For me, the interesting things were (in order of appearance):
- While talking about the resolution test comparing enlargements of the HDSLR output and film scanned at 2K and 4K, the participants said a couple of times that on the web we wouldn't be able to see the differences, but at least on my MacBook I could see a very noticeable difference. This web video isn't so sucky!
- 4K was clearly better than 2K, and 2K was clearly better than digital. That said, no one seemed overly concerned
- The Color tests were interesting - and positive - though the comments from the colorists who worked on the tests suggested that the film was easier to work with, and there was more you could do with the film than you could with the digital
- Similarly, the Green Screen test went pretty well for the video cameras, though the Nikon D3s did the worst (and that might be because it is only a 720P camera)
- In the middle there was a bit of a tirade about how important it was not to work in H.264, and that you must transcode to another codec (like ProRes or Cineform) because of the limitations of the H.264 8-bit RGB colorspace. Unfortunately, they didn't show the footage itself that sparked this. Also, does this hold true for Premiere CS5, which makes a big deal of being able to use native content without transcoding? Clearly it's not a concern to Philip Bloom, who was there, and is planning to edit in RAW with Premiere
- The T2i performed well. Unfortunately, this was not news; when they first mentioned that they were going to include the T2i in some tests back in Episode 1, that was exciting news because so little was known about the camera. But a couple of months later there's been lots of tests and reviews (one today even!) and we pretty much know that it's very close to the 7D. Just goes to show how quickly things change, even in this rapidly changing field!
- They did a "Pseudo RAW" comparison with the 5D as a suggestion of what might happen if these cameras output in RAW. But they don't explain what they were doing (comparing a RAW still to a JPEG still at full res?) or what they are ultimately hoping for (RAW at HD or RAW at sensor resolution?) RAW output sounds appealing, and is clearly what the film makers would like (at full sensor res. too please!) but a camera that can do that is going to be much more expensive. Never mind how you're going to capture/deal with all the data
- The episode also features a music video shot in stop-motion on the Canon 7D by Carlos Lascano. It's definitely worth watching. You can read about the making of it here
- Finally, they once more explain why they didn't include a RED or EX3 - or some regular video camera - in the test, saying that there'd be too many cameras, but I found the answer a little disingenuous for two reasons; Firstly, HDSLRs are replacing video cameras as much - if not more than - film cameras (just ask Philip Bloom, who recently sold his EX1) and secondly, they threw in the T2i (making four Canon HDSLR's in the test), but they couldn't put in a regular video camera?
Despite my griping, The Great Camera Shootout is a great piece of work, and well worth watching. Congratulations to Zacuto for putting this together. Now we know how good HDSLRs can be, and can show this to some of the naysayers. I'm sure people will soon be using these cameras for serious production work, maybe even for network television episodes....oh wait, damn! see how quickly things are changing?!
Zacuto: The Great Camera Shootout 2010