Understanding Digital Exposure Theory

sk66

Advertiser
Messages
8,650
Name
Steven
Edit My Images
Yes
For about the last month I have been posting about topics related to digital exposure theory to my FB page; but I am way too lazy to copy them all here. However, due to many requests I have compiled them into a single PDF. Here's a link to the file on my google drive if anyone is interested.
https://drive.google.com/open?id=1iemxFctZNApqSlcigd-CzQCRtAiqf5s9
 
A useful resource Steven. Thanks for sharing.
 
I updated the pdf today with a small addition to the last section on pixel size vs sensor size:

"Just as it is the sensor size that matters in terms of the amount of light received/collected, it is also the physical output size that determines how it looks. I.e. how much enlargement of the sensor area that must occur. The M4/3 sensor will have to be enlarged 4x as much in area compared to the FF 35mm sensor in order to create the same output image size. Pixel dimensions only matter in terms of digital display; and what that actually means is dependent on the display device/resolution used."
 
Great!
I tried to make it clear and easy to understand.
It is very clearly written, thanks.
For some reason the link does not come up as a pdf on my iPad and can’t be saved as such. Not a problem, I can get round that , but may be for some people.
 
It is very clearly written, thanks.
For some reason the link does not come up as a pdf on my iPad and can’t be saved as such. Not a problem, I can get round that , but may be for some people.
How does it open? It should be in some kind of reader with the option to save/convert.
BTW, I've made small wording changes/additions... but the linked file automatically updates.
 
Last edited:
Very interesting read, thanks for sharing it.

GC
 
How does it open? It should be in some kind of reader with the option to save/convert.
BTW, I've made small wording changes/additions... but the linked file automatically updates.
Well, I’m reading this in Safari on iPad Pro (iPadOS 13) and the link opens directly in a new tab and is readable there with various icons along the top (if you choose to open the link yourself then it just opens in a new tab without the icons but entirely readable). The problem comes if you use the Safari share button, as one usually would. With most PDFs you are offered the choice of saving to iBooks/GoodReader etc but that doesn’t happen with this one and the download/save to files etc just produce an 87kb html file.
As I said, it’s not a problem, just odd behaviour and if you click directly on the link it works fine. I just always choose to open links in a new tab myself.
 
I greatly respect the work and science you have explained in your PDF but as a 'retired-but-hardwired' professional Art Director primarily driven by the emotional aspects of an image, I find that personally I find all this science far too much to want to digest. A precis overview would be a welcome addition to your admirable thesis.

I think that you and I will agree that the 'perfect' photos combine both the emotional and the technical aspects. But I see a camera primarily as an artistic tool rather than just a reproductive machine.
 
A precis overview would be a welcome addition to your admirable thesis.
The basic premise of the entire thing is that exposure is all, and only, about light... imagine that.

The point is to clarify/explain many things that are often misunderstood/wrong... ISO causing noise, ETTR being about bits/data in the last stop, pixel size, ETTL being a form of ETTR, etc... while some of the common explanations/understandings aren't entirely/technically wrong, i.e. dynamic range/data/bits and pixel size, they are effectively irrelevant.

Without actually understanding what is happening you can't really make well informed decisions about what you are doing.
 
Last edited:
Well, I’m reading this in Safari on iPad Pro (iPadOS 13) and the link opens directly in a new tab and is readable there with various icons along the top (if you choose to open the link yourself then it just opens in a new tab without the icons but entirely readable). The problem comes if you use the Safari share button, as one usually would. With most PDFs you are offered the choice of saving to iBooks/GoodReader etc but that doesn’t happen with this one and the download/save to files etc just produce an 87kb html file.
As I said, it’s not a problem, just odd behaviour and if you click directly on the link it works fine. I just always choose to open links in a new tab myself.
This is a peculiarity of iOS -- just tried it on the Mac and right-click and direct click work the same!
 
But that's why you need to digest the science because it helps you learn and understand the craft that's needed to turn that reproductive machine into an artistic tool :)

.... I don't disagree but why not write the science in a more concise and more easily digested form? More people might be more inclined to read it.

I came to this thread expecting to learn something without being blinded by so much longwinded presentation of the science. Such a thesis is fine but a more concise and more easily digested version as an alternative in addition would have been more helpful.
 
Last edited:
But that's why you need to digest the science because it helps you learn and understand the craft that's needed to turn that reproductive machine into an artistic tool :)
There's no real need to understand why X happens when you do Y, just that it does.

I concede that some people are interested in the 'why' but I doubt it helps them make their pictures any better, and I suspect there far more who get confused by trying to understand the science and are held back from making interesting pictures by trying to get their heads round it.
 
.... I don't disagree but why not write the science in a more concise and more easily digested form? More people might be more inclined to read it.

Ah well, that is a different discussion.
 
... I don't disagree but why not write the science in a more concise and more easily digested form? More people might be more inclined to read it.
I tried to make it as clear/simple/short as possible. But if you only make a concise statement it doesn't really prove/show anything. And when you are trying to explain things to dispel myths/misunderstandings you really do have to cover the misunderstanding and why it is erroneous.
 
There's no real need to understand why X happens when you do Y, just that it does.

I think this "depends" on how you are using photography and indeed what you mean by science. If you are struggling to make photographs that look the way you want them to look, then understanding how things work can help you work out why.

But I don't really see papers like Stevens to really be about "why", but more about "how" ie how does changing ISO affect how my photographs look, with just enough why to help visualise what is going on, so you can make intelligent decisions about how you might want to use ISO to help you achieve the image you are visualising.
 
There's no real need to understand why X happens when you do Y, just that it does.
I largely agree with this... but the problem is when what is happening is really Z when you think it is X.

For instance, take a cormorant on the water where there are specular reflections in the ripples/waves. If you underexpose the image (place the specular highlights to the right w/o clipping) absolutely **nothing good can come of it. And if you underexpose the image using SS/Ap the results will be decidedly worse... most understand that as a form of exposing to the right in order to maximize the recorded dynamic range, when in fact it is the opposite.

**you cannot save a specular highlight where there is no detail... you can only make it grey.

Or you decide to buy the new Z6 instead of the Z7; because you believe the Z6 will have better low light performance because it has larger pixels... when it won't; not really.
 
Last edited:
I agree with this... but the problem is when what is happening is really Z when you think it is X.

For instance, take a cormorant on the water where there are specular reflections in the ripples/waves. If you underexpose the image (place the specular highlights to the right w/o clipping) absolutely **nothing good can come of it. And if you underexpose the image using SS/Ap the results will be decidedly worse... most understand that as a form of exposing to the right in order to maximize the recorded dynamic range, when in fact it is the opposite.

**you cannot recover a specular highlight where there is no detail... you can only make it grey.

See. You've lost me already. I don't know what point you're trying to make. Blown highlights can't be recovered. What's new? You make a decision to either preserve highlight detail, or ignore it and let the highlights blow. That's an aesthetic choice, not a technical one.

As @myotis said, a lot (all?) depends on how you are using photography. And I'm out of step with the majority on here in that respect. I'm really not very interested in making perfect photographs. So I'd better leave this thread now.
 
.... I don't disagree but why not write the science in a more concise and more easily digested form? More people might be more inclined to read it.

I came to this thread expecting to learn something without being blinded by so much longwinded presentation of the science. Such a thesis is fine but a more concise and more easily digested version as an alternative in addition would have been more helpful.
Why don’t you do a précis of it then? ;) . Some things are just complicated and some things are hard, sometimes both.
However you can still do things without understanding the underlying technology, maybe by ’understanding’ it intuitively or by trial and error.
Throughout history artists have been obsessed by technology and craftsmanship, much more than most art historians and critics give them credit for.
 
Thanks Steven, I've often felt that descriptions I had seen of ETTR weren't hitting the mark and you have cleared that up.
 
Why don’t you do a précis of it then? ;) . Some things are just complicated and some things are hard, sometimes both.
However you can still do things without understanding the underlying technology, maybe by ’understanding’ it intuitively or by trial and error.
Throughout history artists have been obsessed by technology and craftsmanship, much more than most art historians and critics give them credit for.

.... Why don't I do a précis of it then? - Because it requires the depth of understanding which Steven has in order to break it down and cut it down. I don't have enough scientific knowledge on this subject to do it justice and personally would rather be out taking photographs then writing about it.

Yes, I agree that well exposed images can still be achieved without knowing the science and I think it becomes intuitive through the experience of some trial and error.

One of the things I love about mirrorless cameras is that you can display the histogram in the viewfinder and also see the image change as you change the settings (I shoot Manual-mode RAW files). WYSIWYG (What You See Is What You Get). But of course, technology hasn't advanced far enough yet to show us noise and perhaps may never do so (I'm just thinking aloud).

Just by seeing the graphic shape and extent of a histogram it informs in the same way as seeing the position of watch hands without having to see the numbers. It almost becomes abstract and doesn't require conscious analysis.

But please don't think that I don't greatly respect Steven's effort and intentions by creating this long thesis. It is clearly helping quite a few people here.
 
Last edited:
See. You've lost me already. I don't know what point you're trying to make. Blown highlights can't be recovered. What's new? You make a decision to either preserve highlight detail, or ignore it and let the highlights blow. That's an aesthetic choice, not a technical one.

But that is the entire point. It's not about making technically perfect photographs, just for the sake of making technically perfect photographs, it's about developing your technical understanding to a level where you have the skills to achieve your creative/aesthetic choices. I assume when you said perfect, you meant "technically" perfect, rather than how perfectly an image matched the photographer's aesthetic vision.

In the example Steven gave in his paper, it was considered aesthetically desirable to include both shadow and highlight detail (a common issue with bird photography), and by understanding how iso and sensors work, he was able to retain both. But I don't think we should get bogged down with highlights, as the principle is about mastering your craft so you have control over how your photographs look.

As you will probably know, this was what drove Ansel Adams to become such a good technician, because when he started off he was incapable of creating prints that matched how he '"saw" the subject.

But, and it's a big but, "good" photography covers a very wide spectrum of images and how important technical understanding is to achieving your vision will vary. None the less, I think the greater your understanding of technique the greater your freedom to make unconstrained aesthetic choices.

But, its all a matter of degree, and you actually illustrated it in your example. Your technical understanding about blown highlights, allowed you to make an aesthetic decision on how you might want the highlights/shadows to appear in your final image. Steven's paper is just taking you into a more sophisticated understanding of how ISO and sensors work, which gives you a third choice, of sometimes retaining both highlights and shadows in images where you might have previously thought you would have to lose detail in one or the other.

Again, I don't want to get bogged down in the highlight/shadow thing, its just a good example of how technical understanding affects aesthetic decisions.
 
But that is the entire point. It's not about making technically perfect photographs, just for the sake of making technically perfect photographs, it's about developing your technical understanding to a level where you have the skills to achieve your creative/aesthetic choices. I assume when you said perfect, you meant "technically" perfect, rather than how perfectly an image matched the photographer's aesthetic vision.

In the example Steven gave in his paper, it was considered aesthetically desirable to include both shadow and highlight detail (a common issue with bird photography), and by understanding how iso and sensors work, he was able to retain both. But I don't think we should get bogged down with highlights, as the principle is about mastering your craft so you have control over how your photographs look.

As you will probably know, this was what drove Ansel Adams to become such a good technician, because when he started off he was incapable of creating prints that matched how he '"saw" the subject.

But, and it's a big but, "good" photography covers a very wide spectrum of images and how important technical understanding is to achieving your vision will vary. None the less, I think the greater your understanding of technique the greater your freedom to make unconstrained aesthetic choices.

But, its all a matter of degree, and you actually illustrated it in your example. Your technical understanding about blown highlights, allowed you to make an aesthetic decision on how you might want the highlights/shadows to appear in your final image. Steven's paper is just taking you into a more sophisticated understanding of how ISO and sensors work, which gives you a third choice, of sometimes retaining both highlights and shadows in images where you might have previously thought you would have to lose detail in one or the other.

Again, I don't want to get bogged down in the highlight/shadow thing, its just a good example of how technical understanding affects aesthetic decisions.

If I have a contrasty subject and want detail at both extremes I stick my Nikon in highlight priority mode. That preserves highlight detail. Knowing from experience that this mode is conservative I usually add some positive exposure compensation too. Then fix the shadows in post. It works for me. How and why it works is of no interest. I've never referred to a histogram in my life. If it looks right, it is right.

As for the having a vision and making a picture that matches it I'm more inclined to the Winnogrand approach than that of Adams - photographing things to see what they look like photographed. How cameras see the world has always fascinated me. How they work never has. I often feel I don't belong in TP land! :LOL:

PS TMS has started. Tara.:)
 
Last edited:
But that is the entire point. It's not about making technically perfect photographs, just for the sake of making technically perfect photographs, it's about developing your technical understanding to a level where you have the skills to achieve your creative/aesthetic choices.
Thank you... that is exactly the intention.Having fuller technical knowledge is never a bad thing... but it may not be required.

To put it in terms of the exposure triangle:
If I am a creative artist and I just use a camera to make a record of my creation then I may not need to understand exposure at all. Using the camera on P/auto may be more than adequate.
Or perhaps my only consideration is to make the image darker... I don't need to know why/how each exposure factor works; I only need to know which way to adjust them, and I can use any of them to get the desired result.
But understanding why and how each factor works, and what that means to the final result, allows me to make much more informed choices with greater creativity/control and better results.

If I have a contrasty subject and want detail at both extremes I stick my Nikon in highlight priority mode. That preserves highlight detail. Knowing from experience that this mode is conservative I usually add some positive exposure compensation too. Then fix the shadows in post. It works for me.
You're making the mistake of seeing dynamic range as being dark and light... it's not. It's the number of levels/steps between the darkest and brightest recordable... the dynamic range of a scene almost always exceeds your ability to record all of it (most steps of dynamic range exists in darks/shadows).

In your scenario, which exposure variable is being adjusted? Which one should be and why? In which situation, with which camera? Is adding the exposure compensation actually beneficial, or is it a potential negative? Are you loosing recorded dynamic range, shifting the recorded dynamic range, or extending the recorded dynamic range? And do you have a choice in that?
And most importantly, is there actually any point/potential benefit? If not, then why bother?

I know the answers to all of those questions, and I've done my best to explain them in the pdf. If you know all of the answers then great; and if you don't care, that's fine too.
 
Last edited:
Many thanks for that great piece of work, I have briefly scanned through it, and will do so in more detail over the weekend. Thank you for taking the time to produce this, and more so for sharing it with us all. :)
 
How cameras see the world has always fascinated me. How they work never has.

Fair enough, I obviously want to have more insight into "how" the camera sees the world so I can better match what I "think" I want, with what the camera is going to give me.

But I also have Winogrand moments as well as Ansel Adams moments (in approach, not in the quality of photographs) that's the great thing about photography, it can offer very different things to different people. Or different things to the same person, depending on the day and mood.

I use Nikon's highlight weighted metering for bird photography. I have it set up on the Pv button, so I can turn it off and on as white birds move between shade and bright sunlight. I'm usually giving something like +2/3 stop EC, so it looks as if, in different circumstances, I'm using it a bit like you are.

I don't know how it works either, but it seems reasonably effective when I don't have time to think too much about the exposure, and want to avoid blown highlights.
 
...that's the great thing about photography, it can offer very different things to different people. Or different things to the same person, depending on the day and mood.

(y)
 
I use Nikon's highlight weighted metering for bird photography. I have it set up on the Pv button, so I can turn it off and on as white birds move between shade and bright sunlight. I'm usually giving something like +2/3 stop EC, so it looks as if, in different circumstances, I'm using it a bit like you are.

I don't know how it works either, but it seems reasonably effective when I don't have time to think too much about the exposure, and want to avoid blown highlights.

Since this is for bird photography I can make some assumptions/recommendations. I'll assume you are using a fairly recent Nikon and not a single digit body (D3/D4/D5). I'll also assume the FL, available apertures, and SS means the ISO is rather elevated. And I'll assume you care at least equally about the lights and the darks recorded.

I would set the camera to highlight weighted metering in aperture priority with auto ISO enabled. I would have the minimum ISO set to 400 and the minimum SS set as required. And I would not add the +EC.

The reasoning for this is:
*The sensor is underexposed... it is not sensor/pixel clipping you are concerned with, it is ISO clipping.
*Setting the camera to matrix highlight weighted metering tells it to underexpose when it detects enough light tones that will blow out; but ISO is not light/exposure.
*Pretty much all of the relatively recent Nikons (not single digit) are essentially ISO invariant above ISO 400... it's actually ISO 200 for some (like the D810) but ISO 400 works for all.
*Setting the ISO to a min of 400 prevents the ISO dropping below where it is providing an increase in the SNR in the darker parts of the image (lower noise)... if you don't care equally about the darker parts then set it to base. But I often care at least as much/more about the darker regions and underexposing the sensor at ISO 400 is quite tolerable with most Nikons... it also results in higher SS's which is usually helpful.
*Not setting the +EC prevents the ISO from shifting higher when not enough light tones are detected and the camera doesn't underexpose as much. This prevents the ADC from clipping the highlights and has no penalty within the ISO invariant range (it expands the recorded dynamic range).

If any of my assumptions are wrong then this advice is wrong. And you certainly do not have to do it how I would... manual exposure with auto ISO would work, full manual with you shifting the ISO would work (painfully), etc.
 
I would set the camera to highlight weighted metering in aperture priority with auto ISO enabled. I would have the minimum ISO set to 400 and the minimum SS set as required. And I would not add the +EC.

If any of my assumptions are wrong then this advice is wrong. And you certainly do not have to do it how I would... manual exposure with auto ISO would work, full manual with you shifting the ISO would work (painfully), etc.

Thanks for this, your assumptions are correct. It's a D500, which I use in manual + auto iso with more recently a minimum ISO of 400 (based on the graphs at Bill Claff's site, before that, I used base). I have it set up to show the ISO in the viewfinder and just keep an eye on it when adjusting SS.

The +2/3rds EC isn't a specific component of using HWM, it's just a consequence of how I use HWM. It's just a compensation setting I often seem to end up with to improve shadow areas.

I tried out, and now use, HWM for a very specific situation of when I was surprised by a something like a Great White Egret appearing overhead with bright sunshine reflecting off the white feathers which always ended up with large areas of the bird blown out in all three channels. I have a choice of grabbing a shot, which I couldn't use or adjusting the EC to save the highlights, but missing the shot. Having HWM on the Pv function key allows me to instantly give some correction for the highlights and now saves some shots I would have lost in the past.

It doesn't help getting the bird in the frame and in focus however, I'm still learning that part :-(

But I'm still learning lots of parts, as my foundations in photography come from mainly static subjects, sheet film (mainly B/W) and a spot meter. The way that ISO works in digital is very much something I am still learning and still testing to see how things are affected in practice.
 
As ISO in the digital domain is actually the photosite or channel amplifier gain, increased ISO can increase the noise as the amplifier noise figure is usually gain dependant.
 
As ISO in the digital domain is actually the photosite or channel amplifier gain, increased ISO can increase the noise as the amplifier noise figure is usually gain dependant.

Or indeed reduce noise, depending on circumstances.
 
Or indeed reduce noise, depending on circumstances.
Amplifiers can't have a negative noise figure. They always add noise, they may reduce the signal to a point where it's below the visibility threshold but it's still there.
 
For about the last month I have been posting about topics related to digital exposure theory to my FB page; but I am way too lazy to copy them all here. However, due to many requests I have compiled them into a single PDF. Here's a link to the file on my google drive if anyone is interested.
https://drive.google.com/open?id=1iemxFctZNApqSlcigd-CzQCRtAiqf5s9

Well, I have to say that I try to get by in life without learning anything I don't really need to (an idea stolen from Albert Einstein! Well if the quote was true that is)

And this kind of techie stuff is usually right up my ignore street

So I'm more than pleasantly surprised that I did in fact read it all, understand it due to the OP's easy language, and find it actually interesting; though I do admit to having read snippets along these lines before and hence didn't really learn much new, just clarified a few bits for me :)

Will I alter anything about what I do? Probably not

Does the average tog need to know it? Probably not

Will other techie types argue for/against Steve's article? Probably YES :D

The biggest 'take-away' here for me is this last comment...

"The queston ofen arises, “should I go full frame, should I get a higher resoluton camera, or should I get a
longer lens?” The truth is, it ofen doesn't make much diference."


Which I totally agree with, but wish had been spell-checked lol

Great piece of work Steve, thanks :)

Dave
 
Back
Top