Hi Job:
I've posted this in the forums to un-mangle my test results:
http://community.avid.com/forums/p/113013/654515.aspx#654515
My supposition, based on these tests, is that when you do a
Same-as-source export with RGB levels, the Avid does not actually change
the pixel values, but sets a flag telling the quicktime library how to
interpret the pixel values. When the quicktime library is called to
display the image or serve it up to another application, it checks the
flag, and if it is RGB, adjusts the levels to crush below 16 and clip
above 235.
You can confirm this by checking the following test case. Export
with RGB levels set. It will look crushed/clipped in the finder and
other applications, as you might expect. Re-import that quicktime with
CCIR levels checked. You will get back exactly what you started with,
with no crushing or clipping.
Unfortunately, as my tests show, the handling of the 601/RGB flag
differs in certain cases between DNxHD and ProRes.
Cheers,
--Michael
On 12-09-22 11:53 PM, Job ter Burg (L2B) wrote:
>
> , ProRes exports always behave as though they were exported
> > as RGB, regardless of whether RGB or 601 was chosen in the export
> dialogue.
> >
> > This is consistent with the results you just reported, as well as the
> > odd results I reported earlier in this thread.
>
> Michael,
>
> That doesn't make sense to me. If the APR export would behave like it
> was exported as RGB (which would crush 0-15 and clip 236-255), a
> re-import into Avid would show different levels, and it does not. All
> levels below 16 and over 235 are maintained.
>
> It's hard to follow your test results as they show up mangled up here.
>
> How and where are you determining that the APR export "behaves like RGB"?
>
> J
>
>
[Non-text portions of this message have been removed]
Sunday, September 23, 2012
Re: [Avid-L2] ProRes export weirdness, with same-as-source export (LONG)
__._,_.___
Search the official Complete Avid-L archives at: http://archives.bengrosser.com/avid/
.
__,_._,___
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment