Perhaps naive question, but who needs such fancy metering?

As long as I don't clip the highlights I'm not bothered about ultra accurate metering. I stick to centre weighted and dial in some compensation if I feel it's needed.

Final exposure adjustments ( local and/or global )are done in Lightroom

I've told this tale before, but it's relevant here as well. Wind back many years to a camera club demo I saw, with two photographers showing printing from colour negatives and colour slides respectively.

Mr Colour Slide had tanks of water, thermometers everywhere, kettles boiling and a right old rigmarole it was as well. His results were mediocre.

Mr Colour Negative had a small bottle of developer in his pocket. "This is just the right quantity," he explained, "and by the time I've chosen the negative put it in the enlarger and exposed the print it's just been warmed enough in my pocket to be exactly the right temperature." No hassle, no stressing and perfect results every time.

Experience and practice and looking for the simplest methodology are the moral of this little tale.

----------------------------------
We all find a solution or workround I guess. Funnily enough I couldn't survive without raw and can never fully interpret the relevance of a histogram on the exposure of my small bird in the middle of the frame. To me the sliders are pretty much the same for raw as jpeg, just more effective. Not a criticism, just highlighting the differences in how we deal with similar challenges.

Only procedure difference I've noticed in Editing RAWs versus JPGs is the colour Temp slider in Lightroom is marked Temp for RAW and just an arbitary number for JPGs. IIRC everything else is equal.

In practice the quality of RAWs are MUCH better and thus there's more leeway for judicious editing to squeeze the desired result out of them

I'm assuming you are referring here to centre-weighted metering being 'simpler than you think'. I didn't say it needed to be complicated, my main point about it is when using 'area-restricting' modes, that limit the covereage of what's being metered, it is not being made clear enough that you need to seek out valid mid tone values for it to be accurate.

Yes, the experience comes into it if there's no mid-tones and you have to apply appropriate compensation, but then in that case the centre-weighted mode is not really ideal for that scene, and you would be better zooming in with spot for example.

The multi-pattern is more like taking a range of readings across a much larger area and balancing them, but with the advantage of scene-assessing algorithms to apply the most common corrections (like for backlighting, bright sky etc).

Of course, with experience an individual can get reliable results using any consistent method, but I believe that with the modern aid of the histogram and quick 1/3rd or 1/half stop ec adjustments, anyone will get more accurate and quicker results in Multi-Pattern mode, for the majority of situtions.

No need even for the histogram. The temptation these days is to stop and look at images and even histograms when the better option may well be to make the judgement and carry on shooting. Do the checking at the end when the moment has passed and the shots are in the bag, or maybe once half way through if we need reassurance.

Constant checking is a guaranteed way to miss the action.

You can help your camera a bit though. The K7 / & K5 offer a menu item to link the active focus point to the metering. This offers an automatic bias to the area in sharpest focus, might not be the centre of the frame at all.

I'm not sure why people seem to be connecting so firmly the centre circle of the frame with the right tones to meter off. They would only occasionally be so I would think. The more you limit the area being metered, the greater the risk of meter error, or at least the need to compensate and make tonal judgements.

In isolation when considering metering, I can't accept this as good advice. The histogram is the most valuable tool for judging correct exposure in the field, and I believe its understanding and use should be promoted widely. Our Pentax DSLRs provide very quick and efficient controls for accessing this essential tool, both in RGB and individual channel modes. If people are concerned about capturing DR, then there's no better way of achieving it.

Obviously I'm not saying we should be looking down after every shot, but when setting up our exposure and carrying out metering, then its a most important aid, and should be encouraged.

Indeed, which can be great if you give equal importance to the whole image. But if (which is often the case) the central area is the most important area having its exposure changed because of bright or dark edges isn't helpful.

An advantage for one person is a disadvantage for another though. Like John I find it much easier to predict how unusual lighting conditions will fool CW metering and compensate for that than try to second guess what scene-assessing algorithms are going to do. Showing my age here I guess

The problem Nigel is that the histogram needs understanding. A largely black scene, for example, will have a histogram skewed to the left, and this is where it should be. But how do you know how far to left it should be skewed? How do you measure the actual pixel distribution?

The answer is you don't. You meter accordingly and use -2 stops EV compensation. Or whatever adjustment experience suggests. Because we know what the sort of scene we are looking at will need. We get to know by practising. If necessary, at first bracket a little and see what difference it makes.

As regards metering in the centre of the frame, well you don't do this either. You tie the exposure to the focusing point and use centre spot only. That way, you are taking the reading from the point you are focusing on, after which you recompose and shoot.

But mu advice is not to do what I do. My advice is to understand the principles of exposure and then choose the techniques that best suit the chosen working method. That could be way I do it, but there are other equally valid ways.

Exactly and also the best way to stifle creativity.

When I go out shooting I like to look at life and take pictures of it. I can think of nothing worse than going out and constantly looking at a screen of parameters of graphs.

You should be looking at your extensive point by point checklist.




btw I've never used the histogram in the camera. The only time I use the histogram in Lightoom is when I'm doing some colour balancing and need to see how R, G amd B balance at various luminosities. And even then I use my eye to finalise the setting

Yes, we're not trying to offer a dissertation on the whole principles of metering The skills in assessing exposure mentally (for film - no histogram available!) and applying compensation are well ingrained no doubt and very reliable for those who are used to it. And we are lucky that modern cameras can pretend to be much less sophisticated in order to simulate the older types.

But that of itself does not make centre-weighted the best option now or indeed even the most reliable. the advantage of multi-segment is that it will even out the errors caused by metering non-mid tones, whilst applying the most common adjustments based on the distribution of brightness around the whole scene.

Regarding the use of the histogram, yes it needs understanding, just like metering techniques in general. But it is a far more accurate way of ensuring you record the full range of tones needed than just picking an amount of EC that might have worked on something before.

Understanding exposure and how to meter, fine; I'm all for it.

But a camera metering system that "understands" a scene and gets the exposure right very quickly without muchneed for intervention? That sounds great too! IF it works ;P

It isn't possible to have a 'far more accurate way' than a technique that already gets the exposure right. If you have the right exposure you have the right exposure.

All it is possible to do is to get it right more often. And my experience is that the fancy multi-segment metering and evaluative algorithms get it right less often than the human brain. Possibly helped because it's my goalposts for what is right