8 bit versus 10 bit

I have a G5 3Gig dual core - Intel Mac...with 4 internal 500gig SATA drives. OSX 10.4.9, FCP 5.1.4 amd Quicktime 7.1.5

When I try capturing video in 10 bit uncompressed onto my empty new SATA drives, the video stutters and freezes during capture and playback.

Switched to 8 bit uncompressed and everything is fine.

Am I expecting too much? Should I be able to capture in 10 bit?

And just for my info...how would you describe the visual difference between 8 bit and 10 bit.
Can you see the difference?

Thanks,

Ron

SD or HD?

Are the four drives tied together as a RAID? If not, then you wont get the drive speed you'll need. RAID them together and you'll have plenty of drive speed for SD.

If you're using a lot of motion graphics, particularly with gradients, then 10-bit is good. If you're just cutting footage that was already shot in 8-bit, then there's little to no gain.

Marco Solorio | OneRiver Media | ORM Blog | Twitter | Facebook | Media Batch

A tangential question:

In the past, using dv (8-bit, SD) I've had trouble with pixellation in dark areas when going to dvd. Will going to 10-bit alleviate this problem? I recall trying it once and the output was identical, but didn't look into it after that.

Just wondering if there's some way to stretch out the dark/shadow gamut in 8-bit-originated work.

thanks,

hd

HarryD

In 8 bit footage, red, green and blue all have 256 shades.
In 10 bit these cololors have 1024 shades.

When doing post work like green screening, grading etc 10-bits is far better than 8-bits.
If you are doing no post work, 10 bits is still better but not as apparent.

There is no point really capturing 8 bit footage as 10 boits though as this extra color information will not magically appear out of an 8-bit file.

Johan Polhem
Motion Graphics
www.johanpolhem.com

Marco...

Thanks for your response. I am in fact just using SD. Footage was shot HiRez but downconverted to standard DigiBeta for me to cut with. So is that 8 bit or 10 bit material?

Ron

Right, I know all that. What i'm wondering is - if the file becomes 10 bit after originating as 8-bit, will the 10-bit processing in post smooth things out more than keeping it 8 bit all the way through? Will it add gradations (from 256 to 1024 levels) making shadow-range pixellation less likely when going to MPEG DVD?

HarryD

DigiBeta is a 10bit format. Might be worth trying 10bit capture but all depends on the bit-depth and precision of the downconverter you used.

10bit is good for when you're doing effects and the like, but actually it would be more important to keep the 4:2:2 sampling of the digibeta and use a lightly or not at all compressed codec like prores before going to DVD to keep the noise in the image down.

"Right, I know all that. What i'm wondering is - if the file becomes 10 bit after originating as 8-bit, will the 10-bit processing in post smooth things out more than keeping it 8 bit all the way through? Will it add gradations (from 256 to 1024 levels) making shadow-range pixellation less likely when going to MPEG DVD?"

No it will not unfortunately.
If you add new effects on top of the footage or grade the footage however, you may want to set your project setings (After Effects, Shake, Color) to floating bit point to smoothen out any post work.

Johan Polhem
Motion Graphics
www.johanpolhem.com

Interesting, thanks.

Another thought:

I wonder if ingesting 8bit 4:2:2 footage into a 10-bit sequence or otherwise converting it to 10 bit, then playing with it in COLOR will improve things in the shadow area? I don't yet have FCS2 so i can't test, but I wonder if this will give the desired results?

thanks,

HarryD