HD and Arri Varicon
Posted 13 June 2005 - 07:48 AM
I am about to go into pre prod for a feature in South Africa where most of the testing will take place and where unfortunately there aren?t any Varicons that I know of. Before I stick my neck out and get a Varicon sent out does anybody know of any theoretical reasons why using a Varicon with an HD camera is a bad idea?
I await your collective wisdom with baited breath,
Kindest regards to all the faithful out there,
Joe de Kadt.
Posted 13 June 2005 - 10:44 AM
However, when I was at NAB, Jeff Krienes showed me a new feature he added to the Kinetta where some internal LED's could be dimmed up to basically flash the CCD, similar to what used to be called a "bias light" in the old days of tube cameras. It's a form of latensification; basically the low level of exposure causes light was was too dim to activate the pixel to accumulate enough charge, increasing the speed in the darker areas of the frame. A Varicon would do something similar, improving shadow detail in night scenes. However, you do bring out the inherent noise level and get weaker blacks so it's a trade off.
However, most of the contrast problems with video are in the highlights, not the shadows.
Posted 13 June 2005 - 11:55 AM
Anyway, back to the topic at hand. It?s an idea that has been fizzing around my head for quite a while now. I was in early discussion about this coming project and a couple of films shot by Darius Khondji were mentioned as initial visual references ? the producers were getting at the dark and lurksom characteristics of these films I think rather than specifically the desaturated high contrast nature of ENR and bleach by-pass. Never-the-less doing what Khondji does seemed like a good place to start.
Fiddling with the black gamma and master black was my first approach with the intention of taking the darkest areas back down to black during post. Intuitively though, fiddling with gamma curves in camera just to push it back the other way later seems a bit disingenuous and would really just increase noise in the way you mentioned. |It did work though I found that the blacks whilst less contrasty were becoming milky rather than less saturated. I then tried doing the exact same thing but with a varicon and whilst the results were similar there was a qualitative difference in the nature of the shadows. They seemed more tangible and honest if you?ll allow me to use such nebulous words.
As far as I can tell the theory holds true, pre-sensitizing the chip to it?s equivalent of the d-min. What I found I was getting was essentially an extra stop in the highlights as I was able to stop down as a result of the extra information in the shadows. I haven?t had a chance to view the results carefully enough to see if I am increasing noise in the shadows when I compensate for the low contrast in post (ie pushing the big button on the da vinci board that says ?make this look like ENR), a problem which I?m particularly worried about due to the low bit depth of the compressed data form the tape stock. More thorough testing would be the answer but I?m a bit worried about the extra expense of shipping out a Varicon just for a looks see as we don?t have all that much money to play with.
Still, mustn?t grumble.
Posted 27 June 2005 - 04:46 PM
what i don't get is how a very low light (flasing the shadows) in the d-min can add 1 stop in the high lights(d- max) at the opposit of the curve
what does the curve looks like at the vectorscope?
gama curves and video signals have such a different look that it's amazing to imagin they react just the same?
keep to inform us with your tests
Posted 02 August 2005 - 03:38 AM
The idea about the extra stop was that if I could get extra infromation in the shadows I could stop down and get more info in the highlights, as such increasing the lattitude if the chip. Good theory but can't seem to get it to work yet. Bah.
Thanx for the interest thought.