Good question - and not really OT. Been discussed a lot and I'll only give a broad outline as other know much better. Serena's good on this and hopefully she will reply.
From my point of view I say experiment and see what worked for you. I have the FX1, HDV, and started with the default 100+, which means zebras take you to maximum exposure, only showing fully blown out areas. With this I found I was always over exposing. With my camera, and I think most video cameras, a little on the under exposure side is good as post can bring up shadows, but blown is blown. So I changed to 90 for general well lit shoots and it work good for me. However, you can be a bit more sophisticated as zebras can be used to show more than blow outs. IE skin is best (so the experts say) at 80% luminance, so for face set zebra to 80 and adjust exposure so zebras just appear on faces. It's that sort of a call. With your camera I don't understand "0" - maybe this means OFF. If you can not adjust between 70 and 100 then try 70 for faces. Good luck with replies, but experimentation will teach you more than all the posts in the forum.
Michael
The past few months I have been shooting with mine set to 90. So when I see the zebras, I have a little time to make adjustments instead of being told "I've blown it, fix it quick".
I don't do much "skin" work (unfortunately! - I need to change that, LOL) but I have heard the 80 zebra setting for skin is the best.
Some of this depends upon the lighting. When shooting under stage lighting especially with spots the Zebras are almost useless because if you go by them the video will come out too dark almost every time. With the proper exposure settings under that type of lighting there will be Zebra stripes present even at 100 Ire.
Agree. In film the exposure meter is a guide to the DOP in setting exposure, not an absolute measure. There is no substitute for expertise and experience.
I have that camera. The setting of 70 means that zebra stripes appear at areas with brightness about 70 IRE. The 100 setting shows areas with brightness at or above 100 IRE. This from the manual. They do not say what IRE is, though.
I use the 100 setting always, but rely first on visual judgement of A, the overall brightness in the viewer and B, the perceived dynamics of the shot (meaning: are the brights very bright and the darks very dark). When I'm satisfied I check the zebras, knowing that those areas will be completely blown out. I may still accept that, but only reluctantly. Most of the time I adjust the exposure down.
Which is, however, the biggest shortcoming of that camera. No built in ND filter.
Tor
1- At the low end of the exposure range, you get image noise. This noise obscures shadow detail.
If you shoot underexposed and bring up the brightness in post, you tend to increase noise.
1b- At the highest end of the exposure range, you have clipping. All your detail is lost.
1c- To make clipping less bad, almost all cameras will have knee settings/algorithms/processing. These knee algorithms will 'squeeze' or 'squish' a lot of the highlights down, so that they gradually roll-off before clipping (sort of like audio compression).
A simple knee algorithm would just apply some sort of curve to the R'G'B' channels... you can do this in Vegas via color curves (add a hump-shaped curve to the top end). The problem with this is that it will de-saturate highlights.
*In rare situations like concerts with very saturated lights, these knee circuits will do wacky things and degrade image quality so you might want to turn em off.
**A side result of the knee circuit is that it can make it difficult/impossible to change white balance perfectly in post.
1d- So as you overexpose, you hit the knee circuits with their color shifting / alterations. Past that you hit clipping.
As you underexpose, you get noise.
1e- A lot of cameras will shoot "superwhites", which are values that exceed normal white. If you encode a DVD from there, the DVD player or the TV may clip these values off.
1f- Knowing that, you can experiment with your camera to see what gives you the best exposure. Try shooting a scene at multiple exposure settings to see what works best.
Once you figure out what exposure you like for scenes, try to figure out what the zebras look like when you hit that "sweet spot". Pay attention to what your zebras are set at, and what the image looks like (i.e. on your viewfinder).
2a- DV is stored as digital Y'CbCr values. The Y' component approximates brightness and determines your levels.
For 8-bit formats, 16 Y' is black level and 235 Y' is white level.
Superwhites are Y' values from 236 to 254.
When you convert that to analog, 235 Y' (digital) corresponds to 100 IRE (analog). You actually have some values above 235 Y' / 100 IRE.
"70" in the zebras should roughly correspond to 70% exposure.
So basically, if your zebras are at 100, anything that is zebra'ing is either superwhite and/or clipped (clipped = not even my curves in Vegas will bring em back).
2b- Some people use the 70 zebras to set skin tone exposure. To them, if good area of the skintones are hitting 70 zebras then the exposure is in its sweet spot.
Notice how in Glen's curves, his line is not from corner to corner but he restricts the dynamic range by dropping max white and raising max black. This is what I advocated (either here or in the "Profiles" thread) you do in the FX1/Z1 camera with the 90 zebra setting and Gamma ON. I never have to do this in post as my histogram for my "film" look straddles 40-235. At one time, "when I wus young" about 6 months ago, I used to try to maximise dynamic range by stretching the histogram to go from 0 to 256. I now know this is wrong and is one thing that destroys the nice smooth graduations of the film look.
I don't believe the curves raise max black... they just cut stuff off at 16. If for some reason you have a raised black level... i.e. you have glare from shooting into bright light sources (e.g. the sun), then you can select both points near 16. Use the arrow keys to adjust black level.