Najan wrote:
Can we play too, Peano?
I have an answer but don't want to interfere in your argument.

Sure ... rock on.

Robert Randall wrote:
I'm having a difficult time understanding you and most anyone else that is having difficulty accepting the premise of changing tone and color, mostly because I've been doing it for a living for longer than I care to remember.

The very first time I had to change black hosiery to nude was in 1991, for an ad someone shot for Hanes. I lost the job because I couldn't figure out a way to accomplish the task, while a trade shop did accomplish it. That pissed me off, so I went about learning how to do that specific task. That's about the same time I started working with Dan Margulis, who helped me to learn work flows like that.

Masking that lace took me about three minutes. Convert the file to CMYK using a light black. The cyan channel is almost a perfect mask right out of the gate, and the black channel contains more detail than any of the RGB channels. Other than the fact that the original image supplied is a bit inferior, this little project was a cake walk.

I don't get your apprehension.

This is why I love the forums. Someone steps up, and says "you're wrong", and then I learn. THANK YOU ROBERT, I WAS WRONG.

CMYK hasn't crossed my mind for this kind of thing (and it really does help A TON, as you have another channel to separate things from).

Robert Randall wrote:
I'm having a difficult time understanding you and most anyone else that is having difficulty accepting the premise of changing tone and color, mostly because I've been doing it for a living for longer than I care to remember.

The very first time I had to change black hosiery to nude was in 1991, for an ad someone shot for Hanes. I lost the job because I couldn't figure out a way to accomplish the task, while a trade shop did accomplish it. That pissed me off, so I went about learning how to do that specific task. That's about the same time I started working with Dan Margulis, who helped me to learn work flows like that.

Masking that lace took me about three minutes. Convert the file to CMYK using a light black. The cyan channel is almost a perfect mask right out of the gate, and the black channel contains more detail than any of the RGB channels. Other than the fact that the original image supplied is a bit inferior, this little project was a cake walk.

I don't get your apprehension.

Jakov Markovic wrote:
This is why I love the forums. Someone steps up, and says "you're wrong", and then I learn. THANK YOU ROBERT, I WAS WRONG.

CMYK hasn't crossed my mind for this kind of thing (and it really does help A TON, as you have another channel to separate things from).

Well said Jakov!

Peano wrote:
Sure ... rock on.

Honestly, it's difficult to say which one is a conversion or not. Or, if there is any. Although, from all three pictures there's one detail that attract my eyes and makes me wonder.

It's the button from the left skirt. The occlusion shadow/line between the button and skirt should be more pronounced. The tone and shadow between the button and the skirt texture are somewhat even.

Perhaps I'm wrong, I'm waiting for your answer here to solve this amusing riddle, Peano.

Peano wrote:

Does your experience tell you which (if any) of these skirts was originally black and changed to white in post?

Upon close inspection, about the only tell is the weak black in the skirt on the right, but that's just a guess, and it might not be any of them for all I know.

Peano wrote:
The fact that it's difficult to tell makes my point (and Robert's), that you don't have to sacrifice detail when converting black to white. Here's the original:

I was guessing the left one but wasn't certain at all (and if someone hadn't told me to look I wouldn't have given them a second thought)

The only tip off for me (and it's a similar reservation with the truck) is the shadows.  Notice the difference in shadow density in the pleats on the left vs. the middle skirt.  The truck has a similar problem.. it looks like the "dark" areas aren't dark enough.

Wye wrote:
The only tip off for me (and it's a similar reservation with the truck) is the shadows.  Notice the difference in shadow density in the pleats on the left vs. the middle skirt.  The truck has a similar problem.. it looks like the "dark" areas aren't dark enough.

But that isn't a feature (or bug) of converting from black to white. It's a simple curves adjustment that I didn't make initially.

Peano wrote:

But that isn't a feature (or bug) of converting from black to white. It's a simple curves adjustment that I didn't make initially.

Well.. yes and no.. there's a certain "nature" to the shadows in a white material that is a bit different than that in a dark material.  It's not *just* that the shadows aren't dark enough. There's the darkness in those shadows that, yes, is an overall thing (though I would say the adjustment needed should be only applied in the shadows) but there is also some bounce light that happens as well as the effects of the translucency that a white material will have that a black one won't.

Like I said earlier.. if you had never told me that one of them was black I wouldn't have even had a second thought about the images.. but looking deeper into it the subtle effects of lighting and shadowing are definitely something to be wary of.

Jakov Markovic wrote:
This is why I love the forums. Someone steps up, and says "you're wrong", and then I learn. THANK YOU ROBERT, I WAS WRONG.

CMYK hasn't crossed my mind for this kind of thing (and it really does help A TON, as you have another channel to separate things from).

Earlier in this thread, I believe that Robert mentioned Dan Margulis along with channels. One of Margulis's friends Giuliana Abbiati created Channel Power Tools. CPT costs twenty Euros. Abbiati has provided installation and instructions videos at the previous link.

I have no relationship with Abbiati, other than as a satisfied customer.

Wye wrote:
Well.. yes and no.. there's a certain "nature" to the shadows in a white material that is a bit different than that in a dark material.  It's not *just* that the shadows aren't dark enough. There's the darkness in those shadows that, yes, is an overall thing (though I would say the adjustment needed should be only applied in the shadows) but there is also some bounce light that happens as well as the effects of the translucency that a white material will have that a black one won't.

Like I said earlier.. if you had never told me that one of them was black I wouldn't have even had a second thought about the images.. but looking deeper into it the subtle effects of lighting and shadowing are definitely something to be wary of.

Well, trick or treat, my friend. 'Tis the season. I just knew some pixel-peeper would come along and say "yes, but, yada yada yada." Here's the actual original image:

Peano wrote:

Well, trick or treat, my friend. 'Tis the season. I just knew some pixel-peeper would come along and say "yes, but, yada yada yada." Here's the actual original image:

I don't believe you, sorry.

You can post the raw files and I will gladly admit I'm wrong. But sorry. No. I think you're just trying to save face.

Wye wrote:
I don't believe you, sorry.

Believe, don't believe, it makes no difference. The whole point of this exercise is that you can't really tell whether the color has been changed. You can only guess what the original color of this skirt was:

Peano wrote:
Believe, don't believe, it makes no difference. The whole point of this exercise is that you can't really tell whether the color has been changed. You can only guess what the original color of this skirt was:

You executed your point perfectly. Also, your point applies to many aspects of retouching. I've shown photos to friends that I've retouched, and they'd say things like "It looks like you made her arms too thin" when I had never even touched the liquify tool. Our eyes play nasty tricks.