Where can I get a standard camera test chart, and does printing it myself affect results?

Question

I want to compare smartphone cameras myself in a showroom and noticed that review sites use a standard-looking resolution/color test image. Where can I obtain this kind of camera test chart, and what should I keep in mind if I print one myself for meaningful comparisons?

user37321 · Accepted Answer

Interesting question.

Please google "ISO test charts", or "test charts targets" To my knowledge that one is ISO-12233 "Digital Resolution Still Camera Test Chart".

It is a bit expensive though, so you can find some other camera test chart options on Amazon, or perhaps find some second-hand on eBay.

But as you want to use them in a showroom... You probably need a standardized test.

Some explanations here.

If your workflow needs to be standardized, use some tools for the standardization. If you only want some target to compare, I guess you will be fine with some generic products.

I am not sure if you could commercially use the patterns of the test chart, but if you understand the idea behind them you can prepare some of your own. (Or use the link xenoid posted)

I. For a resolution-only chart, aka Black and white

What needs to be taken into account is the resolution of the file and the resolution of the output, besides the interpolation method used on them.

A real test chart is printed on high-resolution photographic systems, which gives crisp lines, up to 2400-3200 DPI. Probably you can ask if a prepress bureau can either print directly on photographic paper, or print a negative on film and then make a transfer. This is not expensive, but you need to find someone that can print it.

The files must be, either in vector format or at the exact same resolution of the printer, 2400 or 3200 DPI at 1 bit. You should not use a normal 24-bit or grayscale image.

If the image is not photographically printed the image will degrade, depending on the output system.

If the print is lithography and printed as line art, starting with the vector file or the 1 bit image, I am guessing it will lose around half the original resolution on the thin lines due to dot gain. The lines will be a bit wider so the white spaces are a bit narrower. But if it is well printed I think they will work well enough to do some comparation tests.

If the images are printed on a laser-toner or ink-jet printer, I would not risk using a 1-bit raster image, and only use a vector file.

You will have dots of tonner around the line which will give you a slight smudgy edge. Try to find a black-only laser printer at the highest resolution you can find. 600 DPI is normal, try to find a 1200DPI one.

If you find an inkjet one, you probably will have an additional loss of sharpness if the printer is using CMY inks besides the black.

"You can increase" the resolution of your chart... doubling the size. A 600DPI target will look like a 1200DPI at double the distance, but of course, distance is another variable on a lens test.

II. For color targets

No. Do not print your own. Find a Color target of a recognized brand. Datacolor has some economic models.

They are printed with direct paint, matt finish. Any color print will have a deviation on the color. Color is FAR more important than lens resolution if you want to do some commercial work.