agrumpyoldsod
Well-known member
I have only completed a very few vids for posting on YouTube, and am "having" to use DaVinci Resolve 18 (Studio) ["DVR18"] so I can "film" in 12-bit Nikon RAW (or multiple vid types) and I have discovered that when I choose to "Deliver" using the YouTube preset in DVR18 the video quality looks "lousy".
Whereas if I export as ProRes and then use Adobe Media Converter to generate a video in H264 you display on YouTube the resulting video is far higher quality.
Or one can just leave it upto YouTube - ie upload the ProRes 422 HQ version.
Why bother -- well the upload and encoding time YouTube needs is much longer for the file in ProRes and much shorter for the file exported from DVR18 as H264 -- and there is a risk that my internet connection drops out or I get distracted during the upload. [The ProRes upload takes a long time -- still going when this post was saved]
Obviously I keep the videos and workspace at as high a quality as I can while in DVR18 and only "downsample" on export. The same way I work with still images.
What formula/workflow do others use to generate their output and then either use directly or resample for use online ?
Do any just leave it up to youtube to translate the file or do something similar to me -- downsample before uploading?
Here are links to the 3 examples :
Option A) -- select YouTube 4k preset in DVR18 - 480mb file size
Option B) -- select ProRes 422 HQ 8k preset and export from DVR18 and upload the video to YouTube - 44GB file - leaving it to YouTube to resample as it needs to. At least 5 hours to up load this 6 min+ video when in ProRes 422 HQ 8k
Option C) -- select ProRes 422 HQ 8k preset in DVR18 and export then use Media Encoder to resample/downsample to H264 -- 460 mb file size
Whereas if I export as ProRes and then use Adobe Media Converter to generate a video in H264 you display on YouTube the resulting video is far higher quality.
Or one can just leave it upto YouTube - ie upload the ProRes 422 HQ version.
Why bother -- well the upload and encoding time YouTube needs is much longer for the file in ProRes and much shorter for the file exported from DVR18 as H264 -- and there is a risk that my internet connection drops out or I get distracted during the upload. [The ProRes upload takes a long time -- still going when this post was saved]
Obviously I keep the videos and workspace at as high a quality as I can while in DVR18 and only "downsample" on export. The same way I work with still images.
What formula/workflow do others use to generate their output and then either use directly or resample for use online ?
Do any just leave it up to youtube to translate the file or do something similar to me -- downsample before uploading?
Here are links to the 3 examples :
Option A) -- select YouTube 4k preset in DVR18 - 480mb file size
Option B) -- select ProRes 422 HQ 8k preset and export from DVR18 and upload the video to YouTube - 44GB file - leaving it to YouTube to resample as it needs to. At least 5 hours to up load this 6 min+ video when in ProRes 422 HQ 8k
Option C) -- select ProRes 422 HQ 8k preset in DVR18 and export then use Media Encoder to resample/downsample to H264 -- 460 mb file size
Last edited:
