Can't open mp4 with Vegas 13 and have tried "everything"

Kinvermark wrote on 5/14/2019, 9:44 AM

Yes, Cineform is very "scrubbable."

2) Which other programs do you use it in?

Have you tested "Intermediate 422 HQ? I'm just getting my feet wet but JN_'s numbers look impressive if they're on a similar playing field.

https://www.vegascreativesoftware.info/us/forum/codec-xavc-s--115606/#ca716992

@Musicvid

Primarily Davinci Resolve and Handbrake. Handbrake is easier as it is only a destination and can accept several good options. Resolve workflow is trickier as it can be the source or the destination for files and those files can be proxies, intermediates or XML references to originals (which means I need timecode to work too.) Resolve cannot make Prores files (ie Magix intermediate), cannot read or write most kinds of AVI (no VFW support) and is reliant upon MOV container for AVID codecs (no MXF). MOV native support is OK, but limited in Vegas, and nobody really wants to hitch the wagon to QuickTime for windows anymore, do they?

So that leaves cineform - which is somewhat vulnerable.

So, all-in-all a lousy situation for intermediates on windows. Capitalism, pahhhh! What a mess.

Former user wrote on 5/14/2019, 9:52 AM

@musicvid

I think I understand, maybe, or I just assumed it was self evident from my notes sprinkled above and about the tables.

I used a source file, Xavc-I, that was itself rendered from a variety of smaller 4k uhd clips, a few different dslr, mirrorless cameras were used.

Anyone else can use their own source file to do the same tests using the HO and ffmpeg programs. I have supplied the ffmpeg (thanks ffionk) command line that I used to generate the ffmpeg results, included in table text.

While the results I did may not match each other exactly, i.e. HO vs ffmpeg, I think that the trend and pecking order is very similar and stands out, winners and losers.

Another user, using a different source file, I suspect, will get broadly similar results.

Its obvious, but maybe worth saying anyway, the data rate is a broad indicator of quality as well as codec. I couldn’t always get the say, Cpu, Nvenc, Qsv files to have the exact same data rates and same file sizes but even so most people can intelligently parse the results accordingly.

I included the file sizes and data rates because for sure these also need to be taken into account. Its all very well choosing a higher quality intermediate, but is the file size something I can live with. The data rates generally are a very good indicator of how well they will match the source for quality fidelity.

Intermediates and others ...

Intermediates ...

Musicvid wrote on 5/14/2019, 10:13 AM

I used a source file, Xavc-I, that was itself rendered from a variety of smaller 4k uhd clips, a few different dslr, mirrorless cameras were used.

That is a very good approach, and one that may have more general relevance than mine.

While the results I did may not match each other exactly, i.e. HO vs ffmpeg, I think that the trend and pecking order is very similar and stands out, winners and losers.

I also assume that is the case, but I haven't yet tested it using uniform source, so it is still speculation at this point. What I do know from comparing our two charts is that there will be aggregate, if not hierarchical, differences. All quite normal.

If your conglomerate source is available for download, I could use it in another project.

Kinvermark wrote on 5/14/2019, 10:23 AM

@Former user Great information BTW, thanks for all the work testing, compiling, and presenting in a nice table!

Former user wrote on 5/14/2019, 10:27 AM

@Kinvermark Appreciated 👍

Musicvid wrote on 5/14/2019, 11:58 AM

@Former user Great information BTW, thanks for all the work testing, compiling, and presenting in a nice table!

Ditto.

Former user wrote on 5/14/2019, 1:12 PM

"If your conglomerate source is available for download, I could use it in another project"

@Musicvid I thought you might prefer to roll your own! Anyway, if you wish to use it in a test project, knock yourself out! Note that its 745mb in size.

https://www.dropbox.com/sh/cebefr9ru6a3pfv/AADB9T-zWztN0pPijMLgIFBka?dl=0