I’ve got some video together from our little camera test on Friday. The “shoot-out” element of the test is more or less lost unfortunately. I can only get the first 30 seconds of any given XDCamHD clip (trial limitation of Flip4mac MXF), which means most of the footage is of me focusing the shot. The footage off the XL-H1 is largely over or underexposed, as the viewfinder was, as previously mentioned, “gone wonky.”
In any case, I’ve pulled some of the better footage, and have my thoughts, after the jump.
So first, the footage. Each scene contains shots from each camera (with a camera label in the lower left). The links are for 3mbit H.264 versions (at 1920×1080) and also Uncompressed 8bit 4:2:2 Quicktimes. You’ll probably need Final Cut Pro installed to view those.
The clips are very short, but I figure most people go frame by frame for this stuff anyways. Plus, my iMac really doesn’t like rendering uncompressed HD.
So, what can we learn from this footage? First, we can learn that one should plan an event like this a bit more thoroughly – if I’d known about Final Cut’s issue with VBR XDCamHD footage for example, we could have shot all CBR. We also could have gotten ahold of some P2 cards for the HVX-200.
In terms of the footage though, we can notice a number of things. The cameras were more or less in their default settings. Immediately, the HVX-200 footage looks the most “pleasant” – with its supersaturated colors, everything seems obscenely pretty. This has been a hallmark of the *VX cameras, and it certainly continues here.
When you move on to the XDCam and the XL-H1, you notice that 1) they’re much duller than the HVX and 2) they’re much sharper than the HVX. There’s a crispness to them that’s very obviously lacking on the HVX. The F350 (XDCam) and the XL-H1 both have 1440×1080 CCDs, whereas the HVX is significantly less than that. Whether that can totally account for the visual difference or not, I’m not sure.
The XL-H1 certainly looks the most like video. In fact, I think there is too much digital sharpening going on. The XDCam seems to split the difference nicely. I’m still concerned about the fringing, but I continue to chalk that up to the lens.
Having said that both the XDcam and the XL-H1 looked dull compared to the HVX, I must add that a very minor tweak in the 3-way color corrector was all that was needed to bring them into line with the Panasonic. Certainly it’d be preferable to have the footage come out of the camera that way though, as it’s a nice starting point for more intense CC work.
Is it fair to cross compare these cameras? I think it is. Ideally, we would have also had an HDX-900 present, but I couldn’t con one out of anyone.
I think that folks looking at the F350 and F330 XDCam units should take a long look at the Canon XL-H1. While you give up the ability to overcrank and undercrank, and the nice workflow of the XDCam format, you save $20,000. Not a small chunk of change.
That said, I think the XDCam format is amazing. I was skeptical in the past, because it seemed like a stopgap between tape and solid state. However, having experienced the pain that is the Firestore, and the expense that is P2, I think that it may have been a wise choice. The footage off the F350 was very, very good, and it was certainly my preferred camera to shoot with. However, if you’re just looking for an HD camera to shoot very crisp, normal speed video, you may find that the XL-H1 meets all your needs.
Finally, just for fun, here’s a bit of overcrank footage for you. We had to run the HVX-200 in 720p mode, as we were capturing directly into Final Cut Pro and it can’t handle the overcrank in 1080p at this point. The link itself is only 640p, because I don’t have a 720p preset in Media Mill.
The XDCam is running at 60fps, recording to disc at 24p. Note that when you’re doing the 60->24 overcrank on the XDCam, you actually lose half your vertical resolution. So, even though this is a 1080 video, you’re really only getting 540 lines of detail.
Thanks again to Adam for use of his HVX-200, Steve at Z-Systems for the XDCam, and Gary and Mark for helping out.
I’ll post a final update when the 16mm footage returns from the lab.