I was expecting a difference in quality when I went from 8-bit to 32-bit (video levels)

There will very likely be NO difference, for a number of reasons. First, is your source footage 8 bit? (likely). Second, unless you really "push" your colours / luminance around, having 32 bit (floating point) doesn't save anything from clipping. Third, 8 bit 4-2-0 (subsampled) video is very good at perceived image quality compared to higher bit and better subsampling schemes (eg 10 bit 422 from gh5 camera) so that OFTEN (not always) there is no perceptible difference - although there is a technical improvement.

Hi Kinvermark,

Thank you for your reply and the explanation. And furthermore, you are right, I just figured out my source footage is 8-bit too.

By selecting the 32-bit (Full Range) I notice the colors becomes more warm. I kind of like this - is there an effect I can apply to my videos when working in 8-Bit where I can achieve this?

You'll get a lot more control when color correcting/grading 10 bit (or higher) footage when in 32 bit mode, other than that, the difference won't always be noticeable, especially if you aren't outputting to a 10 bit capable fomat that will be viewed on a 10 bit panel. Currently Vegas doesn't offer any sort of HDR output, though you can output HEVC 10 bit (if you can find anything to play it back smoothly), that is currently the only consumer delivery format that supports above 8 bit in vegas, though. I'd love to see full on HDR support added, now that I have a capable TV set and cameras. It makes a nice selling point to be able to tell clients their video is as future proof as possible.

The difference is not in the display but in the fact that when you apply color grading as well as color effects several times in a row (as is usually the case), every one of the effects will reduce the total number of colors in the image/frame when done in the 8-bit mode. That is because 8-bit colors are integers and any change has to be rounded to the nearest integer. And so with each effect added, the number of colors decreases and the quality of the image potentially deteriorates.

But when you do that in the 32-bit mode, you preserve the number of colors as moving from one effect to the other because no rounding to the nearest integer happens in-between. Only after all effects are applied will the values of the pixels be rounded to the nearest integer, so the quality of the image does not deteriorate during the process. That is because 32-bit values are not integers but floating point numbers.

To offer an analogy, suppose you need to add 1/3 and 8/7. If you converted those values to integers before adding them, you would round 1/3 = 0.3333... to 0 and 9/7 = 1.2857... to 1, and add them as 0+1 = 1. If you did not convert them to integers but use floating point numbers, you would get 1/3 = 0.3333... and 9/7 = 1.2857..., and 0.3333+1.2857 = 1.6190 and only then would you round it to 2. And that is just two numbers. Now if you did a number of color effects, the error would potentially increase with each one of them in the 8-bit (integer) mode, but not with the 32-bit (non-integer) mode even though you would eventually round the result to an integer.