Any ideas? My end? Their end?
Did you successfully run the "install prerequisites" section at the beginning of the page?
@@taoprompts So messed around with it a bit, not really sure what what I did or what changed, but it seems to be working now so idk lol
@@hardtruth7555 Glad to hear it works
There is any issue when using a portrait aspect ratio? The ones I tried have a misalign frames, or it can be on the images im using?
It could be both. If you have it working for some aspect ratios and it doesn't work for others it might be because Midjourney does not center the images perfectly.
You could try with a different prompt to see if that works.
@@taoprompts thanks for the reply!
It is the other prompt, I tried a different one and it worked, thanks again!
@@macnamararj glad it's working!
python3: can't open file '/content/zoom_video_composer.py': [Errno 2] No such file or directory
Hey, can you try restarting/refreshing the collab notebook. It seems to be working fine for me.
@@taoprompts Hey! That's an excellent tutorial, but I have the same problem. Isn't there an updated version for this?
@@piccadillysessions Can you try refreshing the page and restarting the notebook?
Hi Tao,
there is some issue on the second step, the images doesn’t loading. could you please help?(V_1.png(image/png) - 1158151 bytes, last modified: n/a - 0% done)
MessageError Traceback (most recent call last)
in ()
14
15 # upload images
---> 16 uploaded = files.upload()
17 for filename in uploaded.keys():
18 dst_path = os.path.join(upload_folder, filename)
Hey, could you please try refreshing your page, and clear any browser cache or cookies?
I haven't seen that error before.
I’m getting the same error , did it get sort out?
@@imaiginarts8698 not yet, still the same
@@user-nr2vy5ng9s Hey I didn't see that you still had the same issue. If the images are not uploading, it could be a browser problem. Google chrome should work.
Do you just keep pressing zoom out 2x, and you get these results, or do you custom zoom every single image?
I just used the zoom out 2x button for these.
Brah.
Can't play the video off the site. Either on mmybdesktop or phone.
It puts it in a .mp4 format that won't play on my desktop.
How can I take this and upload on TH-cam or other sites
Ive tried finding a converter but am not having much luck.
Try this mp4 to avi converter here:
www.veed.io/convert/mp4-to-avi
@@taoprompts worked captain.
Phones video player can play it. I sure appreciate it
@@mr.mikesart7111 Glad you got it working man
@@taoprompts the quality is lower but it works
Is anyone else getting the video and when it zooms out the images are inside one another it's hard to explain.
Hey, Midjourney does not always keep the images perfectly centered during the zoom feature so sometimes this can happen. If you use 1:1 aspect ratio or 16:9 aspect ratio it should work fine with 2x zoom.
If you use 1.5 zoom try setting the zoom factor to 1.333333 and it should work.
thanks! i get the following error: MessageError: RangeError: Maximum call stack size exceeded.
Hey, could you try to use a different web browser, such as Google chrome.
hello amazin gprgram thanks a lot. the first time it was running well. no i have always this message: python3: can't open file '/content/zoom_video_composer.py': [Errno 2] No such file or directory
...
Can you try running it in order over again? It's working fine when I try it.
It looks like you are in the wrong folder location somehow.
i keep getting this error
python3: can't open file '/content/zoom_video_composer.py': [Errno 2] No such file or directory
Hey, did you run all of the blocks in sequential order, Especially the install prerequisites block?
hi yes i fixed it have to run the first prompt then click the restart then redo the first prompt @@taoprompts
Definitely need a follow up video on how to do anything with the saved format.
It doesnt appear sharable, or viewable using standatd phone or computer applications outside of the maker
Hey, is that because you can't play mp4 files?
That should be standard for most operating systems.
You could try a mp4 to avi converter like this one here:
www.veed.io/convert/mp4-to-avi
I did everything, but the first frame remains a block, as if the second frame is missing.. I have image_01 through image_10 ..
I know now I think, it all depends on the dimensions of the image. It doesn't work with every height and width. I'm looking for instagram reels size, so vertical
Hey, I tried running it on 9 to 16 aspect ratio images with 2x zoom and it worked fine. Make sure you have the dimensions entered correctly. For 9 to 16 the height is 1456 and width is 816.
@@taoprompts yess it is good now Thank you so much😘
我收到这样的一个错误:ValueError: could not broadcast input array from shape (922,2970,3) into shape (845,1293,3)
Are you sure you named your images in the correct order? Also do they all have the same aspect ratio?
It can be that is not working anymore? Or there is a new version somewhere? If so can anyone give me the link
What's the error you are getting? It's working fine for me.
@@taoprompts IDK what happened but i had to try it many times with differents google accs before i can use it but in the end it work. So no worrys. Thank you for the reply!
i had many problems, maybe was my Internet connection, but not sure. The pictures not being upload and freezing at 0% in the video composer
@@rustenberg3994 alright, glad to hear you got it working! Google collab can be finicky at times.
I get a warning that they get access to my Google Account, is this method really save??
Yes, its safe. Google Collab is run by google itself.
Also, I'm the creator of the notebook.
I’m not able to get past the 2nd step only , I am unable to upload files only , it always gives error
What's the error?
@@taoprompts is it because I was doing it in Ipad coz it worked smoothly on my phone
@@taoprompts This is the error I get while doing it on Ipad IMG_3655.png (image/png) - 1622761 bytes, last modified: n/a - 0% done
MessageError
in ()
14
15 # unload images
-_-> 16 uploaded = files. upload()
17 for filename in uploaded.keys ):
18
dst path = os path. join(upload folder, filename )
Traceback most recent call last)
y
3 frames
|10
• Share
RAM
Disk
Lus/local/lib/python3.10/dist-packages/google/colab/message.pyin
@@imaiginarts8698 Good to hear you got it working!
How do you do this on a mac
If you run the notebook through a chrome browser it should work fine on a mac.
Ità not working. do i have to a abo to create a video? I thought it`s free.
It is free to use. You may need a google account though.
THANK YOU.
Also, why does the audio in this video have so much bass lol. My sub was goin wild, thought I turned it up by accident or something.
Wait, why does this thing put 4 images on the wrong end of the video? Happened twice so far, so so far it doesn't seem coincidental.
EDIT: Note. My filenames are in lexicographical order, and the modified dates are in order as well as the creation dates. What controls the order???
Can you give me some examples of your image names?
@@taoprompts They were all "test2 (4).png" - "test2 (26).png" (I skipped 0-3)
EDIT: I just seen in one of your responses below that they should have 0's for the single digits, I can try that!
Yeah, try to use "test2 (04).png", that should do the trick, let me know if you need any other help.
Can you do a video over from video to ai video?
Hey, I'm not sure what you mean, do you mean image to video?
@@taoprompts no, i see a lot of video now that turn into cartoon video. are they also made with this software?
@@moniqued8932 I think what you are talking about is combining stable diffusion with controlnet, to create ai videos, frame by frame.
@@taoprompts I only use midjourney😏
@@moniqued8932 well good news is the midjourney team have mentioned they are potentially going to add controlnet in the future, so it's definitely possible.
WARNING: The following packages were previously imported in this runtime:
[PIL]
You must restart the runtime in order to use newly installed versions.
Hey, you can ignore that error, just go ahead to the next section, everything will work.
Зашибись как круто.
WARNING: The following packages were previously imported in this runtime:
[PIL]
You must restart the runtime in order to use newly installed versions.
(restarting run time) no changes
FileNotFoundError: [Errno 2] No such file or directory: 'results/zoom.mp4' tried to refresh the website, but didn't help
Hey,
1. Ignore the error asking you to restart the runtime, the notebook will run fine, just move on to the next cell directly..
2. Make sure to ignore the error from the first block. And make sure to run each of cells in order.
Let me know if that fixes things.