Forums > Jitter

Blackmagic DeckLink HD card + Jitter?

March 14, 2006 | 6:54 pm

Hi,

I know some of you have these in your machines, I was wondering if you
could shed some light on their usage w/r/t Jitter.

Is it easy to bring the card’s foootage into Max? Like as easy as
using qt.grab, or is there somehting else needed?

Is there any latency involved (with the 4:4:4 model)?

Are there any hints/ tips/ things to watch out for when using these cards?

and, from a purely crazy talk perspective : anyone with these cards do
realtime compositing with gl.slab and the hd input?

thanks in advance, and sorry if some of these questions have been
addressed in more detail than my cursory searches revealed.

~scott


March 14, 2006 | 7:23 pm

The decklink cards dont have hardware GL.. so doing openGL output on
them is slow – software mode..

I dont have one, but I picked that up from the list.

v a d e //

http://www.vade.info
abstrakt.vade.info

I LIVE! I LIVE! I LIVE! I LIVE! I LIVE! I LIVE! I LIVE! I LIVE! I
LIVE! I LIVE! I LIVE! I LIVE!

You will not be saved by the Holy Ghost. You will not be saved by the
God Plutonium.

In fact, YOU WILL NOT BE SAVED!


March 14, 2006 | 7:37 pm

well, i’d be using the built-in nvidia card for that render poop. the
deckling would just be for capturing purposes.

the poorly phrased question was about using shaders on hd footage. and
if there was any sort of discernable hit vs standard def.


March 14, 2006 | 7:52 pm

On Mar 14, 2006, at 11:37 AM, Scott Fitzgerald wrote:

> well, i’d be using the built-in nvidia card for that render poop. the
> deckling would just be for capturing purposes.
>
> the poorly phrased question was about using shaders on hd footage. and
> if there was any sort of discernable hit vs standard def.

Provided you have a decent GPU (NVidia 6800, ATI X800 or better),
bandwidth is going to be your biggest bottleneck.

On a Dual 2.5 GHz G5 we’ve been able to decode 2 1280×720 uyvy Phtot
JPEG streams from disk and composite with shaders with room for
additional shader processing maintaining more than 30fps. And a
single 1920×1080 stream with a handful of shader based post
processing @ 30fps. Coming uncompressed might be even faster if the
PCI slots are fast enough.

Let us know your findings.

-Joshua


March 14, 2006 | 9:10 pm

On a somewhat slower G5, I’m doing 2-4 streams (depending on codec)
at 1280×720 (h.264 jpeg or yuv – ordered from most to least costly)
with compositing & effects (approximately 12 instances of
jit.gl.slab) single stream 1080i works too. there is probably room on
the gpu for more. read/decompression seems to be my bottleneck on the
slower cpu, Dual 2.0 G5, ATI x800, 3 disk sata)

-deKam


March 15, 2006 | 3:21 am

thanks everyone.

i’ll let y’all know how it turns out, though it sounds like everything
will work out well.

~s


Viewing 6 posts - 1 through 6 (of 6 total)