RGB images and shadows

[rich apollo]

Working with a client who wants to keep their imagery RGB. "Great," I say.

Doing the conversions isn't a problem, but all the images are product shots in which they've removed the backgrounds and built synthetic shadows. I suggested building the shadows in InDesign using a shape and the feather effect. The client wants to keep the images and the shadows together. I, obviously, want K-only shadows for stability on press.

Attached is an example image.

[Verdant]

As long as they supply you the layered PSD with the shadow on it's own layer you can easily go into the file, select the shadow layer and then make a hue/saturation adjustment layer. Select "Colorize" and then pull the saturation down to 0. This will make the shadow use black ink only. You can verify this by looking at your info pallet while hovering over the image.
If they specifically wanted the shadow to be warmer/cooler you can bring the saturation up a point or two and change the hue slide to add in some cyan or magenta.

[Lukas Engqvist]

Well it realy depends on your ICC profiles if 0% saturation is k only (a Maximum UCR or GCR is necessary for that to happen). It is a tricky situaton, flexibility for the designer versus the stabilty of press. I don't think that press stability is an issue, the shadows will seem neutral enough, compared with the objects.
It is possible to do a CMYK with a K shadow for an RGB smartobject. This is the closest you can get. It is important to change to the right ICC profile in CMYK, but reassigning or changing CMYK profile will keep the K channel intact.
It is a special case, and does require extra work.

[kyle]

Add a spot channel for the shadows. Just be sure to not name it "Black" or "Process Black," because Indesign interprets these as being process black instead of a spot ink (useful for duotones), and won't place the file. The special names are case-sensitive, so you could name the ink "black." Care must be taken to make sure there is not unwanted overlap of the images because the spot channel is completely independent of the composite channels. It could be helpful to make the color of the spot channel a bright saturated color at some point in the editing process for previewing all channels simultaneously. If the color of the ink is changed back to black, Photoshop will automatically rename the channel "Black," so you have to change it again. If you have a lot of existing images that are RGB, you could probably make a Photoshop action to convert in a batch process if the shadows are all on their own layer and the layer is consistently named.

Before placing any of the images with the spot ink in Indesign, add a swatch to Indesign named "black" or whatever ink name is used in the images. Set it to be a spot swatch, with the alternate color space CMYK with values 0/0/0/100. This prevents Indesign from picking up the color of the spot ink from the links, which would be defined as Lab or RGB and wouldn't convert to straight black. If any images are already placed and you don't want to start over, create a new 0/0/0/100 spot swatch and map the ink from the images to it with the ink manager. You cannot convert the spot to process or alias it to process black and keep the spot black-ink-only, as Indesign will substitute the alternate color embedded in the image which at best would be RGB or Lab black.

Export a PDF with the spot ink preserved, then use Acrobat's "Convert Colors..." to convert the spot to process, or do it in your RIP. If you created a spot swatch before placing any images, the alternate color will be picked up from what you defined in Indesign (0/0/0/100), and Acrobat will convert the spot ink to be black ink only. Even though the image is encoded as a single NChannel image, Acrobat seems to do a decent job of combining the two blacks into one. This only works if you are converting the RGB to CMYK when exporting the PDF. If the composite image is left RGB in the PDF, there will be an RGB image with an overprinting spot image on top instead of a single NChannel image, and Acrobat will convert the spot image to CMYK with empty cyan, magenta and yellow. This will render the overprinting attribute useless, since the image now has cyan, magenta and yellow (even though they are empty). Your only chance then would probably be to change the blending mode of the shadow images to multiply with Pitstop, or try doing the spot conversion in your RIP.

This is all based on CS3 and Acrobat 7 behavior. Some of the workarounds may not be necessary with CS4 or a newer Acrobat version.

Alternatively, if the shadows are on consistently named independent layers, you could probably use a Photoshop action list to convert everything to a particular CMYK profile while making the shadows straight black.

[mike.egloff]

Let me pre qualify this,
This will only work if the silo is present
Make a selection of the silo, inverse it, copy to clipboard
delete the background, convert to cmyk
select the black channel, then paste

[disbellj]

Has your pressmen ever printed 4C drop shadows? After seeing how they looked printed, I quit caring if they were 4C or 1C black.

Regards,

Don

Working with a client who wants to keep their imagery RGB. "Great," I say.

Doing the conversions isn't a problem, but all the images are product shots in which they've removed the backgrounds and built synthetic shadows. I suggested building the shadows in InDesign using a shape and the feather effect. The client wants to keep the images and the shadows together. I, obviously, want K-only shadows for stability on press.

Attached is an example image.

[DonRGB]

Rich,
Don Isabel is 100% right. I abandoned K-only drop shadows 30 years ago when I saw how much more realistic 4-color shadows look.

So long as the shadow consists of equal RGB values (and the image space is sRGB, Adobe RGB, ProPhotoRGB, DonRGB4, etc.) then any G7-based profile should produce a near-perfect neutral CMYK shadow. IDEAlliance's free GRACoL and SWOP profiles use high GCR so there's no danger of shadow balance drifting due to ink trapping. You can build an even higher GCR profile from the GRACoL/SWOP data but it's not necessary.

If your press does drift unacceptably in drop shadows, your dot gains and/or G7 balance must be way outside normal tolerances, and the main image areas will also be unacceptable. However contrary to the main argument for k-only, a 4-color drop shadow is LESS likely to draw attention to the small gray balance variations unavoidable in every press run, because the shadow mimics exactly what's going on in the main image. Conversely, because a k-only shadow doesn't drift the same as the main image area, it stands out like a red flag (gray flag?) and is MORE likely to draw attention to otherwise acceptable press variations.

FYI, years ago my mother (not in the trade) asked me why there was a different screen pattern in the k-only shadow of a magazine ad compared to the nice rosette in the tire. The screen pattern alone made the shadow look fake, never mind gray balance. Plus, the black plate was slightly out of register which left a white line between the tire and the shadow. Neither of these would have been problems if the whole image was in cmyk, but back then 4-color shadows were virtually impossible.

Sounds like we need another canyon-trip.

[apepp]

...indeed, k only shadows great for mono printing, in four color printing they can look naff...

...four color black and white images can look much better than black ink only images...

[tmiller_iluvprinting]

I agree with Don I. and DonRGB. Our pressroom and clients would be surprised if they knew how many 4 color drop shadows we have printed in the last 8 years! Long live the RGB workflow! Thanks for all you've done DonRGB!
Best regards,
Todd

[Bret Hesler]

A very dangerous way to make this happen is a device link with R=G=B to K-only (dangerous because any pure neutral pixels in the product will also go k-only)

I will say that I am in favor of a 4/c leading black shadow.

Bret