hi all
I am new to the site and loving it!

I am avidly following up on and tracking down all references to extraction, masks and clipping paths. Most of what I have found so far has been very informative and interesting (some very confusing and way above my head on first reading).
I have noticed something that confuses me, namely:
evryone (or most of you) seems to work in rgb, as opposed to cmyk.
Is rgb better for making masks?
Do you work in rgb because photoshop lets you do more with it, or because the colour gamut is wider?

I work in cmyk, so as to check colours for print (we use a profile also).
I prefer cmyk - I can see the colour seps better in my mind's eye.
Do you suggest that I switch to rgb?

thanks for all the help so far and still to come.
karin

There are few workflow advantages in working with RGB over CMYK, and fewer with each new Photoshop version. There are some theoretical advantages in both directions, but it boils down to: if you're more comfortable in CMYK, especially if 100% of your work will end up being printed using CMYK, stick with CMYK.

RGB is more popular because most consumer and prosumer output devices require it (and, of course, if your outputting for screen RGB is a necessity). RGB has a wider color gamut (more available colors) than CMYK, but if CMYK is your primary output target this could actually be a liability.

And converting from CMYK to RGB is much less problematic than the reverse.

For some images, even if I'm going to work in rgb, I'll send a copy into cmyk. Often you find various colors to be more isolated on one or another cmyk plate. The lack of a black separation in rgb means that each plate simply has more overall information, and hence, less specific colors to focus in on. Further isolating colors in cmyk using calculations in subtract mode is usually far more successful than in RGB.

Also, keep lab in mind. The fact that the a and b channels so specifically target blue/yellow and red/green (well, magenta/green) allows the blending sliders to separate or remove specific colors much more effectively than either rgb or cmyk. If a red region requires masking, removing green from the a channel and blue from the b channel can leave you with your red object isolated in about 10 seconds. Adding an empty layer and then merging the two yeilds a transparency selection that, when moved back to your original file, works easily as well as a mask drawn by hand, or extracted from a channel.

okay, you lost me on the lab colour stuff...
I don't really understand what lab colour is or how to use it.

tutorials?

thanks for the other input, both of you.
k

I work for a printier and he uses cmyk for the printing end of the business for plates and rgb for the color copier. We have an epson geclee large formate that takes rgb better that cmyk. Just my two cent. When I started designing pictures it was in cmyk but the xerox tec told me to use rgb. :nod:

L*a*b* Tutorials:
http://www.dgrin.com and go to the thread: "The Digital Darkroom - Software"
They are taking Dan Margulis' book on working in L*a*b* chapter by chapter and having a discussion on it.

http://www.ledet.com/margulis/articles.html
go to 'Books' and there is Chapter 2 and Chapter 9 for you to check out if you like.

Some people like Dan's writing style, some don't. Some agree with him, some don't.


Blanket statement:
Each color space L*a*b*, RGB, and CMYK have their own strengths and weaknesses to varying degrees.

A few years ago, Don Hutcheson 'converted' me to working in RGB. He had a series of sample files and he made me a believer that RGB was a better general working space than CMYK.
There are color and tonal adjustments possible in an RGB working space that you can't achieve if you try to do it in CMYK.

We ultimately output to CMYK but I do the majority of my work in RGB, then I'll finish out some fine tuning in CMYK.

I've been tinkering in L*a*b* for a little while and find it quite impressive.

thanks for this, Richard.

lots of food for thought.
I get what you're saying about rgb colourspace, it's just that besides being counterintuitive to me (which I'm sure I could train myself out of), I actually work in repro, so I am dealing with images that have alredy been converted from rgb to cmyk. Much of the time I am deep-etching part of an image that I have already colour corrected. Hard to explain.

The final output context is in a magazine; two images are superimposed - one deep-etched, one not, so that the above image (the deep-etched one) has ares that stick out around the frame of the lower one.

Am I making sense- it sounds very garbled to me.
The point is that there are two images - both must have the same colour.
I could convert both back to rgb and then do changes, but will I actually be gaining anything? (Images originally rgb from scans or digital photos)

thanks for your help
k

If it for a magazine they usually work in cmyk. When I first started at the print shop I worked in rgb and sent work to the back room and was told right away that it must be in cmyk.
ydelle

so now you work in rgb and then convert to cmyk?
when do you make colour adjustments?

One can make adjustments in both color modes assuming you are working in a color managed workflow and have accurate monitor and device profiles. That last part can't be emphasized enough.
RGB first, then convert to CMYK and check your CMYK numbers in the questionable areas and adjust to taste.

This link is good reading for those on the fence deciding whether RGB or CMYK is a better general working space. Click on 'RGB Arguments' for the full PDF download.

http://www.hutchcolor.com/CMS_notes.html

If I know it is for print I start with CMYK. if it is for the epson 9600 I work in rgb.

If i have a jpeg, how can i identify if it is a RGB or CMYK? Is there any tool available which determines it?

Typically a JPEG that comes from a camera is sRGB or RGB. It's just what most pro-sumer cameras default to.

I do prepress work and always work in RGB and save my work as a Photoshop RGB file. I will then convert to CMYK as the final step and send a CMYK TIF file out for print. If doing my own page layout, I just drop the PSD file into InDesign and set the print options to convert all RGB to spot when it goes to RIP (or I will set that option when preflighting for an InDesign package). I soft proof on an Epson 900N or Epson 2200 and get very good results.

As far as letting Photoshop handle the RGB->CMYK conversion, it may depend on the graphic. If it has a lot of out of gamut areas, I'll do some of the adjustment myself with HSL and curves/levels. If the out of gamut areas are not prominent or in non "essential" areas of the graphic, I just let Photoshop do the conversion. Either way, those "hot" (out of gamut) areas are going to get converted. It's just the nature of 4 color process printing.

For a simple explanation of color conversion see my video tutorial in the RetouchPRO tutorials forum.

praveenbr, welcome to RetouchPRO.

most graphic editors will have some sort of identifying tool. in paint shop pro it's called 'info'. in photoshop, i'm not sure, but i'm sure it's there somewhere. these will tell you all sorts of information about the image, like image size, image type, what color space it's in and so on. if you're using photoshop, i'm sure you can find it or someone can direct you to it.

you could also use something like irfanview as well. even windows explorer will give you some info if you hit 'properties' on the image.

Why not see what does the Photoshop Help Center say?
Just one quotation from "Correcting in CMYK and RGB" page:

If you must convert your image from one mode to another, it makes sense to perform most of your tonal and color corrections in RGB mode and use CMYK mode for fine-tuning. Advantages of working in RGB mode include the following:

- You can save memory and improve performance because you are working with fewer channels.

- The range of colors in RGB spaces is much larger than the range in CMYK spaces, and more colors are likely to be preserved after adjustments.