[Norman Black]

What does the free MediaInfo utility say about these file(s). MXF is a generic container.

I have no direct experience but Vegas does supposedly support P2 series camera MXF files. Are you on the current VP12 build. 770 was the last VP12.

[Norman Black]

I don’t know about AnotherGUI but here is an ffmpeg command line that creates a standard profile ProRes file.

ffmpeg.exe -i -c:v prores_ks -profile:v standard -pix_fmt yuv422p10le -c:a pcm_s16le -chunk_size 128K output.mov

ffmpeg by itself certainly does not need sqllite.

[Norman Black]

[Stefan Jovanovic] “Could you please explain this part “and I render using x264 via frameserving”?
I’m just trying to figure out what it means.”

I use the “Debugmode Frameserver” utility render driver.
Then one installs Avisynth, or Avisynth+ into the system.
Then I use ffmpeg as the receiver of this whole process. ffmpeg directly support access to Avisynth.

Avisynth is the frameserver. Debugmode framserver is a go between between Vegas and Avisynth and ffmpeg is the encoder which basically receives from Avisynth.

With ffmpeg one can render whatever your heart desires. AVC/H.264 (via x2564), DNxHD, ProRes, XDCAM, Xvid and more. In whatever container file format you want. ffmpeg is a command line tool and not terribly easy to use.

I really only encode AVC for my use. x264 is the best quality and fastest encoder out there for AVC. In its superfast modes it is as fast as GPU encoders at similar quality and in its slower modes it generates better quality. Your choice.

Here is an example of my command script which executes ffmpeg. It is a slower quality encode.


@echo off

title encode crf25

cd d:\renders
d:

REM output the AviSynth commands
echo AviSource("server.avi") > server.avs
echo ConvertToYUY2(matrix="rec709") >> server.avs
REM echo ConvertToYUY2(matrix="PC.709") >> server.avs

c:\systools\ffmpeg32\bin\ffmpeg.exe -threads 0 -i server.avs -c:v libx264 -preset medium -profile:v high -me_method umh -crf 25 -colorspace bt709 -color_primaries bt709 -color_trc bt709 -pix_fmt yuv420p -bufsize 40M -maxrate 40M -c:a libmp3lame -qscale:a 2 -chunk_size 64K output.mp4

pause

del server.avs

REM saved commands for alternate audio output formats and bitrates
REM (mp3 160k vbr) -c:a libmp3lame -qscale:a 4
REM (mp3 192k vbr) -c:a libmp3lame -qscale:a 2
REM (aac 192k vbr) -strict experimental -c:a aac -q:a 2.4
REM (aac 192k) -strict experimental -c:a aac -b:a 192k


[Norman Black]

By 20K, I believe you mean 20 megabits. That is a very high bitrate for HD material. Most every encoder should look the same at that level. This includes MC AVC versus Sony AVC and even those versus x264. Sure PSNR and/or SSIM will show some differences, but the human eye at frame rate, I doubt it.

I once compared MC AVC OpenCL GPU versus my GoPro 1080p30 source file. Both 20Mbit. Pixel peeping a still frame between source and the re-encode was nearly identical. I had to toggle to see any pixel difference.

Back then I had an AMD 5850 GPU. These days I have an AMD 7950 GPU and so MC AVC is no longer an interest to me and I render using x264 via frameserving and encoding is the least of my worries as editing takes all my time.

[Norman Black]

Mainconcept AVC will support AMD 5xxx and 6xxx GPUs for encoding via their OpenCL encoder. 7xxx and newer are not supported by MC AVC but the Vegas video engine itself still works very well with newer GPUs.

GPU encoder algorithms are not typically at the quality level of CPU encoders. It all depends on the bitrate you are encoding to and the source material. Really, unless bitrates are low and you pixel peep you are not likely to notice. Do your own tests.

[Norman Black]

Canon, and most, DSLRs capture video in full range levels. 0-255. Video for playback normally expects video levels, studio, which are 16-235. What you are seeing is your full range levels being encoded and then the playback player thinks they should be studio and it expands the levels to full range and thus your blacks and whites get clipped/crushed.

Vegas does not try to do anything to keep video levels. It leaves that decision and correction entirely up to you, the editor.

Typically if ones source if full range you place a levels, computer to studio RGB, preset adjustment on the output. Also remember that the vegas preview window only properly displays full range levels, in case you grade using that window.

Now I am going to be a little harsh. Does anyone ever search for anything? The number of times this gets answered here and elsewhere is off the scale.

Also, Vegas should probably have an auto levels adjustment of this on import like Premiere supposedly does. Most DSLR output properly marks itself as full range. Vegas should never give up its full user control abilities. Just an option.

[Norman Black]

We really need to know what formats are in the MOV file. The MediaInfo utility previously mentioned will let us know.

If you have Quicktime installed. The bottom line is with these MOV files being given to you is that they contain some video and/or audio format not supported by the default Quicktime installation.

[Norman Black]

[Stephen Mann] “There’s nothing in the GPU API that tells it how to do simple computing, like 1+1. But point it to the data for a series of frames and ask it to compress them into a GOP, then it’s right at home. Theoretically faster than the complex CPU.”

Wow. Educate yourself please.

[Norman Black]

[Stephen Mann] “But if you ask the GPU to add 1+1 it’s hopelessly lost.”

Huh? Math is what GPUs do.

[Stephen Mann] “Also, the GPU has no connection with the system resources such as RAM and hard-disk except what the CPU told it to use.”

Actually anything on the bus has full access to system ram. One can certainly read/write the HD via GPU but this makes no sense. The CPU is best suited to that task. GPUs are actually slow. The only way they beat a CPU is with seriously parallel algorithms. Photo/video image manipulation (effects etc) have many opportunities for seriously parallel algorithms.

[Norman Black]

Vegas uses the GPU for video effects, transitions and compositing/blending of multiple tracks.

All else, including decode, should be done by CPU.

The preview mode Quarter, Half, Full determines at what dimensions the video stream is rendered at. Full being full project size. Half being half size. Half size is 1/4 the pixels to be processed per frame. A big savings in compute.

The preview mode for draft, preview, good, best basically turns off certain features. Many of these features may not be applicable to your source media or project settings.

Here is a link with info on this.
https://www.sonycreativesoftware.com/Forums/ShowMessage.asp?Forum=4&MessageID=95129