Timeline for What settings will preserve quality when I output an edited .mp4 for Youtube upload?
Current License: CC BY-SA 3.0
20 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
Nov 3, 2015 at 22:46 | vote | accept | Reverend Speed | ||
Nov 3, 2015 at 22:44 | answer | added | Reverend Speed | timeline score: 0 | |
Nov 1, 2015 at 21:19 | history | tweeted | twitter.com/StackBlender/status/660929256852234240 | ||
Oct 30, 2015 at 23:42 | answer | added | troy_s | timeline score: 1 | |
Oct 27, 2015 at 10:24 | comment | added | Reverend Speed | @cegaton Might investigate DaVinci, thank you for suggestion. However, am I correct in imagining it offers a better workflow/interface, but I'm likely to encounter the same compression problems? If so, poss just need to jam OBS encoding bitrate as high as I can (currently 8000kb/s, might double?) then suck up the artifacts Blender adds as cost of doing business. Will investigate - if so, will prob provide answer including your link and w/some info on OBS. This satisfactory? Thanks for feedback. | |
Oct 27, 2015 at 10:17 | comment | added | Reverend Speed | @Mentalist The issue here (as detailed in the first line of the original post (after the inserted 'edit')) is the process of editing and reencoding footage in Blender that was initially captured/generated OUTSIDE of Blender. The line in question: "Original file, captured from I Am Alive with Open Broadcasting Software". OBS will only export to .mp4 (individual frames are unavailable as I did not render them in Blender). This is the starting point for Blender;I am unclear on how separating these frames into .pngs would prevent compression Blender ADDS to the images, which is crux of problem. | |
Oct 27, 2015 at 10:02 | comment | added | Reverend Speed | Apologies for delayed response, dealing with heavy cold. | |
Oct 26, 2015 at 6:28 | comment | added | Mentalist | @cegaton Correct. My understanding from the wording of the question and the screen shot is that OP has been skipping the step of rendering to frames and instead rendering straight to a video file, then attempting to edit that in the VSE. The problem being that before the editing stage it's already lossy. | |
Oct 26, 2015 at 4:03 | comment | added | user1853 | @Mentalist note that even rendering in OpenEXR or PNG from the VSE. the upload to YouTube has to be a video file, so those frames will have to be encoded to a video container no matter what... (and then, once uploaded, YouTube will re-encode the file yet again...) | |
Oct 26, 2015 at 0:56 | comment | added | Mentalist | Like others are saying here, don't work from a lossy format. I will add: rendering as individual frames in a format such as PNG is the best way to go. There are several reasons for this: 1. If the animation render is interrupted you can always just resume from the last frame. 2. You can make precision edits in an image editor if you need to. 3. It's easier to trim off unneeded portions of the animation. 4. You can work with lossless formats. If you need more dynamic range, OpenEXR is also a great format, although file sizes can be quite large. | |
Oct 25, 2015 at 18:25 | comment | added | user1853 | An original with less (or no) compression is a good start. Any program you use to edit will need to re-render the footage to get a new file no matter what. Learn about data rates and compression codecs to base your decision. Blender's VSE performance and interface are terrible when compared with dedicated editing software, it's a nice addition to the 3D creation environment, but it's still quite basic and light years away from tools designed just to deal with editing. You might want to try DaVinci resolve lite (free) or pay some money for Adobe Premiere, Vegas or Avid. | |
Oct 25, 2015 at 17:51 | history | edited | Reverend Speed | CC BY-SA 3.0 |
added 558 characters in body
|
Oct 25, 2015 at 17:45 | comment | added | Reverend Speed | Thank you for the link, still absorbing some of the points made there. However, selecting H.264 with Lossless Output selected and the output options as set in the accepted answer on the linked question results in an output almost identical to the second screenshot in my original question post. Is this unavoidable? Is Blender, then, unsuited to editing such footage? Should I use a different program that doesn't need to re-render the footage? Or should I make my initial recordings with far less compression? | |
Oct 25, 2015 at 16:07 | review | Close votes | |||
Oct 25, 2015 at 21:58 | |||||
Oct 25, 2015 at 15:50 | comment | added | user1853 | Possible duplicate of Lossless compression for video with audio | |
Oct 25, 2015 at 15:49 | comment | added | user1853 | The issue is that you are editing from an original that uses a lossy compression codec: mp4. You've already lost some information there in order to get a small file. Every time you decompress and re-compress you'll loose even more information and that will eventually be visible as artifacts (color blocks, blotchy unsharp image, etc). Long story short is that in order to preserve the original quality you'd need to be editing in a more robust codec that is lossless (or visually lossless), that will create much larger files but keep the video from furhter degradation. | |
Oct 25, 2015 at 15:16 | comment | added | Reverend Speed | Remotely related, linked page restates a lot of information from the Youtube encoding page I mentioned in my original question. I did just try rendering to .mov and changing max bitsize to 0 as suggested in your linked page, but seemed to have no effect on the test frame - loss of detail, additional artifacts. I'm looking to eliminate these artifacts before uploading to Youtube. | |
Oct 25, 2015 at 14:37 | history | edited | iKlsR |
edited tags
|
|
Oct 25, 2015 at 14:37 | comment | added | iKlsR | Possible dupe or related. blender.stackexchange.com/questions/24724/… | |
Oct 25, 2015 at 14:13 | history | asked | Reverend Speed | CC BY-SA 3.0 |