-
HELP! dots little dots
I am working on project with 154,000 frames… the job is remove these random dots that appear roughly every 3-4 frames with no specific predictability.
i have been going frame by frame at a slow crawl cloning these small dots out.these dots are on the tape as well as the quicktime i am working on.
my question to you guys is:
how can i speed this process up?
is there another way from the way i am doing it?this process is tedious and i am looking for an alternative
please help
Nick Ito
-
9 Replies
-
if they are bright, you could create an auto mask cutoff matte based on luminance. or try this demo
https://www.redgiantsoftware.com/products/categories/enhancement/film-fix/
or try logoaway if the dots are usually in the same place.
send us a pic.
-
these dots are in random places…will film fix work you think?
if so i may try to talk them into buying this.as far as auto mask. isnce its in random places from frame to frame.. wont it be hard to track? very time consuming remember 154000 frames.
-
Try out the film fix demo and find out. Since its random, logoaway won’t work, but if you send me a pic of the dots, I can see what AE can do. I’ll ask Maltannon about pixelsampler. I’m sure he could make a dot remover since his project is 99% there.
-
not sure ill be on this job… within the hour.
let me know if anything comes to mind.(and i cant release any pics because of a non disclosure.)
-
so i was pulled off .
i did not see eye to eye with the director as far as deadlines were concerned.
i will post up a representation of what i was editing though
-
thanks for seeing it through my eyes…
so this is prety much what i was doing
find those dots (circled)
and clone them out frame by frame.
-
too bad the colors are all different, it seems very close to a “plug holes” command in a color keyer based on pixel size and hue which would work fantastic.
-
alot of them are white…just in different areas of footage throughtout the movie.
you think color keyer is a potential solution? -
I tried several things and clone stamp is still the fastest until someone programs a custom pixel plugger. The reason is, you would have to exclude common large clusters of pixels and only search for 2 pixels in length that are higher in luminance than surrounding area. Maltonnon you listening? There is a pixel sampler he codes all the time. He could become rich and famous.
Reply to this Discussion! Login or Sign Up