3

I'd like to start making prints, but all of my work is set at a resolution of 72. Are there any negative effects to increasing the resolution to 300? I'm very new to this, but from what I've read so far: higher res = nicer looking prints.

I apologize if a similar topic has been posted before, I just wanted to know if there was any pressing reason why I should not be doing this.

13

The DPI of the image itself is not really that important. What is important is how big you will print the image and what kind of press/printer will be used to print it. This is why:

DPI, dimensions and pixels

The dimensions of an image can be specified in 2 different ways.

  • Indicate DPI and dimensions in inches (or cm)
  • Indicate dimensions in pixels (or dots, which mean the same in the scope of this comment)

If you just say "this image is 72dpi" you are just saying that for every inch of image there are 72 pixels, but you are not saying how many pixels the image actually has. It could be a 10px x 10px icon or a 3m x 3m poster. Both of these images, for example, are 72dpi. One is 0.3" x 0.3" while the other one is 3" x 3".

enter image description here

With DPI and dimension in inches you can always calculate the dimensions in pixels. And viceversa. Dimension in pixels = dimension in inches x dpi. The two previous images are, for example, 22px x 22px and 216px x 216px respectively.

Why higher DPI by itself does not necessarily mean higher quality.

Suppose you have an image that was originally prepared at 72dpi and suppose the image dimensions are 2" x 2". Think of the image as a matrix of dots (pixels). Your image is 2" x 2" at 72dpi. It means that, for every row of dots, it has 72 pixels in every inch of the row. Since 72 (the dpi) times 2 (the dimension in inches) is 144. It means your image is, in dots or pixels, 144 x 144.

enter image description here

Now suppose you want to print your image using a regular offset press, which can print up to 300dpi images. Suppose you try to print it exactly at the original image dimension in inches. The press needs 300 dots for every inch to do a good job. The image you are supplying has only 72 dots for every inch, so the press (or more accurately, the software that prepares the plates for the press) has to make up the missing dots. The result will be blurry or noisy because there will be a lot of "transitional" dots that the software will creatively add to fill the missing gaps.

Now, suppose that you want to print the same image but not at its original size. You want to print it at a smaller size: 0.48" x 0.48". Your original image has 72 dots per inch. Your image was 2" x 2" and, as we saw, it means it was 144 dots x 144 dots. It means that it will have 144 dots to print every 0.48". 0.48/144 = 300! The press will have 300 dots to print every inch. The result will be great. The press will have all the dots it needs.

Optimal printable dimensions

You can think of the Width and Height in inches as the optimal printable dimensions at a given DPI. If you notice, when you try to increase the DPI in PS without re-sampling, the printable dimensions of the image decrease. The dimensions in pixels (dots) stay always the same, though. The resulting image will have exactly the same number of pixels. The only difference would be that, in some corner of the file, it will read: "hey, the creator of this file thinks this file will be used at 300dpi, so you know. If you need to show dimensions, please show them considering 300dpi".

enter image description here

If you ask PS to do re-sampling, though, and increase the DPI, then the dimensions in pixels of the image will increase. When you click "OK" PS will add dots in between the ones the image already, the same way the press would have done (take a look at @joojaa's comment below). The resulting image will have more pixels.

So, the same image of 144 px x 144 px can be printed optimally either at 2"x 2" on a 72dpi press or at 1.48" x 148" on a 300dpi press.

If you have huge images at 72dpi and you print them at a smaller size (more or less 1/4 of their optimal size at 72dpi) in press capable of 300dpi, then you are fine because the press will have enough dots to work with. On the other hand, if you have an image that is 300dpi but you want to print it at a dimension that is over its optimal dimensions, then the result will be bad because the press will make up the missing dots.

DPI and LPI

A little digression: We tend to think about presses in DPI terms but, as @Yorik clarified in his comment, this is inaccurate. If a press that can print images up to 300 dpi this does not mean that for every line of the final print, the press is capable of printing 300 dots in every inch of the line. Thanks @Yorik, I did not know that! So I went and researched about it...

The resolution of the presses is indicated in LPI (lines per inch). If you take a look at a printed image with a magnifier, you will see a set of different dots (one set per ink). By using different sizes of dots the press creates the illusion of tones. Bigger dots create the illusion of darker tones. Smaller dots create the illusion of lighter tones. This set of dots is called the screen and the process of turning an image into a screen of dots is called halftone screening.

The dots are usually arranged in a diagonal line (there are other methods out there, but enough digressions). The closer the lines can be packed (the finer the the screen) the greater the illusion of the tones. The resolution of a press is indicated by the number of lines it can pack in one inch. A 150LPI press can pack 150 lines in one inch of the halftone screen.

Here is a simulation of a B/W image printed in 3 different presses, each one with twice the LPI from the previous one, and then magnified. These are exaggerated. The actual halftone screens are way finer than this. I have highlighted the lines of dots in each of the simulations. There is only one screen in this case because the image only needs black ink. The screens are usually rotated an angle to confuse the eye.

enter image description here

The top right simulation creates a better illusion of tones. The lines of dots on the screen are packed more closely. It has a higher LPI.

Now, if you are going to print an image in a press capable of 150LPI, the first impulse would be to prepare the image so there are 150 dots in every inch, hence prepare it as a 150DPI image. If you do that the press goes ahead and renders each one of your pixels as a dot. The dots are small, but not that small that the eye cannot notice them. This fact combined with the fact that the screens are rotated results in an image that reminds a bit of pixel art, where diagonal lines look jagged, sort of like stairs (this is called "staircasing," the "jaggies," or "pixelation.")

If instead of that you provide an image that hast twice DPI than the LPI of the press, then each of the dots of the halftone screen will represent a tone from your image instead of a single pixel. The result is a smoother image.

This page has excellent examples: http://the-print-guide.blogspot.ca/2009/04/image-resolution-for-printing-lpi-vs.html

The general guideline is to provide images with a DPI that is 1.5x to 2 times the LPI of the press. That is why for presses that are 150LPI, the images should be 300DPI.

  • Very well explained! The simple answer is to build your files in exact intended size as possible. Period – Nomadme Feb 24 '15 at 21:56
  • +1, good description. very minor nitpick: the press (the plate) doesn't have "300 dots," but rather 100, 133, or 150 lines-per-inch halftone screens. We want 300 dpi so as to be 2x the line screen size. – Yorik Feb 24 '15 at 22:07
  • You might want to say, more forcefully, that resampling the image is the same action that the printer is doing when making up pixels in between. Just that you have control over the how. But you dont magically get that much better results. Now you have this in parens so it can get lost. – joojaa Feb 25 '15 at 0:16
  • @Yorik: Great correction. Edited and added a section about it. Thanks a lot! – cockypup Feb 25 '15 at 1:47
  • @joojaa: Added a reference to your comment : ) – cockypup Feb 25 '15 at 1:48
2

It's not so much as "higher than 72 dpi is nicer looking prints" The reality is that anything under 200 dpi will come out pixelated and low quality. Usually it s recommended to use between 240-300 dpi but not under 200.
The simpler way is to enlarge the image by using: Image>Image size> change the resolution to 240 dpi and choose: bicubic smoother option (best for enlargement) and obviously change the colour mode to CMYKenter image description here:

I would never recommend that though, The best way to work for print is to rebuild your files for print entirely. Change the mode to CMYK and take care of every element separately, in a 300 dpi document and in a print software like indesign.

2

There seems to be this assumption in the graphics world that everything is printed CMYK. There are photographic printers that use RGB. We have six of them at the lab I work at. We also have two CMYK printers that do a fine job of converting from sRGB to their internal CMYK profiles, so we leave images in sRGB when printing to those as well. If CMYK is needed, there's more to it than simply "switching to CMYK." We'll leave that to another post.

But getting to the OP's question: Increasing the DPI (or more accurately, 'PPI' - pixels per inch, rather than 'dots per inch') by itself doesn't affect the output. You still have the same number of pixels if you're not interpolating.

If you are increasing the PPI while maintaining or increasing the dimensions (width/height in inches/cm/whatever), then quality will decrease as the number of pixels grows. You're not adding more detail to the image, but rather adding more pieces of information about the same detail you started with.

  • Humans move at glacier speed. So once they learn something it often takes momentous effort to change that. But they are being a bit over pragmatic. I too believe in profile to profile conversion. Its just that some commercial printshops do not and many of the designers have bad experiences. This may be outdated since they dont bother to update a system that they know works, to one that at least previously did not work. they dont really consider the archival aspect. I like this answer even if it is a bit opinionated :P – joojaa Feb 24 '15 at 19:58
  • 1
    Whenever I get bad color in a proof, it is invariably because I accidentally placed an RGB image. IMO proof-time color conversion is just gambling with other people's money. – Yorik Feb 24 '15 at 20:09
  • To be clear, I'm not suggesting people stop sending CMYK to their printer/print shop. Just pointing out that CMYK is not always used for printing. Without knowing explicitly what profile and color-space your printer/shop expects files to be converted to, and without soft-proofing your work in that profile on a properly color-managed system, you are indeed gambling. – digijim Feb 24 '15 at 20:14
  • 2
    @york its not really the same thing. If your print shop expects you to do the color conversion then you have to do the color conversion. But there's no technical reason why they couldn't do as good job at profile to profile conversion as you do (or better if they are not using cmyk process). Printers are (excruciatingly) slowly changing so some printers indeed do produce good quality rgb pictures by now. mainly because it allows them to accept Joe averages jobs directly bigger markets and all that. But again printers change at glacial speeds for same reasons as all humans. – joojaa Feb 24 '15 at 20:40
  • 1
    @joojaa: too many "ifs". – Yorik Feb 24 '15 at 21:58
0

The full answer to your question is several pages long. Here is an attempt at a short answer.

My definitions: DPI = dots per inch. I use this to refer to the number of physical dots the output device is capable of producing. PPI = pixels per inch. I use this to refer to the number of image points the output device is capable of producing. These two values are almost NEVER the same.

The 72 you see in the Resolution box in the Image Size dialog is an arbitrary number that is put there by the software or camera firmware that created the image and should be ignored.

You are only interested in the numbers in the "Image Dimensions" part of the dialog. This is what tells you what size you can print your image.

You need to know the PPI for the device you are going to print your image on. For most inkjet printers you can use the value 200 PPI.

You take the number of pixels your image is wide and tall and divide by this number.

For example. My D90 creates images 4288 pixels x 2848 pixels. On a typical inkjet printer these images could be printed at 21.44" x 14.24". In general you can print an image any where from 50% to 200% and get good quality.

If I want an 8x12 on a 200PPI printer I would resample the original image to:

12x200= 2400 pixels 8x200= 1600 pixels.

I generally use the CROPPING function to re-size images since I can put in the exact PIXEL x PIXEL dimensions I want. Never use INCHES x INCHES here because if you have the wrong value in the Resolution box you are in real trouble. If you are printing on a 200PPI device and you crop as 8"x10" at 72PPI you will get a horrible result!

I find that if you send the print an image that is pre-sized to the correct number of pixels for that printer you will get a sharper print that if you allow the printer to do the re-sizing for you.

(I've printed literally millions of digital images in the last 10 years at work)

0

Another option is to reduce your printing size. For example, if you're printing a 72ppi image that's A4 size, reduce your image or printing size to, say, A5 (half of A4) or A6 (1/4 size of A4). That would double (A5) or quadruple (A6) the dpi of your image.

But, of course then they'd be printed smaller than your initial target size.

  • Thanks @ZackSaucier for editing and cleaning it up :) I know I left asterisks there (for BOLD, but I then used HTML so I messed it up), but I was at work and was busy too. – Lorgen GR Magpantay Feb 25 '15 at 16:20

protected by Mᴏɴᴋᴇʏ Jan 10 '18 at 18:44

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.