problems w/ CMYK chapter

[dpnew]

Hi,

First, I have the errata for step 3 in "Making a Luminosity Mask" on p.139, which adds the directions "Press Crtl + E to merge the Luminosity and Black layers". The Black layer is a 50% gray layer. To create the mask, you start off with the Luminosity layer created by the Hidden Power tool Split Luminosity which already merges the luminosity channel with a 50% gray layer to commit the luminosity. Then you make a levels adjustment on the Luminosity layer. Finally, you merge into a 50% gray layer. That last step has me stumped. What change needs committing at that point that requires merging into a gray layer? As far as I can tell, that step doesn't commit any change that wasn't already committed, and therefore has no purpose.

I've done all the steps for CMYK separations at least 10 times, and I can't get my final image to match the plates in the book or what I got when I used the Hidden Power tool "CMYK Process". When I used the CMYK Process tool, for any Curves prompts I just clicked OK and did not make any changes, and for any Levels prompts, I always slid the right slider to 128, and the results from the tool closely match the pictures of the plates in the book.

The CMY separation is very straightforward, and each channel looks correct after I do it, but I repeated those steps anyway just in case I was inadvertently doing something wrong. As for making the saturation mask and the luminosity mask, I've done it enough times that it also seems straightforward, and my masks seem to match the pictures in the book, but when I transfer the mask to my original file, and do the inverting, and copying and the placement in each channel, the image I get is pretty bad. I checked all the blend modes, and they were correct. Is there a tricky step where people frequently make mistakes? I can't figure out what's wrong.

[edit]
I performed an experiment to narrow down my problem. First, I used the CMYK Process tool to create the separations. Then, I opened a new file and duplicated the background image from the first file, and I made the saturation mask and the luminosity mask according to the instructions in the sections "Making the Saturation Mask" and "Making the Luminosity Mask". After creating my saturation mask and luminosity mask and merging them together, I copied the merged mask layer, called the Black layer, into the original file where I used the CMYK Process tool to create the separations. I did that in order to compare my Black layer to the one generated by the CMYK Process tool, and they are not the same: mine is slightly darker. I have no idea why that is the case. To find out if I could see anything different the tool was doing, I watched the CMYK Process tool as it created the masks and the only thing it did differently is it added a Curve at one of the steps when making the masks, but I didn't make any curve adjustments to that curve when I was prompted, so I don't think that should make any difference. To make absolutely sure, I tested that by adding a Curves adjustment layer without making any adjustments to my Black layer, but it remained the same.

At this point, I have a hard time believing anyone can get the same plates as the CMYK Process tool or the ones pictured in the book following the steps in the book.
[/edit]

[Richard_Lynch]

At this point, I have a hard time believing anyone can get the same plates as the CMYK Process tool or the ones pictured in the book following the steps in the book.

hmmm...I am not sure what benefit it would be to the core of the book and my reputation to include procedures that are a lie. I can tell you this: if you run the CMYK action, you get a CMYK separation. You do it all in Elements. You do it by following the instructions, which are very long, and in the end, quite technical. The plates you see are a result of me following my own steps.

Without standing over your shoulder and watching you go through the 150+ odd steps, it would be difficult to tell which is not being executed as I thought i wrote it down. On the other hand, i hope I suggested it in the book, you have an action in some 6 different parts to accomplish the end result. Going through each of the segments of the exercise and confirming the results against the actions would be a great way of determining where the difference occurs.

There is quite a bit of theory involved here, and this shows just one possible way of making the separation. I added a few opportunities to customize the result in the action, but the curves provided there do not have any settings (or they are the default, so no change occurs).

If you have the free tools, a curious thing to do might be to slow down the action playback so you can watch the steps of the action as they occur. This can be a great learning tool.

As for the other procedure, again, something must not be translating from what i think I wrote. The Black layer should not be gray but for a moment, and this could be the key to the difference in your result. Chances are there is something that is not getting done. The point of steps 1 and 2 (and the step added via the errata) is to commit the luminosity. That is, the luminosity only appears to show the tone at this point, you need to merge the layers so the luminosity is actual rather than virtual (or so the layer is all tone and no color), and to convert the layer mode to Normal 9just changing it may not produce the same results).

Q: when you merge the Luminosity and Black layers, you get a gray result? If so, then either the luminosity layer is incorrect, or the layers are not properly ordered.

This is, BTW, the most difficult conceptual material in the book. Most professional Photoshop users would have no clue as to how to go about making this separation...they would not know the theory behind it or where to begin. If you can understand the process it is far more important than getting the results to match any of those shown in the book exactly. While I understand that desire, the goal (at least my goal in the writing) was to show an example of what the process entails. If you understand this, you understand far more than most professional digital artists (or those whom I have recently been interviewing for a position opening), and in the long run this can lead to a greater understanding and better results in ALL of your print work.

Can be done, cross my heart.

[dpnew]

Hi,

Thanks for the response.

hmmm...I am not sure what benefit it would be to the core of the book and my reputation to include procedures procedures that are a lie.

lol. I wasn't suggesting that you were a scoundrel--only that there might be a mistake in the book somewhere.

Going through each of the segments of the exercise and confirming the results against the actions would be a great way of determining where the difference occurs.

Yes, that's a good idea, and I'll try it tomorrow, although I have at least one problem narrowed down.

If you use the CMYK Process tool to create the separation, the Black Layer is on top. It only takes 15 steps(and some of those are very simple) to recreate the Black layer from the background image. If you or anyone else could, would you recreate the Black layer following the 15 steps in the book starting on p.138 and compare it to the one created by the CMYK Process tool? It's hard to make a mistake on those 15 steps, and I've done them numerous times, and I compared the Black layer I came up with to the one the CMYK Process tool creates, and they are not the same.

If you have the free tools, a curious thing to do might be to slow down the action playback so you can watch the steps of the action as they occur. This can be a great learning tool.

Yes, I do have the free tools, and that is a great feature that I didn't know was included--however, my computer is so slow, I can actually see what's occurring at a reasonable speed already.

Q: when you merge the Luminosity and Black layers, you get a gray result? If so, then either the luminosity layer is incorrect, or the layers are not properly ordered.

No, I don't get a gray result. I'm just merging a grayscale layer into a 50% gray layer, the "Black" layer, which does nothing. I have learned that if you have a duplicate of your background image and set the blend mode to luminosity, that change is not committed, and to commit the luminosity, you have to merge it with a gray layer(or black or white). However, when you use the Split Luminosity tool, the tool commits the luminosity of the background duplicate layer for you by merging it into a gray layer. One way to tell that's the case is if you look at the blend mode of the Luminosity layer created by the Split Luminosity tool, the blend mode is Normal, so you know that layer's pixels aren't interacting with another layer below it to give you the grayscale result. Another way is to watch the Split Luminosity tool action as it plays, and you can see it commit the luminosity by merging the Luminosity layer with a 50% gray layer.

Since the Split Luminosity tool was used to create the Luminosity layer in the example way back in the section on "Making the Saturation Mask", there is no need to merge the Luminosity layer into a gray layer to commit the luminosity--as was added by the errata. Instead, all you need to do is change the name of the Luminosity layer to "Black". (There is an intervening Levels adjustment, but that certainly doesn't require merging into another layer to commit the adjustment.) I believe Step 1, on p.139 should be eliminated, the errata calling for a merge should be removed, and step 3 should direct the readers to change the name of the Luminosity layer to "Black" after the Levels adjustment.

If you understand this(CMYK separations), you understand far more than most professional digital artists...

I certainly have the mechanics down after the number of times I've repeated the example, but my understanding of what's going on is still gelling. I'm compiling a post about what I think is going on confirmed by some experiments, and I was hoping someone could review it once I post it.

[dpnew]

Going through each of the segments of the exercise and confirming the results against the actions would be a great way of determining where the difference occurs.

I did the steps by hand and compared them with each of the Hidden Power tools. In order to make sure each of my plates was exactly the same as the Hidden Power tools' results, I copied my plates into a separate file where I was performing the separations using only the tools, and then I toggled each of my plates on and off with each of the tools' plates, and mine were exactly the same.

That is until the CMYK Black tool. As stated in my previous post, the problem is trying to match the result of the Black layer. It's 15 steps, starting with the background image, and I have done it so many times I have the steps memorized. It seems to me there must be a mistake in the book in those 15 steps. I've done it exactly according to the steps in the book with no success, and I've also tried altering the steps a little bit: by splitting the luminosity by hand instead of using the Hidden Power tool, etc. to see if that would make a difference, but I can't match the result of the CMYK Black tool. I also tried using a Levels correction on my Black Layer, to see if I could use the middle slider to get it to match the Black layer produced by CMYK Black, but that didn't work either. They are close, but they are not the same: the Black layer produced by following the steps in the book is darker than the one produced by the Hidden Power tool(either CMYK Black or CMYK Process), and that creates a noticeable difference when you look at the final image with all the separations--because the difference is tripled when you invert the Black layer and use the inverted layer in each separation.

[edit]
I just thought of something: i'll step through the CMYK black tool and see what it's actually doing.
[/edit]

[dpnew]

There is a mistake in the book!

It's a miracle! I'm saved! Wooohooooo!!! I can finally make CMYK separations that match the CMYK tools!

Initially, I played the CMYK Black tool with long pauses, and the only thing I saw different than the steps in the book were the insertion of two Curve adjustment layers, but since I always clicked on them without making any changes when I ran the tool, those extra steps the CMYK Black tool inserted shouldn't have made a difference.

All the same, I dutifully tried adding those steps in when doing the CMYK Black steps manually, and as I knew there shouldn't be, there wasn't a difference in my result, so my Black layer did not match the Black layer produced by the tool once again.

So, I played the CMYK Black tool action a couple of more times to make sure I didn't miss a step, and just when I was about to throw in the towel--I noticed the blend mode the tool used before merging the Saturation layer(the saturation mask) and the Black layer(the luminosity mask): the tool uses SCREEN! The book directs us to use LIGHTEN! Bingo! With the Screen blend mode, I can make the Black layer match what the tool produces. I looked up the difference between Screen and Lighten blend modes, and there is a significant difference, as toggling on and off those different blend modes also demonstrates.

Problem solved.

[dpnew]

still can't get correct full separation

Hi,

Now, I can't get the final 6 steps under the section "Remove Color Under Black" to work. I've watched the Hidden Power tool perform those steps, but I can't locate the error yet. I have found an additional error in the book: the book says the last six steps of the separation--in the section "Remove Color Under Black"-- can be replicated by the Remove Black Color tool, but when I try to run that tool, I get an error(it's looking for layers that haven't been created yet). The Apply Black tool, however, does perform those steps, but it's use isn't mentioned anywhere. After using the Apply Black tool, the Remove Black Color tool can be used to merge the Black Copy layers into the channels, which doesn't seem like a very functional tool to me.

[dpnew]

Done.

I finally got a full CMYK separation. Step 2, on p.140, under the section "Remove Color Under Black" is what tripped me up. It says:

2. Open Levels and change the white Output slider to....128.

"the white OUTPUT slider" is the key phrase. In all the other steps, the reader has been instructed to move "the white slider to 128". Well, the white slider and the white Output slider in Levels are different. I would hazard to guess that I'm not the only one that has gotten tripped up by that one. I firmly believe there should be some clarification there. I have never used the Output slider in Levels before, however, I read somewhere, maybe in the newsletter, that a very experienced PS user had used the Output slider only a handful of times in their whole career. I think there should be something in the errata for that step pointing out the difference between the two sliders. It's not an easy distinction to catch.