Hello,

I hope you guys can help me. I recently made an ad in adobe illustrator
which has a greyscale image with an orange to transparent gradient on
top made by creating an opacity mask. Well on the screen it looks fine
and so I sent it to the comercial printer. However when the comercial
printer gave us the proof it shows a sort of lighter hue band in the
area where the orange gradient ends and meets with the grey image
below. In other words, instead of the gradient blending smoothly with
the grey image below (as it shows on screen!) there is like a lighter
strip that just almost ends abrouptly and looks bad where the gradient
ends.

This same thing had happened to me with almost the same ad in the same
place when I took it to Kinkos to print. However this time the gradient
and the greyscale image were one rasterized bitmap image exported from
photoshop. The guy in kinkos was telling me that it is a problem that
appears because the image has an area of just k and sudddenly it goes
to color and it not always goes smoothly. Well I told him that in my
printer at my office that didn't come up, but he told me again about
the k to color problem which was odd to me because this time it was
just one image like a photo so I don't see how that could be my
problem. Anyways I suggested he try printing it from another printer
there at kinkos and he did and this time it came out without any
problems. So that solved it that time.

But my current case is that the proof from the comercial printing is
showing this band again, so my question to you all is this: Is this a
problem that the Comercial printer should solve or is there something I
can do to correct this in the illustrator file?

Any help/suggestions will be really appreciated.

Thanks,

Aluxe

stuart b. henlis

Aluxe,

A former friend of mine used to work at a Kinko's in Philly as a "graphic
designer" there. He still is a very talented designer. To get to the point,
he would tell me about the total lack of skill of the "designers who worked
with him. He even told me that Kinko's was for amateurs and off the street
work. So, Kinko's is not a commercial printer. It's nothing more than a
glorified quick printer. In fact, UPS has taken them over and done away with
most of the printing. Please take your work to a real commercial printer.
They know what they're doing.

Good luck for the next time,

Stu

mike russell

It would be helpful to know what type of printer it is. Is it a laser
printer? And it would also help to know what type of file you are
providing - CMYK? Third, are you using a custom profile to generate the
CMYK file, or leaving this to Kinkos.

Few people would provide even a guess without knowing both of these, but der
Curvemeister loves to step across the plate and swing at pitches like this
one, so here goes.

If the answer to both of these is yes, then look at the K plate and look for
a loss of density or other banding at the point in question. Since the
problem involves orange, look carefully at the Y plate - monitors are poor
at representing light, saturated yellows. Whichever channel the problem is
in, a good way to hide the transition is by adding a strategic amount of
noise.

Anopther possibility: custom profiles are a famous source of gratuitous
banding - get hold of the profile in question and run it on some gradients,
and see if it introduces its own banding.
--
Mike Russell
http://www.curvemeister.com

I think my story was kind of confusing, the proof I got was from a real
comercial printer and I mentioned Kinko's because this same thing had
happened to me about a month ago also with Kinkos. But my current
problem is happening with a real comercial printer, so is that
something they should fix?

Mike thanks a lot for all your suggestions. About which printer it is,
I don't know, it is a comercial printing not Kinkos. Like I was telling
Stuart I only mentioned Kinkos because the same thing happened to me
about a month ago at Kinkos, however my problem now is coming from the
proof that I am getting from the comercial printer.

Now as to your questions, the orange gradient I am making is a pantone
color and the bitmap image below is cmyk. What you are suggesting about
checking the K plate, is it something I should check in adobe
illustrator or something that the comercial printer people should be
resolving? And could you explain me a bit more about the adding noise
part? Is this something I have to do or the printer? Thanks very much
for all your help.

mike russell

My previous discussion assumed a CMYK image. I think the problem is
with the spot color.

If I understand correctly, you are overlaying a semi transparent spot
color gradient on a CMYK image.

This gets complicated because the behavior of the spot color when
printed on top of the other colors is unknown. For example, it depends
on the opacity and dot gain of the spot color.

I don't think the proof is reflecting reality, but lets see what others
think.

leeb

Not being an Illustrator expert it's hard for me to understand this. You
have an orange to transparent gradient? What does this mean? It's opaque
at one end and becomes transparent at the other? That implies some sort
of graphic mobius strip. At what point does it cease to be transparent?
How would the program decide where to put that point? Do you mean you
have an orange gradient on top of a grayscale image that is transparent?
Like an orange that overprints a grayscale image?


Ah, I think the fog is clearing. You want the grayscale image to fade
into the orange blend, decreasing density as it approaches the solid
part of the orange blend. Am I right? And then the orange blend
overprints the grayscale?

But one issue I'm not clear on. How many colors is this job supposed to
be? Is it a two color job with Black + Orange? Is it CMYK? Is it CMYK
+Spot? These details are very important in how to address this problem.


What kind of printer was used? Was one a Postcript printer and one not?
That would created very different results since one would use the screen
drawing mechanism and one would use Postscript. A very common problem we
experience with people who own inkjet printers but don't invest in
Postscript rips. Keep in mind the world of commercial printing is based
on Postscript/Acrobat. No devices in commercial printing use the
computers windowing mechanism to render images.

Iluustrator warns about using spot colors with transparency but it has a
check box that says "don't show this dialog again" which most people
check and then forget about it.

The solution to your problem (or statement that there is no solution)
requires that we know about the color of the do***ent.

1) Black plus spot
2) CMYK
3) CMYK plus spot.

I would have to know the answer to really know what to tell you.

Even how to proof it depends on the answer to that question. You might
be experiencing a problem where a CMYK proofer is trying to simulate
spot colors and the final output will be fine. I don't know.

How many colors is the job supposed to be?

tnsmith44

Well, I'm far from being an expert, but the first thing I would check would
be the color profile. You might have to try a couple of different ones. If
you work with a regular printer, check with them for recommended settings.
It's best to work with one or two printers so you know how they work. No two
bureaus are going to print exactly like each other.

You might also add just a touch of gausian blur after you build the
gradient. That tends to soften the transition and reduces banding,
especially if you're going to a small laser printer.

Skywolf.

_+arrooke

I would guess that the proof (method of proof) is where the problem lies. It
sounds like the digital proofing used has limitations as to fine degrees of
output, specially so far as converting the gradient to such a light colour.
I would bet that final output from RIP (for press) will be ok. Just my 2
cents.
Keith.

tacit

I think I know exactly what you're talking about. If so, what you are
seeing is an optical illusion; the "stripe" at the end of the gradient
appears because the gradient ends abruptly and the human eye has a trait
called "lateral inhibition," which exaggerates edges in a way similar to
the way unsharp masking works.

I have had similar problems with Quark and Illustrator created gradients
which end at an image. The solution? Create the gradient in Photoshop,
which will add a small amount of noise to dither the gradient, which
helps get rid of the appearance of the bright "stripe" or line where it
ends. (Your inkjet print does not shop this optical illusion in part
because it, too, uses dithering to produce color.)

--
Art, photography, shareware, polyamory, literature, kink:
all at http://www.xeromag.com/franklin.html