Why no automatic exposure to the right w/o blown highlights?

[stephen_asprey2]

<p>I think this is why we film refugees who have been dragged kicking and screaming into digital, wish for the good old days!</p>

[peterblaise]

<p>.</p>

<p>"<em>Getting it right</em> " in camera is a photographer's artistic decision, and cannot be automated as an option "ETTR Exposure To The Right" rule since no programming foresight can anticipate what light coming through the lens is subject light the photographer wants, and what light is miscible and can be allowed to saturate. Is UV and IR important or not? How much? Does the photographer want detail in the water's reflection of the sun, or detail in the water's reflection of the sky? Hence the paucity 1,000:1 or 100,000:1 dynamic range sensors to match everyday scenes. If we're stuck with recording scenes with today's limited dynamic range sensors (~5-10:1), then we still have to be there as photographers making artistic decisions.</p>

<p>"<em>ETTR when done correctly</em> " <strong>can't</strong> be done "correctly"! ETTR can't be locked down as exactly only one way of doing ETTR, considering in-camera live histograms are JPG based, and who knows what metering variables versus sensor variables don't match, not just number of points used in metering versus points in the sensor, but also differing response curves, and so on. And if specular highlights are "empty", does one ETTR specialist discard them while another ETTR specialist includes them? ETTR is "just" another artistic photographer tool to interpret and apply by experience of one's own camera and metering and choices. Hey, we're back to artistic choices again!</p>

<p>"<em>ETTR defers the zone placement to the processing stage rather than the exposure stage</em> " is pretty much the promise and problem here, and still, the photographer must choose their compromise:<br>

- capture highlight detail and bring up shadows <em><strong>and noise</strong> </em> later,<br>

- capture shadow detail and <em><strong>fake the highlight detail</strong> </em> later.<br>

Decisions, decisions. ARTISTIC decisions, and that's why the copyright goes to the PHOTOGRAPHER, not to the gear or the gear designers.</p>

<p>"<em>Film refugees dragged kicking and screaming into digital, wish for the good old days</em> " -- what's wrong with turning our camera around, slapping on our http://www.expodisc.com/ and setting our exposure for our interpretation of the <em><strong>incident light</strong> </em> and our assessment of the scene, then turning the camera around, taking off the ExpoDisc, and shooting? Digital has nothing to do with it EXCEPT the misuse of immediately available little tiny JPG UFHOs Unidentified Flying HistOgrams.</p>

<p>Photographer -- it's a tough job, but somebody's gotta do it!</p>

<p>.</p>

<p>PS - "<em><strong>transferred</strong> to the Casual Conversations Forum</em> " -- cool tool! <br>

"Admin" -- it's a tough job, but somebody's gotta do it! <br>

Thanks, Lex.</p>

<p>.</p>

[bernardwest]

<blockquote>"ETTR when done correctly " can't be done "correctly"!</blockquote><p>

 

Of course it can, and it IS done correctly numerous times a day by photographers around the world. The principle of ETTR is this: 'Overexpose' a scene by a X stops over what you the photographer would consider a 'correct' exposure, and then in post-processing, reduce the exposure by X stops to 'normalize' the image. All this worry about tone curves and gamma and whatnot mentioned by posters above is irrelevant. As long as you have a capable raw converter which works directly on the raw sensor data in a linear fashion (eg. Lightroom or DCRaw; but not Canon's DPP), your 'normalized' image will look exactly the same as your 'correct' exposed image, except that it will have significantly better shadow noise and detail.<p>

 

<blockquote>ETTR can't be locked down as exactly only one way of doing ETTR, considering in-camera live histograms are JPG based, and who knows what metering variables versus sensor variables don't match, not just number of points used in metering versus points in the sensor, but also differing response curves, and so on. And if specular highlights are "empty", does one ETTR specialist discard them while another ETTR specialist includes them? ETTR is "just" another artistic photographer tool to interpret and apply by experience of one's own camera and metering and choices.</blockquote><p>

 

No argument there. That's the beauty of digital however. Have an automatic ETTR setting on your camera with some user defined preset parameters. If you don't like the ETTR it spits out at you, then you can do it yourself via the lcd histogram. But I would be willing to bet that the camera processor and the exposure sensor would do a much better job of it than the camera operator trying to do the same thing on possibly a luminosity histogram of a gamma encoded, white balanced, JPG no less!

[mt4x4]

<p>To the ORIGINAL POST:<br /> <br /> Nikon DSLRs already do this.<br /> <br /> This is what the blinking highlights in image review are all about. All you have to do is relegate to <a href="http://en.wikipedia.org/wiki/Chimping">chimping</a> once and a while.</p>

<p><img src="http://farm1.static.flickr.com/167/478452393_0baa8a3f60.jpg?v=0" alt="" width="500" height="379" /> <br /> <strong>Chimpers</strong></p>

[peterblaise]

<p>.</p>

<p>Okay, for our archives, let's define the word "correct":</p>

<ul>

<li><a href="http://www.google.com/search?q=define%3Acorrect">http://www.google.com/search?q=define%3Acorrect</a> </li>

</ul>

<ul>

<li><a href="http://www.etymonline.com/index.php?search=correct&searchmode=none">http://www.etymonline.com/index.php?search=correct&searchmode=none</a> </li>

</ul>

<p>... and it seems <em><strong>we agree</strong> </em> regarding ETTR even if we think we disagree -- since there is no ONE "correct" way, then all ways are equally correct, and all ways are equally <em><strong>in</strong> </em> correct, and therefore, the use of the word "correct" is useless in this thread when trying to define how to use ETTR Expose To The Right "the correct" way -- unless we mean "according your your own artistic choices and whims", in which case, "correct" is not an absolute term. After all, what does saying "use ETTR correctly <em>according you your own criteria</em> " mean other than "there is NO "correct" way to use ETTR, but pick a way that serves you appropriately and don't get in your own way following rules you don't understand or that you misunderstand, rules that then work against your photographic goals". Argh!?</p>

<p> </p>

<blockquote>

<p>Define ETTR by <strong>Bernie West</strong> : "<em>... The principle of ETTR is this: 'Overexpose' a scene by a X stops over what you the photographer would consider a 'correct' exposure, and then in post-processing, reduce the exposure by X stops to 'normalize' the image ...</em> " Got it. Whatever PLUS you did in-camera over camera-recommended exposure, do an equivalent MINUS in Raw development. Uh-huh.<br>

----------<br>

Define "ETTR" by everyone else:<br /> <strong>John Shaw </strong> <a href="http://www.johnshawphoto.com/tips/workflow_outline.html">http://www.johnshawphoto.com/tips/workflow_outline.html</a></p>

<table border="1" cellspacing="0" cellpadding="0" align="center">

<tbody>

<tr>

<td width="300" valign="top">

<p><em><strong>JPEG</strong> </em></p>

</td>

<td width="300" valign="top">

<p><em><strong>RAW</strong> </em></p>

</td>

</tr>

<tr>

<td width="300" valign="top">

<p><em><strong>Centered </strong> </em> histogram; no clipping.</p>

</td>

<td width="300" valign="top">

<p>Histogram <em><strong>weighted to the right</strong> </em> side; no clipping</p>

</td>

</tr>

</tbody>

</table>

</blockquote>

<blockquote>

<p>... hmmm. Does "weighted to the right" mean ETTR?</p>

</blockquote>

<blockquote>

<p>----------<br>

<strong>Andrew Rodney</strong> : <a href="http://www.digitalphotopro.com/technique/camera-technique/exposing-for-raw.html">http://www.digitalphotopro.com/technique/camera-technique/exposing-for-raw.html</a> "<em>... You must [</em> MUST! YOU HAVE NO CHOICE, RESISTANCE IS FUTILE!<em>] attempt to fit the dynamic range of the scene within the dynamic range the capture device can record ... Expose to place as much data within this linear-encoded RAW image without losing highlight <strong>values you wish to reproduce</strong> [</em> <strong>= PERSONAL WHIM/ARTISTIC CHOICE</strong> <em>] ... </em> <em>ETTR presents a few problems, one being that the LCD camera preview, including the histogram and clipping indicators, isn't based on the linear RAW data. Instead, this preview is based on the rendered gamma-corrected JPEG your camera is set to produce, even if you don't save that JPEG and only shoot a RAW file! If your goal is to produce the best possible exposure for RAW, using the ETTR technique, the feedback on the LCD could steer you in the wrong direction ...</em> " So Rodney has yet to incorporate artistic whim in the calculation, instead using the camera as a presumed scientific recordation device, not an artistic expression device.</p>

</blockquote>

<blockquote>

<p>----------<br>

<strong>Rags Gardner</strong> <a href="http://www.rags-int-inc.com/PhotoTechStuff/ETTR/">http://www.rags-int-inc.com/PhotoTechStuff/ETTR/</a> "<em>... ETTR ... promoted ... as a replacement for traditional exposure metering ... premise is that you can validate camera metering by simply reading the histogram in the camera’s preview window ... based on some basic misunderstandings about digital photographic technology.</em> <br /> <em>- the premise that each bit level in a digitally encoded image represents an exposure stop</em> <em></em> <br /> <em>- the premise that all digital cameras capture light in a perfectly linear fashion</em> <em></em> <br /> <em>- the premise that the histogram represents the raw image data captured by the camera</em> <br /> <em>... briefly address each of these [misunderstandings] ...</em> "<br /> See also: <a href="http://www.rags-int-inc.com/PhotoTechStuff/">http://www.rags-int-inc.com/PhotoTechStuff/</a></p>

</blockquote>

<p>Oh, you all can Google search for more and more and more on ETTR than it makes any sense to quote here!</p>

<p>Me? I've been reading and re-reading the 30-or-more pages that make up a small book here, by <strong>Emil Martinec</strong> : <a href="http://theory.uchicago.edu/~ejm/pix/20d/tests/noise/">http://theory.uchicago.edu/~ejm/pix/20d/tests/noise/</a> I won't even try to pull a quote.</p>

<p>It's really fascinating, and I got the point in exploring the <em><strong>scientific versus artistic use of our cameras</strong> </em> . If we are not on the same page here, it may look like we disagree when in fact, we're just on different pages, and probably agree, even if we use different exposure techniques to accomplish our own personal whimsical artistic -- and scientific -- photographic goals.</p>

<p>.</p>

[lovcom_photo]

<p>Speaking of blown highlights, I find it exceedingly funny or silly when I hear about digital photographers avoiding blown highlights at any cost, and for all their compositions. They will see another's great landscape, but pooh pooh it because 1/10th of a cloud is blown...how rediculous!</p>

<p>This policy of "no clipping" is at best foolish.</p>

<p>The dirty secret is: Blown highlights are a reasonable compromise in order to get the subject exposed propely, and provided the blown highlights are not distracting, and not taking away from other elements of the composition. You got to be judicious about it, and never, never annal.</p>

<p>Often, a good comp is ruined because the photographer was too annal about making sure not even one pixel is blown. Often these types show under exposed flat images, lacking detail in the shadows.</p>

<p>If you're going to shoot digital, then accept the fact right up front that exposure is going to be compromised. What part of it is going to "pay"? The shadows? The highlights? A little of both? Until the DSLR makers provide us much wider DR, this is compromise we all have to make. </p>

[bernardwest]

Peter....<p>

 

I said<p>

<blockquote>what <b>you the photographer</b> would consider a 'correct' exposure</blockquote><p>

 

You said<p>

<blockquote>the use of the word "correct" is useless in this thread when trying to define how to use ETTR Expose To The Right "the correct" way -- unless we mean [aha!] "<b>according your your own artistic choices and whims</b>"</blockquote><p>

 

We are in agreement. You are making this more complicated than it needs to be. The question was about a setting on the camera to have the camera do ETTR automatically. It would be easy to implement, and could be turned off if one doesn't like the results, therefore, why NOT have it? Simple, really.

[peterblaise]

<p>.</p>

<p>The reason ETTR cannot be automated is because ETTR requires choices only an informed decision maker can make because YOUR ETTR is not MY ETTR is not the next person's ETTR and so on. There is no "correct" ETTR, so if we agree that there is no "correct" ETTR, then there is no "one" ETTR that any so-called photographic scientist can program into a camera system. G'luck tryin ' though.</p>

<p>Anyway, for blown highlights, I often fill them with a small percent noise so when printed, people think there's detail there. Works for me.</p>

<p>.</p>

[bernardwest]

<blockquote>The reason ETTR cannot be automated is because ETTR requires choices only an informed decision maker can make because YOUR ETTR is not MY ETTR is not the next person's ETTR and so on. There is no "correct" ETTR, so if we agree that there is no "correct" ETTR, then there is no "one" ETTR that any so-called photographic scientist can program into a camera system. G'luck tryin ' though.</blockquote><p>

 

The point is the individual can set parameters through the menu to customize the camera's implimentation of ETTR. It's still an artistic decision because you can review the results straight after.<p>

 

Anyway, the real point that you are missing is this: The user can't make an informed decision about ETTR from looking at a histogram (often only luminosity) of a jpg on a tiny lcd monitor. The jpg is white balanced (the major reason why jpgs blow highlights that aren't blown in the raw data), and the blinkies on the screen are usually a green weighted luminosity reading (i.e. you could be totally blowing the red channel, but the blinkies won't show it). So attempting to do ETTR this way is basically just guess work and experience. And this isn't anymore artistic or accurate than having the camera processor do the assessment on the LINEAR raw data.

[lovcom_photo]

<p>But Bernie, parameters on a menu often mean the processing is for the ENTIRE image. This is why JPG and in-camera settings like sharpening, contrast, saturation, noise reduction SUCK. It is big bang. All or thinging. Bad.</p>

<p>Peter nailed it. It is far better to let the human being decide how far to the right to expose. And a trained human can do this far "better" then any camera ever made. This is the same argument for going with raw over jpg....raw means the humand decides and NOT the machine.</p>

<p>And I realize the in-cam histogram is based on the teeny-tiny jpg gen'd for the purpose of the LCD, but it is a good approximation, and over time one can "calibrate" their perception for a particular camera model to ascertain how blown it really is.</p>

[peterblaise]

<p>.</p>

<blockquote>

<p>Earlier: "<em>... It is far better to let the human being decide how far to the right to expose ...</em> "</p>

</blockquote>

<p>I don't think manual photographer's decisions are "better" than automation ever could be <em>per se</em> , I just think ETTR can't be predicatively or accurately automated to everyone's or anyone's criteria or satisfaction -- we have do many differing and competing definitions of ETTR right here in this thread to consider any one of them "the one".</p>

<p>.</p>

[bernardwest]

<blockquote>Peter nailed it. It is far better to let the human being decide how far to the right to expose. And a trained human can do this far "better" then any camera ever made.</blockquote><p>

 

The point I am trying to make is that the human can't make an informed decision because they don't have access to the linear sensor data. The camera does. Anything you do to get the "best" ETTR for yourself, will be at best a 'guess'. Therefore, I say you have little to lose letting the camera have a go at it itself.

[lovcom_photo]

<p>Disagree Bernie...the in-camera histo is more then enough to ascertain good exposure, even if based upon the jpg image on the back LCD is not representative 100% of the raw image.</p>

<p>Peter, the human made manual choices are exceedingly better then the machine's.</p>

<p>How can a machine know the elements of a composition? It doesn't, and it never will.</p>

<p>See what I mean?</p>

[peterblaise]

<p>.</p>

<p>Dan, you don't mistakenly thing we disagree, do you? I'm the one suggesting that an auto-ETTR setting on a camera would mean nothing without a detailed explanation in the owner's instruction manual as to what ETTR choices of the many available that camera's programmers chose.</p>

<p>More important to me would be COMPLETE and ACCURATE information provided to the photographer so they could decide for themselves, and sadly, the data we now get is so approximate, and unique model to model, that all we are learning is the quirks of one model, and we cannot take that experience to our next camera. Oh well.</p>

<p>However, the camera marketers do disagree, and want us to believe there is an image database inside our cameras that new scenes get compared to, so really, we don't have to think, and we don't have to be afraid to buy a DSLR.</p>

<p>I won't share all the details of my historical exploration of fuzzy logic patents as applied to other cameras since the mid 1990s that successfully used metering pattern dynamic range information, and focus point information, and lens focus distance information to come up with a fuzzy solution on the fly, also incorporating high frequency movement in the auto focus video stream to adjust shutter speed for subject movement, and low frequency movement in the auto focus video stream to adjust shutter speed to avoid camera shake, although already incorporating lens focal length in the equation -- dynamic programming that anticipated non-specific groups of complex subject situations have been around for years!</p>

<p>On a pixel level, the baton has been passed on to <strong>Apical Imaging, UK</strong>, see<br /> <a href="http://www.apical-imaging.com/autoexposure">http://www.apical-imaging.com/autoexposure</a><br /> and browse their web site for some amazingly efficient chip-level exposure toys, some as small as 90 kb!</p>

<p>.</p>

[peterblaise]

<p>.</p>

 

<blockquote>

<p>Earlier: "<em>... human made manual choices are exceedingly better then the machine's ...</em>"</p>

</blockquote>

<p>This is where I disagree in principle, though I agree in fact, sadly, because I find many cameras to be designed by marketers, not photographic engineers. </p>

<p>MY joy in lerarning any new camera is in my (sonetime arduous but always rewarding) reverse-engineering the mind of the photographic engineers who brought their genius and skill to me through the mechanical and elctronic programming choices in their camera designs. I marvel as I finally figure out the benefits of first one feature, then another, and eventually there is no camera, no gear here, just seamless photography, vision recording and playback. </p>

<p>Some call it ergonomics, but it's more than just putting a few controls at hand, er finger. It's also including and coordinating controls whose features and benefits make immediate sense to photographic situations. On my camera, for instance, I've set the back thumb dial for plus/minus compensation and the front forefinger dial for program shift, essentially giving me aperture or shutter priority and exposure compensation at my finger tips on demand without moving my eye from the viewfinder or using my other hand. Then "click". </p>

<p>And so on to the next visual experience to record and playback. Hence my call for total system consideration, rather than one feature or trick like ETTR.</p>

<p>.</p>

[lovcom_photo]

<p>Peter, do you really want a machine to make decisions on saturation, sharpening, contrast, noise abatement, curves, levels, B&W conversion, exposure compensation? If you do, then never shoot raw.</p>

<p><br />Clearly I am not talking about ergonomics here. Remember, machines are not creative. Only humans are.</p>

<p><br />What if you want those tweeks to be surgically applied? How does the camera know what and where are the elements of a composition? How does the camera know that you what the model's right and NOT left cheek saturated? How does the camera know that you want to sharpen just the eyes and teeth and not the hair? I can go on and on with millions of "what ifs". A camera often applies tweeks "big bang" over the whole comps.</p>

<p>As to ETTR, well that is one of the most important "tricks" one can do with DSLR's. You minimize it too much. Until DSLRs with 3-4 more stops of DR are available to us, ETTR is your best friend (assuming you shoot raw of course). I would never want the camera to figure out exposure beyond standard, or in other words, I would never want the camera to ETTR for me automatically. This is not to suggest that I find no use for metering modes...but when it comes to adjustment of exposure, it is far better for the human to do this.</p>

[g._v.]

<p>How's this for an option? Learn how to use an exposure meter, either the one in the camera and/or an incident meter.<br>

An exposure meter tells you more information than a histogram and in skilled hands it is faster and more accurate when it comes to determining the exposure/dynamic range of your subject.</p>

 

[mt4x4]

<blockquote>

<p>How's this for an option? Learn how to use an exposure meter, either the one in the camera and/or an incident meter.</p>

 

</blockquote>

<p>I'm with G.V. on this one. Expose the photo how YOU want it to look, not how the camera deems best. If we all let our cameras pick exposure 100% of the time, people looking at photos would think there was no such thing as white snow.</p>

[bernardwest]

<blockquote>Disagree Bernie...the in-camera histo is more then enough to ascertain good exposure, even if based upon the jpg image on the back LCD is not representative 100% of the raw image.</blockquote><p>

 

I don't think you know how unrepresentative the jpg is of the sensor data. If you did, you would understand that "creative" considerations in the application of ETTR come a distant second to "technical" considerations.

[peterblaise]

<p>.</p>

<p>Dan I think we disagree about different things, and you have misunderstood me, perhaps I have not been clear, let alone succinct.</p>

<p>1 - I never suggested that I think any camera can (or should) make (or take credit for) artistic choices. However, I am suggesting that any camera's automated choices be well documented, and or turn-off-able. No camera has ETTR documentation, and ETTR itself is not unambiguously documented "out there" as a science, but, rather, ETTR is an ersatz "feel good" target of DnR/SNRs (pronounced "dinner sinners": Dynamic Range over Signal to Noise Ratio maximizers) who can't agree with each other how to get there. ETTR has nothing to do with tone/chrome/zone placement in the final print or display image, nor does it assist in knowledgeably managing tone/chrome/zone placement in each system module getting there.</p>

<p>2 - I don't underestimate the general principles and usefulness of ETTR-type decisions IF that is supportive of a photographer's goals. However I also don't overestimate any camera's ability to accurately support any photographer's intended use of ETTR as assessed on any in-camera histogram. None do. Each is whimsical, and experience gained in surmounting ETTR to one's own personal standards on one camera may or may not be experience that is transferable to the next camera.</p>

<p>Great thread, Jacob. This is a subject that needs careful and exhaustive scrutinizing. Google should lead many web searchers to photo.net because of this thread. I also thank everyone for their patience as curmudgeons wake up and try to get verbal at all, let alone try to bring clarity to our sharing what we know in ways that are useful and referential.</p>

<p>.</p>

[Tim_Lookingbill]

<p>The main problem with ETTR is you can't consistently fit or gauge accurately the dynamic range of any given scene to the capabilities of any given camera sensor by pushing exposure. All you have to go by is the flashing blinky's on the LCD. And sometimes they're too small too see on the screen that by the time it's examined in the raw converter the interpretation of the data by that converter (ACR in particular) has blown out more data than realized.</p>

<p>I've been trying to ETTR shooting outdoors and it's just a big PITA readjusting exposure to gain what little benefit ETTR brings that it slows me down considerably to the point that I've lost quite a few shots especially those of white birds or other animals that move in brightly lit scenes futzing with the exposure adjusts.</p>

<p>I think what I'll do is take the advice given to me here about a year ago and just set the camera to Auto Exposure/or Av @ f8 and metering to Center Weighted and just shoot. Even though my camera's metering tends to underexpose, I don't see any of those 4096 levels, half of which are supposedly captured in the first stop, editing the image whether I ETTR or not.</p>

<p>Frankly, I get better shots making sure the incamera histogram shows the highlight tapered end point just off and away from the right by 1/8th if dividing the entire graph into eight sections. This is of course allowing for spectrals to spike, but even that is hit or miss especially with clouds and chrome metal work on cars. Some you want blown but some you want retained and it's damn near impossible how far you've gone viewing the LCD with or without flashing blinky's and histogram.</p>

<p>And practicing until familiar with the DSLR's exposure latitudes doesn't work either because of the linear nature of the sensor. One scene seems to fit nicely using one exposure setting going by what's on the LCD preview, blinky's and histogram. Turn just slightly changing the makeup of scene within the frame and BANG! all of a sudden you've got blown highlights or severe underexposure with the same exposure.</p>

[peterblaise]

<p>.</p>

 

<blockquote>

<p>Earlier: "<em>... white <strong>birds or other animals</strong> that move in brightly lit scenes <strong>futzing with the exposure</strong> adjusts ...</em> "</p>

</blockquote>

<p>=8^o THAT's the problem - keep those birds and other animals from futzing with your camera, and then you'll be okay! PITA or PETA, you choose! I'll have mine on wry. ;-)</p>

<p>.</p>

[lovcom_photo]

<blockquote>

<p>I don't think you know how unrepresentative the jpg is of the sensor data. If you did, you would understand that "creative" considerations in the application of ETTR come a distant second to "technical" considerations.</p>

</blockquote>

<p>Actually I do know very well how the histogram and the corresponding raw image correlate.</p>

<p><br />You can learn this over time. It's about learning one's camera. How the jpg based histo corresponds to the actual raw, even if that histo is not based on that raw.</p>

<p><br />In application, and practice, ETTR's creative aspects come first, over the technical. And because one is shooting raw, one usually has sufficient wiggle room during raw conversion and post processing.</p>

<p>Tom, thousands use ETTR all the time, and see the differences and benefits over just going with the camera's own metering. The reason you can't see it is because you are too new to it, have not mastered it, and perhaps need to fine tune your post processing procedures. ETTR means more details in the shadows, yet the highlights don't have to be sacrificed.</p>

<p> </p>

[mt4x4]

<p><strong>News Flash!</strong> <br /> <em><br /> Nikon announces new generation of DSLRs that not only automatically set ETTR photos without blown hilights, but also go out AND TAKE THE PHOTOS FOR YOU!</em></p>

<p>Here is a sneak peak at the new camera!</p>

<p><img src="http://blogs.govexec.com/fedblog/images/wall-e.jpg" alt="" width="400" height="589" /></p>

[Tim_Lookingbill]

<p>Dan,</p>

<p>I respectfully disagree with you and my name is Tim, not Tom. I've been researching digital imaging and photo restoration for ten years, studying, shooting and processing raw a little over a year. I'm not a newbie and do know my way around making an image look good processing an image. I base my conclusions about ETTR on pure observation shooting real world scenes and processing in ACR and other raw converters. One of the main problems shooting ETTR is not knowing how a raw converter will interpret blown detail.</p>

<p>ACR is one of these converters that interprets the data differently over other converters because its tools don't really show through ACR's odd truncated previews the effect of editing all those levels. I'm assuming it renders previews this way for the sake of speed because a competing raw converter actually renders pixel for pixel the sensor's linear data but is painfully slow editing in showing all those levels.</p>

<p>Another cause of not being able to see all these levels is that the wider the dynamic range of a scene captured the less levels there are available to be distributed in rendering this amount of detail which brings me back to my original point about not being able to gauge the dynamic range of a scene that allows knowing how far to push exposure to the right. It's just not that simple determining this out in the field and knowing for sure what data has been captured exposing this way.</p>

<p>If you've ever underexposed a wide dynamic range scene or any scene for that matter and increased exposure gradually in 1/3 stop increments and watched the right end point of the incamera histogram gradually move closer and closer to the right but abruptly and unexpectedly spike to the right with the next 1/3 stop increase, you are seeing the linear behavior of a sensor. And you'll never know just by looking at a scene at what exposure setting will cause this abrupt jump toward full saturation. You can try to calibrate your exposure for this, but it's like predicting the position of the gauge on a dwell and tach meter. Measuring photons in outdoor sunlight in any given scene is impossible to gauge and predict for.</p>

<p>I've included some screenshots of a scene I shot using ETTR whose dynamic range was beyond the 12bit A/D converter processor of the camera. I posted a topic a while back on this subject over at Luminous Landscape and was told by a very knowledgeable member using a raw processing analyzer that I exposed this image quite well.</p>

<p>The first screenshot is of rock detail and the lack of levels demonstrated by posterization using ACR's curve tool. Note that ACR's mandatory noise reduction has nothing to do with the posterization. The second is the same rock detail using another raw converter, Raw Developer, which shows pixel for pixel preview and curve tool which DOES show more levels making up the detail in the rock than ACR. It's still not the amount of levels you'ld expect in the first stop of brightness for such a scene which causes me to conclude that the dynamic range of that scene contributed to the lack of levels seen in the preview. It's for this reason I don't trust ETTR shooting outdoors.</p><div></div>