From: bugbear on
NameHere wrote:
> On Thu, 04 Feb 2010 14:14:45 +0000, bugbear
> <bugbear(a)trim_papermule.co.uk_trim> wrote:
>
>> CMYK is a redundant colour space. It's equivalent (pace gamut concerns)
>> to RGB.
>
> Wherever did you read this? Must have been one of those "net-truths" that
> are so popular. The "net-truths" that float to the top of a Google search
> because it's the most plausible and popular, but wrong, explanation for
> those who can't think very clearly. The vast majority that don't want to
> try to understand nor take the time to educate themselves on anything more
> complex. This comprises the majority of all Google search-hits within the
> first 3 pages of them which are offered. Someone's feeding you some pretty
> good manure and keeping you in the dark. Do you feel like a fungus yet?
>
> When converting my images to CMYK from their RGB sources it's easy to see
> how many of the colors are not a match and get shifted. The reverse also
> true. Have you never done this in any editor with even something as simple
> as a Granger Calibration Chart? The shifts and obvious gaps between the two
> color-spaces are astounding.
>
> Here's an example to show you, starting with a 3000x3000 16-bit RGB Granger
> Chart as the source. Sorry, I don't have a CMYK Granger-like chart handy to
> show you the reverse. It can only be approximated in CMYK anyway. And I've
> already wasted far too much of my valuable time trying to educate you to
> begin with. Any CMYK to RGB conversion of anything would, of course, show
> far less disastrous results. Though, come to think of it, the CMYK to RGB
> is already implied in the right-panel because the CMYK space has to be
> converted back again to JPG's RGB and your monitor. This would explain the
> serious gaps in the colors re-presented.
>
> http://farm5.static.flickr.com/4071/4332259033_c0dced9de7_o.jpg
>
>
> 3 discreet units of information vs. 4 discreet units of information. There
> can only be a rough approximation between the two. And you want to argue
> about 8 vs. 10 vs. 12 senosr bit-depths while ignoring something even more
> simple for an example? Trying to go from RGB to CMYK is like trying to go
> from RGB to reconstruct Bayer's RGGB. Though I'll admit, with slightly
> less difficulty. Have someone explain to you why you can't reconstruct the
> sensor's RGGB data from the resulting RGB file and you'll start to
> comprehend why CMYK and RGB are not equivalent. I can't be bothered with
> trying to educate someone on something so rudimentary. I've tried that in
> the past and it was a painstaking task not unlike trying to teach a worm
> how to flip a light switch.
>
> I wasn't going to address the rest of your comments because I was trying to
> spare you the embarrassment. So I'll just address this one to show you why
> I didn't bother. You can thank me for not addressing the rest of your post.

Go and look up the word "gamut".

BugBear
From: NameHere on
On Tue, 09 Feb 2010 10:18:09 +0000, bugbear
<bugbear(a)trim_papermule.co.uk_trim> wrote:

>NameHere wrote:
>> On Thu, 04 Feb 2010 14:14:45 +0000, bugbear
>> <bugbear(a)trim_papermule.co.uk_trim> wrote:
>>
>>> CMYK is a redundant colour space. It's equivalent (pace gamut concerns)
>>> to RGB.
>>
>> Wherever did you read this? Must have been one of those "net-truths" that
>> are so popular. The "net-truths" that float to the top of a Google search
>> because it's the most plausible and popular, but wrong, explanation for
>> those who can't think very clearly. The vast majority that don't want to
>> try to understand nor take the time to educate themselves on anything more
>> complex. This comprises the majority of all Google search-hits within the
>> first 3 pages of them which are offered. Someone's feeding you some pretty
>> good manure and keeping you in the dark. Do you feel like a fungus yet?
>>
>> When converting my images to CMYK from their RGB sources it's easy to see
>> how many of the colors are not a match and get shifted. The reverse also
>> true. Have you never done this in any editor with even something as simple
>> as a Granger Calibration Chart? The shifts and obvious gaps between the two
>> color-spaces are astounding.
>>
>> Here's an example to show you, starting with a 3000x3000 16-bit RGB Granger
>> Chart as the source. Sorry, I don't have a CMYK Granger-like chart handy to
>> show you the reverse. It can only be approximated in CMYK anyway. And I've
>> already wasted far too much of my valuable time trying to educate you to
>> begin with. Any CMYK to RGB conversion of anything would, of course, show
>> far less disastrous results. Though, come to think of it, the CMYK to RGB
>> is already implied in the right-panel because the CMYK space has to be
>> converted back again to JPG's RGB and your monitor. This would explain the
>> serious gaps in the colors re-presented.
>>
>> http://farm5.static.flickr.com/4071/4332259033_c0dced9de7_o.jpg
>>
>>
>> 3 discreet units of information vs. 4 discreet units of information. There
>> can only be a rough approximation between the two. And you want to argue
>> about 8 vs. 10 vs. 12 senosr bit-depths while ignoring something even more
>> simple for an example? Trying to go from RGB to CMYK is like trying to go
>> from RGB to reconstruct Bayer's RGGB. Though I'll admit, with slightly
>> less difficulty. Have someone explain to you why you can't reconstruct the
>> sensor's RGGB data from the resulting RGB file and you'll start to
>> comprehend why CMYK and RGB are not equivalent. I can't be bothered with
>> trying to educate someone on something so rudimentary. I've tried that in
>> the past and it was a painstaking task not unlike trying to teach a worm
>> how to flip a light switch.
>>
>> I wasn't going to address the rest of your comments because I was trying to
>> spare you the embarrassment. So I'll just address this one to show you why
>> I didn't bother. You can thank me for not addressing the rest of your post.
>
>Go and look up the word "gamut".
>
> BugBear

Yes, you might want to do that. A good thing you left a note to yourself to
remind you.