Return-Path: <pyfave@gmail.com>
Received: from mail-pb0-f41.google.com ([209.85.160.41] verified)
  by media-motion.tv (CommuniGate Pro SMTP 4.2.10)
  with ESMTP-TLS id 4720538 for AE-List@media-motion.tv; Mon, 21 May 2012 09:36:50 +0200
Received: by pbbrp2 with SMTP id rp2so6193394pbb.28
        for <AE-List@media-motion.tv>; Mon, 21 May 2012 00:38:46 -0700 (PDT)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=gmail.com; s=20120113;
        h=mime-version:in-reply-to:references:date:message-id:subject:from:to
         :content-type;
        bh=oc0JJiB37otnMynQdTLDkyOWF6QiCGmSicy7/AacY1o=;
        b=GBmKWGTOOMmH6W7X3QOFMDybEO4e+NmHRCT2Riv+qiuC6TOT6JbUs8vgOcEt6q/fkf
         MdWUKiZ6vlWduWoy2lsVy26d3qXW0Tgd56YQaKDlvxk5aUOX1ZsEkEBOWuCriU9W1nOB
         Y0f4qY7IA6tc15Xm8N4qH95VZTYyv4CNfzVlFqYVVwxgGdg2LsDf87a5argfg9JIsQuz
         e4+/6kGsKRt0NZEcRiYgkxk0opbxPuQPemrVaiOwNUBKo2iHWdh596ZqcTCRUtHHQuFU
         uTdkOXfVF/yJa3f1TWkU4my0+szbGxC/Tr2SbQPwzs6azMDaGf4Sf9lK/F9lTlTw61HK
         48IQ==
MIME-Version: 1.0
Received: by 10.68.132.101 with SMTP id ot5mr1594071pbb.91.1337585926242; Mon,
 21 May 2012 00:38:46 -0700 (PDT)
Received: by 10.142.233.13 with HTTP; Mon, 21 May 2012 00:38:46 -0700 (PDT)
In-Reply-To: <list-4716830@media-motion.tv>
References: <list-4716830@media-motion.tv>
Date: Mon, 21 May 2012 09:38:46 +0200
Message-ID: <CAAQHdpDJ799u=vtKuk=B+qaKY7NCebASO_md762utD_zjKxskw@mail.gmail.com>
Subject: Re: [AE] Perfect luma gradients in YUV for Blu-ray
From: Py Fave <pyfave@gmail.com>
To: After Effects Mail List <AE-List@media-motion.tv>
Content-Type: multipart/alternative; boundary=047d7b15a677fd82b204c086fd2f

--047d7b15a677fd82b204c086fd2f
Content-Type: text/plain; charset=ISO-8859-1

i think the problem might come from the fact you are using 8 bit images .

you have 256 luma values and hd is a lot more pixels

but i don't know if you can encode 16 bits to bluray .
is hdmi 8 bits only?

in after effects mostly the workaround is to use 4 color gradients with a
kind of "noise" for wich you have a slider control in the filter .


"...PERFECT gradients in Y, no skipped values, no repeated values "..
i think this is actually impossible.

anyone knows more about it?



2012/5/17 Sixtus Beckmesser <sbeckmesser@gmail.com>

> 1. My assignment: to make some test patterns that will appear on Blu-ray,
> the authoring system of which requires 8-bit YUV 4:2:2 files of some kind
> (I've been using AVI files as rendered by CS5.5 and now CS6 Cloud)
> 2. What I'm starting with: RGB still frames (usually TIFFS) with pixel
> values corresponding the digital video standard (16-235). To be sure that
> each pixel has the expected value, these still frames were calculated
> algorithmically.
> 3. The problem: YUV files made from such still frames of black-to-white
> RGB gradients end up with repeated pixel values or skipped pixel values
> (the gradient is not "monotonic"). I am assuming that this is because of
> round-off effects in the formulas used to get from RGB to YUV (I may be
> wrong on this).
> 4. Restrictions as to the solution: a. must be software based, b. must be
> free or very cheap, c. must produce the EXACT results I need. i.e. PERFECT
> gradients in Y, no skipped values, no repeated values -- and in a file
> format usable by the Blu-ray authoring system.
> 5. I guess what I need is a file-conversion system in which I can simply
> designate the contents of one color channel of an RGB signal (which are
> identical in black-to-white gradients) as Y and which will stick in the
> appropriate values in for U and V (0 in the case of monochrome test
> patterns). Anybody come across something like this? Is there a way to do
> this with anything contained in CS5.5 or CS6?
>
> --Sixtus
>



-- 
Pierre-Yves Fave
0612994425
http://www.5data.net

--047d7b15a677fd82b204c086fd2f
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: quoted-printable

i think the problem might come from the fact you are using 8 bit images .<b=
r><br>you have 256 luma values and hd is a lot more pixels<br><br>but i don=
&#39;t know if you can encode 16 bits to bluray .<br>is hdmi 8 bits only?<b=
r>
<br>in after effects mostly the workaround is to use 4 color gradients with=
 a kind of &quot;noise&quot; for wich you have a slider control in the filt=
er .<br><br><br>&quot;...PERFECT gradients in Y, no skipped values, no repe=
ated values &quot;..<br>
i think this is actually impossible.<br><br>anyone knows more about it?<br>=
<br><br><br><div class=3D"gmail_quote">2012/5/17 Sixtus Beckmesser <span di=
r=3D"ltr">&lt;<a href=3D"mailto:sbeckmesser@gmail.com" target=3D"_blank">sb=
eckmesser@gmail.com</a>&gt;</span><br>
<blockquote class=3D"gmail_quote" style=3D"margin:0 0 0 .8ex;border-left:1p=
x #ccc solid;padding-left:1ex">1. My assignment: to make some test patterns=
 that will appear on Blu-ray, the authoring system of which requires 8-bit =
YUV 4:2:2 files of some kind (I&#39;ve been using AVI files as rendered by =
CS5.5 and now CS6 Cloud)<div>

2. What I&#39;m starting with: RGB still frames (usually TIFFS) with pixel =
values corresponding the digital video standard (16-235). To be sure that e=
ach pixel has the expected value, these still frames were calculated algori=
thmically.</div>

<div>3. The problem: YUV files made from such still frames of black-to-whit=
e RGB gradients end up with repeated pixel values or skipped pixel values (=
the gradient is not &quot;monotonic&quot;). I am assuming that this is beca=
use of round-off effects in the formulas used to get from RGB to YUV (I may=
 be wrong on this).=A0</div>

<div>4. Restrictions as to the solution: a. must be software based, b. must=
 be free or very cheap, c. must produce the EXACT results I need. i.e. PERF=
ECT gradients in Y, no skipped values, no repeated values -- and in a file =
format usable by the Blu-ray authoring system.</div>

<div>5. I guess what I need is a file-conversion system in which I can simp=
ly designate the contents of one color channel of an RGB signal (which are =
identical in black-to-white gradients) as Y and which will stick in the app=
ropriate values in for U and V (0 in the case of monochrome test patterns).=
 Anybody come across something like this? Is there a way to do this with an=
ything contained in CS5.5 or CS6?</div>
<span class=3D"HOEnZb"><font color=3D"#888888">
<div><br></div><div>--Sixtus</div>
</font></span></blockquote></div><br><br clear=3D"all"><br>-- <br>Pierre-Yv=
es Fave<br>0612994425<br><a href=3D"http://www.5data.net" target=3D"_blank"=
>http://www.5data.net</a><br>

--047d7b15a677fd82b204c086fd2f--