Forum Moderators: not2easy
The project involves creating images for a commercial printer who wants customers to design things like business cards online. I've got the structure of the system in place and it "works", but my client says the image quality is poor. Here's what my code does: I create a main canvas image at 350x200 pixels, which is based on my working principle that 100 pixels per inch is the same as 100 dpi and we have a 3-1/2x2 inch target. (This concept seems to be wrong, but I can't find a different/better method online.) I fill the 'canvas' with a base image and subsequently draw True-type text on it based on user input. At the end, I save the image to a file.
As I say, my code "works", but the client reports low quality results from his 4-color commercial printer when I try to increase the image quality for the final output. I do this by multiplying the canvas size and other image attributes by a scaling factor of 4 to a target of 400 pixels per inch as the image is created. It looks great on my screen, but the client isn't happy with the results. I'm beginning to suspect we have a resolution/dpi/ppi mismatch, and that his printer or other intervening software is attempting to scale the image on its own because he says the text comes out fuzzy, and its razor sharp at my end when I view it at full size. By that I mean, he says the printer outputs at 400dpi. So, I give him a 1400x800 pixel image on the assumption that it should print a 3-1/2 by 2 inch image with good quality. And that's where I'm sure I'm going wrong here.
I've tried outputting to different file formats, but my knowledge there is almost nil. Customer suggests using .eps so that scaling of the text is not a real issue. I haven't been able to test that much. Mostly what I'd appreciate is some pointers on the overall approach. Am I OK figuring pixels per inch is equivilent to dots per inch? I can live with high quality .jpg's if I can make the final output work, but any advice on creating .eps files with Image Magick would be really great as well. Thanks a bunch!
I can sympathize with both you, and the printer. As the manager of a pre-press team, I have to deal with my share of low res JPG’s, and let me tell you, it’s not real fun. First off, if you are using Image Magick, I doubt that it is separating the black text from the rest of the “image”. Meaning that the black text will print in all four colors, resulting in a minor “blurred” effect if the printer can’t keep his registration true on the press. Another thing is that an image (Yes, even at 400 dpi) regardless of how it looks on the screen, 'would be' vectors (such as text) will not come out looking as clean as they do on your monitor because they have been rasterized. I suggest you look into PDF output with Image Magick or the GD library (Not sure if either supports PDF though). What an EPS/PDF does is save the fonts as ‘vectors’ rather than ‘rasters’.
So I think you may have one, or both problems:
1) Image not separated out in CMYK, rather left in RGB making the text not just ‘black’
2) Texts aren’t vectors, therefor succumb to the same rules as basic images.
Personally? I would use the image given as a ‘template’ and actually take the time to put the card file together in a worthy program of a printer. Ie. InDesign, Quark, FreeHand, Illustrator, etc. Trust me, the printer and the client will be much happier!
-- Zak
I am very grateful for your response. I can get Image Magick to output CMYK seps on the .jpg's, and have left a switch in my code to turn it on and off. But the results are the same in terms of sharpness. I know the text is a raster image, and that's why the client is pushing for .eps files, but I haven't had any luck with them either. I intend to work on that a bit, too, but I'd be really grateful for some references to read on dpi issues in general from the printer's perspective so I'm working from a better foundation. I need to know if a 1400x800 image will actually come out at 3-1/2 x 2" on a 400dpi printer or if I'm just way off the beam here. Thanks again!
-- Zak
I'm not sure if you've solved your problem yet, so I figured I'd offer these additional comments.
As Zak pointed out, using the web interface for the template of the design and using a worthy program for the final output file, I think, will give your printer the best results.
Some years back, I put together a web interface for the company I was working for. Basically, it let the web user select the specs (i.e., type size, font, print area, etc.) they wanted. It probablly wasn't as robust as the impression I'm getting about yours. But, it was capable of giving the user an idea of how the layout would come out looking with regards to how the text would fit in the print area.
What I did, at that time was, I wrote a vb5 program that referenced the pagemaker 6 library and translated the design instructions to pm commands and had pm create a file with text, specs and positioning info. Then it saved it as a pagemaker file and printed the output for production. I took advantage of the scripting capabilities of pm. So it went from the users input to the file and printer without any interaction from me or anyone else.
Illustrator and Corel are vector programs in contrast to raster programs that output only bitmap formated files. These vector programs (and probably many others) also have scripting capabilities. It shouldn't be too hard to pick up the scripting in Illustrator and create complete designs using text and placing graphics then saving it to an eps or pdf file. All within your own program or script for hands-off processing, to yield the output files for the printer.
You don't necessarilly have to separate the cmyk files for the printer for full color processing as long as you only use cmyk color in the design. Output the file to the proper ris driver that your printer uses. Currently I output to the Scitek ultra bridge 3 ppd for one of my printers for full color printing.
