DPI, PPI and Photo Units

Fundamental Photography Concepts

Today I’m going to cover a subject that can be quite confusing: DPI.

As you may know, among the many fields stored in the metadata of image files (EXIF), there is one that records the DPI value of a photo. So, what exactly does that mean and what is it used for?

DPI stands for Dots Per Inch, and it describes a physical measurement because it refers to inches.

This value expresses the definition — not the resolution — at which a device is capable of displaying or reproducing an image. It may refer to a screen, a print on paper or any other medium used to display the image.

In simple terms, it tells us how closely packed the pixels or dots are.

Imagine we have a mobile phone screen with a resolution of 1920×1080 (Full HD) and a 50-inch television with the same resolution of 1920×1080.

Both devices have exactly the same resolution, but the mobile phone has a much higher pixel density and therefore greater image definition.

The same principle applies to printing.

Imagine a 15×20 cm photographic print and a large billboard. They may both display exactly the same image, but if we look closely at each one, we will not normally see the individual dots on the 15×20 cm print, while on the billboard they become clearly visible.

In the case of the billboard, we would actually begin to see the colour printing pattern itself.

It is therefore obvious that the billboard has a much lower DPI than the small photographic print.

And now comes the interesting part.

A typical question:

Q: What size do you want me to send the photos for the magazine?

A: Send them at 300 DPI.

Wrong!

If someone tells us 300 DPI, they are probably referring to the printing resolution used by the magazine’s printing process.

But that information alone is incomplete.

We also need to know the physical size of the magazine page — or, even better, the size at which the photograph will actually be printed — so that we can calculate the minimum pixel dimensions required.

We have three parameters:

• Physical size in millimetres
• Pixels
• DPI, or dot density

If we know two of them, we can calculate the third.

It is also useful to remember:

1 in = 25.4 mm

or

1 mm = 0.03937 in

Some formulas for converting between units:

DPI / 25.4 = DPMM
(dots per millimetre)

1. DPI and mm → pixels

(dpi / 25.4) × mm = px

2. DPI and pixels → mm

(dpi / 25.4) / px = mm

3. mm and pixels → DPI

px / (mm × 0.03937) = dpi

Example

A 300 DPI printer producing an image 100 mm / 10 cm wide:

(300 / 25.4) × 100 = approximately 1200 px

So we would need an image around 1200 pixels wide.

For this reason, simply storing a DPI value inside a JPG file does not really tell us much by itself, because when we save the photograph we do not necessarily know at what physical size it will eventually be printed.

I never normally configure that DPI metadata value.

In many cases, the best approach is simply to provide the image at its maximum available resolution and allow the printer’s interpolation algorithms — either through the printer driver or its internal software — to resample the image and calculate the colour dot patterns required for the final print.