[Tig] 8bit vs 10bit for TV

Carl Skaff carl at stopp.se
Fri May 4 07:41:55 BST 2012


Hi Cem

I think actually that a few years ago the Flint/Flame/Fire/Inferno/Smoke
was either 8 or 16bit (on Linux) when it comes to setting your project.

That's a why they always used to ask for 8bit (or convert 10>8bit) so they
didn't have to work in 16biy since that would be to heavy.

But now since a few years (maybe 2010) they got a third option, 10bit. But
I suppose if they got used to 8bit they want to stay in that land.

Or am I wrong?

/Carl




(Sent from mobile device)
____________________________
*Carl Skaff*
*Colorist*
Stopp Stockholm
Office +46 8 50 70 35 00

*Stockholm | Los Angeles | Linz*
www.stopp.se

4 maj 2012 kl. 00:06 skrev Cem Ozkilicci <cemoz101 at yahoo.com>:

Hi Carl,

I don't think I have encountered an online artist asking for material in
lesser bit depth. Seems very odd, considering that even Digibetas are 10
bit. (Yes, 4:4:4 RGB 8 bit vs. 4:2:2 10 bit YUV has been an ongoing debate
for years...! Let's not go there...)

Furthermore, I suspect that working from 8bit files, banding could easily
be an issue, especially if you have gradients, vignettes or have used
Gaussian blurs on certain parts of the picture.

I know that the earlier Smoke's on Mac were restricted to 8 bit so it would
make sense if you are on an older Smoke, but the Linux based Autodesk
machines still function internally and output at higher bit depths.

However, I too would be curious to hear if any other colourists have
encountered such requests...

Cheers!

Cem

Colourist
Storyline Studios / Oslo






  ------------------------------
*From:* Carl Skaff <carl at stopp.se>
*To:* "tig at colorist.org" <tig at colorist.org>
*Sent:* Thursday, May 3, 2012 11:20:24 PM
*Subject:* [Tig] 8bit vs 10bit for TV

Sohonet www.sohonet.co.uk sponsors the TIG.
Support from Digital Vision www.digitalvision.tv
Support from Blackmagic Design www.blackmagic-design.com
Support from Dolby Laboratories www.dolby.com
====

In grading we strive to get realtime performance. A few years ago HD was
possible in a non-linear suite. But we preferred SD if we needed to to a
lot of blurred keys.

A couple of years ago HD got as fast as SD used to be. And we sent had to
bother if the client wanted it delivered in HD or SD.
Now I don't care anymore. I do everything in HD wether the client needs it
it not.


Now we try to grade in 4K and are in some cases possible to to that. If I
is a TV-commercial we can grade in HD and then just change the resolution
when we start the render. The rendertime won't be days, maybe a few hours
at the top if it is a lot of material.



BUT

I seem to get some requests from online-guys that would prefere to get the
HD-DPX as 8but instead of 10bit.
Because they are working in 8bit anyway and just have to re-scale it
otherwise.
In Sweden we still deliver everything in SD, but at least we grade in HD so
they can downscale during online (so they can zoom if needed).


I'm I the only one getting requests to deliver 8bit HD-DPX? Or are other
colorist nagged on by online-people to deliver smaller files?


There seems to be a rule:
If it is already graded, there is no need for 10bit!


Feel free to comment either Pro or Against 8bit.


(note. this is only for non-vfx/greenscreen jobs that are targeted for TV.
Transmitted as 16/9-SD)

/Carl





(Sent from mobile device)
____________________________
*Carl Skaff*
*Colorist*
Stopp Stockholm
Office +46 8 50 70 35 00

*Stockholm | Los Angeles | Linz*
www.stopp.se
_______________________________________________
http://reels.colorist.org
http://tig.colorist.org/wiki3


More information about the Tig mailing list