I’ve been looking at the Handy tokens on IVDB, which are great but the main downside is the inability to crop/change the funscript file which also means you can’t create custom clips of the videos with a matching script.
I was curious if there was any way to re-encode the video to use a super low bitrate on portions of the video that I would normally skip such as the intro, or portions where there’s no action etc. That way you’d keep the original video length and can use the matching token/script, but also save hard drive space.
I tried rendering out blank video with a low bitrate to see if I could combine the two in Avidemux (which in theory I could crop the original video and then place in the blank to keep the video length the same), but all I got was a blank video for the entirety.
Not sure if there’s any software or if it’s even possible to do what I’m thinking of. I know you can set a variable bitrate, but I’m not sure if there’s any way to set that for certain sections of a video?
Any help would be appreciated. These VR videos are really taking up a lot of space so I’m trying to cut them down without losing access to the matching script.
To my knowledge there are not applications that can encode an entire video using different encoding settings in different parts of the video.
I’m not sure if it is possible to concatenate (merge) two videos with different settings without reencoding the individual parts to a common setting. I guess you have to go through the options of ffmpeg to determine if that is possible.
Should 100% be possible, used to do this years ago when there were divx fast and low motion codecs, encode separate sections then merge, the decoder was the same for both variations of the codec.
Usually not much call for it now though, encoders are good at deciding where to allocate bitrate. especially if you use a two pass system.
I doubt it would be worth the time and effort for most people.
If you did want to do it though, encode your blanks with the same codec, ensure the resolutions match and its probably a good idea to match frame rates too, although some encoders can do variable frame rates.
Alternately transcode the videos on whole to a bitrate / resolution you can be happy with, lots less messing around.
I did another test and got it to work. I must have had different versions of h265 for each file or something on my first try. I tried one clip with h265 and combined it, it gave me an image but it made the original scene showed still images this time. So then, I rendered out a clip of the original scene with the same codec in Handbrake, this time with a higher bitrate and the two clips combined just fine.
Probably will be more hassle than it’s worth lol. Especially when I’m working with 8K files. I think in my head I ignorantly hoped a blank file would render out super fast, but obviously it still has to render out an 8K video even if it is all black.
I just don’t really use my Handy for missionary/doggy sections, so when over half the scene is stuff I would skip through it feels like a waste of HDD space.
Thanks for the help! I’ll probably just do this for scenes that are 20GB+.
I still wouldn’t really recommend it, but there is a better process.
Reencoding the whole video isn’t required, and its a quality lossy process and CPU / GPU intensive.
Consider something more like this:
Create a 1 min blank, create a 1 second blank
Find the range of the scene(s) you want to remove / keep, think of it like a chapters list, use a tool that allows you to keep this down to frame accuracy to keep it all in sync.
by now you should have a chapter list something like…
(2) 15001 - 35000
(3) 35001 - 60000
From here, save a copy of the original full audio stream, don’t reencode
Save out chapter 1 and 3, copy video stream, don’t reencode
Mux a a combination of 1min blanks and 1sec blanks till the is the same frame count as chapter 2, you might need to delete out a few frames if your 1s goes over, easier to probably just discard a partial second so it lines up to a second mark too e.g. adjust the chapter boundary so it lands on a whole second in step 2
join the original 1, new blank 2, original 3 coping video streams
mux with the original full audio stream from earlier.
Verify the output is as expected before deleting the original and temporary files
You should be able to reuse your 1min / 1sec template blanks in other videos of the same resolution / codec and frame rate.
I haven’t touched video encoding for a long time so specific software is hard for me to recommend specific software, I know FFmpeg can do all the stream copy / extraction and merge functions, you can probably find a GUI for it.
if you use handbreak and do a conversion with h.264 or h.265 this is something already done automatically. less movement + less bitrate. h.265 saves more space, but takes longer to encode. the end resulting file usually shows the average bitrate or total bitrate but im sure theres a program out there to see bit-rate throughout the video
Ah yeah, that’s smart! I am having issues with getting the codec correct though so I don’t have to reencode the main file. It says it’s H.265, so I encode the 1 min/1 sec clip in h.265 (I’ve tried all 4 h.265 presets in Handbrake), but when I mux it together the main file just stays black after the 1 sec/1min.
VLC says it’s specifically MPEG-H Part 2/HEVC (H.265) (hev1). I’m not sure if there’s any way to specifically select that version of h.265? I’m not too familiar with codecs so not sure if it’s something I’m doing wrong or I need to use a different program.
Edit: Just checked the codec for the 1 min file and it’s hvc1. I’ll look online for how to encode in hev1 but hopefully that should work. Thanks a lot for your help!
Edit 2: I’m honestly not sure what I’m doing wrong. When I clip out a section in avidemux, it changes the codec from hev1 to hvc1. So I compared the codec information of the main clip and the blank clip and they’re identical except for the bitrate. But it still won’t play the video correctly after the initial blank clip.
I think I will just have to at least reencode the part I intend on watching. It’ll definitely save some time from having to encode an entire video though.