I use spot metering almost exclusively. Once you understand it, you're exposures can be spot on. (the following is from posts I've made in other threads on the subject).
(in terms of spot metering and exposure)
1. Your meter assumes everything it sees is 18% (nominal) gray. It doesn't see color and has no real way of knowing what you are pointing the camera at (face recognition technology aside).
2. If you meter a White object. A white flower, a white shirt or blouse for instance and allow the meter to Center (think 0 EV compensation), The camera is assuming that object is in fact, 18% gray.
3. If you meter a Black object, the cameras meter will again, assume you are metering an 18% gray object.
The result of 2 and 3, will be under exposure and over exposure respectively if no further adjustments are made.
Some say to use the Histogram and expose to the right (ETTR). Some people live by this as if it's a gospel hard fast rule. I disagree. What *I* say to do, is look at your scene (unless you're just after a snapshot), decide What you want to present in the photo (final composition aside), and expose for that.
In the case of #2 for instance... Your whites and highlights will be a couple stops under exposed. BUT, all of the detail you could want, will be there. However, in the shadows and blacks, you will Lose detail. It may be recoverable in post processing with some masking, curve adjustments, dodging and burning, etc but for the sake of this discussion, assume it's gone.
In the case of #3 however, the exact opposite is true. That is, the photo will be Over exposed by a couple stops and your highlights will be blown out. These, as we all know, are far less recoverable than shadow detail. It's simply gone. This is why I am not an advocate of ETTR as a hard fast rule (and why I cringe whenever I see someone recommend it to a new shooter).
The difference from black to white and how far that scale will stretch, is Dynamic Range. Properly defined as a scale from black with detectable detail to white with detectable detail (pure black and pure white on the extreme ends of that scale). There are a lot of factors that affect it but our earlier cameras weren't capable of very much range. Enter HDR. The purpose of it is to stretch the dynamic range in our photos. The way it's accomplished is to mix exposures taken at a centered meter reading and under/over exposed images. Properly done, it's a very effective tool. Not properly done, we get these glowing images that are fun to look at but are nothing like the actual scene. Later cameras are getting more and more capable of stretching that DR scale and keeping details on the ends of the scale. The K5 is phenomenal in that respect.
Now that I've sufficiently confused you even more, here is a simple exercise.
1. Set your camera to RAW+JPG (so you have the RAW to recover that once in a lifetime shot).
2. Set the cameras Processing to show you a b&w image with no further corrections. No highlight or shadow corrections (which I think the K-x is capable of). Nothing. You want Lens, Meter, Sensor, and that's it.
3. Turn off any EV Compensation you have set up and turn off auto ISO (M mode does that automatically).
4. Set your camera to Spot meter.
5. Take a photo of a black and a white piece of paper (or similar but make sure both are in the frame), taking your exposure reading off of either one using the green button to center the meter reading. Lighting doesn't matter here, use a tripod or support if needed. Take a Second photo metering on the other object.
6. Look at the resulting B&W images. Yes, the images will be different. The photo metered on the White paper will be under exposed by a couple stops and will appear at whatever % gray the cameras meter is calibrated to. The photo metered on the Black paper will be Over exposed by the same amount (roughly).
7. Now, Take another shot, metering on the white object, only this time, Hit the green button to center the meter, and then adjust your exposure to let in 2 stops more light. Either open the aperture a couple stops or slow the shutter a couple stops (or simply use EV comp +2 if you're using an A type lens).
8. Take a 4th photo, this time metering the Black object (green button). Then, however you want to do it, adjust your exposure 2 stops by closing the aperture a couple stops or increasing the shutter speed a couple stops (EV comp -2).
9. Now look at the resulting photos from 7 and 8. They will appear Very similar if not exactly alike. You can adjust the compensation until they Are identical and the +/- range of your exposures, will be the dynamic range your camera is capable of (for all practical purposes).
It all sounds a lot more complicated than it really is but if you step through it, the clouds will part (with complete detail), you'll see the leaves in the dark forest, and you'll be able to rule your exposure world.
and...
This is My way. There are many ways of handling exposure, mine is only one of them.
What I do, is use Spot metering in M mode.
I decide what about a particular scene is important to me. If I'm just going out for snap shots, I'll point the camera (in M mode) at the concrete sidewalk or something of a similar shade, and hit the green button. That will get me close. If I want to change something to favor shutter speed or aperture, then hit the AE-L to lock in the exposure value (EV). That will allow me to change one (aperture or shutter speed) while the camera adjusts the other accordingly.
If there is no concrete colored object to meter off of then I'll look for something else to meter from and make the necessary adjustments. That could even be a white cloud in which case, hit the green button and make adjustments to allow in 2 more stops of light. Do so by opening the aperture more, slowing the shutter, or boosting the ISO... or some combo of the 3. I hope you can see what I'm getting at here. If you do a quick run through the exercise I outline above, you will.
Now, All the semi-Auto modes do is the same thing the Green button does (for metering). In a semi auto mode, that same exposure value (EV, Remember that?) will always be used unless you tell the camera to do otherwise. Enter EV Compensation. Set it to plus two (EV +2) and whatever I told the camera was white by metering off of it, will appear white in my photo. If I am in Av, the EV Comp will adjust the shutter speed to make the adjustment. If I am in Tv, the EV Comp will adjust the Aperture (if possible). Example just for the sake of the numbers.
ISO 80, Av, f4, meters at 1 second aiming at a white object (my front door inside). Setting EV Comp to +2 will take my shutter speed to 4 seconds (each full step doubles the shutter speed). Now, that white door will be white no matter how many times I look at it. If I take my aperture to f8, the camera will meter the shutter speed at 4 seconds (two stops difference). EV Comp will take that shutter speed to 16 seconds.
Using that method requires even a basic understanding of what you are looking at. You have to learn to take your Color glasses off and understand what you are looking at in a black and white world because that is all the Camera sees. More to the point, the camera Meter assumes Everything it is evaluating is 18% (nominal) grey. You mentioned the Histogram but may notice that until now, I have not. Regardless of metering method, the Histogram is useless unless you understand what you are looking at and have some understanding of how to expose for it. You can then use the histogram basically to check your work. It will give you better information than just looking at the photo on a tiny screen at that point.
Using the different metering modes also requires some basic understanding of what they are telling you. Spot metering takes a reading of a small portion of the center of the frame, ignoring the rest. Center Weighted, gives priority to the middle of the scene but also takes into account the rest. Matrix tries to look at everything and make a best guess based on the range of readings. For someone living in auto modes, matrix is probably the safest way to go.