Thomson Grass Valley is planning to release its new entry into the digital film camera race, called Viper FilmStream, by the end of this month (August, 2002). Digital Media Net's Charlie White talked with the man in charge of marketing the Viper project, Mark Chiolis, Marketing Manager, Acquisition and Production Markets, at Thompson Grass Valley. Chiolis has worked with companies like BTS, Philips and Thompson since 1992, and prior to that he was a production manager at a television station in northern California. In this extensive interview, Chiolis gives us a detailed look at the camera that many think represents the next step in the evolution of a viable digital cinema capture system.

DMN: Tell us about the Viper FilmStream camera and what makes it unique.

Chiolis: The Viper actually was born from people sitting around a table saying, "You know, we have all this extra capture power in the camera, and we're losing it just putting out the 4:2:2 bitstream out the back of the camera. Is there a way we can utilize this and make something out of it? The camera engineers went back to the factory, they wrote papers and drew on white boards, went out to the pub and had six or seven beers, and they came up with the fact that the amount of raw data coming out of the CCDs is something around 2.5Gbit per second. And what has been going on the recorders in the high-def format is somewhere around the 160 to 240 mark (megabits). So you're losing a lot of data between tape and what is captured on the sensors. One of the things we decided to do was to take a look at another recording format. The only thing that can capture that much raw data would be disk. So now that we have that much raw data coming out, what do we do with it?

So now you have a whole solution that says, we're going to take the raw data coming off the sensors from the Viper, and they're the same sensors that we use on our high-def camera, the LDK 6000, and in the standard-def oversampled camera, the LDK 5000, so we like to say there's a little bit of Viper in all our cameras. What we're doing is taking the 12-bit linear signals right out of the CCDs and each of the CCDs is nine million pixels.

Most of our competitors use IT and FIT, and some of our cameras, we use those as well. It's just a different way of capturing. But we're the only ones to use Frame Transfer. We use those in the higher-end cameras. What it does is, instead of having to clock up the signal as part of the imaging device, on the IT and the FIT, in between the imaging pixels they're called transfer channels. That’s how the image is clocked off, so you're not utilizing the entire area of the CCD for imaging -- part of it is used for transfer gates. And what that does is that gives you a better chance of getting aliasing between the imaging pixels, so you'll see that coming out of the final signal.

Anyway, what we do is capture almost the entire image area of the CCD, and then we clock that off en mass as a frame. The reason we can do that is we have a mechanical shutter just like a film camera. That allows us to do a couple of things. One, when the shutter blade comes around, it completely blocks the CCD imager, so we can transfer the entire image off to the storage area. We can also purge the entire CCD, so we get no vertical smear from one frame to the next. Sometimes you'll see on very highlight-specular points, vertical smearing. And because we're actually purging the CCD frames, you're not going to get that with Viper.

1 2 3 4 5 6 7 8 Next