I would like to know if is it possible to increase the brightness or gamma of a video without full re-encoding, doing it a quick (and maybe dirty) way? For example some versions of ffmpeg allow certain special bitstream filters described here - presumed we have a h264 video - : http://forum.doom9.org/showthread.php?t=152419 which can alter some of the video characteristics without re-encoding the video. Some containers also can store such meta informations about the video gamma and/or brightness, as far as I know matroska, and QuickTime, and maybe there are some video players too which can deal with this information.
I already know that a similar modification, namely cropping the video is possible without a full re-encode merely by setting the cropping information at container level for H264 bitstream. When such a cropped video is uploaded to youtube, on the server while conversion takes place it is cropped at the values which were set in the bitstream and without real re-encode. So this could be a quick fix for unneeded logos and black bars.
Now I would like to apply some similar trick to adjust gamma (preferably) or brightness before uploading videos to a video sharing website. It also would be important that ffmpeg must honor the metadata when the conversion at server side takes place. Those video sharing servers usually use ffmpeg, I think.
+ Reply to Thread
Results 1 to 3 of 3
-
-
In general, a brightness/contrast/saturation/gamma/etc. changing filter gets applied to the video content as such, it will change the video as a whole. It is not just "metadata" to be interpreted and applied by the decoder or renderer while displaying the video. I could only imagine two exceptions:
a) The decision between full or limited range would have an impact on the contrast range. But only by one discrete factor, in a meaning that you could fix it if the flag was set with the wrong value (e.g. too deep or too shallow black and white limits). There is no finetuning.
b) Some advanced custom colorimetry may have an impact when the player (and monitor) is able to handle it. But its purpose is not personal customization, especially not during a format conversion. -
IIRC, the Cineform codec has the capacity for storing & retrieving/acting upon "active metadata" such as this.
Whether an app can read from or write to the encoded file properly is another matter. Pretty sure only Cineform studio and a few $$$$ specialized post-production apps can write. Similar rarities for playback/decode.
Note that this type is embedded within the encoded video data, not as part of the container's.
The problem as well as the benefit of container-level metadata is that it can be easily divorced from the core data.
Scott
Similar Threads
-
How to Increase Audio Level (Volume), Once Extracted - ?
By elling in forum AudioReplies: 5Last Post: 21st Jun 2016, 14:27 -
What Is Video Level and How To Change Video Level
By Aashik Alam in forum Video ConversionReplies: 3Last Post: 12th Jan 2016, 07:26 -
MeGUI - Black Level booster for anime/how to increase black edges?
By kairukun in forum Video ConversionReplies: 2Last Post: 5th Nov 2015, 01:44 -
Sound level increase...
By ziggy1971 in forum AudioReplies: 13Last Post: 7th Feb 2015, 20:31 -
need to put content of FLV container into MP4 container
By arraboy in forum Newbie / General discussionsReplies: 3Last Post: 29th Jun 2012, 19:52