M8 manual metering for highlights.

[lars_bergquist]

Advertisement (gone after registration)

I'm with Sean on those (few) occasions when neither straight metering nor incident will work. Before I got my SF24D I used test-and-chimp for bounce flash with a non-swivelling little Metz flash and a Nikon cord. It works.

 

But for continuous light, incident gives you that certainty that no other method can produce.

 

The old man from the Age of No Meters

[ChrisC]

.....I am not clear on one point. Are you saying that your technique gives you good results out of camera, or are you using curves in RAW on most images?....

 

Guy - I'm saying that by metering the highlights as I described I get, as Sean says [Hi Sean] :

 

........ the histogram nearly kisses the right side of the graph.

 

But more reliably, more efficiently, and without the slightly apprehensive histogram chimping that I had before I changed technique. Optimising the recording Levels as in the 'expose to the right' principle is by my reckoning the best result. If by a good result out of camera you mean an un-tweaked image, that's not how I work, and the image mid point will need setting in the RAW conversion, after which many post-production tools become fair game. I hope this answers your question.

 

But why not try it if, like me, you had uncertainties with the M8's mid grey metering? Try setting your meter to + 1 2/3, if it's sunny outside; try manual metering off the brightest subject [maybe that white fluffy cloud], see what you get on your histogram. Because the metering is mid-point centric and not highlight centric, you will likely have to increase shutter speed a tad if it is a less contrasty day. It works for me.

 

............... Chris

[lars_bergquist]

[ ... ] But more reliably, more efficiently, and without the slightly apprehensive histogram chimping that I had before I changed technique................ Chris

You are still relying on average approximations, a.k.a. guesswork. It will probably work more often than not. Is that enough for you? If so, you're welcome to it.

 

The old man from the Age of No Meters

[guywalder]

Hi Chris,

thanks, I thought you must be working the files in post processing. Personally I prefer to keep pp to a minimum, so unless there is a large expanse of sky that would otherwise blow out I usually meter off my main subject. Although I always set my exposure manually and will meter of whatever part of my subject is appropriate.

 

Hi Lars,

your crusading enthusiasm for incident metering is touching, but its becoming rather repetitive. I presume you are not familiar with the idea that there is no such thing as correct exposure, there is only ever an exposure chosen by the photographer to provide the interpretation he/she choses. One of the joys of the medium is that any and all of its 'rules' are flexible...

[ChrisC]

You are still relying on ... guesswork... you're welcome to it.

 

Lars - Try it. Don't try it. It actually doesn't matter to me either way because I'm not crusading, I'm sharing an experience which I thought others might find useful. If others aren't interested that's fine too. But I'm curious that you equate incident metering, then transferring the reading to a [supposedly] fast handling camera, and jiggling the exposure because of your own 'reading' of light conditions, followed no doubt with a secondary exposure if your first blows the highlights, with; accuracy. Yet my reported experience of reliably hitting an optimum histogram is; 'guesswork'. I'm surprised that you didn't recognise immediately that what you wrote was patronising.

 

 

.... I thought you must be working the files in post processing......

 

Guy - In the context of this discussion, please don't read too much into that fact. I believe in interpreting files by building accents in the image through post production - it comes from my past exhibition work, and far too long spent darkroom printing, The only thing you should note is that, as Sean said, I too want my histogram to nearly kiss the right side of the graph. I hate blown highlights, and don't want them in my work.

 

I'm surprised no-one responded to the idea of a highlight-centric metering option to set the brightest metered highlight at the highest histogram level that's printable.

 

................ Chris

[sean_reid]

I'm surprised no-one responded to the idea of a highlight-centric metering option to set the brightest metered highlight at the highest histogram level that's printable.

................ Chris

 

It's an interesting idea. The metering would have to be pretty sophisticated and very accurate. A camera that meters from a live view feed might work well with this.

 

Cheers,

 

Sean

[jevidon]

Advertisement (gone after registration)

Chris,

I'll throw my oar in the water only for the purpose of commending you as I would commend anyone that has a refreshing new approach. I understand that you put the idea out, not as an end-all solution to the vagaries of lighting, but only as a novel approach which seems to work for you and you intended to share it. Bravo. I'll be giving it a try. Maybe it will work and maybe it won't. But I won't put you down for putting out there. That is what forums are for. BTW I will also try the approaches that some of your "critics" spoke of, although I really don't think they were criticizing. so much as being, perhaps, a trifle blunt in expressing their opinions. Or, maybe not, since a forum by it's nature does not permit knowing one's true intentions. Thanks for the suggestion.

Jim Evidon:)

[lars_bergquist]

Hi Lars,

your crusading enthusiasm for incident metering is touching, but its becoming rather repetitive. I presume you are not familiar with the idea that there is no such thing as correct exposure, there is only ever an exposure chosen by the photographer to provide the interpretation he/she choses. One of the joys of the medium is that any and all of its 'rules' are flexible...

Any repetitiveness has been caused by repetitive misunderstanding of what I wrote. I simply wanted to make myself clear.

 

No, there is no absolutely 'correct' exposure. Neither have I claimed that. But any intelligent interpretation has to be founded on some fact, not on a wild guess. And the digital sensor's total lack of exposure latitude is a fact that can't be ignored.

 

After all, both Edward Weston and Ansel Adams got themselves light meters when these became available. They did not like guesswork, either.

 

The old man from the Age of the Hand Meter

[giordano]

I'm surprised no-one responded to the idea of a highlight-centric metering option to set the brightest metered highlight at the highest histogram level that's printable.

 

IMO this approach is extremely useful if you're using a narrow-angle spot meter. With the M8 it's much less so because the brightest important highlight tends to be smaller than the area covered by the meter.

 

Maybe the M9 could have a built-in, non-TTL 1 degree spot meter as well as the averaging meter ... and an incident light meter for the Old Man and me:o

[ChrisC]

Sean, Jim, John - Thanks

 

.....Maybe ..TTL 1 degree spot meter as well as the averaging meter ..........

 

Possibly, providing it were switchable to 'highlight-centric'. My experience of many years of 10 degree metering [i started with 5, but found 10 degrees faster and better for [b]me[/b]] leads me to think that even 10* would work fine with intelligent aiming. How about a highlight option that just uses the area of the rangefinder patch? Whether or not there are insurmountable engineering problems with the last suggestion; I have no idea.

 

.................... Chris

[nhabedi]

...metering the brightest highlight so the red metering dot is showing plus the right hand arrow, and then dialling back four clicks [two stops] for an alternative way of getting a good exposure. I tried this for a while, and then decided I may as well set my meter to +1 2/3 exposure compensation for simple reading of brightest tones to give my exposure without having to dial back the shutter speed wheel.

 

Coming a bit late, but here's a (probably dumb) question from a digital newbie.

 

I never used this technique myself but it sounds interesting. What I don't fully understand is why you're compensating by 1 2/3 stops. When I read your description I thought that you'd have to compensate more by 3 to 4 stops given that a digital sensor can capture 7 stops.

 

Why only 1 2/3 stops? Is it because you're deliberately "underexposing" because you know that you can resurrect the dark areas on the PC? Or is it because you're adjusting for the fact that the M8 meter is not a spot meter, i.e. with a real 1 degree spot meter the correction would be more than 1 2/3?

 

Thanks,

Edi.

[pgk]

I'm surprised no-one responded to the idea of a highlight-centric metering option to set the brightest metered highlight at the highest histogram level that's printable.

 

................ Chris

 

The idea is good but the reality is that depending on the scene the photographer still has to determine exactly where this is and whether some highlight clipping is acceptable and if so how much - all dependant on image content.

 

As an example can I quote a similar problem from underwater photography which is effectively taking this problem to an extreme level? Underwater sunbursts were relatvely simple with film (due to high intensity recirocity law failure which tonally compressed exteme highlight detail), but aren't with digital because depending on the intensity of the sunburst through the water, the photographer MUST determine how much clipping is acceptable, because there will be clipping.

 

OK, this isn't relevant to the M8 but the principle is. Any extreme highlights will pose a problem and by far the easiest method of dealing with this is to zoom in on the relevant area when 'chimping' and reset exposure accordingly. If you do this in manual mode I cannot see how it differes in speed of execution nor difficulty compared with using a hand held meter or even a spotmeter. It might of course go against the grain.....