From: bugbear on
NameHere wrote:
> On Thu, 04 Feb 2010 11:38:59 +0000, bugbear
> <bugbear(a)trim_papermule.co.uk_trim> wrote:
>
>> NameHere wrote:
>>> And it won't do a lick of good if you can only present it in 6-bits worth
>>> of resolution on your monitor or even less bit-depth on paper.
>> You appear to be suggesting that prints only support 32 gray levels.
>> What are you using - 24 pin dot matrix?
>
> 6-bits per each channel of CMYK. That's a 24-bit color depth resolution.
> 16,777,216 colors.

CMYK is a redundant colour space. It's equivalent (pace gamut concerns)
to RGB.

> There's a good reason that people resort to Piezography for B&W prints.
> Using various shades of gray inks to achieve more than 64 gray levels
> without dithering. (6-bits is 64, not 32)

Indeed. But you said "less bit-depth[than 6]" on paper.
5 is less than 6 (stop me if I'm going too fast for you)

>
>> AMD does 10 bit video cards - welcome to 2010.
>
> All fine and dandy. Now tell me how many distinct hues your human eye can
> discern faithfully.

The human eye is not a trivial light meter - with adaption
it can perceive a truly massive range of intensities.

>> Further, since processing involves arithmetic, which involves
>> rounding, having extra colour resolution prior to processing
>> is advantageous.
>>
>> BugBear
>
> This is why I said that 10-bit depth per sensor channel is more than is
> needed for the final display. 12 and 14 bit camera sensors are nothing but
> a marketing ploy that is directly targeted to major fools. Rarely do I need
> to do much post-processing to images from my P&S cameras because I know how
> to use them properly in the first place.

If you're happy for your images to printed as captured, that's dandy.
But if you're doing anything interesting, such as contrast curve
adjustment, or pananoramic stitching, you're STILL
going to want deeper data to allow a little room for intermediate
rounding errors.

> They also do a fine job of
> retaining the full dynamic range of the sensor in the JPG file to begin
> with.

This is trivial; here an alogorithm:

RAW 0 -> JEPG 0
RAW MAX -> JPEG MAX

You speak as if this is some kind of achievment.

> I've compared the JPG and eventual RAW manipulated data. There's so
> little difference between the two that it's difficult to best the JPG file
> by starting with RAW. You can spend upwards of half an hour or more trying
> to tweak anything extra out of the RAW sensor data. It's mostly a huge
> waste of time because the camera's programming does it so well in the RAW
> to JPG conversion to begin with. Those who have less capable cameras will
> have to resort to post-processing the RAW data to fix what their camera
> failed to deliver in the first place. Or, more commonly, what they as
> snapshooters failed to do in the first place.
>
> How difficult is this for you to comprehend? It seems to be frying your
> brain so far.

I seem to be comprehending more than you, and better too.

"There seems to be much confusion on the internet about the
relevance of image bit depth on recordable dynamic range."

http://www.cambridgeincolour.com/tutorials/dynamic-range.htm

Read and learn.

BugBear
From: Ray Fischer on
bugbear <bugbear(a)trim_papermule.co.uk_trim> wrote:
>NameHere wrote:
>>
>> And it won't do a lick of good if you can only present it in 6-bits worth
>> of resolution on your monitor or even less bit-depth on paper.
>
>You appear to be suggesting that prints only support 32 gray levels.
>What are you using - 24 pin dot matrix?

Do you believe that your printer can do more than 64 levels?

>AMD does 10 bit video cards - welcome to 2010.

Video cards are not monitors.

>Further, since processing involves arithmetic, which involves
>rounding, having extra colour resolution prior to processing
>is advantageous.

During processing it's useful. Afterward? Not so much.

--
Ray Fischer
rfischer(a)sonic.net

From: Ray Fischer on
bugbear <bugbear(a)trim_papermule.co.uk_trim> wrote:
>NameHere wrote:
>> If a value of 0 from a sensor's
>> photosite is converted to 0 in the JPG file, and a value of 1024, 4096, or
>> 16384 (10 to 14 bit dependent) from a photosite is converted to a value of
>> 255 in the JPG file, then the JPG file represents the whole dynamic range.
>
>Agreed. However colour resolution may have been lost.

So what?

--
Ray Fischer
rfischer(a)sonic.net

From: Paul Furman on
bugbear wrote:
> NameHere wrote:
>> I see the correct color balance and exposure on the LCD or
>> EVF *before* I take the shot.
>
> Since colour perception is dependant on ambient light,
> I think you're full of ..it.

I can't see someone routinely adjusting white balance, contrast &
saturation for each shot. I do that for focus stacking somewhat because
I don't want the huge raw files but for general photography there's no
way to really get a good look at things still you get home.

Of course it's possible to take nice photos with a canned jpeg output
from the camera but more flexibility is more flexibility.

I agree 14 bit is overkill. I played with it and could see no value.
From: LOL! on
On Thu, 04 Feb 2010 14:14:45 +0000, bugbear
<bugbear(a)trim_papermule.co.uk_trim> wrote:

>NameHere wrote:
>> On Thu, 04 Feb 2010 11:38:59 +0000, bugbear
>> <bugbear(a)trim_papermule.co.uk_trim> wrote:
>>
>
>> They also do a fine job of
>> retaining the full dynamic range of the sensor in the JPG file to begin
>> with.
>
>This is trivial; here an alogorithm:
>
>RAW 0 -> JEPG 0
>RAW MAX -> JPEG MAX
>
>You speak as if this is some kind of achievment.

Ask any DSLR owner who worships RAW and you'll find out that it is very
much some kind of major achievement. Apparently none of their cameras are
capable of something so simple. They're always clamouring how they get two
or more stops of dynamic range out of the RAW data compared to the JPG file
their camera can produce. Go ahead, ask them. They're even stupidly willing
to spend an extra $100-$200 for the required software needed to repair what
their camera's firmware failed to do correctly in the first place. Then on
top of that they waste even more valuable hours of their life trying to
correct all the errors from their camera's firmware on every snapshot they
take.

LOL!