HD-SDI capturing

[TheReverendDave]

Originally Posted by petereit

Click here for a search for "dropped frames" on the BlackMagic forum on CreativeCow.net.

Oh man... I think you hit gold!

I had done lots of google searches and had checked CreativeCow hits multiple times. Unbelievable that my desktop screen resolution would cause delay lag and dropped frames! Changed from 1680x1050 to 1280x800 and capture is syncing great in FCP now. This is GREAT as we are capturing straight to ProResLT, and using the line-in input on the Mac. We are doing our first in-service test right now, but it all appears like it's gonna work great. Right now, were capturing directly via SATA with no issue. RAID seems unnecessary for ProRes acquisition.

My current plan is to get a couple of these and record via eSATA to drive, and then move that same HD to our iMac for instant editing, and just put in second HD to record the next service.

Now that we got our system in place, I'll try to take some pics and make a small write-up on how/what we're doing. I know I would love to see other's setups and this can totally help others looking to IMAG and or HD capture acquisition. Thank you all, and especially you Mark, for checking this thread out and helping!

Check out this weekends service at http://www.crossing-church.com

[petereit]

Originally Posted by bladeaudio

Anyone using the new blackmagic box that converts HDSDI directly to H.264?

I would think that for most applications, capturing to H.264 would be an acceptable solution. (That is what most new mid range cameras are recording to, not to mention all the DSLRs)

The biggest problem with H.264 is that it is a delivery codec designed for linear streaming -- not an editing codec. If you're capturing straight to web or BluRay (no editing) then this is a killer tool and will save you a boatload of compressing time.

But if you're doing anything with your video in post, the H.264 codec actually causes problems since the processor has to constantly seek back to the last full reference frame and linearly process all the partial frame updates in between to assemble a complete current frame. It's very processor intensive and ultimately results in at least three generations of compression of your footage before you get a final product.