Preparing RGB-scanned B&W images for print

This is a bit of a narrowed-down redux of a post I submitted a few weeks ago (Thanks to Mike Russell for responding before; some lingering questions remain)

I am preparing a CD insert layout in which I am including several images shot in B&W (on film) and scanned in from prints. (Resolution is OK at the effective enlargement I am using them).

The scanned image files were delivered to me as RGB TIFF’s. At this point, I simply converted to CMYK in PS-CS2, did some rudimentary contrast/brightness adjustment, scaled to the proper layout size and imported them into the Illustrator layout.

But I am somewhat concerned that the CMYK "color build" of the images is not ideal for print;the separations are not the same, and the brightness/contrast adjustments have left some images with a subtle hue shift relative to others. Additionally, close examination of some images uncovers some subtle and not-so-subtle color shifts or certain details, e.g guitar strings that end up with a bronze tinge in a *black-and-white* image, subtle rainbowing at high- contrast edges, etc.

So, here’s today’s question: should I smash these images to greyscale to remove the color artifacts and then rebuild them so that the image’s CMYK separations track proportionately from white to my CMYK black? Or is there some better approach that can be recommended as a generic formula to apply to black-and-white images intended for full-color offset printing?

So, here’s today’s question: should I smash these images to greyscale to remove the color artifacts and then rebuild them so that the image’s CMYK separations track proportionately from white to my CMYK black? Or is there some better approach that can be recommended as a generic formula to apply to black-and-white images intended for full-color offset printing?

The reason for the conversion is this:

Converting from RGB to Grayscale and the back to RGB will produce a neutral RGB. That is, each channel will be the same. This assures a pefectly (theoretically) neutral RGB image.

BUT…

There’s one gotcha here.

When converting to CMYK from the RGB image you want to use heavy GCR. This will make your black printer the fullest and save you the grief of gray balance shifts on press.

This is much more important than you might think. I’d suggest exploring how you can get MAX GCR.

Without a heavy GCR your black printer will stop short of the highlight. (Skeleton black.) This will mean that your cyan will extend further to the highlight than the magenta and yellow. You’ll get color fringing this way when trying to print a black and white CMYK image.
Your pressman will love you and commend your expertise when printing the images this way.

wrote:

So, here’s today’s question: should I smash these images to greyscale to remove the color artifacts and then rebuild them so that the image’s CMYK separations track proportionately from white to my CMYK black? Or is there some better approach that can be recommended as a generic formula to apply to black-and-white images intended for full-color offset printing?

You should convert the RGB to Grayscale then then back to RGB then convert them to CMYK.

Since (I’m assuming) the end result will be a CMYK printing job there is no real advantage to having them in Grayscale. They’ll print better as CMYK.

The reason for the conversion is this:

Converting from RGB to Grayscale and the back to RGB will produce a neutral RGB. That is, each channel will be the same. This assures a pefectly (theoretically) neutral RGB image.

BUT…

There’s one gotcha here.

When converting to CMYK from the RGB image you want to use heavy GCR. This will make your black printer the fullest and save you the grief of gray balance shifts on press.

This is much more important than you might think. I’d suggest exploring how you can get MAX GCR.

Without a heavy GCR your black printer will stop short of the highlight. (Skeleton black.) This will mean that your cyan will extend further to the highlight than the magenta and yellow. You’ll get color fringing this way when trying to print a black and white CMYK image.
Your pressman will love you and commend your expertise when printing the images this way.

In comp.publish.prepress Lee Blevins wrote:

: There’s one gotcha here.

: When converting to CMYK from the RGB image you want to use heavy GCR. : This will make your black printer the fullest and save you the grief of : gray balance shifts on press.

: This is much more important than you might think. I’d suggest exploring : how you can get MAX GCR….

Thanks. This is just the sort of information I needed — and it makes plenty of sense! The gotcha you describe is just the sort of problem I feared.
Burrowing down into the Photoshop options for conversion, I notice there is a "heavy GCR and a "Max GCR" conversion; the latter appears to leave no CMY gray component at all (at least by the curves in the "custom CMYK" window). "Heavy" starts to replace with K at about 25% and crosses over at about 50%, while the default "Medium" has the breakpoints at about 45% and 85%. You mentioned both "heavy" and "MAX" above
So which one is more appropriate (given you haven’t seen the images)? My intuition would say "heavy", but I’m not the prepress expert. Actually performing the conversions on the pre-desaturated image shows the "MAX" setting leaving nothing but the faintest whispers in the CMY channels (probably all of the color artifacts I complained about 😉
I also left the dot gain curves alone, and assume I should do so unless otherwise advised. Dot gain is set to 20%, with the curves showing C pushed to 50% >> 74%, versus 50% >> 70% for the others.
Bob