The most recent videos, specifically
"How to turn paintings into photos (and vice versa) with style transfer AI workflows.", and "Weird black and white blobs might be the secret to controlling AI images"), showcase images in the thumbnails that aren't actually featured in the videos themselves. While the concepts are explained well, I personally find it misleading to include images that aren't actually included.
#Dislike the clickbait of the most recent Invoke videos
cloud shoal
tough marlin
CC @pure sorrel
pure sorrel
Thanks for the feedback! Ideally, the thumbnail image would use exact imagery from the session it’s pulled from, but since these sessions are recorded live, we don’t typically plan ahead for specific visuals. This means that during editing, we sometimes decide to swap out imagery to make the title card more engaging or better communicate the video’s focus.
In the black-and-white blobs video, the thumbnail reflects the concept of using abstract shapes to control AI images, even if the exact image differs and in the Renaissance art video, we chose a more recognizable image to highlight the concept. I though the workflows remained consistent with the thumbnails, but understand how this could feel misleading, and I'll aim to align them more closely in the future.
Thanks again for your input—it helps us find the right balance!
cloud shoal
Thanks! I totally get what you're going for, and it's definitely getting the message across. With the blobs one in particular I was like 'OHH neat dragon!' only to realize I didn't get to see it full size in the video
One possible suggestion would be even just to include the thumbnail images overtop while some explanation is being given, i.e. quickly edited in. Just a thought
pure sorrel
Yeah, can think about how to incorporate the thumbnail image into the video in the future, or maybe link to it in the description?