In this article I intend to give you some simple advice how to look for the best affordable digital snap shot camera that will give you great quality pictures. Some of the charts that you can download from the following pages are also good for testing high end cameras, but most discussion here does not apply to those.

There are several sites dealing with this topic with much more detail. Some of the links to tutorials are given on the links page. This one gives you some general advice what to consider when buying a digital camera.

I assume, that before the shopping, at least you already know the price range and the approximate size of the camera. You may shoot for quality and not care a lot for the camera size or you may search for a tiny, credit card- size camera. The main specifications that jump at you immediately when you look at the cameras in a store is the pixel resolution, and the pricing goes up with the number of available pixels. Unfortunately, it is not necessarily the case with the picture quality. A 2 meg pixel camera can sometimes make nicer pictures than a 3 meg bargain. Here's why.

The image sharpness and quality depends on a number of factors, but the lens quality and matching it the light sensitive element (CMOS or CCD sensor) are the keys. If the sensor has its pixels very densely packed, the lens sharpness must be awesome. Otherwise adjacent pixels will receive the same blurred 'point of light'. In other words, low quality lens will function acceptably well only with larger sensors with low density pixel distribution (practically non-existing nowadays). The problem is that the cheaper the camera the smaller the sensor is (small sensors are cheaper). Small sensors have also more noise, less sensitivity and make blurred photos at low apertures (diffraction effect). Another problem is that the manufacturers usually do not explicitly list the sensor size unless it is unusually large. What's the solution?

It is always a good idea to read camera reviews, but you need to read a few on a given model because the reviewers may have quite a different opinion. Consumer reviews sound good but are quite chaotic and very subjective and you need to read more of them. Phrases 'razor sharp' or 'excellent picture quality' mean nothing (especially if expressed from a novice camera owner) unless the data is backed up with numbers or test images. Numbers are easier to interpret and harder to cheat on, but even many professional reviewers prefer 'natural' image displays because it's less time consuming, more subjective, so it's easy to cheat and get away with it. They call it 'real field tests' and present many arguments why charts are bad, but let's face it - a camera or lens that passes well resolution chart tests will perform accordingly well in the field. So, you've read the reviews or not and you are in a photo store and lots of unfamiliar cameras are presented to you.

For small snap-shot cameras, the easiest thing is to look at the lens physical size. Don't buy cameras with the smallest lenses among their class and high pixel count. It is simply not a good match. There are few exceptions. Some pricey 3M pix cameras have lenses of good quality despite their small size. A 5M SONY DSC-T3 performs reasonably well (Karl Zeiss lens), excellent - considering its ultra compact size. Most 2M pixel cameras have adequate lens size/quality. If you see exactly the same lens in a 4M pixel model, it is a cause for concern. This approach usually works well with the brand name cameras, where manufacturers have good reputation in optics quality (although I start seeing some risky moves to impress potential customers with a low price and high pixel count cameras).

Once you purchase a camera, you can make a test, described below, and if you should find a large mismatch in expectations and reality, you may still return such a lemon even the same day. The test takes only minutes, provided that you have your test chart ready.

The main purpose to use this chart is to determine whether the pixel count on the sensor matches the line count on an image recorded by it. It doesn't have to match it exactly, but a 25% discrepancy I find unacceptable. This is equivalent of selling de facto a 2M pix (resolving power) camera and claiming 4M pix resolution on it.

This chart is also useful in measuring lens distortions and sensor noise in 3 color channels. The sensor noise is a very important factor and will be discussed here.

 ________________________