Learning the hard stuff about DSLR settings

-super generic version

Ok so you've switched from your simple camera where you pointed and shot and got a good enough shot most of the time, to a much more expensive Prima Donna DSLR which reacts to your attempts to fiddle with it settings by puking up bad photos.
Pro's use DSLR all the time to take good shots, sometimes even the same model as yours so why does your camera suck and not theirs?

All camera are a balancing act between the Aperture at the front letting light in and the Sensor at the back catching it. 
All photos are a balance between whats happening outside the camera and what happening inside the camera.

User-friendly cameras make the choice for you, pro camera's let you choose configurations that the camera would think are unbalanced, but which suit the scene.

So lets dive in

What the aperture is doing out the front

The aperture's the adjustable pinhole in front of the lens.  Since it blocks light it's measured as a fraction using the letter f.  So the lens F-Stop setting f:1 means the aperture is as wide as the lens and blocks nothing, while f::64 is a small hole blocking 63/64's of light reaching the sensor.

Put simply, it takes 64 times longer to collect enough light for your photo at  f:64 than at f:1, and if the subjects moving you'll get 64times more blur.

The aperture also affects the focus.  Lenses focus on a point and give a bubble of sharpness, the aperture affects how deep that bubble goes.  It's referred to DOF(depth of field).
So if I was to focus on someones eyes so they are sharp, a wide aperture like f1 would give me such a shallow DOF bubble that the nose and ears would be blurred. But f64 would give me a sharp nose, eyes, ears and mountains in the background.

So already have two mutually exclusive factors, the amount of stuff you can focus on in one time vs how blurry it gets if it moves.

Portrait photographers tend for lower f-stops f5.6 since they want the face sharp but the background out of focus. 
While landscape photographers shooting a sign in front of a mountain, would go for f64, since they want both in focus and neither item are likely to move.  (And to prevent the camera moving they'd put it on a tripod)

What the sensor is doing out the the back

All the little sensors on the big sensor chip collect light, count it as a fraction and give the numbers to a digital file.  This file is the photo. 
There are typically three sensors per point, one each for -Red, Green and Blue, so each point in the finished photo has three numbers which are combined to give the RGB colour on your computer screen.

So R(0/256) + G(0/256) + B(0/256)  =RGB(0,0,0)   = Pure Black
So R(256/256) + G(256/256) + B(256/256)  =RGB(256,256,256)   = Pure white
So R(256/256) + G(0/256) + B(0/256)  =RGB(256,0,0)   = Pure Red

To figure out the fraction your camera uses, just read the sales guff or the manual for the bit count.
24 bit colour photos (JPG) have 8 bits per colour.
so 2^8 =256 different shades of each colour

36 bit colour photos (TIFF, RAW format) have 12 bits per colour
so 2^12  =4096  shades of each colour.

More than 256/256 is "too bright", less than 1/256 is "too dark".

The sensor has an sensitivity ISO rating, this corresponds to how much time it needs to catch enough light.  This is a multiplier based on a fraction. 
So what takes a  sensor set to ISO 100,  1/100th of a second to capture a certain fraction of light, would take the sensor set to ISO 3200, 1/3200th of a second.  This reduces the time to take a photo, but because the raw info is being multiplied any inaccuracies also get multiplied. 
So taking a photo of runner, the low ISO like 100 would appear smooth but blurred; while a high ISO like 3200 would freeze them, but would appear more blotchy and speckled.

To limit how much light gets counted by the sensor, there is the shutter blocking access to it.  It flicks open when you take a photo, for a brief period of time. 
The longer it's open the more time there is for light to hit the sensor, but also the harder it is for the photographer to keep the camera steady to avoid blur.
So sports photographers tend to shoot with a shutter speed, 1/800th of second or faster to freeze the moment.
While a landscape photographer on a pitch black night, might hold their shutter open for 30 seconds, to make the shot 3000 times brighter, than if it was shot at 1/100th's of a second.

Now the shutter speed and the ISO are inversely  linked, so if you want to keep the photo at the same level of brightness but reduce the shutter speed by half, then you need to double the ISO and vice versa.

Also the aperture and the shutter are also inversely linked, so if you want to increase your aperture (getting deeper DOF but reducing the amount of light it lets through) then you need to increase the time that the shutters held open so it can catch more light. and vice versa.

Lens length is linked to DOF so a wide angled [short] lens gets the same DOF with a wide aperture as a telephoto [long] lens does with a narrow aperture.
Finally lens length is linked to shutter speed blur, a hair width of a tremble will be more magnified the longer a lens is.  So the longer the lens, the faster the shutter speed should be to reduce shake. 
Firm tripods or leaning against something robust, cancel a lot of camera-shake so they let you use a slower shutter speed.

Finally you can increase the amount of light by using a flash or artificial lighting,

Amateurs practise until they get it right. Professionals practise until they can't get it wrong

Rough guide to fault fixing
(stuff in italics mentions post processing with computer software)

Photos too dark
Use a wider aperture, slower shutter, flash and/or higher ISO
Switch the camera from JPG 24bit to RAW 36bit, take the photo a bit dark, then use the extra detail the extra bits give you to brighten the photo on a computer afterwards.  E.g. there are 40 shades under 41/4096, but only 9 under 10/256

Photos too light
Use a narrower aperture, faster shutter and/or lower ISO
Using RAW mode, though better than JPG,  doesnt work as well with bright pictures as it does with dark ones.

I want to shoot a range of brightnesses for an HDR (High Dynamic Range) compilation
Use a tripod. Keep the aperture and lens fixed to maintain the shot's layout and only change the shutterspeed and/or ISO

You want the subject in focus, but everything else more blurry
Use a wider aperture (compensate with a reduced shutter speed or ISO)   
Use a longer lens
Spend time in an imaging program, to paint in blur

You want everything in focus
Use a narrower aperture (compensate with a slower shutter and/or higher ISO) 
Use shorter "wide-angle" lens

You want to reduce camera shake
Use a faster shutter (compensate with a wider aperture, flash and/or higher ISO)
Use a tripod

You want something moving to be sharper
Use a faster shutter speed (compensate with a wider aperture, flash and/or higher ISO)

You want to reduce the grain / increase the smoothness and detail
Use a lower ISO (compensate with a wider aperture, flash, and/or slower shutter)
Shoot RAW to get the most detail and then use a variety of software to reduce the appearance of grain  

2006/10/10 -Text