The Optimum Digital Exposure

[Fine_Art]

The article makes perfect sense.

Why don't our cameras, especially any over $500, have an ETTR exposure mode built in? Surely it is just a software addition to the metering reading?

People should be able to specify 100% within the matrix or 99.5% within the matrix in ETTR mode. The manufacturers should understand exposure enough to build this in.

[markd61]

It seems that I have been doing this for a few years now. I evolved into this out of a desire to maintain shadow detail.
 Never mind that a final image had deep info-free shadows!
I found that my files were naturally being "overexposed" by about a stop on average yet my processed images seemed very good. Rarely do I have blown highlights and if I do it is in scenes where the outdoor sun has introduced backbreaking contrast.

I agree that it seems that the camera manufacturers "should" be able to have an exposure matrix that could detect the maximum exposure and the minimum exposure and create an exposure that would then be optimal. However as a layperson these things always seem to be a matter of a weekend's worth of effort rather than the gigantic task they actually are.

[Ray]

The article makes perfect sense.

Why don't our cameras, especially any over $500, have an ETTR exposure mode built in? Surely it is just a software addition to the metering reading?

People should be able to specify 100% within the matrix or 99.5% within the matrix in ETTR mode. The manufacturers should understand exposure enough to build this in.

I suspect that part of the reason is that the inexperienced amateur would get confused to sometimes see such dark images produced by his camera. Often, the effect of a built-in ETTR exposure mode would be to cause a significant underexposure of the most important part of the image in order to avoid blown highlights in less significant parts of the image, such as patches of sky visible through gaps between the branches and foliage of a tree.

The experienced photographer will tend to make an assessment regarding which parts of the scene he wants to be as noise-free as possible. He may decide to sacrifice less significant detail in a white wall or sky in order to get more pleasing and noise-free results for a subject in the shade. A camera in ETTR mode could not make such decisions.

Such a feature could also be very negative for certain brands of cameras, such as Canon, which have as much as 2 EV lower dynamic range than Nikon at base ISO. It would be just too easy for even a complete amateur to shoot the same high-contrast scene with a Canon and Nikon DSLR, with both cameras in ETTR mode, and see how much noisier the shadows are in the Canon images. So there would clearly be a big disincentive for Canon to introduce such a feature, unless it were to simultaneously raise the DR performance of its cameras. 

Also, if Canon were to significantly improve the DR of its cameras, there would be less need to develop an ETTR mode, just as there is already a reduced need for Nikon to do so, because of the wide DR of its current DSLRs.

[Torbjörn Tapani]

My favourite post on ETTR comes from Pekka Potka, it works with EVF and I first read about it in this post about E-P3: http://www.pekkapotka.com/journal/2011/12/20/expose-to-the-right-ettr-with-e-p3.html?rq=ettr

No need for bracketing.

Maybe Pekka Potka is not so well known but I can recommend his blog http://www.pekkapotka.com/journal/ 

[JohnBrew]

Interesting. I've been trying to work with this idea in mind lately, however, I find it very camera specific. Or sensor specific, if you like. On my Leica M8.2 (CCD) you absolutely cannot overexpose, because if the highlights are blown - they are forever blown. With my D800 (CMOS), which has a totally different sensor, I usually overexpose by a 1/3 stop.

[michael]

The article makes perfect sense.

Why don't our cameras, especially any over $500, have an ETTR exposure mode built in? Surely it is just a software addition to the metering reading?

People should be able to specify 100% within the matrix or 99.5% within the matrix in ETTR mode. The manufacturers should understand exposure enough to build this in.

I have been evangelizing for an ETTR exposure mode now for some years, whenever I meet with camera company product development people. In Live View mode the camera knows exactly what every pixel is reading. Piece of cake. Easy as pie.

One should even be able to specific a % of "blown" pixels to accommodate specular highlights.

What happens when I proposethis? Some nod sagely and say yes, good idea..."we will study", which in some parts of the world means..."we didn't think of it so fuggetaboutit". Incredibly there are some that actually just don't get it. They just nod, look like they understand, but you know from their eyes that they don't. Just incredible.

Arde we surprised then that we still have cameras from major manufacturers that don't have blinkies or zebras for judging ETTR exposure oneself, and some that don't even have histograms on instant review. No names, because I don't want to start a pissing match, but you know who they are!

Michael

[michael]

Interesting. I've been trying to work with this idea in mind lately, however, I find it very camera specific. Or sensor specific, if you like. On my Leica M8.2 (CCD) you absolutely cannot overexpose, because if the highlights are blown - they are forever blown. With my D800 (CMOS), which has a totally different sensor, I usually overexpose by a 1/3 stop.

John. You can't "overexpose" a sensor. Once the wells are full the photons fall on the floor. It's a brick wall.

But some cameras don't give accurate histograms. Most are not based on the raw data but on an sRGB JPG. This is how you lose a stop or more of your data.

You need to run some tests and compare what the camera says is overexposed with what really is when the histogram is viewed in a raw processing program.

Michael

[BernardLanguillier]

Interesting. I've been trying to work with this idea in mind lately, however, I find it very camera specific. Or sensor specific, if you like. On my Leica M8.2 (CCD) you absolutely cannot overexpose, because if the highlights are blown - they are forever blown. With my D800 (CMOS), which has a totally different sensor, I usually overexpose by a 1/3 stop.

John,

I don't think this has to do with CCD vs CMOS. This probably is just about how the ISO value was calibrated.

Overall this article is a good summary, but I am a bit unclear what is new relative to the original ETTR concept?

Yes, we can use bracketing in order to increase the chances we get one image optimally exposed (from an ETTR standpoint). Now, with the latest Exmor sensors, where is so much DR available anyway that under-exposing one stop relative to ideal ETTR has a fairly limited real world impact.

Cheers,
Bernard

[deejjjaaaa]

> I have developed an exposure method

...

> Method #1 - Using a Spot Meter

http://www.rawdigger.com/houtouse/lightmeter-calibration

no ACR no LR no books, no "OneZoneTM Digital Exposure Method" - thank you

> The “Blinkies” on the back of your camera occur about 1 stop before the highlight warning in your software – highlight “Clipping”!

blinkies (or zebra) in my setup for E-M1 and A-7 are happening exactly where rawdigger says raw clipping is (with << 1/3EV precision)  ... but granted I live with not so nice OOC JPGs to get there.

[BobD]

I have been evangelizing for an ETTR exposure mode now for some years, whenever I meet with camera company product development people. In Live View mode the camera knows exactly what every pixel is reading. Piece of cake. Easy as pie.

What happens when I proposethis? Some nod sagely and say yes, good idea..."we will study" ... Incredibly there are some that actually just don't get it. They just nod, look like they understand, but you know from their eyes that they don't. Just incredible.

Michael

Micheal, I totally agree this can so easily be built in to as a menu choice in digital cameras.  I can only re-new the discussion and hope to move the ball a little further to that end. However, observing your pleas to camera manufactures over the years with such little response (and incredibly that they "just don't get it") I can only wonder if we will ever see it.

Years ago, when I worked in R&D at Polaroid, I witness one of the great visionaries Dr. Land work for years and years on instant movies... code name "Sesame". Whenever we thought we were getting closer to a viable product... it was "not ready yet" and again more years were added to the development of Sesame. It was finally ready for market about 2 years after the introduction of Betamax and VHS tape!

I appreciate you relentless pursuit of pushing this industry forward. Your recent video stating that there has been very little new in the development with digital cameras is so true. An "ETTR exposure mode feature" in digital cameras is easily doable. We'll see.  My hope is it doesn't become a digital "Sesame"... my guess is, if it does happen, it will probably be heralded as a brilliant NEW idea!

Thank you for all you efforts and the forum you have created for voices and thoughts to be shared.
Bob

[Chrisso26]

I have my Sigma Merrill set to overexpose +0.3 for every shot. Isn't this effectively ETTR.
On the BM Pocket Camera I have zebras set to 95%. I set my exposure so the brightest part of the image is not quite triggering the 95% zebras. I know it's not exactly accurate, but my graded results have been pretty good so far.

[Fine_Art]

Part of the package the manufacturers need to figure out is:
the photog still gets to adjust +/- ev from ETTR for things like chrome, clouds, etc.
their jpg processing engine should produce a "normal" exposure from the ETTR data.

They may be ignoring the input thinking about their jpgs for casual users.

[Fine_Art]

I suspect that part of the reason is that the inexperienced amateur would get confused to sometimes see such dark images produced by his camera. Often, the effect of a built-in ETTR exposure mode would be to cause a significant underexposure of the most important part of the image in order to avoid blown highlights in less significant parts of the image, such as patches of sky visible through gaps between the branches and foliage of a tree.

The experienced photographer will tend to make an assessment regarding which parts of the scene he wants to be as noise-free as possible. He may decide to sacrifice less significant detail in a white wall or sky in order to get more pleasing and noise-free results for a subject in the shade. A camera in ETTR mode could not make such decisions.

Such a feature could also be very negative for certain brands of cameras, such as Canon, which have as much as 2 EV lower dynamic range than Nikon at base ISO. It would be just too easy for even a complete amateur to shoot the same high-contrast scene with a Canon and Nikon DSLR, with both cameras in ETTR mode, and see how much noisier the shadows are in the Canon images. So there would clearly be a big disincentive for Canon to introduce such a feature, unless it were to simultaneously raise the DR performance of its cameras. 

Also, if Canon were to significantly improve the DR of its cameras, there would be less need to develop an ETTR mode, just as there is already a reduced need for Nikon to do so, because of the wide DR of its current DSLRs.

You are right for Canon. However, if an Exmor using company like Nikon does make an exposure toggle "Expose grey/ Expose ETTR", they will take even more customers from Canon. It is in all their best interests to not be late to the party once it starts.

[BobD]

Part of the package the manufacturers need to figure out is:
the photog still gets to adjust +/- ev from ETTR for things like chrome, clouds, etc.
their jpg processing engine should produce a "normal" exposure from the ETTR data.

They may be ignoring the input thinking about their jpgs for casual users.

The default can be for the "casual user".  However, there are so many pages of menus - if the have a menu item for for "back focus" there can surly be a menu item for ETTR!

[BernardLanguillier]

I appreciate you relentless pursuit of pushing this industry forward. Your recent video stating that there has been very little new in the development with digital cameras is so true. An "ETTR exposure mode feature" in digital cameras is easily doable. We'll see.  My hope is it doesn't become a digital "Sesame"... my guess is, if it does happen, it will probably be heralded as a brilliant NEW idea!

Bob,

What is your view about the value of the highlight priority metering mode of the D810?

Since it is linked with the spot meter mode of the camera, it seems like a pretty good solution, isn't it?

Cheers,
Bernard

[ErikKaffehr]

Hi,

In my view the article has a very low noise to fact ratio, but I also feel that it has several serious issues. The foremost issue is that it ignores the fact that both ACR and Lightroom do excessive highlight recovery, and the histogram on LR (and I presume ACR) takes that recovery into account. So you can get a pretty good histogram in LR or ACR, where much of the highlight information is actually blown out.

The only histogram that can be trusted is a raw histogram like the ones in RawDigger, an excellent learning tool.

My experience is that the histograms on both my Sonys and the P45 are quite OK.

Exposing based on spot metered highlights may be a good idea. But those highlights may be hard to locate and to measure.

I would say that an exposure program based highlight preservation make may a lot (or at least some) sense.

Optimal exposure strategy may differ between sensors. Canon sensors may need a different exposure strategy than Sony Exmore sensors, due to different handling of pre ADC analogue gain for different ISOs.

Best regards
Erik

[Torbjörn Tapani]

Erik, not really I think. ETTR is the same for every sensor. Meaning to gather as much light as possible. ISO and shadow priority is a different matter.

[Torbjörn Tapani]

I have my Sigma Merrill set to overexpose +0.3 for every shot. Isn't this effectively ETTR.
On the BM Pocket Camera I have zebras set to 95%. I set my exposure so the brightest part of the image is not quite triggering the 95% zebras. I know it's not exactly accurate, but my graded results have been pretty good so far.

Not exactly. ETTR might mean you underexpose from camera metering in some rare case. It all depends on relevant highlights.

[bjanes]

In my view the article has a very low noise to fact ratio, but I also feel that it has several serious issues. The foremost issue is that it ignores the fact that both ACR and Lightroom do excessive highlight recovery, and the histogram on LR (and I presume ACR) takes that recovery into account. So you can get a pretty good histogram in LR or ACR, where much of the highlight information is actually blown out.

The only histogram that can be trusted is a raw histogram like the ones in RawDigger, an excellent learning tool.

I agree that the RawDigger histogram is the preferred tool for histogram analysis, but one can obtain a reasonable approximation of the status of the raw RGB channels after white balance in ACR or LR by setting the color space to ProPhotoRGB and using PV2010 with the sliders on the main panel zeroed out (set to 0).

My experience is that the histograms on both my Sonys and the P45 are quite OK

A major limitation of the histograms on my Nikon D800e is that they are based on the JPEG preview image, and the widest color space is Adobe RGB. The luminance histogram may not detect channel clipping in the blue or red channel, since it is heavily weighted for the green channels. The RGB histograms may show saturation clipping in the AdobeRGB space when the raw channels are intact. For daylight white balance, the red multiplier is near 2.0 and the red channel histogram shows the red channel values after white balance, effectively moving the histogram to the right by one stop. When photographing red flowers, one may need to underexpose by several stops to eliminate clipping in the red histogram. The camera needs to offer ProPhotoRGB as one of the camera renderings.

Regards,

Bill

[Bart_van_der_Wolf]

When photographing red flowers, one may need to underexpose by several stops to eliminate clipping in the red histogram.

I agree with Bill, flowers can have amazing saturation thus presenting not only overexposure risk, but some also risk of underexposure, e.g. with near zero Blue and very high Red channel signals (at the same time).

RawTherapee at least offers a solution by allowing to change the per channel pre-demosaicing signal levels, I wish other converters allowed such manual control.

Cheers,
Bart