Thanks Scott.
24bits RGB equals 4.1.1 in the HLS system.
How many bits per pixel are there in High Definition (4.2.2) images ?

Actually 24 bits and 4:1:1 have nothing to do with each other. 4:4:4, 4:2:2, 4:2:0, 4:1:1 can all be 24 bits or 16 bits or 30 bits or 36 bits. Confusing, ain't it?

4:1:1 is a sampling ratio. Pixels are grouped into sets of 4, and only one chroma sample will be made for that set of 4 pixels. This information can still be stored in 24 bit detail or amounts of bits. 4:2:2 means that two samples are taken for those 4 pixels, and this information can still be stored in 24 bit detail or other amounts of bits. The number of bits merely specifies the accuracy with which the sample is stored.

These video parameters being camera hardware dependent, I suppose there is little one can do to change them afterwards.
DV cameras have CCD capture cells that vary from model and brand.
One tends to overlook the importance of the camera's ccd capture cells and of the electronic engine that controls their subsequent burning on tape... As a Movie Studio user I am concerned by the type of camera that will generate the film at the basis of the process.
I am really interested to understand more of what can really improve the quality of an image, before one has to try and improve it with FX effects in MovieStudio...
What is the initial pixel value parameter that will really make a difference in the end, when the aim is to make 360x240 WMV9 streaming files at 300kbps ?
My camera's manual, -a plain (but recent) Sony camera (DCR-HC18), with a 3mm type 1/6 CCD-, states 'about 800K pixels of which 400K are actually useful'.
Can you help me interpret this information ?
I know I could spend much more money on a camera, but I have not yet been convinced that I could get really better image quality for the specific use I make of it.
Learning from your reply that pixel values are grouped in dots of 4 with the same chroma value... is there a way to calculate the real number of different pixels needed to create a sharp film of a given pixel size in a multimedia environment ?
Thanks.

300Kbps WMV9? Don't worry about it. Your old 8mm Brownie made better images than that.

All NTSC DV is 24 bit 4:1:1. All NTSC DV camcorders are going to record in that format. DV is about 25,000Kbps, so when you encode to 300Kbps you are throwing out about 98.8% of the video signal information. For the remaining 0.2% that you are keeping it isn't worth worrying about any of those details.

320x240 is about 75K pixels. Even the 400K usable pixels are vastly more than you are keeping. Once again, it's really not worth worrying about.

What you need is a camcorder with good optics, good color balance, good light sensitivity. All the electronics behind it aren't going to make much difference at all once you compress to 300Kbps.

Thanks for the information.
But in which circumstances may the 25Mbps DV flow become critical in a multimedia app ? At what point is the full flow needed ?