I don't think Levels makes sense unless you are working with your video scopes.

Open your waveform to see how the Input and Output settings affect your video. You want to use levels to pull the video within legal limits.

I think the basic way to work with Levels is to use the input/output sliders to bring your black and white points just so the blackest black is at the bottom edge of your waveform and the whitest white is at the top edge. Then you adjust your gamma to your taste.

In Photoshop, the UI for the levels tool is a histogram with a black point slider, a white point slider and a midtones slider. The way you work with it in photoshop is drag the black point to just where the histogram graph begins to go up. You do the same for white. Then you adjust overall color with the midtones.

Here's a tutorial on how to do in Photoshop, it may help you understand what is going in Vegas. The Photoshop UI, is much better than Vegas, because you see the change in the scope right in the tool.
http://www.cambridgeincolour.com/tutorials/levels.htm

Very good explanation for a very confusing UI [Vegas, not PS]. What has always bothered me is that to EXPAND the dynamic range of the video you COMPRESS the input START/END settings, this goes against the grain of everything I've ever learned and I can't seem to reconcile it in my mind. Imagine having the functions of the accelerator and brake pedals in your car work exactly the opposite they are meant to. Yes, I always use the waveform scope when making these adjustments. And after years with Vegas, I still haven't fathomed the utility of the output sliders

What has always bothered me is that to EXPAND the dynamic range of the video you COMPRESS the input START/END settings, this goes against the grain of everything I've ever learned and I can't seem to reconcile it in my mind.

Think about zooming in on a photo in order to be able see more detail. I think it works like that.

Sorry but that makes sense to me.
It didn't at first and now I cannot remember why it didn't.

Now if only I can work out how to use the Convolution Kernel.

Bob.

"Think about zooming in on a photo in order to be able see more detail. I think it works like that."

Thanks, that helps. Guess it's like labeling an audio level control as an "attenuator," the larger the value the lower the level. Guess this old dog is just having a hard time learning new tricks.

Now, do you have any light to shed on the "output" controls? Just when would I use them? and am I right in thinking that, according to my interpretation of the help file info, that it has no effect on the rendered video?

You use output or input depending on what you are trying to do. For example sometimes your waveform values would outside your legal limits. In this case, you would use your INPUT sliders. (I think it's input, I'm actually not at my desk, I'm watching the Spurs/Suns game). You would narrow your input values to bring the video back to 0 and back to 100.

Other times your video has a narrow dynamic range, perhaps it goes from only 20-80 on your waveform. In this case you narrow the output range to move the output to 0-100.

When I use levels I just watch the waveform and see if it's moving the way I want. I don't really think about which slider I'm using.

If you have video that's StudioRGB i.e 16 to 235 and you want to convert it to ComputerRGB i.e. 0 to 255 then Levels is your friend.

Start/End Input = 16,235
Star/End Output = 0,255

Input luma 16 becomes 0 at output
Input luma 235 becomes 255 at output
Everything in between is scaled accordingly.
Anything in the video outside the Start/End input range is clipped.
So if by chance the video contained superwhites they still end up at 255, the same as legal white at 235.

Bob.

Convolution Kernel:
Suppose in the middle row you put in 0.25, 0.5, 0.25
This will blur the image (horizontally).

You can think of it as this: You take the center pixel of the original. You take 0.25 of that, and put that to the pixel to the left. 0.5 of the original pixel stays there. And then 0.25 of the original pixel will go to the pixel to the right.

2- If you have the "normalize" box checked, then the FX will make sure that everything adds up to 1.

Suppose in the middle row you put 1, 2, 1. Without the box checked, the left pixel will get 1X the original pixel and the center pixel will get 2X of the original pixel. Everything will be a lot brighter.
If the normalize box is checked, nothing gets overdistributed.

3- If you have -0.25, 1.5, and -0.25 then you'll get a sharpen filter.

4- Sometimes, the convolution kernel can approximate the erode tools in programs like Combustion. You can use it for cleaning up mattes.

Here's a tutorial on using Levels in Vegas:
http://www.vasst.com/resource.aspx?id=a7a8c403-64dc-420d-97d0-90d2f8de9fc1

Input Start and Input End are analogous to the top part of Photoshop's levels filter, while output start and output end are analogous to the bottom part of Photoshop's levels filter.

Thanks,
obviously one needs to spend a bit of time with it but that's given me a starting point.

Bob.