[Rick Anvican]

Hi Steve,
did you say something with regards to this in a previous thread about that Vegas has reached its limit? I was wondering if there were speed differences between Vegas 11 and 12 as well.

RickAVC

[Rick Anvican]

I’ve had a few BSODs from Nvidia drivers (WHQL assured) in the past, so I’ve stuck with working ones and wait for big updates, there was a GPU bake method that could resurrect broken that have symptoms like yours with the broken BIOS screen, but the Quadro’s an expensive one, probably best not to even if it’s not under warranty,
anyway all the best saving for the Mac Pro!

[Rick Anvican]

Thanks for your help.

I’ve read from one of your posts saying that your Quadro card is getting replaced, what problem caused it to fail if you don’t mind sharing? Could invest in a workstation card in future…

Media players like MPC-HC and VLC tend to convert sRGB to Computer RGB colors and it seems that only importing video files into Vegas would show their true colors, according to the scopes there are times when there are out of range colors and luminance for brief periods…

[Rick Anvican]

Actually I was amused when I saw the Pixel Format setting for my first project in Vegas and was tempted to change it, but I left it at the default 8bit because logic tells me that I’m working with over 10 million colors and not 256 colors so that should be plenty.

I rendered that particular part of the project with artifacts at 8bit instead because there wasn’t any visible banding on those two clips anyway and the color grading was done, after that smart-rendered the pre-rendered segments, so all was fine.
Did notice slight color changes (shift in hue) to the original grading when switching pixel format so probably it’s best to do some testing with source footage to check signs of banding due to masking, effects, etc., and then put the project together only after deciding on pixel format.

The 8bit pixel format had me wondering what causes the visible banding, whether is it because of a slight reduction of colors in the sRGB palette which I doubt or is it that the rounding errors are quite significant, I thought that Vegas had a 10bit mode at first…

[Rick Anvican]

Hi John,
Colors in the project were graded with the sRGB range in consideration, I’ll try just putting Levels on those events with out of range colors, otherwise regrade them to legal colors.

Pixel Format set at 32bit(Video levels) does indeed reduce banding with Levels on, as well as on some events with slight posterization (artefact in the source), guess it’s worth a try, but I’ll just check that the plugins I’m using won’t cause problems during render.

I’ve always used 8bit in Pixel Format for my projects and only this one has stumbled me with the visible banding due to effects and feathered masking, perhaps it’s just the nature of the source footage I’m using.

Thanks.

EDIT: Just finished rendering with Pixel Format set at 32bit(video), it surprisingly took roughly the same amount of time (about 40 mins with Cedocida DVCPRO50) when Pixel Format was set at 8bit, I thought it would have taken 3 hours to complete.
There was one glitch where two events composited with “Screen” mode had patches of bright yellow and red in the render (events were in its original form with no effects/transitions applied), Levels(sRGB) was applied and colors after that according to the video vectorscope were in the legal range and luminance was bordering on 90 IRE(sRGB) but the artifacts remained. This only happens on 32bit(video) and not 8bit.

[Rick Anvican]

I’ve heard somewhere, probably on the forums that Reduce Interlace Flicker doesn’t always work as expected, the Guassian blur tricked worked well anyway, thanks again John.

[Rick Anvican]

Thanks John, this blur trick kind of reminds me about Reduce Interlace Flicker in Vegas…

[Rick Anvican]

Thanks for the tips guys, I’ve heard about the Gaussian blur trick some time ago, was thinking of applying linear blur (horizontally).

[Stephen Mann] “Also, gently inform the client that red on black is probably the worst color combination for text on video.”
Stephen, I have informed the client about the color combination, but I’m just doing what was requested. We’ve negotiated and opted for red but with slightly less saturation than before which reduced the color bleeding a bit, and as everybody knows, DV does not like overly saturated reds.

The final delivery format was to be DV(NTSC) but I’ve tried rendering to DV(PAL) first then re-rendering the DV(PAL) version to DV(NTSC), reds were retained in the calligraphy in exchange of more color bleeding on the black BG, not a good idea but the calligraphy did visually appear in red more often than if rendering straight to DV(NTSC) where at least half of it was gray…

[Rick Anvican]

Thanks for clarifying Stephen, I thought it meant 8 bits per primary color for the pixel, I’ve read that common projects should stay at the default 8-bit pixel format, in terms of reducing the banding, would you guys recommend me switching to 32-bit instead or stay on 8-bit and reduce the banding by another means?

[Rick Anvican]

Well Stephen M., I did mention the dominant color on the shots of the source video was dark blue, perhaps this could link to how the human eye has lower sensitivity to blue than green or red, I had other shots where yellow was the dominant color but banding wasn’t present, “bit-budgets” and such I guess.

Thanks Stephen C., good reads.