Project Properties: Full Range vs. Video Levels?

xstr8guy@sbcglobal.net wrote on 5/11/2009, 7:22 PM
Ok, we've got a new choice in V9 with Project Properties. Now we can use 32-bit floating point (full range) and 32-bit floating point (video levels). Which is best and when should we choose other options? And what's the difference between the two?

The Help topic says this...
"32-bit floating point (video levels) is recommended when working with 10-bit YUV input/output or when using xvYCC/x.v.Color media."

I'm not a total newb but I'm not a techie either and that means absolutely nothing to me. I'm ediing HDV/M2T files. I have no idea if they are 10-bit YUV blah blah blah.

Comments

farss wrote on 5/11/2009, 7:32 PM
"I'm ediing HDV/M2T files. I have no idea if they are 10-bit YUV"

They're not 10bit.

As for the rest of it I'll take the 5th.

Bob.
xstr8guy@sbcglobal.net wrote on 5/13/2009, 5:52 PM
No one else has any thoughts on this?
farss wrote on 5/13/2009, 9:47 PM
Seeing as no one else is putting their hand up and at the risk of being wrong I'll try to help but I'm not certain what you specifically want to know but here goes.

The original 32bit float was confusing, it could cause levels to change and you most likely needed to wrangle that.
The new 32bit float works the same as it does in most other NLE's. It works the same as 8 bit but with greater precision. In other words, switch between 8 bit and 32 bit (video) and you should see almost the same result or most likely exactly the same result.

Any 32bit float processing creates much more demand on the CPU and likely uses more RAM. As your source is 8bit and your output is 8 bit you're unlikely to gain any advantage from switching to a 32bit mode.

Bob.
Brad C. wrote on 7/17/2009, 10:25 AM
Bob,

I always look for your words of wisdom in threads like these.

I too have been working with different in's and out's lately both with and without 32bit float VL. When compared side by side to a similar project at the 8bit level, the differences are fairly minor to the eye although there does appear to be a little more color information at times. Of course I'm probably wrong, but it's just what I see. This is all just viewing it on my PC.

The big difference I've noticed is when uploading content to the web. If I have a project that I've rendered in both 8bit and 32bit float, and subject them both to the web (to Vimeo for instance), the 32bit version looks WAY better. I had never been happy with the way my videos were coming across at Vimeo until I uploaded a 32bit rendered version of a video. Gone was the horrible (to my eye) pixelation in darker areas (i.e. shadows etc.), and artifacting in some motion. Vimeo is one of my favorite places to post, but it's also the worst at dumbing down HD video, IMHO. At least it was.

The big downside, as you suggested, is that 32bit is quite taxing on the computer. I have the processor to back it up (3ghz quad) but my RAM is limited to 3gig because of the OS. I haven't been successful at having 32bit work with full color grading from Bullet without it crashing on me.
ECB wrote on 7/17/2009, 10:50 AM
Here are some basic definitions and tips out of Help:

Pixel format
Choose a setting from the drop-down list to indicate whether you want to perform video processing (compositing, scaling, previewing, rendering, and most video plug-ins) using 8-bit or 32-bit, floating-point arithmetic.

8-bit
Performs video processing using 8-bit arithmetic and in the video (studio RGB, or 16-235) color space.

32-bit floating point (video levels)
Performs video processing using 32-bit arithmetic and in the video color space.

32-bit floating point (full range)
Performs video processing using 32-bit arithmetic and in the full-range color space.


The 32-bit floating point settings allow greater precision for processing video, but require significantly more processing power than working with 8-bit video.

Tips:

32-bit floating point (video levels) is recommended when working with 10-bit YUV input/output or when using xvYCC/x.v.Color media.

I use this:
When using 8-bit input/output, the 32-bit floating point (video levels) setting can prevent banding from compositing that contains fades, feathered edges, or gradients.

Video plug-ins and media generators that do not support floating-point processing are indicated by a blue icon in the Plug-In Manager and Plug-In Chooser with this icon in the Video FX and Media Generators windows.

If you're creating a 32-bit project, you can increase performance during editing and playback by using the 8-bit setting during editing and switching to 32-bit floating point (video levels) before rendering.

Compositing gamma
When you choose 32-bit floating point (full range) from the Pixel format drop-down list, you can choose a compositing gamma value.

1.000 (Linear)
The default setting when you choose 32-bit floating point (full range) from the Pixel format drop-down list.

2.222 (Video)
Processing in 8-bit video is always performed using a setting of 2.222