View Full Version : High Resolution Printing

11-25-2003, 06:00 PM
How does one go about setting up for high resolution output (say, 300 dpi)?
I need to output some files for printing in a magazine and not sure how to set up the math.

Thanks for any advice!
- Brad

11-25-2003, 06:13 PM
very easy actually, just multiply your desired dimensions by the dpi. for example, if you want a 4" x 6" printed image:

(4"x300dpi) x (6"x300dpi) = 1200 pixels x 1800 pixels. they'll probably want the images in cmyk format as well (which you can do in photoshop) so you might ask.


11-25-2003, 06:20 PM
Thanks for response Chris. What considerations need to be taken with regards to images maps when talking print resolution?

11-25-2003, 06:22 PM
you could also use print assistant in Layout. it is in the plugins list in the edit menu area.

11-25-2003, 06:32 PM
that's a good question, and as my design and print knowledge is much deeper than my lightwave knowledge i'll let someone else answer it :)


11-26-2003, 04:00 AM
>>>What considerations need to be taken with regards to images maps when talking print resolution?

Obviously the same if you don't want to see pixels and jaggies. Just look how large the imagemaps are seen on the final picture and you know how large you have to make them.

But take care about print resolution renderings, i have no problem getting away with 150dpi instead of 300dpi. The best is that you make a test with the printer (the person) who prints your images.



05-02-2004, 02:20 PM
the basic math calculation makes sense, for example, a 10'' x 10", 300dpi final print size equals 3000 pixles x 3000 pixles, right? So, when i render this bad boy up and save the file, then load the image into photoshop the image is being loaded as a 72 dpi file sized at nearly 41"x 41"(caught me off guard). So if anyone can steer me into the right direction on getting a file from lightwave into photoshop at 300 dpi i would be thankfull.
On a side note, Bryce 5 can do this very well, with all kinds of options for exporting the required resolutions.
oh ya, heres the project ive been learning on.

05-02-2004, 02:42 PM
You can use the normal resize dialog in PS, just make sure it doesn't resample the image. Since I only have a version in German here, I'm not sure of what the option is called in the english version (the one right at the bottom, with the bicubic etc next to it. Just turn it off and it will set the image to the new size, without actually changing anything on a pixel level).

05-04-2004, 11:52 AM
make sure you're not saving out to a lossy format like jpg because sometimes it will scale everything to 72dpi since it's mainly a web format so if you save as tif, tga etc that might help.

05-04-2004, 01:20 PM

Here is a picture showing what Lightwolf is talking about.


05-05-2004, 05:11 AM
*phew* thanks meatycheesyboy :)

05-05-2004, 06:32 AM
important is only the image size in pixels, if you have eg. 1000x2000 pixels they are the same regardless DPI settings in Photoshop, 72 DPI is just the monitor resolution, so Photoshop uses this value to display images.

And by that way, to avoid useless render times, AFAIK the proper setting for offset print is 225 DPI.

05-05-2004, 10:49 AM
Originally posted by meatycheesyboy
Here is a picture showing what Lightwolf is talking about.

Impressive! Thanks for the nice diagram!!!

05-05-2004, 12:15 PM
Digimasa, Lightwolf, ingo, and fig all made excellent points. So, allow me to sum-up and start my own little tirade.

Electronic pre-press was part of my job for nearly 10 years. It amazed me then and still amazes me now how little the topic of resolution is understood, even by professional graphic artists (actually that statement extends to nearly the entire print production process.) Everyone should have to work in a print shop for at least 1 year before they can get a graphic arts degree.

A history lesson:
The human eye is an amazing instrument. No optical device invented by man to date is as subtle or sophisticated as the human eye.

The first attempts that man made at printing, were in single colors. Mostly a reddish color that was derived from clay, or black from the ashes of burned wood. If you wanted to print something in another color, you had to wait a few more thousand years for new pigments to be invented. But basically, the system was: if you wanted to print a color, you needed an ink for that color. Early lithographs sometimes used 20 or 30 colors.
In order to print all the range of colors that your eye can see, it would take hundreds of millions of colors of ink. Likewise, to print the entire range of values (that's black & white kids) would take nearly two hundred. Obviously, this is impractical for a commercial printing press. So, printers are left with a dilema: how do we make it look like there are a bunch of colors and values, when we can only print a few at a time?

Part of the solution was discovered by a sub-section of the Impressionist artists known as pointillism (mostly Seurat.) If you put two dots of paint next to each other, and stand far enough back, your eyes (really your brain) will mix them together, and you will see one dot and the color will be a mixture of the two. This is called optical mixing. The most famous example of this is a painting in the Art Institute of Chicago, called Sunday Afternoon at La Grande Jatte (sorry I can't remember the proper French title.) It is a huge painting made up of millions of tiny little dots of color. If you've seen Ferris Buehler's Day off, you know what I'm talking about. If you have the means, I highly recommend going to see it. It is so choice.

Prior to that, Isaac Newton proved that visible light can be split into three primary colors with a prism. Those colors are Red, Green, and Blue. (RGB: sound familiar?) When mixed together, they add up to white. This is called additive color.
When white light bounces off of things, some of it gets abosorbed, and some is reflected into your eyes. The reflected part of light is what you see as color. If something looks to be a certain color, then it is absorbing certain kinds of light and reflecting others. This is called subtractive color.
Cyan absorbs red light and reflects everything else. Magenta abosorbs green light and reflects everything else. Yellow abosorbs blue light and reflects everything else. So, theoretically, If you have solid cyan on top of solid magenta on top of solid yellow, everything should be absorbed, and it will look black.

So, combining the two, if we print little dots of cyan, magenta, and yellow inks, then we should be able to print all the colors in the world. One problem solved. Sort of.

Theoretically, this is sound and should work. However, in the real world it does not. If you add solid C,M,and Y you get an ugly muddy greyish color instead of black. So, why not just add black ink? Ta-daaaa!!!


OK, second problem: even this mixture was not enough to print every color the human eye can see (I told you it was an amazing instrument didn't I?) But, for most purposes it was close enough. (More on using special mix colors to fill in the gaps later.)

All right, now on to part 2: how do we break the image up into a bunch of little dots? Well, the pointillist artists put them there one by one, by hand. Again, obviously not practical for high volume commercial purposes. Then, someone (sorry, I don't know who) had a brain storm: what if we take a photograph of the art through a screen door?

Early printing plates were made just this way: by photographing reflective art through a mesh screen with a special camera. This broke the image up into little dots. How close together the lines of the grid on the mesh were, determined the resolution of your image. If the mesh had 50 lines per inch, that was your resolution. 50lpi. 50 lines (per inch) on the screen broke the image up into 50 little dots. This was called a half-tone, because everywhere there was screen, the paper was white, everywhere there was a hole, the paper got ink. So, half was paper, half was tone.
Then a filter was placed in front of the lens in order to cut out certain kinds of light. This is how color separtions were done. The art was photographed 4 times: 1 for cyan, 1 for magenta, 1 for yellow, 1 for black.

The only problem with this scenario is that your eye can see the dots at 50 lpi.
Eventually, after improvements in printing technology and photography, the mesh screen was replaced with a celluloid filter that had the half-tone pattern on it.
Also, the resolution of the half-tone got finer. Today, most commerical lithography presses print between 100 and 150 lpi.
However, Different printing processes and paper types require different resolutions.
Newspapers, which are printed on rougher paper by a process called rotogravure, are printed between 60 and 85 lpi. Billboards are printed at around 10 lpi. But you're usually really far away from a billboard, so you can't see the individual dots.

05-05-2004, 12:34 PM
Part two:

Advance forward to the computer age.
Images are now digitized and made up of little dots of their own called pixels. The number of pixels per inch that make up an image determines its resolution. NOTE: it is very important to keep the concepts of lpi and ppi separate (to confuse things ppi is often called dpi)

What is really important is the number of pixels. An image that has 2400 pixels x 3000 pixels would be 8x10 at 300 ppi or 4x5 at 600 ppi, but it has the same amount of information.

So, you ask, what does this have to do with me?

Well, if you are going to print a digitized image, your ppi has to be converted to the printers lpi. A problem arises because you basically have two grids on top of each other. This can cause a unwanted side-effect called a Moire pattern. Whenever two grids cross, you start to see wavy lines and patterns. This is bad, especially in print. (I supposed it would be good for a Grateful Dead type background)
To avoid this, printers started telling people to make their images twice the count of the printed line screen (commonly 150 lpi) This gave you a nice round number of 300 ppi, which was easy for everyone to remember.

What is wrong with this, was that printers were really trying to avoid is the crossing of two grids, and they way overshot the mark. All you need to do is make sure that no points are shared when the two grids overlap. At a 2:1 ratio, all of the points will behave, but it is not completely necessary. All you really need to be completely safe is for your pixel count to be 1.5 times your line screen. A 200ppi pixel count for an image printed at 150 lpi will usually work.They also left out that the rule is 300 ppi at THE PRINTED SIZE. So if your image shows up in photoshop as 72 ppi, but the pixel count is 2400 x 3000, then you're ok to print it at 8x10 at 300 ppi. (It's a FORMULA. It only works when all the pieces are there)

05-05-2004, 12:40 PM
Part Three:

Now, if you're printing at a different line screen, your resoultion needs change. 133 is also a common resolution for offset litho.

It also pays to keep in mind the viewing distance from the printed material. 150 is used for magazines that will be read 8-12" away from your eyes. The dots have to be small so that you don't notice them.

If you're doing a sign, and no one will get closer than 20 feet away, then it is absolutely unnecessary for the resolution to be that high.

What usually happens is some client gets the notion that higher number=better quality, and your stuck dealing with unrealistic file sizes and wasted rip time to satisfy someone who doesn't know any better.

05-05-2004, 12:48 PM
Sorry for the extra long post, but I told you it would be a tirade.

300 dpi is the QWERTY keyboard of electronic pre-press. That is, it is a dumb rule that someone thought up to solve a problem that became law because people didn't know enough to change.

For those that didn't understand the above: The keyboard you have in front of you is called a QWERTY for the arrangement of letters across the top. The reason for that arrangement had to do with a mechanical flaw of the original typewriter. The letters were spaced out to keep the mechanical arms of the typewriter from sticking together. That is the reason for the arrangement, not because it is a good layout, or even a smart or logical one.
When the electric typewriter was invented, it used a metal ball, so the issue of the arms went away. But no one re-designed the keyboard. Now, typewriters are almost non-existent, and we use keyboards with no mechanical parts, and we have the same stupid QWERTY layout, because "that's the way it has always been."

Thanks for listening. I hope someone learned something from all this.


05-05-2004, 01:38 PM
Wow, This forum is great!!
Iím new to CG and want to make a serious career out of it. Thanks to you all, Iím learning so much every day.

Just wanted to say: Thanks to all for great information

05-06-2004, 01:32 AM
Great History
I have missed a part about the connection among output lpi, image dpi, and halftone screens. Because ink has only one color (C, Y, M, or K), printers use halftone screens to display the 'grayscale' levels of that color, and need an 8x8 halftone cell to make the 256 levels in the 8-bit image. Film output for print from an imagesetter will be about 2400dpi, but it can only represent an image with 2400/8 8-bit pixels per inch. The 300dpi on an image Standard is an artifact of the resolution used by imagesetters at the dawn of digital prepress. Anyway, most print techniques can barely handle the half of 300dpi input.

Also, most image file formats do not store any info about the "dpi" of an image, since it is only meaningful for scans. PShop loads images at a default of 72dpi, that is approximate screen resolution.

05-06-2004, 06:58 AM

lpi= lines per inch for commercial printing
dpi= dots per inch for digital images.

Basically output lpi is the halftone screen.
The most commonly used ones for commerical offset lithography are 133 and 150. (At least in the U.S. I'm not sure what they use in Europe and Asia) Some fancy art books will use 175 or 200. When a continuous tone image, like art work or a photograph, is converted for print, it becomes a pattern of little dots,

An imagesetter is a fancy laser printer that prints on film. It prints its own pattern of dots, which is different than the pattern of the halftone screen. In order to prevent the two patterns from interfering, the resolution is extremely high, i.e. 2400 dpi.

You could also think of it this way:
An imagesetter prints only in bitmap mode. Since there is not dithering in bitmap mode, to make dots look round, the pixel count has to be extremely high.