Am I missing something?

[Paul Verizzo]

Yes, I've ruffled a lot of feathers about my paper/ink profiling opinions. Mostly along the lines of subjectivity trumps all that tech objectivity.

But I'm always willing to learn and I am well able to say, "I was wrong," if confronted with good evidence. So, here's the situation: In my testing of prints in harsh environments, and using my several color with corresponding B&W test images, something becomes very obvious. The subjective eye may see little or no changes in the color portions, but very obvious changes are happening in the monochrome portions.

Which then leads me ask, if a monochrome portion of a test image is "spot on," would not the color portion be, too? For instance, the Kodak model image here, http://1drv.ms/17BXQ1D has a small circle of 18% gray scale. If, in your appropriate program, it shows up as about 130 levels of RGB, isn't your color spot on?

 

[Smile]

As you know the printer linearity is what makes good ICC profiles, the trend to make ICC profiling to fix printer problems was not why this was invented. Sure the modern software can compensate allot of deviations but good linearity (when ink curves programmed into printer and actual ink used) is when quality profile can be made.

So single small circle 18% gray scale can't be used as ICC profile quality gage.
As you know there are many delta E formulas by year, and by industry. You just have to choose one that matches your industry and use it you can't compare different ones.

Also the synthetic profile accuracy and real printed hard proof comparison is very different. The synthetic may be very good but for instability with printer the hard proof would never pass certification. Such problems can't be revealed by synthetic testing.

 

[Paul Verizzo]

@Smile: I humbly admit that was a few thousand feet over my simpleton head.

So let me re-ask, if I may. If a standard grey scale consisting of equal part Red, Green, and Blue (set aside the 18% part for now) prints out equal colors, isn't that a perfect match?

I think I understand that a step tablet of each color might show variance at different densities. That would explain color shifts in, say, shadows vs. highlights. No?

But my question inevitably leads back to: If I'm printing perfect neutral B&W, doesn't the color image follow?

Thanks for your expertise.....

 

[mikling]

Depends on printer model as well. Your test will fail on a Canon OIG based printer such as the Pro-10.

 

[Smile]

@Smile: I humbly admit that was a few thousand feet over my simpleton head.

So let me re-ask, if I may. If a standard grey scale consisting of equal part Red, Green, and Blue (set aside the 18% part for now) prints out equal colors, isn't that a perfect match?

I think I understand that a step tablet of each color might show variance at different densities. That would explain color shifts in, say, shadows vs. highlights. No?

But my question inevitably leads back to: If I'm printing perfect neutral B&W, doesn't the color image follow?

Thanks for your expertise.....

That depends on printer how it mixes inks, what colors are used. If it's a true RGB printer (for example is laser based photo lab for example) then it calibrates internally the strength of each RGB laser beam and can print perfect BW images for example Same amount of light from each laser activates the paper and makes perfect BW. RGB in RGB out so to say. But that is where the simple things end.

If you want to ask such printer to print in color and match it to say monitor, other printer, printing standard etc. you have to make ICC profile for it (remember I said ICC profile is recipe for how printer prints) with the profile now one can do conversion from source ICC profile to the destination being printer profile and get same predictable color, over and over again. That is until something changes that in turn makes profile not accurate anymore.

Such condition inaccuracy is detected by RGB certification, and gives unambiguous answer to current printer printing quality. It is best to have two certification results to compare one to another if trending is important.

Since No Inkjet or laser printer prints with RGB inks, in turn they also need ICC profiles for even BW prints. If printer has BW ink carts and uses only these in BW print mode then this may not be the case for neutral BW but gradations can still be blocked depending how the conversion is done internally by printer driver.

I think I understand that a step tablet of each color might show variance at different densities.

True.

That would explain color shifts in, say, shadows vs. highlights. No?

If your printer BW with color inks and adds not equal amounts at different densities then, True.

But my question inevitably leads back to: If I'm printing perfect neutral B&W, doesn't the color image follow?

We all know what perfect BW is you have to define term "perfect color" to get answer to this one.

 

[RogerB]

Which then leads me ask, if a monochrome portion of a test image is "spot on," would not the color portion be, too? For instance, the Kodak model image here, http://1drv.ms/17BXQ1D has a small circle of 18% gray scale. If, in your appropriate program, it shows up as about 130 levels of RGB, isn't your color spot on?

OK @Paul Verizzo I'll bite.

Yes, you are missing something. A colour space is just that - it's a 3-dimensional space that contains all the colours. In the colour colour spaces that we use the line joining black and white represents all the shades of grey. If that line is perfectly straight then all your greys will be neutral. Unfortunately it tells you nothing about what is happening in the rest of the space. You can bust a gut to get perfect B&W but it doesn't follow that you get anywhere near perfect colour.

It's a simple matter in Photoshop to modify the colours while keeping the greys perfectly neutral. In other words, keeping the black-white line nice and straight but twisting the rest of the space around it. Here are a couple of random examples (left and right images) of the effect on the PDI test image (centre). If you use your R=G=B criterion you will find that the greys in all images are perfect.

The K.I.S.S approach has its merits, but is really challenged by this 3-D problem. If you don't believe me, just try editing one of these "distorted" images to get a perfect match to the original - including keeping perfect greys!

BTW, the Kodak image that you linked, having a very limited colour gamut, is much more tolerant of colour errors. Here's how it looks with the same "distortions" as the right-hand PDI image. I suspect that some people might be happy with that.....

 

[Roy Sletcher]

Yes, I've ruffled a lot of feathers about my paper/ink profiling opinions. Mostly along the lines of subjectivity trumps all that tech objectivity.

But I'm always willing to learn and I am well able to say, "I was wrong," if confronted with good evidence. So, here's the situation: In my testing of prints in harsh environments, and using my several color with corresponding B&W test images, something becomes very obvious. The subjective eye may see little or no changes in the color portions, but very obvious changes are happening in the monochrome portions.

Which then leads me ask, if a monochrome portion of a test image is "spot on," would not the color portion be, too? For instance, the Kodak model image here, http://1drv.ms/17BXQ1D has a small circle of 18% gray scale. If, in your appropriate program, it shows up as about 130 levels of RGB, isn't your color spot on?

Hi Paul,

Ruffling feathers is good. Makes us re-examine our core beliefs for validity before responding to others comments as I am doing now.

I also know from experience that colour and colour rendition is the most difficult subject to discuss with words. Ideally it needs an accurate colour managed environment so we can present accurate visual examples to bolster our empirical logic. Something not possible on a web browser and the compressed images of most websites such as this.

I am not going to try and change your mind, but present some facts for your consideration. Food for thought if you like, and you can make up your own mind

FIRST - if you have a couple of minutes watch the following video from a British colour scientist called Beau Lotto. This directly contradicts your subjective colour is accurate colour theory. I would be interested in your comments - especially regarding the desert scene comparison about 5½ minutes in.

https://www.youtube.com/watch?v=mf5otGNbkuc


SECOND - One can find innumerable examples of colour vision distortion o various complexities. One example as follows:

You have probably seen this before, and know it is a trick question. How much darker is square A than B?

Subjectively a blind man can tell that square A is much darker than square B,

Or is it?

See the following screen shot from my home computer.

When I take the file into Photoshop and put the eye dropper on each square they both register 101,101,101. A medium grey slightly lighter than mid grey.


This is NOT THEORY - You can duplicate these results at home!

To use one of my favourite homilies "Who are you going to believe - Me or your lying eyes?"


Enough for now - A good start.

Roy Sletcher
"We look with our eyes. We see with our brains. We understand with our minds." Dr Harry Wachs

 

[The Hat]

Now that you‘ll taken out that darn checker board this thread could go on for years.
I see with my eye and use my brain to define what exactly I am looking at.. The Hat..

 

[RogerB]

FIRST - if you have a couple of minutes watch the following video from a British colour scientist called Beau Lotto. This directly contradicts your subjective colour is accurate colour theory. I would be interested in your comments - especially regarding the desert scene comparison about 5½ minutes in.

https://www.youtube.com/watch?v=mf5otGNbkuc

Roy Sletcher
"We look with our eyes. We see with our brains. We understand with our minds." Dr Harry Wachs

Excellent video - well worth watching!

 

[stratman]

Now that you‘ll taken out that darn checker board this thread could go on for years.

"It's like deja-vu, all over again." - Yogi Berra