preparing artwork for photopolymer

I know this might tick off the purists here but I feel like I’ve got a missing link. I need a tutorial or online class to help me take my artwork from sketch to ready to send off to be made into a photopolymer plate. I know it’s complicated and really is a job for a designer but…..there has to be a simple step-by-step way for me to learn to do this! My sketch book is full of great ideas, I have Illustrator on my laptop, I have a tablet - and I keep beating my head against the wall trying to make it happen. Can anyone shove me in the right direction?
Thanks, Mb

Log in to reply   30 replies so far

There’s three ways to go about it, each are easy to do badly and once you understand the basics, just practice, google and fill in the blanks of how to do them better:

1) Raster. Scan the image you have in your sketchbook, import it to photoshop and transform it into a bitmap (image->mode->bitmap.) It’s now ready for output, although most printers will prefer AI, EPS or PDF files. Imagesetters usually rasterize images do 2880x2880dpi, so if you scan it to less than that you’re sacrificing resolution. 300dpi is not ok no matter what anyone tells you.

2) Auto-trace. Plunk your scan into Illustrator and use the trace function. Adjust the settings in AI to get the best result, make sure you have it set to output one colour and ignore white. Helps to bitmap the file in photoshop first.

3) Manual trace. Plunk your scan into Illustrator and use pencil, line and bezier curve tools to recreate it in vectors.

Never rasterise text.

If you don’t know the difference in-between raster and vector, do some research. If you don’t understand the steps above you have work ahead of you before you worry about prepping for photopolymer. If any of the steps seem difficult or tedious, google them to see if there are any tricks. Normally, there aren’t.

Always output to a press-quality PDF or version of AI compatible with your printer’s computer. Output is a whole other story, hopefully your printer (or platemaker if you are printing) will be patient enough to help you through it.


If you want, I can teach you private lessons through skype. You can e-mail me if this sounds like something you would like.


I would suggest purchasing the book by Gerald Lange: Printing digital type on the hand-operated flatbed cylinder press.

Join PPletterpress on Yahoo which you’ll find resources and answers to photopolymer question.

Not long ago someone would draw an image or take a photograph and send it out to have a copper plate made and then print from it on a letterpress printing press. Now days you can draw the image and send it to a photopolymer platemaker to scan in and make a plate. Or, draw the image > scan in at 1200 ppi > Auto Trace in Adobe Illustrator > send to a photopolymer platemaker for a plate to print from.

Technology continues to advance the process in which we can obtain images to print from. Copper plates or Photopolymer plates still need to be mounted type high to print on a letterpress printing press. I’ll carve and make photopolymer plates on my AV Orbital VIII. Gotta love technology.

Thanks for your help folks.
Enrique - so kind of you to offer…. I’ll contact you if I still can’t work it out.

Remember to work with spot colours if you are going to need multiple colours printed.

Hi Mbeeson,

I cheat. Vector Magic seems to produce better vector images than the AI trace function, but AI may have improved since I used it. It seems that you have to purchase a subscription, but you also can buy token packs that work out to less than $2.50 per trace. You need to go to the “Pricing” tab and scroll down to “What happened to the tokens?”

EDIT: If that link doesn’t work, I was referring to


I do this all the time. Raster graphics actually work pretty well if you ink your images very well. Personally, I do a step by step process…

1. Scan the inked image (must be black)
2. Bring it into Photoshop and adjust the levels to get the brightest white/blackest black you can manage. At this stage you should also erase any smudges and make any repairs to the image.
3. Switch the image mode to Bitmap (pure black and white)
4. Save as a Jpeg
5. Import into Illustrator
6. Use live trace, adjust the settings - the default almost always looks terrible
7. Make sure ignore white is checked. If it wasn’t remove any white areas. Expand (button at the top of the screen ususally) it once it’s live traced.
8. Use the smooth tool to fix any tiny mistakes.
9. Select the image and re-group it if you had to ungroup it to remove white areas.
10. Set the fill color to pure black #000000
11. Export as PDF or EPS

I also make my own vector designs right in illustrator. Of course, type (unless it is hand drawn) should always be this way since bitmap text never works well. You can always skip the illustrator part and send the Bitmap jpeg to the platemaker, it often works. I prefer to have it as a vector so I can make size adjustments and smooth certain areas.

The attached image was done that way. I assure you, no one would be able to tell it was live-traced. You just have to make sure you do a really good job inking, it all starts with the drawing itself.

image: owl.jpg


Artwork needn’t be in vector mode in order to be made into films which can produce plates; Imagesetters actually RIP, or Raster Image Process, a vector image into slices- or lines of code- and then fire a laser on/off one line at a time in sequence when outputting film. So, the film processor has to re-translate the language of the art BACK into dots after you’ve converted it to these smooth shapes.

In effect, if your image is hand drawn, vector kind of complicates things as it forces a computer program to interpolate and interpret the data and produce complex paths/shapes. This is going to sound counter-intuitive, but sometimes these shapes actually end up being ‘simpler’ than what you had already drawn because the limits of the coding actually get in the way.

*Conversely; where vector REALLY comes in as a necessity is type- type never seems look as crisp when it comes from a raster source. But you’re not using type that matters in that way I take it?

There’s not really a wrong way if you pick between vectoring and bitmapping provided the bitmap is of a high enough resolution, and if vectorizing the art gives a pleasing result and the outcome is what you want, that’s great- but I prefer to forward high resolution bitmap files to my film output vendor when hand-drawn images are involved, and I follow a workflow that I’ll outline below.

My workflow process for bitmapping is to:

1. Import scanned grayscale art into photoshop. Make sure the image is set to “grayscale”.
2.Clean art with eraser tool or remove things that shouldn’t be there.
3. Use the levels tool to boost the blacks to become 100% density and turn the whites to 100% clarity, observing the edges of anything that appears to be shaded in so doing; I have found setting the black to 50, the gray to .60, and the white/highlights to 240 to be a pretty good all around contrast for work that is executed on bright white paper; if you’re drawing on natural or more warm tinted papers, you’ll need to adjust it a bit on the white end of things and the midtones will need to change as well.
4. Save your work, and create a new document from the current state by clicking the icon on the ‘history’ window.

5. Convert image in the new state to a bitmap.
Depending upon how powerful your computer is and the scale of the artwork, you’ll probably need to pick a resolution that works and is able to be processed in a reasonable amount of time; the higher the res, the more computing power you need.
But basically, at the bitmap dialogue box you input a resolution, and a bitmap style.
Style: If you want the image to be Black, white, no dots, and true to the inking you see on the screen, then you can select “50% threshold”. This converts your black and white art to flat lines of on/off code at a very high resolution; anything that is less than 50% gray is deleted, everything that is darker than 50% gray becomes black!

Resolution: This is the DPI that the software will convert your image to; basically, it’s going to turn it into a lot of points on the graph. With 50% threshold, the points are on or off- they do not turn into dots. So, it is to your advantage to pick the highest resolution your machine can handle. 2880. Incidentally, this is one of the maximum DPI settings on a lot of imagesetters, and it is going to make a very large file and your computer will take a while to crunch through it. Also, it’s not the max resolution an imagesetter can perform.

Then I click “OK” on the dialogue box and let it chew through the lines of code until it is complete.

Some go all the way up to 3600 DPI and I don’t know if any go higher than this. It’s not a matter of dots at that point though, it’s a matter of smoothe curves with nearly stepless edges so you don’t really need it because 2880 is already kind of “overkill” and you’re not even likely to have a computer that can do this.

But, a 5x7 card drawn, scanned high res, contrasted, bitmapped 2880, and forwarded to a good film processing/plate output service will give you a really great plate, without the need to vectorize it!!

conventional practice for the use of bitmap lineart in general is at least 1200 at 100%. Higher can be better, but that’s usually fine. The image will look bad when looking at it in photoshop close up, but zoomed out/printed it’s fine.

Another thought on the above process. Photoshop now has a layer adjustment called “Black and White”. Users of recent versions of CS know this because every time you try to convert your image to grayscale it says “hey, why not use the Black and White layer adjustment?”

Applying this layer adjustment to a color image gives you a powerful way of adjusting how different tones will convert to a black and white image. Then you can convert to bitmap with 50% threshold. Otherwise, I just use curves or layers on the image to pump up the contrast as close to black and white as possible, make a few adjustments manually, then convert to bitmap at 1200+ dpi, 50% threshold.

Everyone has posted really good info here. I forgot to mention… I think some of it matters with how you process the plates. With Boxcar (a professional service), I can never tell the difference between a vector and a raster graphic that I’ve sent off.

However, when I make my own photopolymer, and make my own negatives, the raster graphics sometimes come out grainy around the edges. This is particularly true of Xante laser negatives. It has to do with how the printer translates the artwork when printing. When I’ve had linotronic output for my negatives, the difference is once again negligible. I just like working with vectors. Do what is best for you.

Oh wow….I’ve got a lot to process. What a great community of friends! :)


I’m quite sure I am in the minority here regarding images, but I would never use vector for this. I enjoy how computer software allows us to do stuff, I just don’t want it to make the decisions. One has an incredible number of options in Photoshop for rendering raster imaging exactly as one would want it, or as it should be. Illustrator will give you the Illustratored version, which I guess is fine for most folks, but sometimes quick and easy just doesn’t do it.


I don’t know if you’re in the minority…half the posts on this thread are about how to use hi-res bitmap images!

I don’t know about never using vector, but when I’ve sent art out for platemaking, It’s always been individual separations converted to 2400 DPI bitmap TIFF. If it’s something I set up from scratch, I’ll usually be working from either InDesign or Illustrator with only the last few steps being done in Photoshop. If it’s a scan from some hardcopy, I’ll usually just leave it in Photoshop whole way through the prep process.

Some other important things are to gain an understanding of how solid inks interact in spot color layouts and to make sure things like your blacks are pure black (rather than RGB or some CMYK rich black mixture) so that the artwork will separate properly when it is time for them to. Whether you provide pre-separated art or not, the art still has to separate properly in the end.

I want to clarify to Mb and everyone else, that the private lessons would be free of charge. I know all these comments are great advice, but it can be hard to translate if you are not familiar with Illustrator or Photoshop. So if anyone wants, I could show them a step by step guide through skype of my process wich basically covers what’s been talked here.

Enrique dear - you are so kind. I hope you’re still willing after I get a chance to digest, process, and experiment with all of this information….and find myself still too stupid. I’m teaching full time right now so I have to work on this when I get a minute here and there. It’s slow going but I am a determined girl!

Of course, send me an e-mail and we can arrange something for sure! :) I’ll be glad to help you get the nitty gritty of this process.

I just want to add that provides tutorials on many software packages for a reasonable price. I haven’t used it myself yet, but I’ve heard from many sources that the instruction is very good.


Just want to say that this is an excellent thread which really should be pinned as an FAQ!

I’ll certainly recommend it to designers and clients.

Best for 2014 to you all!

I might add that since my last post on this thread, I have subscribed to, and it is indeed excellent. In short order, I learned InDesign well enough to create an entire 330-page book the right way. I say “the right way” because I’ve seen some pretty clunky InDesign files.


Silly question: When you finish the image, do you make it backwards, or do they do that for you (if you send it to Boxcar)?

Does no one use pen and ink anymore? Am I the only one left who doesn’t use a computer for artwork?

I use type-high woodblocks cut with gravers, and occasionally linoleum cut with knives and gouges.


Boxcar does the reversal for you. They have a bunch of useful information on file types, artwork preparation, etc. if you’re getting plates made with them.

I think Pantherapress’ list above is a good one. I’d like to add a bit.

I’ve added a preface note and a “0” and then a bit of my information to 1, 2, 4, 6, and 8.

One thing that is important to understand on the computer is that the scale of your digital image on your computer screen doesn’t really translate very well. I THINK I read somewhere that Boxcar Press (where we get our photopolymer plates made) says the thinnest line should be at least 0.25 points. I think it is a good idea to draw a small square that is 0.25 square and note both how big that appears and what the % scale of your image on your computer is.

0. Sketch with the understanding of where you are headed. The more clearly black (or at least very dark) and with less tonal variation the better the drawing will scan. It is worth trying your sketch with both a good, soft pencil and then the same drawing using a sharpie. You may like the softness the pencil adds to the image.

1. Scan the inked image (must be black). I think scanning in color and then coverting to greyscale gives you better resolution of the image.

2. Bring it into Photoshop and adjust the levels to get the brightest white/blackest black you can manage. At this stage you should also erase any smudges and make any repairs to the image. White dots that are too small will fill in. Lines that are too small will not reproduce or well in a photo polymer plate.

3. Switch the image mode to Bitmap (pure black and white)

5. Import into Illustrator

6. Use live trace, adjust the settings - the default almost always looks terrible. This is more complicated than it sounds. Try your first one at all of the default options listed. I’m on a Macintosh with CS5.5 and the two default options I generally like the best are “lettering” and “grayscale.”

7. Make sure ignore white is checked. If it wasn’t remove any white areas. Expand (button at the top of the screen ususally) it once it’s live traced.

8. Use the smooth tool to fix any tiny mistakes. The ‘pencil’ tool is also a good way to widen an important line that is too thin.

9. Select the image and re-group it if you had to ungroup it to remove white areas.
10. Set the fill color to pure black #000000
11. Export as PDF or EPS

Boxcar has a decent explanation of the process, but I personally like the numbered listing that is going here. Boxcar could use more detail and some headings down through the text to help you locate specific issues.

Very informative thread!

this is very informative. I work primarily in photoshop, and need just a bit of clarification here, is it that is MUST be a vector, or that vectors produce the best results?

I have experience with screen printing and understand the need for a pure black to stop out the light in development, and I understand that jpegs would tend to give you fuzzier pixelation, among other inferior qualities, but if I convert that jpeg to a bitmap, in which it is now a high contrast pure black and white image, would that not be sufficient to go to print?

I know how to import that into illustrator, and fuss about with the live trace settings, but i still feel like when i look close enough it has that “live trace” look, and I have not noticed this being the case with letterpress work I have seen in the past.

so what would be the issue exactly if I just sent in a bitmap, .tiff?

Depending on the platemaker you may be able to send the .tiff if they accept that. The difference between the vector and bitmap images is going to be similar to the picture I’ve posted here. If it works. The vector image lines are exact lines and curves are exact curves as described by the math coprocessor of the computer. As much as you zoom in they will not get fuzzy. The bitmap on the other hand at some level of zoom is going to make curves look like stair steps because the image is made of dots. The highest resolution bitmaps can be indistingushable for your purposes or may not be depending on how they are created, saved, processed and printed.

image: image quality.JPG

image quality.JPG

I’ve been making Film in house for a long time. While I still have Rips (Raster Image Processors) which would accept .tiff,
the Trend for the last Decade and a half is PDF.
But we also have a Rule:

Garbage in - Garbage out.
If the File is not correct - none of what I have can correct it, unless I have to completely rebuild it.
If you build your File in Illustrator or Indesign, we don’t run into the Problem of Bit mapped Files like you have them from Photoshop.

I am not sure if your plates are image only or exactly what you are creating. There was a lot of very good info to help you move forward. I did not see this mentioned but also ask the company you are getting your plates made from what they need. If you create a file in the best way for them to work with rather than what you think should work you will be a step ahead. We also open all files to be plated in Illustrator, convert any text to outlines and view the document in outline mode. In outline mode you can see any unused or hidden info which we delete. We want our files to contain only what they need to make a plate and no more. As a final step we flatten the transparency with using high resolution.