I took this on a whim while visiting some truly amazing people at Edinburgh University this week. The building is called St John’s Land and something about the interplay of the building’s symmetry juxtaposed against the offset drainpipes appealed to me. As you know architecture is not my specialism, but variety is good for a photoblogger I suspect. I will resume normal portraiture sevice soon

Make sure that your photograph has a subject.

This may seem obvious, but in terms of having a well composed photograph it’s important that the viewer can work out what the subject of the picture is.

In the following picture I intended the line of birds to be the subject of the picture. There are few other distractions that will confuse the viewer about my intention. By that rule, this is a good picture.

Ask yourself what the subject of the next picture is? I know! I’m not sure either. We are often tempted to capture a view because the light is nice, or the vista is big and appealing. The simple truth though is that it will not be a good picture if the viewer is not clear about the subject. This picture fails this test and won’t be hung an anyone’s wall.

In the following picture, the subject has no ambiguity, it’s the Erskine Bridge in case you care. Subject clear – decent picture.

Obviously I’d like to pretend I didn’t take the next picture because I think it’s poorly composed; sadly though we all know I did! What is the subject from this Oban scene? I don’t know and I took it! This is my dirty laundry from my hard drive that should never be seen.

and again, is the subject clear? Yes, a pair of Wayfarer dinghys. A well composed picture.

And finally, a bright scene of festive illuminations in George Square in Glasgow. Full of colour and excitement, but ultimately another hard disk clogger. Yes, I know you are seeing the compositions now with the scales fallen from your eyes; no clear subject, so nice colours but bad composition.

Summary: In general make it clear what the subject of your photograph actually is.

Warning: All rules about composition are guidelines. You can break them occasionally, and still produce something good, but in general you will get more good pictures, more often by internalising the rules. Then we earn the right to break them artistically, but not before.

Next week: We’ll look at avoiding distraction from the subject.

Remember that not all compacts will have all functions, but if your compact has metering modes they might include the following. Their symbols might be a variation of these:

• Evaluative or Matrix mode

• Spot Mode

• Centre weighted or Partial Metering mode

This all sounds a bit complex, and you are probably wondering how it links to exposure. Well putting it simply, your camera will be in “evaluative” mode by default. When your camera looks at a picture, it splits its sensor into an imaginary grid of sections. (often 35). It looks at the brightness across all of them, and tries to make them spread evenly between the brightest, (near white), and the darkest, (near black).

If the scene has a person in it, and the camera fails to prioritise the brightness of that grid section, then the face would be too dark or too light. If there was a sky behind filling half the sections, then the camera will tone down the overall brightness to keep the average brightness from being too high.

In the picture above the girl’s face may be the most precious thing in the universe for you, but to the camera it is just about one percent of its sensor surface and not even a whole grid section. If its intelligent programming spots that this is a person, then it might prioritise that grid section, if not then tough cheese. (Most cameras still get this wrong a lot of the time).

This is where spot metering and to a lesser extent, centre-weighted metering come in. If you switch to these modes, you tell the camera precisely which section or sections of the grid to treat as a priority. It will then set the brightness based on what is at that section.

In the following centre-weighted, all sections are used but the red, centre ones carry more importance than the outside ones.

This means, that in spot mode, you simply hold the centre focus point over the object you need to be well exposed (a face perhaps), and the camera will meter from that section alone in spot mode, or it will prioritise that general area in centre weighted mode. This will ensure that the object you focussed on gets the right brightness and detail regardless of the rest of the scene.

If I was just getting to grips with my compact, and it had all 3 modes, I would simply use evaluative or spot mode. Choose spot mode when the exposure is wrong in its default evaluative mode, select the focus point, and that will fix your exposure. A bit more complex than the simple exposure compensation, but for many people depending on the camera, this is more suitable.

Full frame or 1.3x frame like the 1DMk3

A normal size, not built in sidegrip like the Canon 1D’s (Too big and heavy)

Good noise performance at 1600 ISO or beyond

Fast autofocus, with good frame coverage, like the Nikon D700 or Canon 1Ds Mk3

Reasonable frame rate, although I almost never use the 6+ fps the 40D can give. 3 or 4 would do.

Live view, for macro

A good screen for reviewing fine focus

In body Image Stabilisation. (For my prime lenses)

So what’s my problem? The only upgrade path Canon has given me is the new 50D. This is essentially my 40D on steroids. It has 15 megapixels as opposed to my 40D’s 10 megapixels. They claim to have redesigned the sensor to improve noise, but the serious reviews like DP review’s say that the sensor is overpopulated with pixels. The result is no noise improvement in the face of Nikon’s D300 which is a great camera, and beating the 50D in the reviews. They are £900 and £850 respectively, so direct competitors. If I was starting again, I would buy the Nikon; its better in lots of ways, the trouble is amateurs like me can’t suddenly change our cameras since we have a collection of lenses worth more than the camera. The Nikon has much better noise performance, but why? I think the sensor on the Canon is slightly smaller (APS – C 22.2 x 14.8mm) than the Nikon DX sensor (23.6 x 15.7). Sensibly the Nikon has pushed resolution to 12 or so megapixels, Canon went for 15 megapixels in a smaller sensor. You don’t need to be a former Physics teacher (which I am) to know this is madness without some secret technology breakthrough. The future may well be black silicon, but for the moment both companies are using similar technology and Canon therefore seem to have allowed the marketing people to go for the headline pixel count, on a smaller sensor.

I admit there is much to like about the 50D. There is the high res screen, the compensation for front or back-focusing lenses, the resolution in good light, but serious commentators are saying, it doesn’t amount to significantly better picture quality than I have at present.

So my next choice from Canon, the recently released 5D Mk2. This ticks most of my boxes, although we will have to see the reviews over the next few weeks as it is newly released. It is full frame, with 21 megapixels, and I think I would really like one, but its £2000. Hopefully the much larger sensor can support the greater pixel density and result in a better noise performance at 1600 ISO than I have at present. So to update my image quality significantly, I need to go from my current £600 ish camera, to a £2000 camera. What are you playing at Canon? Nikon could do me a better camera at £900, and their full frame, 12 odd megapixel dream D700, which lets you photograph miners at work by the sparks of their pickaxes costs £1600. Get your act together Canon, this is an all or nothing upgrade path! Right now if you are buying into a new DSLR, I’d say, buy a Nikon.

Why do I care? I’ve just sold my second 40D, bought to see me through a repair under guarantee, and I’m feeling ready to divert the funds into an upgrade. If I had a Nikon, I could afford it, with Canon, I can’t.