Bijan Sabet asks:
There is a street in our town where all of the homes have a bazillion xmas lights. I thought that would be good practice. […] I cranked up the ISO to 1600 which is the max on my RebelXT. I had the f/stop at 9. […] What else could I have done (minus a tripod which I’ll never carry).
You need to let in enough light to expose the sensor enough to produce a good image without a lot of noise. Sensors work as light collection devices: the longer the shutter is open, the more light gets in, and the sensor accumulates enough photons at each pixel to discern the different colors and brightness levels of your image.
Understanding what ISO 1600 and f/9 mean will answer this question for you.
ISO 100 is the baseline for sensitivity, which has an inversely proportional relationship to the dynamic range of your image (the difference between black and white). At ISO 100, the sensor uses its full range. At ISO 200, it only uses the bottom half, so what’s considered “brightest white” is the equivalent of half-brightness at ISO 100. The result is that the sensor can capture the image in half of the time (after accumulating half of the photons) as it would at ISO 100. Unfortunately, by reducing the sensor’s range, you also increase noise and decrease color fidelity and contrast. You should always use the lowest ISO number possible to achieve the shot-speed you need. And as a general guideline, I only go to ISO 800 on my XTi if I’m desperate, and I never use 1600 because it’s unacceptably noisy and “flat” in color.
The aperture (“f-stop”) number, f/9 in your case, expresses the diameter of the iris in your camera as the bottom half of a fraction: f/9 is short for 1/9, f/2.8 is 1/2.8, etc. See Wikipedia’s diagram. Because it’s a fraction, lower numbers mean more light. And because this is expressing the diameter for a circle, and the amount of light is determined by the area of that circle, f/2.8 is twice as much light as f/4. (There’s also an important focal depth-of-field relationship, but we’ll talk about that in a future post. Basically, f/2.8’s focal plane is much shallower than f/9’s.)
The challenge of night photography is that you either need a long exposure (with a tripod, which you didn’t have at the time) or you need to let in a lot of light. By setting your aperture to f/9, you severely restricted the amount of light that came in — so you had to crank up the ISO to compensate. As long as the depth of field is big enough, which I think it would have been for that shot, there’s no reason for you to narrow the aperture (raise the number) any more than necessary from your lens’ widest of f/2.8, which lets in 10 times as much light as f/9 and would allow you to drop the ISO to a more reasonable level.
If the depth of field is enough, you can even do better than that. No Canon zoom lens is “faster” (supports a wider aperture and therefore allows faster shots) than f/2.8, but there are plenty of faster primes (lenses that are fixed at a certain length and do not zoom — you zoom with your feet). The most common and affordable of these is the “Thrifty Fifty”: the 50mm f/1.8, which is a steal at about $75. If you’re serious about it and aren’t as concerned with the price, consider splurging for the $300 f/1.4 USM version, which not only lets in almost twice as much light, but is much better built, has a much faster autofocus motor, and is much easier to manually focus. Keep in mind that f/1.4 is four times as much light as f/2.8, and 41 times as much light as f/9!
Ultimately, though, even the fastest lens won’t be better for night photography than a $20 tripod.