I was going to do some testing to figure this out but I thought I'd see if anyone had already done so. Also if there is already a thread on this please point me to it (I didn't have any luck searching).

We are currently shooting some dramatic works using cheap consumer mini-dv cameras (we are waiting for the right 720p camera to come along). My biggest complaint about the output from these things (or at least the only one I can't work around) is color noise. So we're thinking about shooting in B&W to eliminate this (seems obvious enough right?).

So I would imagine that this would eliminate the "mosquito" effect, but I'm wondering, if I use the B&W setting on the camera, does this happen before the compression stage and if so, should it also reduce compression-related artifact such as keystone-blocks, etc.?

Has anyone experimented with this or have any technical insight that might help us choose the best option?

If not, and if anyone is interested, I'll proceed with the tests and write up the results.

Thanks in advance;


Jason J. Gullickson
Producer
the second society
http://2soc.net

HA!

I found the thread; thanks guys.

Jason,

Can you post a link to the thread? Not sure what it has to say, but if you want B&W, it makes sense to shoot B&W, since the full 25MBps DV bitstream can be devoted to the B&W image, instead of being split between luma and color channels. More efficient compression, better image all around.

Josh

Wouldn't the B&W mode be just another digital effect that happens after the compression? I can't see any camcorder having a true "B&W" mode, please correct me if I' wrong.

Wouldn't think it'd be applyed after compression - that'd require a full uncompress-process-recompress cycle. It'd probably be after ADC before compression.
But - The DV standard means it'll be recorded to tape at a set resolution, frame rate and colour depth. Whether or not you apply a 'B&W' digital effect, the DVstream will be written as colour, and will be read by your NLE as colour.

Jason,

B&W will not noticeably affect mosquito noise - or any other compression artifact. Chroma noise is eliminated right along with the color, regardless of where it's done. JVC HD-1, HD-10 HDV camcorders use a very unusual color chip (single) to improve mono noise at the expense of chroma noise. This is generally a pretty good compromise as low light scenes generally don't have much chroma anyway.

Shoot what you want for the final product. If a low light scene is noisy, try REDUCING the chroma, you'll reduce noise, and no one will hardly notice!

This was the thread I found yesterday that seemed to answer my question:

http://www.dvinfo.net/conf/showthread.php?t=41738&page=1&pp=15

I wish I had a block diagram of my camera so I knew at what point in the signal flow these things happen, it would make these guesses so much easier!

In any event I suppose the proof is in the pudding, and all technical analysis aside, the only way to know is just to shoot a few test shots and compare them.

Thanks for the info everyone. I'm hoping that HDV cameras eliminate the need for such loop-jumping.