To ETTR or not to ETTR...?

Started 3 months ago | Discussions
OP Jeepit Regular Member • Posts: 168
Re: example
1

TomFid wrote:

Jeepit wrote:

TomFid wrote:

Many pieces on this topic that I have read states to have the ISO as low as you can get it. Interesting yours was at 200 & 640.

200 is base for my EM5ii. 640 was just carelessness.

I'll try again later.

Here's an example. I shot a sequence at f7.1, ISO 200, varying shutter speed for exposure compensation.

Sequence as imported into LR. Note that I mislabeled the top right shot - it's +1EV.

The 0EV exposure would make a decent jpeg but looks a little under to me. OTOH the +1 shot is a little washed out.

The +2 shot roughly pinned the camera histogram to the right edge without visible clipping. The +3 shot looks pretty hopeless.

You did this by using the EC dial and bumping up the brightness?

In LR, I used the exposure slider to roughly normalize everything to look like the +1 shot (i.e. the +3 shot got -2, etc.). Then they all look similar. It turns out that the jpeg histogram is too conservative, and even the +3EV shot is not clipped:

Same sequence, after rough leveling using the LR exposure slider.

However, if you zoom in, they look pretty different:

Left: the metered exposure (no compensation). Right: the +3EV shot.

The +3 shot is much cleaner in the mids and shadows, as expected (illustrating knickerhawk's point above).

On a day that's not so hazy, you wouldn't get away with this much compensation.

Thank you for the response...this was whatI have been hoping someone would do ...give a visual exemplar  This aids in the learning process alot.

 Jeepit's gear list:Jeepit's gear list
Olympus E-M1 II Olympus E-330 Olympus E-M1 Olympus Zuiko Digital ED 12-60mm 1:2.8-4.0 SWD Olympus Zuiko Digital ED 18-180mm 1:3.5-6.3 +16 more
OP Jeepit Regular Member • Posts: 168
Re: All depends on your expected output

knickerhawk wrote:

Jeepit wrote:

Pixnat2 wrote:

Jeepit wrote:

Pixnat2 wrote:

You shoot landscapes in difficult light and are adept of squeezing out all the DR and shadow details you can in post processing?

You can't stand a bit of noise in your shadows?

In those cases, mastering ETTR could be helpful.

But for casual shooting, I wouldn't worry about it. Our modern cameras have excellent metering system and enough DR to cope with nearly all situations.

That said, it's a nice technique to learn and use in some cases.

Thanks for your post. What would those cases be?

You're welcome.

Lanscapes with high contrast (deep shadows and strong highlights) is probably the main case where ETTR can be useful.

In this scenario, camera tend to protect highlight for the JPEG, thus it underespose a bit. This will introduce noise in the shadows which will be visible when you try to recover your RAW files. In a landscape picture, you would lose fine details in the shadows and get a mushy result.

ETTR will burn your JPEG, but less noise will be present in the shadows. If correctly done, you could recover your highlights, and you'll get a landscape with clean shadows and fine details.

Peronally, that's the only case where I use ETTR.

Landscapes was one of the main questions I had regarding ETTR.

Yes, landscape is generally the best use case for ETTR.

What about street shooting in b/w, where there are shadows?

Wildlife?

Use cases involving fast moving subjects are tricky because metering on your critical highlights (or otherwise monitoring highlight blinkies and the histogram) is hard to control in a timely way. Plus, action shots are more often shutterspeed constrained and perhaps also aperture constrained, which limits your flexibility for pushing up exposure.

With respect to B/W, there are additional pluses and minuses. On the plus side, you're less worried about adverse color issues from blowing one or even two of the channels. In general B/W street shooters aren't so concerned about blown highlights either (as long as it's not affecting the subjects' faces). Plus, the improved tonality resulting from ETTR done right can be very important in some B/W work. However, lots of street shooters like the gritty/grainy look that's reminiscent of fast film. For them, ETTR might be counterproductive, especially when considering the noted exposure limitations.

Duly noted...thanks more info i was hoping for.  Learning so much on this topic...this is the way this board should be...an educational tool.

But all pictures with high contrast, where you need a wide Dynamic Range, would benefits form ETTR.

 Jeepit's gear list:Jeepit's gear list
Olympus E-M1 II Olympus E-330 Olympus E-M1 Olympus Zuiko Digital ED 12-60mm 1:2.8-4.0 SWD Olympus Zuiko Digital ED 18-180mm 1:3.5-6.3 +16 more
Barry Twycross Senior Member • Posts: 1,127
Re: To ETTR or not to ETTR...?

Jeepit wrote:

Never had to do this back in the film days; so how important is it now? Do you or don't you ETTR (expose to the right)?

In the film days, I relied on negatives having good highlight preservation and resistance to overexposure. So I exposed past the right.

For slides, I'd tend to expose a little under, so not quite to the right.

It seems my current method is just about equivalent to ETTR, I expose what I think is good, and if I get a highlight clipping warning, I reduce the exposure. It works out at much the same thing. I'm also shooting the lowest available ISO (extended downwards as much as the camera will let me), unless forced to change up.

 Barry Twycross's gear list:Barry Twycross's gear list
Panasonic GX850 Panasonic Lumix G Vario HD 12-32mm F3.5-5.6 Mega OIS Leica Nocticron 42.5mm Panasonic Lumix G Vario 35-100mm F4.0-5.6 ASPH Mega OIS Panasonic 20mm F1.7 II +3 more
Pixnat2
Pixnat2 Veteran Member • Posts: 5,718
Re: All depends on your expected output
1

Jeepit wrote:

Pixnat2 wrote:

Jeepit wrote:

Pixnat2 wrote:

You shoot landscapes in difficult light and are adept of squeezing out all the DR and shadow details you can in post processing?

You can't stand a bit of noise in your shadows?

In those cases, mastering ETTR could be helpful.

But for casual shooting, I wouldn't worry about it. Our modern cameras have excellent metering system and enough DR to cope with nearly all situations.

That said, it's a nice technique to learn and use in some cases.

Thanks for your post. What would those cases be?

You're welcome.

Lanscapes with high contrast (deep shadows and strong highlights) is probably the main case where ETTR can be useful.

In this scenario, camera tend to protect highlight for the JPEG, thus it underespose a bit. This will introduce noise in the shadows which will be visible when you try to recover your RAW files. In a landscape picture, you would lose fine details in the shadows and get a mushy result.

ETTR will burn your JPEG, but less noise will be present in the shadows. If correctly done, you could recover your highlights, and you'll get a landscape with clean shadows and fine details.

Peronally, that's the only case where I use ETTR.

Landscapes was one of the main questions I had regarding ETTR.

What about street shooting in b/w, where there are shadows?

For me, street shooting is more about getting the mood than ultimate details like in landscapes. I never ETTR for that kind of shooting. In fact, I don't fiddle too much with settings, because I try to be as connected as possible with my environnement. The decisive moment or seeing the detail is much more important than the perfect exposure, IMHO. So I put my camera in Auto ISO, A mode and just adjust aperture when needed, or sometimes even in P mode and just pressing the shutter.

But again, why not, all depends on your priorities.

Wildlife?

I don't do a lot, but I think a fast shutter speed is often the essential setting for wildlife, thus ETTR is not very compatible. And like street shooting, getting the right moment is much more important than the perfect exposure.

-- hide signature --
 Pixnat2's gear list:Pixnat2's gear list
Panasonic Lumix DMC-GM1 Panasonic Lumix DMC-GM5 Sony a7R II Fujifilm X-T2
XRF Forum Member • Posts: 78
Re: To ETTR or not to ETTR...?

It seems that the number of actual answers to "do you use ETTR" are less than the number of technical arguments, which I myself am guilty of contributing to. Still want to discuss the technical side but first I should address the actual question.

Sometimes. Since ETTR involves taking a more active role in selecting your exposure, it takes more time. Therefore, I use ETTR mainly when I am shooting in situations where I have a large amount of time to take the shot such as landscape. If my shutter speed is high enough and there is plenty of light, I may use a slight increase in exposure compensation (normally +1 EV) for dynamic subject like wildlife, but I won't fully ETTR in case they move into a situation where overexposure is risked. I do not bother with ETTR at ISO values other than base ISO.

Now as to why. Given a perfect sensor, the noise level in any area of a photo is the square root of the number of photons hitting the sensor during the exposure. Quadruple the amount of light and you double the signal to noise ratio (SNR). Electronic image sensors have a hard cut-off to the number of photons they can record before the electron wells are full and you get clipped highlights. The best possible SNR occurs just before this happens. Since the data from an electronic sensor needs to be decoded and processed before it is displayed, the number of photons corresponding to a particular output lightness value can be practically anything the processor or camera company wants it to be.

When a camera meters for a photo it has no knowledge of what the actual incident light is, so it makes an estimate. First the camera assigns a weight to each region of the sensor depending on the metering mode, and possibly adds a vignetting compensation factor for those regions. Next the camera averages the incoming light, and selects the exposure based on the assumptions that (1.) the scene is evenly lit, and (2.) the average amount of reflected light in the scene is a set value, normally 15% or 18%. The second assumption is the reason you get underexposures in scenes filled with reflective materials such as snow, and overexposures in scenes filled with non-reflective materials. Some manufactures may also be implementing smart exposure functions now such as ignoring sections of the image that are much brighter or darker than the rest.

Where ETTR becomes useful is the relation between sensor saturation and what the camera company sets as corresponding to middle grey. The ISO standard for digital camera sensor offers three methods of determining what the base ISO of a sensor is. Of those, one is based on SNR, one is based on sensor saturation, and the last basically allows the company to do whatever they want. The saturation method is the only one that actually ties the amount incoming light to the ISO value. The saturation method states that the sensor should saturate at 3 1/2 stops above middle grey, that gives 2 1/2 stops between middle grey and white, and 1 stop extra for specular reflections and emitting light sources. It also states that at 100 ISO a sensor should saturate at 0.78 lux·s. Now take a look at

http://www.photonstophotos.net/Charts/PDR.htm#Ideal%204/3,Olympus%20OM-D%20E-M1%20Mark%20II

notice how the E-M1 mk.II has better than ideal performance. That is not because the senor is perfect, far from it, remember the CFA is removing a good portion of the light, it is because camera manufacturers do not use the saturation definition of ISO. By the saturation definition 200 ISO may actually be 100 ISO or 50 ISO, the camera just sets middle grey further from saturation. The extra dynamic range comes from extra highlight headroom, not better performance in the shadows. To take full advantage of the dynamic range, you must bring the sensor as close as possible to saturation as you can. Both ETTR and expanded low ISO ranges help to do this. The advantage of ISO based ETTR is that even though you are loosing dynamic range, you are gaining slightly in shadow performance, and if you are not clipping highlights, you used dynamic range is increased.

miketala Regular Member • Posts: 320
Re: ETTR rules

Adielle wrote:

TN Args wrote:

Adielle wrote:

Clipping / squeezing highlights is the worst thing to do in MFT as far as I've seen. True for GX8, E-M1 and G6, so I'd expect it to be similar in other models. Far better to be careful and slightly underexpose than to overexpose.

ETTR is the art of NOT clipping or squeezing desired highlights. Once you understand it, that's obvious.

Level 1 ETTR: JPEG histogram.

Level 2 ETTR: raw histogram. Once you learn to do this, best IQ in terms of noise and colour is attained.

Cheers

When people say "expose to the right" I understand it as having a bias towards possibility of overexposing than towards underexposing, and I believe that's what is actually meant. I don't support this bias at all, and I'm saying that it's far better to have a bias towards underexposing, because any obvious loss of quality is much less likely in that situation.

I don’t think so.   I see it as choosing to expose as much as possible without clipping.   Because m43 will often underexposed.

timo Veteran Member • Posts: 5,440
Re: Have I got it right?

FingerPainter wrote:

timo wrote:

This thread seems to have got super-complicated. Have I misunderstood ETTR?

My understanding is that the further up the histogram you go (i.e. to the brighter end), the more useful data the file can contain, and that includes finer distinctions of tonal values.

Yes, as long as you move to he right by increasing exposure - the amount of light per unit area falling on the sensor.

In many cameras, you can also make a (usually much) smaller improvement by moving up the histogram by increasing the ISO setting for the first few stops above base ISO.

So for the purposes of taking the photo, you push the exposure up as far as you can SHORT OF blowing the highlights to the point where they are unrecoverable.

Yes.

The image will look overexposed at that point.

The image will probably look too light. It won't actually be exposed too much. It is useful to avoid calling an image "overexposed" when it might be too light for a reason other than too much exposure. The mis-use of the term "overexposed" is a holdover from roll-film, where there was often a close to one-to-one mapping between lightness and exposure. This is not the case for digital images.

Then in post-processing, you bring the 'exposure', or 'brightness', or whatever you want to call it,

Call it "lightness". "Exposure" is the amount of light that fell on the sensor per unit area. You cannot change that after capture. "Brightness" is a property of an image displayed by a light-emitting device, like a monitor or television. You can increase the brightness of an image displayed on your computer monitor by increasing the brightness setting of the monitor without affecting the image file. Prints do not, AFAIK, have a brightness, Just a lightness.

back down to whatever reflects you visual intentions. That way you will get more subtle, more accurate, shadow tones. And better S/N ratio,

Yes

provided you haven't had to adjust the ISO upwards, which would have defeated the purpose.

Adjusting ISO upward only defeats the purpose if it is done instead of increasing exposure. If you got the maximum exposure you can tolerate (not too much motion blur, too shallow DoF or too much lens aberration blur), then incrasiog ISO can make a samll improvement to SNR on many cameras, at least for the first few stops above base ISO.

If, using the camera's metering system and with your EV dial set to 0, your shot is already just about triggering blinkies, you are already 'exposing to the right', and you don't have to adjust anything.

In fact, when desired highlights are blown at the metered exposure, ETTR requires negative EC.

ETTR is most relevant in situations where there are no important extreme highlights, and you have a lot of shadow detail that you want to maintain.

Hmm, given that ETTR is essentially exposing for the hightlights, I don't think it becomes less relevant in situation where there are no important extreme highlights. It just doesn't give as much SNR improvement. It is still relevant because ETTR is as much about preserving desired highlight detail as it is about increasing SNR. Fundamentally it is about maximizing usable data. It does this both by trying to increase light capture and also by trying to avoid throwing away data in highlights.

What have I missed?

See above?

Yes, that all makes sense. Like many of a certain age, I'm conditioned by hauling nearly transparent slides out of the plastic box, holding them up to the light and saying 'wow that's over-exposed' ... Another factor is the use of the term 'exposure' for a slider in Lightroom. But sure, 'lightness' does express what one means.

-- hide signature --

Tim
'If I were you, I wouldn't start from here ... '
http://timauger.smugmug.com
http://timauger.blogspot.sg/

 timo's gear list:timo's gear list
Fujifilm X30 Pentax K-5 Pentax K-30 Panasonic Lumix DMC-GX7 Panasonic Lumix DMC-GX8
TomFid Veteran Member • Posts: 3,115
Re: example

Jeepit wrote:

You did this by using the EC dial and bumping up the brightness?

Correct - the difference between this and last night's failed experiment was that there was plenty of light, so varying EC varied the shutter speed, not the ISO.

Thank you for the response...this was whatI have been hoping someone would do ...give a visual exemplar This aids in the learning process alot.

Great!

OP Jeepit Regular Member • Posts: 168
Re: To ETTR or not to ETTR...?
5

I would like to take this opportunity to express my gratitude to all who have responded to this thread.  ETTR was confusing at first, but after reading and researching the topic my understanding of why its important and when its important has increased twofold.

To my thinking this is how this message board should run as in to use it as an educational tool.  I was happy to see less trolling, less verbal jousting and bickering.  I'm on this board daily but I usually don't post unless i have questions, want to further my learning process in photography, or give praise to images that speak to me.

So well done to all who have taught myself and others about ETTR!

Thanks

Rick

 Jeepit's gear list:Jeepit's gear list
Olympus E-M1 II Olympus E-330 Olympus E-M1 Olympus Zuiko Digital ED 12-60mm 1:2.8-4.0 SWD Olympus Zuiko Digital ED 18-180mm 1:3.5-6.3 +16 more
jimkahnw
jimkahnw Regular Member • Posts: 248
Re: To ETTR or not to ETTR...?
2

The one takeaway from this thread should be the need for constant practice. Learn what equipment you need and how to work it in relation to the situation you want to depict; learn what post-capture processing is required to realize your vision. It takes a while. Henri-Cariter Bresson said your first 10,000 exposures would be your worst--but that was in the days of film, before motor drives and one-hour processing; adjust upward for digital.

In photography experience is the best teacher.

-- hide signature --

JimK
It's not the camera; it's the photographer.
--Anonymous

 jimkahnw's gear list:jimkahnw's gear list
Olympus E-M1 Olympus OM-D E-M10 Olympus 12-100mm F4.0
jimkahnw
jimkahnw Regular Member • Posts: 248
Re: example

The example also demonstrates what a fool's errand we are all on. There's little difference between the 0EV and the +3EV, at least in the mid- and 3/4-tones. I know with a little more processing, there would be none. However, what about the highlights? That's the whole thing about ETTR; increase exposure without clipping--recover in post. In this example, what does the sky look like for the +3EV exposure? I would bet that the highlights are unrecoverable.

I still think ETTR is a misconception that leads to overexposure and clipped highlights. Our 4/3 sensors are fantastic at preserving highlights and shadows. We can get great images without fretting over the shadows to preserve highlight details. I'm continually amazed how much 3/4- and full-tone values can be lifted in an image exposed for the highlights.

-- hide signature --

JimK
It's not the camera; it's the photographer.
--Anonymous

 jimkahnw's gear list:jimkahnw's gear list
Olympus E-M1 Olympus OM-D E-M10 Olympus 12-100mm F4.0
Tom Axford Veteran Member • Posts: 6,233
Re: example
2

jimkahnw wrote:

The example also demonstrates what a fool's errand we are all on. There's little difference between the 0EV and the +3EV, at least in the mid- and 3/4-tones. I know with a little more processing, there would be none. However, what about the highlights? That's the whole thing about ETTR; increase exposure without clipping--recover in post. In this example, what does the sky look like for the +3EV exposure? I would bet that the highlights are unrecoverable.

I still think ETTR is a misconception that leads to overexposure and clipped highlights. Our 4/3 sensors are fantastic at preserving highlights and shadows. We can get great images without fretting over the shadows to preserve highlight details. I'm continually amazed how much 3/4- and full-tone values can be lifted in an image exposed for the highlights.

I agree broadly with what you say and I often play safe rather than risk blowing out important highlights.

However, I do not see this as a negation of ETTR.  If I am not sure about the precise light levels, then it's best to leave a margin of error.  I am still trying to achieve ETTR, just playing very safe about doing so.

I see ETTR as more a principle - a recognition that:

1. If highlights are clipped, then they are lost forever.  So take great care not to do so.

2. The best image quality (lowest noise) is achieved by the maximum possible exposure.

ETTR simply amounts to trying to achieve both of these at once.

jimkahnw
jimkahnw Regular Member • Posts: 248
Re: example

You understand the importance of careful exposure control. ETTR refers to the histogram and if you look at yours, you'll see that the best exposures have a pixel value distribution that is not necessarily dominant on the right side of the graph. There is no such thing as the "perfect" histogram, as each subject is different and consequently will have a unique pixel value distribution.

I'm lazy and impatient: I set the exposure compensation to -.3EV, ISO high enough to avoid camera shake and subject motion and watch for overexposure in the live histogram. I use a middle f/stop for the lens to maximize image quality; the 4/3 sensors offer great depth of field. I don't worry about bokeh. Works every time. Some examples here: jimkphoto.com.

This thread has demonstrated how hung-up we are on following a certain technique to achieve the optimal exposure. I think we've made it needlessly complex. The scientists and engineers have enabled the camera manufacturers to simplify the process and made it simple for photographers to make great pictures. Coming from a film background, where I depended on a hand-held light meter and processed both color and black and white, I assure you the camera marvels we can put in our pockets are far, far easier to use than any analog camera. And, I would never trade Photoshop for a darkroom. Never.

-- hide signature --

JimK
It's not the camera; it's the photographer.
--Anonymous

 jimkahnw's gear list:jimkahnw's gear list
Olympus E-M1 Olympus OM-D E-M10 Olympus 12-100mm F4.0
TomFid Veteran Member • Posts: 3,115
Re: example

jimkahnw wrote:

The example also demonstrates what a fool's errand we are all on. There's little difference between the 0EV and the +3EV, at least in the mid- and 3/4-tones. I know with a little more processing, there would be none.

I think the difference here is fairly small, but this is basically a snapshot. If this were a trickier shot, and you wanted to bring up shadows or otherwise increase contrast, the 0EV shot's noise would become more apparent, not less. You could use noise reduction to make the 0EV shot acceptable, but it'll never be as good.

However, what about the highlights? That's the whole thing about ETTR; increase exposure without clipping--recover in post. In this example, what does the sky look like for the +3EV exposure? I would bet that the highlights are unrecoverable.

Not so. There is no clipping in the +3 image. As I pointed out, that might not be true on a day that's less hazy, but then +3 would be a bad choice.

I still think ETTR is a misconception that leads to overexposure and clipped highlights.

If you're clipping highlights you care about, you're doing it wrong.

Our 4/3 sensors are fantastic at preserving highlights and shadows. We can get great images without fretting over the shadows to preserve highlight details. I'm continually amazed how much 3/4- and full-tone values can be lifted in an image exposed for the highlights.

True. But if Oly announced that the EM5iii was going to have 3 stops better noise performance, people would FREAK OUT. There would be a riot at B&H. And yet, here it is, available for the price of a small amount of thinking. If PanOly were willing to step away from the meter-the-grey-card mentality, it could be made automatic and riskless.

OP Jeepit Regular Member • Posts: 168
Re: To ETTR or not to ETTR...?

jimkahnw wrote:

The one takeaway from this thread should be the need for constant practice. Learn what equipment you need and how to work it in relation to the situation you want to depict; learn what post-capture processing is required to realize your vision. It takes a while. Henri-Cariter Bresson said your first 10,000 exposures would be your worst--but that was in the days of film, before motor drives and one-hour processing; adjust upward for digital.

In photography

experience is the best teacher.

I couldn't agree more.  Thank you for your words and insight.

I's been the last few years that I personally become more serious about learning how to USE a digital camera as apposed to back in the day...film camera.

 Jeepit's gear list:Jeepit's gear list
Olympus E-M1 II Olympus E-330 Olympus E-M1 Olympus Zuiko Digital ED 12-60mm 1:2.8-4.0 SWD Olympus Zuiko Digital ED 18-180mm 1:3.5-6.3 +16 more
bobn2
bobn2 Forum Pro • Posts: 61,751
Re: example
1

jimkahnw wrote:

The example also demonstrates what a fool's errand we are all on. There's little difference between the 0EV and the +3EV, at least in the mid- and 3/4-tones. I know with a little more processing, there would be none. However, what about the highlights? That's the whole thing about ETTR; increase exposure without clipping--recover in post. In this example, what does the sky look like for the +3EV exposure? I would bet that the highlights are unrecoverable.

I still think ETTR is a misconception that leads to overexposure and clipped highlights. Our 4/3 sensors are fantastic at preserving highlights and shadows. We can get great images without fretting over the shadows to preserve highlight details. I'm continually amazed how much 3/4- and full-tone values can be lifted in an image exposed for the highlights.

You post illustrates the crux of the problem discussing ETTR. The technique itself was based on a false premise (that giving more exposure would result in the image being encoded in more digital 'levels' which would give better image quality. In fact, any benefit is due to maximising exposure, and nothing to do with digital encoding. However, despite being based on a fallacy, the idea and name has stuck, and I'm not sure why, because as a technique, it isn't a very sound way of maximising exposure.

The reason it's stuck, I think, is that it forms a sort of a link between two very different models of exposure management.

Model 1 posits that the purpose of exposure management is to control the output lightness of the final image. This idea is so entrenched that many people do not even acknowledge that there is a distinction between exposure and final image lightness. It is from this model that there often used terms 'underexposed', 'correctly exposed' and 'overexposed' derive. They generally mean that the final image is deemed to be not light enough, a perceptually pleasing lightness or too light.

Model 2 posits that the purpose of exposure management is to ensure the maximum possible information about the scene in the raw file, and if there are constraints that limit how much information be captured, that those are managed to the best effect for the desired output image. In general this means maximising exposure, except where that is in conflict with practical or aesthetic requirements. The terms 'under', 'over' and 'correctly' exposed would have completely different meanings within this model of exposure management.

ETTR provides a method (though not a great one) for using the tools and vocabulary of Model 1 to achieve at least some of the results of model 2. That, however leads to it being fundamentally inconsistent and contradictory.

And, by the way, no sensor is particularly 'great' at preserving highlights or shadows, and mFT ones are certainly no greater at it than others. The question is more about the exposure you chose in relation to the capabilities of the sensor.

-- hide signature --

263, look deader.

knickerhawk Veteran Member • Posts: 6,262
Re: example
2

bobn2 wrote:

jimkahnw wrote:

The example also demonstrates what a fool's errand we are all on. There's little difference between the 0EV and the +3EV, at least in the mid- and 3/4-tones. I know with a little more processing, there would be none. However, what about the highlights? That's the whole thing about ETTR; increase exposure without clipping--recover in post. In this example, what does the sky look like for the +3EV exposure? I would bet that the highlights are unrecoverable.

I still think ETTR is a misconception that leads to overexposure and clipped highlights. Our 4/3 sensors are fantastic at preserving highlights and shadows. We can get great images without fretting over the shadows to preserve highlight details. I'm continually amazed how much 3/4- and full-tone values can be lifted in an image exposed for the highlights.

You post illustrates the crux of the problem discussing ETTR. The technique itself was based on a false premise (that giving more exposure would result in the image being encoded in more digital 'levels' which would give better image quality. In fact, any benefit is due to maximising exposure, and nothing to do with digital encoding. However, despite being based on a fallacy, the idea and name has stuck, and I'm not sure why, because as a technique, it isn't a very sound way of maximising exposure.

While it's certainly true that Reichmann got the reason to "ETTR" wrong in his articles, it's pretty rare these days to see improved digital encoding raised as the reason to ETTR. For instance, I don't recall seeing that argument made anywhere in this thread, although admittedly, I haven't read every post here with that in mind. We've come a long way in understanding how to optimize raw exposure (in no small part thanks to your efforts and others and buttressed by Martinec's paper). For better or worse, though, the terminology has stuck even as the meaning and application of the terminology has evolved.

The reason it's stuck, I think, is that it forms a sort of a link between two very different models of exposure management.

Model 1 posits that the purpose of exposure management is to control the output lightness of the final image. This idea is so entrenched that many people do not even acknowledge that there is a distinction between exposure and final image lightness. It is from this model that there often used terms 'underexposed', 'correctly exposed' and 'overexposed' derive. They generally mean that the final image is deemed to be not light enough, a perceptually pleasing lightness or too light.

Your "Model 1" is really just the JPEG exposure model. It serves its purpose pretty darn well as a working mental model for JPEG-only shooters, especially for the majority of whom are interested in doing little or no postprocessing. However, it breaks down (indeeed, is even conceptually counterproductive) as soon as one starts thinking in terms of optimizing raw exposure. The frequent problem I see for photographers trying to bridge the two models is how they approach shooting JPEG+raw while using the JPEG exposure model. They then express surprise that their renderings from the raws aren't all that much better than their OOC JPEGs and declare that shooting raw is much ado about nothing!

Model 2 posits that the purpose of exposure management is to ensure the maximum possible information about the scene in the raw file, and if there are constraints that limit how much information be captured, that those are managed to the best effect for the desired output image. In general this means maximising exposure, except where that is in conflict with practical or aesthetic requirements. The terms 'under', 'over' and 'correctly' exposed would have completely different meanings within this model of exposure management.

ETTR provides a method (though not a great one) for using the tools and vocabulary of Model 1 to achieve at least some of the results of model 2. That, however leads to it being fundamentally inconsistent and contradictory.

I agree, but the real obstacle here is the cameras we have today are almost completely designed to reinforce the JPEG exposure model. In that regard, would you be as critical of using "ETTR" as a succinct short hand for optimized raw exposure if cameras with actual raw histograms started being released?

And, by the way, no sensor is particularly 'great' at preserving highlights or shadows, and mFT ones are certainly no greater at it than others. The question is more about the exposure you chose in relation to the capabilities of the sensor.

tomhongkong Veteran Member • Posts: 3,756
Re: To ETTR or not to ETTR...?

bobn2 wrote:

tomhongkong wrote:

Some facts

None of the below is a fact.

Most people use an average auto exposure setting which is based on averaging the light measured and having an 18% greyscale exposure correct. That means that bright portion of a bright image is going to be underexposed (of course if you are using spot or centre weighted exposure that may not apply). A predominately dark image might similarly be underexposed

With most digital sensors there is around one stop of 'headroom' for highlights, when the auto exposure is used. Even blinkies typically start flashing at a maximum setting of 125% so there is plenty of headroom left when they are blinking.

If you don't believe this, try taking a photo in RAW with 'blown' highlights and reducing the exposure in PP. You will be surprised at how much latitude is built in to your camera, which is not used when taking JPEG. Having a feeling for how much you can overexpose is a good thing.

So there is a lot of scope for 'overexposing' based on what the camera meter believes.

Getting the right exposure (and I don't mean correcting a dark image by winding up brightness with a higher iso, which does nothing for exposure) is important to improve the SNR and reduce noise, and to capture more detail in the whole image, in particular in the shadows. To be safe, and underexpose "to protect the highlights' does the reverse.

Of course if you only use JPEGs, and I accept the convenience of this strategy, you pay the price in reducing potential DR.

Just try it

Tom

Perhaps as this forum's self professed expert, you would like to correct my errors, paragraph by paragraph.

I am always willing to learn

tom

OP Jeepit Regular Member • Posts: 168
Re: example

jimkahnw wrote:

You understand the importance of careful exposure control. ETTR refers to the histogram and if you look at yours, you'll see that the best exposures have a pixel value distribution that is not necessarily dominant on the right side of the graph. There is no such thing as the "perfect" histogram, as each subject is different and consequently will have a unique pixel value distribution.

I'm lazy and impatient: I set the exposure compensation to -.3EV, ISO high enough to avoid camera shake and subject motion and watch for overexposure in the live histogram. I use a middle f/stop for the lens to maximize image quality; the 4/3 sensors offer great depth of field. I don't worry about bokeh. Works every time. Some examples here: jimkphoto.com.

This thread has demonstrated how hung-up we are on following a certain technique to achieve the optimal exposure.

I think we've made it needlessly complex.

There is complexity to this but if one person can learn a new technique or skill I'll mark that down as a positive.

The scientists and engineers have enabled the camera manufacturers to simplify the process and made it simple for photographers to make great pictures. Coming from a film background, where I depended on a hand-held light meter and processed both color and black and white, I assure you the camera marvels we can put in our pockets are far, far easier to use than any analog camera. And, I would never trade Photoshop for a darkroom.

I too miss the b/w darkroom!

Never.

 Jeepit's gear list:Jeepit's gear list
Olympus E-M1 II Olympus E-330 Olympus E-M1 Olympus Zuiko Digital ED 12-60mm 1:2.8-4.0 SWD Olympus Zuiko Digital ED 18-180mm 1:3.5-6.3 +16 more
tomhongkong Veteran Member • Posts: 3,756
Re: To ETTR or not to ETTR...?

tomhongkong wrote:

bobn2 wrote:

tomhongkong wrote:

Some facts

None of the below is a fact.

Most people use an average auto exposure setting which is based on averaging the light measured and having an 18% greyscale exposure correct. That means that bright portion of a bright image is going to be underexposed (of course if you are using spot or centre weighted exposure that may not apply). A predominately dark image might similarly be underexposed

With most digital sensors there is around one stop of 'headroom' for highlights, when the auto exposure is used. Even blinkies typically start flashing at a maximum setting of 125% so there is plenty of headroom left when they are blinking.

If you don't believe this, try taking a photo in RAW with 'blown' highlights and reducing the exposure in PP. You will be surprised at how much latitude is built in to your camera, which is not used when taking JPEG. Having a feeling for how much you can overexpose is a good thing.

So there is a lot of scope for 'overexposing' based on what the camera meter believes.

Getting the right exposure (and I don't mean correcting a dark image by winding up brightness with a higher iso, which does nothing for exposure) is important to improve the SNR and reduce noise, and to capture more detail in the whole image, in particular in the shadows. To be safe, and underexpose "to protect the highlights' does the reverse.

Of course if you only use JPEGs, and I accept the convenience of this strategy, you pay the price in reducing potential DR.

Just try it

Tom

Perhaps as this forum's self professed expert, you would like to correct my errors, paragraph by paragraph.

I am always willing to learn

tom

I see an error in the last sentence of my first paragraph.  The last word should be 'overexposed' not 'underexposed'

Sorry

tom

Keyboard shortcuts:
FForum MMy threads