These are form my two most recent projects. I was DP on one (the one in the house) and Director on the the other (the one in the gray set). These are straight from the camera, no grading. They haven't been touched other than being converted to RGB PNG files.
I thought people might like to see what Uncompressed looks like.
Also, here is a clip with a chip chart so you can see some stuff in motion. (24p from UNCOMPRESSED in Lagarith to keep file size down) Most peoples machines wont be fast enough to actually play it at full speed but it will give you enough frames to play around with.
LT_ChipChart.avi
Here are some frame grabs in PNG format. (RGB but otherwise totally un-processed) The stuff in the gray room is all Cinemode. The stuff in the house is "TV" mode. We kept everything at 1/48th sec exposure and usually aimed to keep the aperture as wide open as possible. (between f2.4-1.8)
LT_And_I_Will_Call_You_Mini-Me.png
LT_button.png
LT_Calibration_Compete.png
LT_Eris_CU3.png
LT_Eris_CU5.png
LT_Eris_John_TS.png
LT_Eris_Mike_Full.png
LT_Eris_Mike_OTS.png
LT_Future_Stack.png
LT_Im_Down_Here.png
LT_Jazz_hands.png
LT_Obligations.png
LT_Refcube1.png
LT_Sit_Down.png
LT_Understand.png
LT_You_Gotta_Be_Kidding_Me.png
My work flow is:
1. Capture live "24p" UNCOMPRESSED with the Intensity over HDMI
2. Remove 3:2 pulldown with custom scripts staying in 4:2:2 YCC color space though out (no color space round-tripping), re-compressing to Lagarith or HUFFYUV. (I generate SD res proxies for off line editing at the same time, in HUFFYUV)
Then after that edit, color grade, etc.
The biggest down side with the Intensity is you need to operate tethered of course. But as it turns out that doesn't seem to be much of a problem when you are working on a set. You need SOME sort of "Video Assist" station anyway, and the capture station can really do double duty in that respect.
Considering I still use the "free-ware" method of removing 3:2 pulldown, shooting UNCOMPRESSED isn't much of a hassle since the system of 3:2 pulldown removal I use is similar and the resulting file sizes are the same. But I have the benefit of starting with 4:2:2 color and a better quality source.
We typically generate between 500 and 700 gigs of uncompressed video data a day shooting live uncompressed. Once I process it, removing the 3:2 pulldown and compressing it lossless with Lagarith it ends up being around 200-300 gigs of data.
Considering the cost of storage these days that seems quite a bargain. (Especially when you compare the price of hard drives to film or even high end tape like HDCAM-SR) Even HDCAM SR is not uncompressed and it's about 112 bucks for 64 minutes. Film? It depends on if you use new stock or re-cans. But just developing it would cost $350+ and that is without any kind of transfer to video. 200-300 gigs of data takes up maybe $50 bucks of storage. even with a back up for redundancy, you are still looking at a cost comparable or cheaper than "SR" tape (only UNCOMPRESSED!). It's definitely cheaper than film, even 16mm.
I'm not trying to say the HV20 makes pictures anywhere near what the equipment that would typically be used with SR tape came make. Or that it can honestly compare to film. Only that it's a very economical work flow. (the quality not being half bad either!)
But... If you want to see some film scans at 2k you can down load some here (look for "Output Samples (downloadable files)" About half way down the page:
http://www.lasergraphics.com/us/pages/director.htm
I'd say the stuff out of the HV20 seems "sharper" if not as clean in the final analysis since the S/N seems worse on the HV20. The "noise" or grain on the film scans is definitely more uniform in each color channel. But over all I would say the HV20 material hold up pretty well, wouldn't you?
I'm going to NAB next week and I hope that Cineform has more news regarding their HDMI DVR. That may free me from the tether.


Reply With Quote

