How do you know you're on a Stu shoot?
Wednesday, March 15, 2006 at 11:17AM If you see something like this, it's a good sign:
The Orphanage wrapped a shoot last week for a new spot, and I got to play with this helicopter mock-up suspended on this giant forklift. Once again DP Ketil Dietrichson shot for me with the Viper, this time using DigiPrime lenses. The spot will be finished HD, which is becoming quite common for me these days.
After this discussion on the CML, I had been wanting to try the Viper with a corrective magenta filter. I don't have any issue with the green cast of the raw Viper filmstream output, but correcting for it on the lens should result in a broader dynamic range more evenly shared across R, G and B. So far I'm sold on this approach — the images we made look great, and the dynamic range will be put to the test with some sun glints off that helicopter mock-up. I'll post some images when the spot is done.
Here's an article that goes into more detail on the rationalle behind the magenta filtration.
Here's another with sample images, although I quibble with the author's characterization of the green cast as a "problem." One of the great strengths of raw images is their lack of built-in white balance, or any other subjective processing.
Reader Comments (11)
We're using a Viper camera as well, and I was wondering if there is a certain colour-correction setup in After Effects that would automatically remove the green tint.
Right now we're doing it manually every time using grey cards and so on, but a factory-approved calibrated version wouldn't be too shabby, I suppose.
I made my viper-to-linear correction in Fusion, since it's just not possible to implement equasion-level operations in After Effects (which is a bummer). It's something I'd love to make public at some point if I get them time, since I very much want to support Thomson in their openness about the camera's storage LUT.
Ah right, are these specifications available openly online, or would I need to contact Thomson about that? I've seen the Filmstream --> 12 bit Excel-based LUT, but this obviously won't remove the green cast, since it basically just converts the 10-bit data to the original 12-bit data from the AD converters.
I got the math from some very helpful folks at Thomson.
if there's a helicopter stu isn't far behind...
Hi Stu,
I've just finished a greenscreen shoot using the Viper and have been looking online for solutions to colour correcting the footage in Fusion. Do you still plan to make your viper-to-linear solution public? My current solution is to colour correct the footage towards magenta, but if there's a mathematically correction solution, I'd be really grateful if you'd be willing to share it.
Thanks,
Jon
I got the math from some very helpful folks at Thomson.
Is it really such a big deal, the math for removing the green-cast, that they're so protective of it? Thomson is publishing their log-lin and lin-log math (that is Stu-linear-light right?). So why not share the rest of it? You really can't do without it to get a proper white balanced picture.
Aren't you allowed to share it? If not, who at Thomson do I have to talk to? ;)
Yeah, I second that.
Why is it such a secret how to get rid of the green cast? It seems that most of the people working with Viper footage are doing a whole lot of guess work, and therefore the image must suffer from it if it isn't mathematically correct.
Yeah, why is it such a secret how to remove the green cast?
It seems to me that many people working with Viper footage are doing a whole lot of guesswork in order to remove that green cast and thus resulting in less than ideal image.
Why can't the math behind it be published so people can work on images that are mathematically correct?
And Stu,
Why aren't you answering these comments?
When do you estimate the time is right to make your conversion public? Do you need clearance from Thomson?
Hi Marijn,
Sorry to not be replying. I have a day job that keeps me rather busy.
The Vipr inverse LUT is public, you'll find it and many other lovely things here:
http://www.cinematography.net/CINEFILE.HTM
The maths behind the LUT are not secret, Thomson gave them to me freely. All I had to do was ask.
The reason I haven't posted any Fusion files is that I don't use Fusion much anymore, so I don't have them handy.
So my original comment still stands: "It's something I'd love to make public at some point if I get them time, since I very much want to support Thomson in their openness about the camera's storage LUT."