How do you measure "correct exposure"

[Dubaiian]

I have been taking pictures for a couple of years now and mostly use manual, Av or Tv and use the viewfinder meter to tell me when something is correctly exposed. I also understand things like centre exposure, spot exposure, average exposure etc in camera. Additionally I usually check what Lightroom thinks in the develop menu. This seems to produce pleasing results most of the time.

My question comes from recently starting to use a Tilt Shift lens to capture some of the modern architecture in Dubai (where I live) as I like to see the straightened buildings but of course once you tilt or shift the in camera metering is meaningless. So the questions is, should I be using a scientific way to measure exposure or is the naked eye on the LCD and or Lightroom to see whats pleasing enough.

Excuse me if this is asked anywhere else, but I could not see it.

 

[pgriz]

Use your histogram to tell you how you're doing. Try to avoid having (m)any pixels on the right. I use ambient light reading (handheld meter) to tell me what the exposure should be based on the dominant light source. Then I spot-meter the highlights in the image to see if any will exceed my camera's dynamic range. With this data in mind, I will set the exposure and then examine the resulting image's histogram. If the bulk of the pixels are on the right half, and none are clipped (ie, right against the right side), then I have a good exposure and can extract most info from the image when processing from RAW.

You need to follow the process to see what a "good" exposure in camera translates to in Lightroom.

 

[Buckster]

...but of course once you tilt or shift the in camera metering is meaningless.

I'm interested in this aspect of the OP's question. Why would the camera metering be meaningless? Shifted, tilted or not, it's still the actual image through the lens projected into the camera's metering system. So, what would render it "meaningless"?

 

[KmH]

"Correct exposure" is an artistic choice, but there are some guidelines.

Because of the way digital images work - http://www.adobe.com/digitalimag/pdfs/linear_gamma.pdf - one of those guidelines is ETTR - Expose To The Right.
Exposing to the right - Wikipedia, the free encyclopedia
ETTR
Optimizing Exposure

DSLR light meters are calibrated based on the assumption that a majority of scenes average out to having a reflectance equivalent to a 12% to 18% neutral gray.
Some scenes fall outside that average which causes the light meter to allow an under exposed or over exposed scene though the meter indicator was centered.
Photographer knowledge/experience is used to know when exposure compensation is appropriate in that kind of a situation.

For that reason many photographers use a hand held incident/reflected/flash light meter.

Understanding Camera Metering and Exposure
Understanding Digital Camera Histograms: Tones and Contrast
Understanding Digital Camera Histograms: Luminosity and Color

 

[amolitor]

...but of course once you tilt or shift the in camera metering is meaningless.

I'm interested in this aspect of the OP's question. Why would the camera metering be meaningless? Shifted, tilted or not, it's still the actual image through the lens projected into the camera's metering system. So, what would render it "meaningless"?

Ditto.

 

[Dubaiian]

...but of course once you tilt or shift the in camera metering is meaningless.

I'm interested in this aspect of the OP's question. Why would the camera metering be meaningless? Shifted, tilted or not, it's still the actual image through the lens projected into the camera's metering system. So, what would render it "meaningless"?

Very good point Buckster and focuses my question. This is my query. If I use a standard f2.8 lens the in camera meter shows me the central point to match my exposure dependant upon settings. But. If I use my 24mm TSE without adjustment it does the same, as soon as I tilt or shift, th meter goes haywire.


Should I expose based on no T/s nd the. Move the lens?

 

[KmH]

Should I expose based on no T/s

Yes.

 

[Buckster]

Should I expose based on no T/s

Yes.

Why does it "go haywire" Keith?

 

[amolitor]

If I use my 24mm TSE without adjustment it does the same, as soon as I tilt or shift, th meter goes haywire.

What do you mean by "goes haywire"?

The metered exposure may well change when you tilt or shift, for several reasons. So, if you meter-match in manual mode, and then tilt/shift, the exposure you dialed in may well be wrong now. Probably not wildly so, but wrong.

You need to tilt/shift, set the composition up the way you want it, and then meter-match.

 

[Derrel]

Should I expose based on no T/s

Yes.

Why does it "go haywire" Keith?

Both Canon and Nikon specifically recommend that metering be done with their tilt/shift lenses in non-tilted positions. I do not own a tilt/shift lens, so I cannot say why, but the longstanding wisdom is that tilt/shift exposures are best measured with the controls zero'd out, at neutral. I just visited the Nikon USA website, and the professional panorama expert's article on using the PC-E 85mm lens strongly suggested metering and doing a test shot with the lens not shifted...

I dunno...I bet HelenB knows....she's a smart cookie AND she uses tilt/shift lenses all the time.

 

[amolitor]

Huh, that's interesting.

What if you shift the sun into the frame? Or out of it? I can definitely see metered exposures going all over the place, as you introduce more or less of Dubai's sky into the frame.

For these applications maybe you ought to meter the subject, and let the sky fall where it may. On the other hand, maybe the OP means that meter starts yelling at him and blinking wildly, in which case something else is going on. Possibly the T/S lens is saying "I AM SHIFTED" and the camera is saying "OK ALL METERING BETS ARE OFF BLINK BLINK BLINK BEEEEEEEP" or something. I have only one camera that tilts and shifts, and it doesn't use electricity.

 

[jowensphoto]

OP, are you free lensing, maybe? That is the only thing I can come up with other than what Derrel said.

 

[cynicaster]

OP, you'd do well to learn how to read histograms, assuming you haven't already. In a generic low/medium contrast scene, you're probably going to find that a mound in the center of the histogram that trails off at the edges will get you a nice result most of the time. But the histogram won't always be a mound. A very high contrast scene, for example, might generate 2 or 3 "pillars" on the histogram. Sometimes, it is possible to discern with reasonable certainty what parts of the image correspond to the discrete pillars--which is great info to have in the field, to ensure you get the shot right.

Ever have those moments where you're all excited about a certain photo you took, can't wait to get it into the computer to pretty it up, only to end up horrified that the picture looks pitch dark on the big screen? Histogram reading is not a panacea for such letdowns, but it is leaps and bounds more reliable than judging results by simply looking at the preview on your LCD.

"Correct exposure" is an artistic choice, but there are some guidelines.

Because of the way digital images work - http://www.adobe.com/digitalimag/pdfs/linear_gamma.pdf - one of those guidelines is ETTR - Expose To The Right.
Exposing to the right - Wikipedia, the free encyclopedia
ETTR
Optimizing Exposure

IMO, every photographer--even hobbyists--should do their best to understand the ETTR principle and when it can be used to advantage. Unfortunately, I think a lot of people's faces glaze over the moment they bump into terminology like "digital levels" and "bits", but I think the layperson need not even touch on those terms. The important take-away can be sufficiently communicated in rather plain English: "The digital camera's sensor can only record a certain range of brightness, which is less than the human eye can see. At the upper end of that range, the digital camera is capable of capturing smaller differences in brightness than it is at the lower end of the range, therefore more overall detail can be recorded in a digital image by using the upper end of the range as much as possible, without getting blinkies on the preview. Due to this additional captured detail, darkening a bright exposure in photoshop is preferable to brightening a dark exposure."

I'd suggest that the ETTR method doesn't deliver some ambiguously defined "correct exposure"; rather, in the context of the overall workflow from the shutter press to printing, ETTR delivers the "most versatile" exposure.

EDIT: it seems there were a pile of replies that popped up while I was typing so this response seems a bit incongruous, but I'll leave it anyway.

 

[Derrel]

I did a bit of quick link-looking. Here are some web pages that might be of some interest on this issue of wonky metering with T/S lenses.

Metering with a Tilt-Shift Lens - Photography Stack Exchange


Tilt shift lens help


Canon TS-E 45mm f/2.8 Tilt Shift Lens Review


Tilt shift - the Canon TS-E 24mm 3.5L - a review

 

[timor]

Maybe after all the best thing for OP would be a handheld spot meter and better understanding of light metering process instead of relying only on software in cameras. Looks like with all the sophistication of modern DSLR, they still could be easily fooled.
Derrel, thank's for links.