logo tiffs printing badly when using quark and PDF

[MarkXero]

Twice now I have received print jobs back with errors where a logo is placed over a solid block of colour. I'm using Quark XPress 6.1 on PC. The logo is basically a TIFF of white text, and it is placed over a variety of different coloured background boxes.

The first time we got an error, the logo was a greyscale TIFF (white text, black background), coloured in Quark to match the same CMYK spec as the box behind it.The background of the Quark frame was transparent.The second time, in trying to avoid the problem, the logo was a CMYK TIFF with the background colour matched to the same CMYK spec as the background of the box behind it. Again, the background of theframe was transparent.

The first time was a litho print job, the second was digital. In both cases I generated a PDF using the inbuilt Adobe PDF setting of Quark, but obviously with relevant settings, and in both cases the error was not apparent on the PDF. It alsoappeared as though the error was more to do with screening (? which I must admit I don't understand very well) than with actual colour settings. That is to say, under an eye-glass the error appeared to have more to do with the print process than the inks used.

What really confuses me is that in both cases, similar jobs sent at the same time to the same printer using the same TIFF file placed in the same kind of frame and output as PDFs using the same settings printed absolutely fine.

Has anyone else come across this problem? Firstly, any explanations? Secondly, any solutions? Placing a CMYK TIFF on a matching CMYK background box in Quark is a fairly staple part of a lot of my design work and it worries me that I can never be totally certain that the colours will match.

[PrintDriver]

Pardon for asking, but why are you using the logo as a tiff and not an eps? Logo not vector?

PrintDriver is a grande format digital print dude (bigger than a proofer, LOL). His advice/opinions may not apply to the 4color/offset/web world of printing

[defjoe]

it's missing a clipping path and are you assigning the path in quark?

that NEVER works.

'I will become the most powerful Jedi ever!'
'I'm the damn designer, bitches!'

Check out my indie comic book!
www.assassinsguild.net/ (http://www.assassinsguild.net/)

[Jason Fraker]

^^ what they said. You should build your clipping path in photoshop, and save the tiff with alpha channels, then use the alpha channel in quark when you specify clipping.

-Jason

A lie can travel halfway around the world before the truth can even put on its shoes.
-Mark Twain

[Broacher]

Hold on--- why would you create a clipping path for type? You might just as well reset or recreate it in pure vector in AI or whatever. I'm assuming that's why we're talking raster approaches here-- that is, there isn't a vector version available.

Now with greyscales, or cmyks, you are always going to get screened type edges because of halftoning, no matter what the resolution of the graphic. This means if the piece is produced at lower line screens (such as in a newspaper) the edges on small type really suffers.

But, if this logo is pure line (one colour solid, no gradients or feathering), and you don't have a vector, then I'd use a high rez (2400) black and white and spot colour it to 'Paper'. BW files are self-masking and do NOT suffer from halftoned edges. They're also competively very tiny-- even more so when embedded in a PDF which use CCITT4 compression.

[MarkXero]

Thanks all for the replies, but the problem isn't the white type. I'm not making the logo background transparent; I'm creating a normal rectangular TIFFwith aCMYK blue colour as the background: so all round the edge of the TIFF is the same blue. I'm then placing this TIFF in an image box over a bigger,empty image box with the same apparent CMYK background colour.When I output the PDF, you can't seethe edge of the TIFF, but I'm getting a difference between the two blues when the job comes back from print.

[Jason Fraker]

If it were spot color, I would suggest that you used an uncoated PMS color in one part and a coated PMS color in another. I have had similar problems in Quark, and I ended up just making a clipping path and not trying to do what you are trying to do. I hope someone here can help you more than I can, but I think I just gave up...

-Jason

A lie can travel halfway around the world before the truth can even put on its shoes.
-Mark Twain

[Broacher]

The black and white TIFF tagged white would give you more flexibility, better output, and smaller files size.

But as to the problem with the mismatched blues--- sounds like a colour management problem. Are you using profiles anywhere along this line?

[MarkXero]

I don't think it's colour management. Under an eye-glass, the flat blue in the background box appears just that - completely flat. But the area of the TIFF which is meant to be the same blue appears lighter because it has 45º thin white lines across it - hence my assumption there was some kind of screening issue. But why would it appear on one job and not on another when both use the same TIFF and the PDFs are produced with the same settings? And how on earth do I stop it appearing on a future job?

It concerns me because even if I get hold of an EPS of the logo to use on future work, how do I know another CMYK TIFF on another job won't do the same thing - a map, or graphic or suchlike? I'm trying to understand what has caused the issue in the first place. Neither of the two printers who produced the jobs can offer me a decent explanation as to what has happened - it's all been guess work and 'might-be'.

[Broacher]

I've got a bell ringing 'PDF' in my head. Specifically, a flattening issue. Certain things can create strange output situations. For example, if you try to create a PDF with a page containing a GIF with transparency (I'm just using this as an example-- don't ever try this at home!), Acrobat 'interprets' the transparency by slicing the whole image into pixel wide 'slices' and applies little tiny clipping paths to each slice.

I've seen this happen in certain combos of other software and Distiller too. Here's a way to check for that scenario: zoom right into the edge of that trouble placed TIFF, and with the Object Touchup tool (it's usually hiding right under the Text Touchup tool), carefully select an edge. If this thing is 'sliced' you'll see it right away as a tiny selection-- otherwise, it's the whole bitmap.

If that's not it, I'd like to take a peek at one of these troubled PDF pages in my own 'Broacher's Home for Wayward PDFs', if you're so inclined to email me one.

[MarkXero]

Sadly I've only got Acrobat 6 standard, and the Object TouchUp tool is one of the ones they've helpfully put into professional only, so I can't check for the slicing. I'm more than happy to send you a copy of the PDF to look at – so long as you don't mind receiving an 11MB file. I'd recreate a smaller PDF for you, but I have a sneaky suspicion that a different PDF might not have the same issue; particularly if this is as quirky an issue as I think it is. If you still want to take a look let me know an address to send it to.

[Magnus]

This has happened to me before as well. We print cards out digitally on our Xerox with a logo on a 'black' backround...to make bleeds for cards I used a larger box in quark of 100% black. When printed, you can see a definite difference in each of the 'blacks'...I think the problem occurs when the RIP on the printer gathers the information on the 2 colours, if they are different in anyway then there is going to be trouble. I found that increasing the colour of the box around the logo in Photoshop eliminated the need for the box in quark...thus all the colours were coming from the tif file, and not the tif and quark.

Also, why are you converting to pdf first then outputing?

"It's not cheating if you win."

- Bill Mears, 5th Dan Muso Jikiden Eishen Ryu Iaido

[Broacher]

Black mismatches are more common-- especially since not everyone realizes that most 'rich blacks' use more than just black ink. My guess is that logo's black was a different mix than that 100k black you used in the Quark frame. (One of ID's niceties is an eyedropper that can 'pickup' the true black in your image's background and quickly apply it to a frame.)

[Jason Fraker]

MarXero, couldn't you just crop the problem area of the PDF and email that to good old Broacher? Surely that wouldn't be 11mb.

-Jason

A lie can travel halfway around the world before the truth can even put on its shoes.
-Mark Twain

[Broacher]

Jason, nice idea-- but it ain't that easy. Cropping a PDF does not 'throw out' anything-- easy to prove by comparing the cropped file size to an uncropped file size. (Wish my bank account worked that way)

[MarkXero]

Well, I tried. Both the uncropped 11MB PDF and the cropped 7MB PDF bounced back. Thanks anyway Broacher.

[Broacher]

Mark,

While a simple cropping won't reduce the file size, but (and I'm not absolutely sure on this), I think Acrobat 6 Standard will throw away stuff if you do a 'SaveAs' to a new filename with Optimization on. Worth a try. And if that doesn't work-- you could try 'refrying' the cropped PDF by 'Printing' it to a new PDF, or even Saving the cropped version as an EPS and redistilling that.

I guess you don't have AI or CorelDraw then, eh? If you did you could likely open up that PDF directly to peek at its innards (or most of them).

[thelobst]

Instead of using a cmyk tiff from photoshop with the same cmyk values, covert the file to black and white and assign the colour of the box within quark.


When you import the greyscale tiff into quark you can change the black to the required blue. Make sure that the black is 100% black before you import. This should remove any colour differences as the colour is not in the original tiff


edit: just read your post again and you already tried this.

[Broacher]

>>covert the file to black and white and assign the colour of the box within quark<<

If you convert without any adjustment for resolution, you'll get very blocky edges. To use black and white this way you should start with a high, very high resolution -- typically 2400 dpi original or vector intepretation (through opening an EPS in PShop), and, assuming there's no gradients or tints, one without anti-aliased edges.

[imagemaker]

Had this problem on a digital printed billboard. Though the Quark background color of the banner was the same CMYK mix as the background for the placed logos, you could see where they were ‘mortised’ in by a slightly different hue.

Out of curiosity-

Have you tried printing the same CMYK mix independently from PhotoShop, Quark and Illustrator? I mean a 2'x2' square, in three different documents.

How do the colors compare?

How about all three 'mixes' out of each program. In other words, place the Illy and PS files into the Quark file and print it. Then put the Quark and Illy into the PS and print from PS, and finally put the Quark and PS into Illy and print that one. You'll have to export the Quark color to a .eps or .pdf, but you can still get an idea whether the three programs are interpreting the same. Is there color consistency?

The other thing to do is to adjust the color in PS to match the color Quark is printing. One may need to a 60 /30/10/5 mix and the other may need to be 63/32/8/4 to look the same. Look at a color chart from the printer to see EXACLTY what one mixture looks like, then adjust the other mixture to match.

Ingredient wise, we may all be cooking the same cake, but if my Land o’ Lakes butter is a little different than your Challenge, I need to do something to compensate in order for both our cakes to turn out the same.

Theoretically, we shouldn’t have to deal with this kind of stuff, but there are just WAY too many variables.

--Ciao for Niao--

Ethan

[MarkXero]

Thanks again all for your replies. I've still not really got to the bottom of this, but seeing as how the TIFF, QXD and PDF have now gone round the world to the techies at Xerox, Pindar and a heap of other places without a lot of success yet, it looks like it's not going to be as easy to suss out as I thought it might be. The sensible money is on a Quark > PDF issue, but that seems to be as far as I can get.

One thing it has thrown up is a difference between printing to PDF from Quark and exporting layout to PDF. Apparently exporting creates a better PDF as printing can sometimes miss elements (Or something?!), although I'll hold judgment until I get an exported job back from the printers.

Team Xero - invisible to radar (with any luck) (http://www.teamxero.com/xerum)

Post Edited (MarkXero) : 11/29/2004 3:11:31 PM GMT