[opendtv] Re: RGB mania ('translated' for Prin)
- From: "Alan Roberts" <roberts.mugswell@xxxxxxxxxxxxxx>
- To: <opendtv@xxxxxxxxxxxxx>
- Date: Sat, 15 Jan 2005 09:42:07 -0000
Thanks John. Here's what I normally recommend for shooting HD drama, where
the footage will always be graded.
I set the knee to capture as much contrast as is feasible, up to 600% in
some cameras. Setting knee point and slope depends on the DoP and how he
wants the pictures to look. I set the maximum exposure to be at around 106%,
about level 249 in 8-bit terms. This still allows some space for ringing,
but since we usually shoot will detail low or off, ringing rarely happens
anyway. The concept of peak white is irrelevant in this case, exposure is
set with an 18% card, or just by looking at a monitor. The idea is to get
the scene contrast to the colourist, who will then make the limiting
decisions, which include ensuring that no significant areas of picture
exceed notional 235.
If you seen any HD drama shot by the BBC, you've seen the effect.
Alan
----- Original Message -----
From: "John Golitsis" <john@xxxxxxxxxxxxxxxxxx>
To: <opendtv@xxxxxxxxxxxxx>
Sent: Friday, January 14, 2005 2:49 PM
Subject: [opendtv] Re: RGB mania ('translated' for Prin)
> Allow me to perform the translation this time ;-)
> ===========================================
>
> The problem we have run into in the TV world is the interpretation of
> the coding range - specifically the "meaning" of black and white. This
> is something that Alan and I have kicked around...Take the meaning of
> 'white'
>
> In an 8 bit ITU-R BT.601 system the code for 'white' is 235 [decimal] -
> all seems clear and simple.
>
> However - if you look at the taking characteristics of the camera - and
> specifcally its optoelectronic transfer function: there are at least 2
> meanings of white. The first is established by the knee, the second is
> the area which is used to cope with specular highlights. The position
> of the knee - and the slope of the compression for highlights are both
> under control of the camera man.
>
> So what do you do? Do you set the knee [I'll call this the mid to large
> area white] for 235, and allow speculars to go up to
> the link clip limit. Or do you set the speculars for 235?....or do you
> make them both the same?
>
> At the display end, a conventional TV [not a broadcast monitor] - will
> allow the small area luminance level to reach say 450cd/m2. However
> the large area light output will be limited to say 120cd/m2. So a
> tellie - as the ability to have a stab at reproducing specular
> highlights.....and does this by the miracle of beam current limiting.
> It also manges to interpret small areas of code 235 as specular. This
> is probably a good thing.
>
> Now - you try doing that on either a PWM based or light valve display.
> Interestingly a plasma panel does have a related small/large area white
> characteristc - based on overall power limiting, and Panasonic have
> engineered some dynamic APL based processing into their professional
> projectors. However this combination/mixed melange of display
> characteristics in the field just makes things even worse, we have NO
> chance of achieving a controlled transfer between broadcast monitoring
> and home viewing.
>
> In addition to this - because BT.601 does not distinguish between
> specular highlights and medium/large area white...we do not have any
> simple mechanism to signal to the display that this is a specular
> region...which has had this compression slope applied by the camera.
> To compound matters - if you even tried to use this specular region
> then a) Some alarm in some gamut legaliser would trigger and someone
> would wind in the clippers. And/Or b) BECAUSE we did not codify the use
> of this region, and because we did not say that all digital displays
> should have a soft clipper......then...if the broadcaster did not apply
> the soft clipper - the display would probably hard clip - which would
> be even worse.
>
> So - this is a subtle and unintended consequence which is due to a
> simplistic interpretation of the coding range [based on what worked for
> CRT's] and the switch to digital matrix displays.
>
> The whites are relatively straight forward to explain!
>
> Television is just plain broke...
>
> Prin
>
> Prinyar Boon
> Snell & Wilcox Ltd
> Liss Mill
> GU33 7BD
>
> email: prinyar.boon@xxxxxxxxxxxxxxx
> Tel: +44 (0)2392 489 280 direct
> Fax: +44 (0)1730 894 341
> www.snellwilcox.com
>
> -----Original Message-----
> From: opendtv-bounce@xxxxxxxxxxxxx [mailto:opendtv-bounce@xxxxxxxxxxxx
> g]
> On Behalf Of Alan Roberts
> Sent: 14 January 2005 12:16
> To: opendtv@xxxxxxxxxxxxx
> Subject: [opendtv] Re: RGB mania
>
> Not quite right Jeroen. ITU.601 etc all include specifi
> cations for the
> coding of RGB as well as YUV signals. They form a cl
> osed set. The rules
> for Y also apply to RGB. Indeed, the coding of YUV c
> omes from a digital
> calculation using digital RGB signals, each of which must
> conform to 601
> coding rules. At no point is there any flexibility in
> the coding of Y,
> since it normally doesn't exist in analogue these days.
>
> The headroom in the coding range is specifically for the
> RGB signals
> rather than Y, the extra range that Y has as a resul
> t is almost
> accidental.
>
> Alan
>
> ----- Original Message -----
> From: <jeroen.stessen@xxxxxxxxxxx>
> To: <opendtv@xxxxxxxxxxxxx>
> Sent: Friday, January 14, 2005 11:51 AM
> Subject: [opendtv] Re: RGB mania
>
>
> Hello,
> 3.
> With digital RGB "graphics" signals (e.g. DVI out) it
> is common
> to have black 3D 0, white 3D 255, but with digit
> al YUV "video"
> signals (e.g. ITU-R.656 or SDI) it is common to have
> black 3D 16,
> white 3D 235. The Y signal has headroom that the
> RGB signals have
> not. This headroom is reserved for the ringing due t
> o anti-
>
> Greetings,
> -- Jeroen
>
>
>
>
> ----------------------------------------------------------------------
> You can UNSUBSCRIBE from the OpenDTV list in two ways:
>
> - Using the UNSUBSCRIBE command in your user configuration
> settings at
> FreeLists.org
>
> - By sending a message to: opendtv-request@xxxxxxxxxxxxx wit
> h the word
> unsubscribe in the subject line.
>
>
> This email and any attachments is confidential, may be l
> egally privileged and is intended for the use of the a
> ddressee only. If you are not the intended recipient, pl
> ease note that any use, disclosure, printing or copying
> of this email is strictly prohibited and may be unlawful
> . If received in error, please delete this email and a
> ny attachments and confirm this to the sender.
>
>
> ----------------------------------------------------------------------
> You can UNSUBSCRIBE from the OpenDTV list in two ways:
>
> - Using the UNSUBSCRIBE command in your user configuration settings at
> FreeLists.org
>
> - By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
> unsubscribe in the subject line.
>
>
>
> ----------------------------------------------------------------------
> You can UNSUBSCRIBE from the OpenDTV list in two ways:
>
> - Using the UNSUBSCRIBE command in your user configuration settings at
FreeLists.org
>
> - By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
unsubscribe in the subject line.
>
----------------------------------------------------------------------
You can UNSUBSCRIBE from the OpenDTV list in two ways:
- Using the UNSUBSCRIBE command in your user configuration settings at
FreeLists.org
- By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word
unsubscribe in the subject line.
Other related posts:
