Image Science Printing FAQ

(Frequently Asked Questions - Answered!)

On this page you will find more in-depth discussion about issues related to printing

Pixels Per Inch vs Dots Per Pinch

PPI - Pixels Per Inch

DPI - Dots Per Inch

Unfortunately, in the world of imaging, PPI and DPI are used almost interchangeably. They are two quite different things and the confusion of the two leads to more confusion than almost any other single thing in digital imaging.

The first things to understand is that in digital imaging, the only thing that really counts about a file is how many pixels are in a file. Terms like megapixels, DPI and file sizes in megabytes only confuse the issue. In the end, all digital images are simply X pixels by Y pixels big (by Z bits of colour data but we can ignore that for now).

This is the only absolute measure of the quantity of information in a file (nb it has nothing to do with the quality of information in a file!)
In almost all cases, unless you are talking about physically how your printer is laying ink down on the paper, you are actually dealing with PPI - pixels per inch.

Dots Per Inch is an old printing term and has almost no place in modern digital imaging. DPI is a measure of how many tiny, tiny droplets of ink a printer is laying down in its dither pattern to form one inch of a print. Most Epsons, for example, operate in Photo mode at 1440 DPI (can be as low as 720 DPI and as high as 5760 DPI). You tell the printer which mode to print in in the driver - this is usually camouflaged by the use of modes 'Photo', 'Best Photo', 'Photo RPM' or 'Fine', 'SuperFine'. In the bigger printers you can usually choose the DPI directly as, say, 2880 DPI. Incidentally, almost all printers operate best for general photographic usage at around 1440 DPI. The higher DPI modes like 'Photo RPM' are useful if you're printing really high key (i.e. all light toned) shots, but next to useless for general printing. In fact, they are worse than useless, they are positively damaging - way too much ink is layed down on the paper, resulting in seriously impaired shadow detail. And of course the more ink you use, the more ink you pay for (an Epson R800 running in 'Best Photo' mode is less than half as expensive to run as one left in 'Photo RPM' mode for instance, and in 99.9% of photos, you'll get worse quality from the 'Photo RPM" mode!).

Pixels Per Inch is the key term. It is a description of the logical number of pixels from your original image (X pixels by Y pixels, remember) that will be used to tell the printer to print one inch on paper. Assuming a sharp original shot with good technique (see resolution discussion below), the higher the PPI, the better the quality print you can achieve - this is testably true even well beyond most claims of 360 PPI being the most you need ... 600 PPI images can easily be seen to be much sharper again if this data is available at good quality from the original file).

PPI is a logical term - changing the PPI of a particular file does not in any way affect the file itself - it is simply a decision about how many pixels of the available pixels you will use to print an inch on page. You can choose any number you like - from 1 to infinity. The de facto standard for high quality, photographic printed images is 300 PPI - that is, for each inch of the printed image, there must be 300 source pixels to use.

This is why the 'resample' check box, in the Image Size dialogue, is the single most important (and dangerous) control in Photoshop! When you resample an image, you are actually changing the number of pixels in your image (i.e. changing the value of X and Y) - adding some or throwing them away. You should only do this if you are making an explicit and informed decision to do so, because no single other thing will affect the quality of information available to you from your file as this!

An example will make all this clear:

Say we scan a 35mm tranny at 4000PPI - this will result in a file that has 5400 by 3600 pixels.

We now want to make a print of 12 by 8 inches. This means the PPI we have available from our file for a print of this size is 5400/12 = 450 PPI (or 3600/8 = 450). We are choosing to use 450 pixels to represent one inch of our print.

The printer driver will now translate those 450 logical pixels into 1440 physical dots per inch (DPI) and produce a very high quality print for us.

If we wanted to know the maximum print size we can achieve at good quality, and we know from experience that given a sharp original outputted on an Epson inkjet, 240 PPI is sufficient, we can calculate the maximum print size by taking the total number of pixels available to us (5400 on the long edge) and dividing it by the PPI required to give us, in inches, the size of the print:

5400/240 = 22.5 inches
3600/240 = 15 inches

So, our final print will be 22.5 by 15 inches, with 240 pixels used to represent each inch. And of course, the printer will actually use its 1440 dots per inch to actually print that image to paper.

Hopefully that is clear - it can be a bit confusing at first. Just in case, you can read another version of the same stuff here to get a different perspective on it.

How much resolution do I really need to achieve a high quality, sharp print?

The generally accepted figure in the publishing and photographic industry is 300 PPI. However, this tells very little of the really story of resolution and image sharpness/quality.

The key issue here is resolution - that is, how much fine detail is recorded by the capture system, and can be seen in a print. It's also important to separate detail itself from apparent sharpenss (ie edge definition) - as poor use of the unsharp mask so often demonstrates, fine detail and sharpness are two very different things.

The first thing I would say is that very often pin sharp results just aren't required. A fair percentage of images simply do not rely on fine detail in any critical way to achieve their expressive goals. That said, most images do benefit from fine detail, especially when viewed for extended periods - fine detail rewards the eye as it moves through the image, and allows your image to work more successfully when viewed from closer distances. Fine details gives your image greater longevity and interest.

The first and most important thing that will affect your image's sharpness (and thus its quality when enlarged) is your photographic technique. No amount of spending on fancy lenses will save you if your basic technique is poor.

The old maxim of being able to hand hold at (1/focal length) is fine for adequate results with small prints. The reality is, however, that even when using a 35mm camera with a 50mm lens, there will typically be an obvious visible difference in sharpness between shots taken (hand held) at 1/60th, 1/125th and 1/250th. Fundamentally, the faster shutter speed will result in a sharper image, and most people grossly over-estimate their ability to hand hold their cameras if shooting for anything but the smallest prints. There is no subsitute for steady support, so the first step to take is to always shoot from a stable platform (e.g. tripod) if at all possible. If that is not an option, aim for a faster shutter speed if possible.

The second most important thing that will affect the final resolution of your image is focus and depth of field. Most people just aren't that good at focussing really well, and with small camera systems, it's very hard to focus properly because fundamentally it is very hard to see your subject through a small hole!

In any picture, there is only one plane of absolute focus (parallel to the film plane in all cameras without movements). Areas is front and behind this plane of absolute focus may appear acceptably sharp through depth of field, but are not truly in focus. When calculating what is 'acceptably' sharp, many people use systems like hyperfocal distance, circles of confusion etc. While these are useful aids, be aware that most of these systems/tables were created long ago when imaging systems and lenses were of lower quality - film and digital have come a long way, as has lens design (with higher quality lenses, anyway), and so modern systems are capable of capturing and printing far higher detail than ever before. With this in mind, the goal should be to take advantage of this higher recording ability, and thus to use a greater depth of field in most situations. Blur is easy to add later, detail lost however is lost forever.

The third determining factor is the quality of your recording system - that is, the quality (and size) of the system you are using to capture your image. Fundamentally, bigger pieces of film (or bigger sensors) record more fine detail. This is why two cameras both having 8 megapixels may be far apart in quality terms - the digi-cam with its small sensor has far less resolving power than the SLR with a large, high quality sensor. The ISO of your system is also important - with film, the smaller grain allows for recording of finer detail, and with digital, less noise is created at lower ISOs. Finally, the quality of your lenses is of course very important - but less important than the elements mentioned so far. Buying a better lens will not help you take sharper, more richly detailed images unless you address all of the above points first.

The second to last important factor after technique, focus, and capture system resolution, is how carefully you deal with your image in Photoshop. Poor photoshop technique can very quickly ruin fine detail in an image - poor retouching, not understanding the image size dialogue box, and bad sharpening can all be catastrophic to an image's fine detail.

The final factor, of least but still significant importance, is the printing system's ability to resolve detail and it's native resolution (PPI). This is where 300 PPI comes into play. Of course, 300 PPI of unsharp, out of focus data will result in an unsharp, out of focus print with no detail. But if you're managed the process properly up until the point of printing, then the ideal situation is to send the data to the printer with no further scaling by the printer's driver or RIP system. Some popular printers and their native resolutions:

Beyond these resolutions there is very little increase in resolved detail. N.B not none, just very little!

As a good default, 300PPI is a sensible figure to use if you don't know the resolution of your printing system, and in general will result in very high quality results. In many ways, it is best to flip this discussion on its head and ask - how large can I print my files? Assuming you have shot a truly sharp image using good technique and high quality systems, then depending on your capture system, you will get a truly sharp print using the follwing formulas:

I.e. the lower number you divide by, the less pixels you need to make a sharp print. This is because the quality of those pixels is higher in those systems - more of those pixels are being used to actually record detail, and less to record noise/grain etc.

There is in fact much more to the story of resolution, detail and sharpness, but these are the basics. Master these are you will be well on your way to sharp, richly detailed prints.