[Ian Stuart Forsyth]

Is ETTR that relevant these days with ISO invariant cameras?

The answer to this question is simple, Yes
An invariant sensor does not in its self decrease noise and or increase DR , what gives us DR of the current cameras is mainly how much light we can capture. With ETTR we are maximizing how much light we can capture.

It is just as important in decreasing noise in most of the areas within your photograph.

One of the reasons is how light data is stored for us to use, one of the problems is that for a given scene how much light that is captured varies over the image.
If we break it down into zones 10 highlights being the near FWC that area shows the least amount of noise but as we start to look at areas in the photograph that fall beneath this was start to see a trend.

Areas that fall into zone 9 are made up with 1/2 as much light as the FWC, but still show very little noise
Zone 8 these areas are only made up of 1/4 as much light as the FWC
Zone 7 1/8 as much light
Zone 6 1/16 as much light
Zone 5 1/32 as much light as FWC now we can start to see some of the results of this and this starts to become important to the photographer as this is where their I mid tones start to fall
Zone 4 1/64 as much light as FWC
Zone 3 1/128 as much light as FWC

So when we get into Zone 1 we are sitting at 1/512 as much light as FWC

Here we are dealing with very small amounts of light, this will degrade IQ and color information. If you ETTR you can double the amount of light at that lower limit and this can help improve the IQ

---

Last edited by Ian Stuart Forsyth; 03-10-2021 at 09:52 PM.

[GUB]

The answer to this question is simple, Yes
An invariant sensor does not in its self decrease noise and or increase DR , what gives us DR of the current cameras is mainly how much light we can capture. With ETTR we are maximizing how much light we can capture.

It is just as important in decreasing noise in most of the areas within your photograph.

One of the reasons is how light data is stored for us to use, one of the problems is that for a given scene how much light that is captured varies over the image.
If we break it down into zones 10 highlights being the near FWC that area shows the least amount of noise but as we start to look at areas in the photograph that fall beneath this was start to see a trend.

Areas that fall into zone 9 are made up with 1/2 as much light as the FWC, but still show very little noise
Zone 8 these areas are only made up of 1/4 as much light as the FWC
Zone 7 1/8 as much light
Zone 6 1/16 as much light
Zone 5 1/32 as much light as FWC now we can start to see some of the results of this and this starts to become important to the photographer as this is where their I mid tones start to fall
Zone 4 1/64 as much light as FWC
Zone 3 1/128 as much light as FWC

So when we get into Zone 1 we are sitting at 1/512 as much light as FWC

Here we are dealing with very small amounts of light, this will degrade IQ and color information. If you ETTR you can double the amount of light at that lower limit and this can help improve the IQ

All good but I just thought I would add that ETTR was only ever relevant at base iso. As such it is totally compatible with utilising iso invariance.

[BarryE]

The answer to this question is simple, Yes
An invariant sensor does not in its self decrease noise and or increase DR , what gives us DR of the current cameras is mainly how much light we can capture. With ETTR we are maximizing how much light we can capture.

It is just as important in decreasing noise in most of the areas within your photograph.

One of the reasons is how light data is stored for us to use, one of the problems is that for a given scene how much light that is captured varies over the image.
If we break it down into zones 10 highlights being the near FWC that area shows the least amount of noise but as we start to look at areas in the photograph that fall beneath this was start to see a trend.

Areas that fall into zone 9 are made up with 1/2 as much light as the FWC, but still show very little noise
Zone 8 these areas are only made up of 1/4 as much light as the FWC
Zone 7 1/8 as much light
Zone 6 1/16 as much light
Zone 5 1/32 as much light as FWC now we can start to see some of the results of this and this starts to become important to the photographer as this is where their I mid tones start to fall
Zone 4 1/64 as much light as FWC
Zone 3 1/128 as much light as FWC

So when we get into Zone 1 we are sitting at 1/512 as much light as FWC

Here we are dealing with very small amounts of light, this will degrade IQ and color information. If you ETTR you can double the amount of light at that lower limit and this can help improve the IQ

Ian, my question was too simple and a bit too much of a throwaway. I understand the sensor's light gathering ability is biased towards the highlights and ISO invariance comes into play at high(er) ISOs. As I use base ISO, ETTR is helpful as you point out, however, for some who use higher ISO settings, then it becomes more nuanced. Anyway, I will continue to ETTR when I'm shooting slow on a tripod and when I'm not I'll rely on post recovery.

Another advantage with ETTR is that it slows me down. With the image composed, time is taken to check how the light is distributed via the histogram. A few tweaks in exposure often leads to a change in composition and more time actually thinking about what is being shot. All good ...

[swanlefitte]

A bit tangent but current, as in just came in and everyday this week. I have been shooting cardinals of late. ETTR gives me good color and an eyeball but a blurry bird. ETTL gives me clear feather and beak detail with bad color and no eye. It is an interesting experiment. It shows me how knowing my subject and good technique still carries the day. Most people around here don't even know a cardinal song and can't find them except at a bird feeder sadly.

[BarryE]

A bit tangent but current, as in just came in and everyday this week. I have been shooting cardinals of late. ETTR gives me good color and an eyeball but a blurry bird. ETTL gives me clear feather and beak detail with bad color and no eye. It is an interesting experiment. It shows me how knowing my subject and good technique still carries the day. Most people around here don't even know a cardinal song and can't find them except at a bird feeder sadly.

Don't I just hate it when folk in warmer climes bang on about their colourful and interesting birdlife. UK birds are mostly LBJs ** and variants of. Please be considerate

** LBJs, in case you don't know are Little Brown Jobs.

[TonyW]

A bit tangent but current, as in just came in and everyday this week. I have been shooting cardinals of late. ETTR gives me good color and an eyeball but a blurry bird. ETTL gives me clear feather and beak detail with bad color and no eye. It is an interesting experiment. It shows me how knowing my subject and good technique still carries the day. Most people around here don't even know a cardinal song and can't find them except at a bird feeder sadly.

Yes, but cardinals are usually very slow as they are retired. Try going for the much younger and normally more agile bishops or archbishops 🤗

Seriously, what you are describing cannot happen with ETTR or even ETTL. So I cannot understand what you mean or what you may be doing.

ETTR is meant to be used to achieve optimal exposure and maintain the fullest DR of your system. It is only relevant at base ISO and only then when your scene dynamic range sits within your cameras dynamic range. Anything else and blown details is just overexposure. ETTL is just underexposure in comparison.

[GUB]

A bit tangent but current, as in just came in and everyday this week. I have been shooting cardinals of late. ETTR gives me good color and an eyeball but a blurry bird. ETTL gives me clear feather and beak detail with bad color and no eye. It is an interesting experiment. It shows me how knowing my subject and good technique still carries the day. Most people around here don't even know a cardinal song and can't find them except at a bird feeder sadly.

That is why I feel ETTR is a bit of a dogma. ( I presume you have saying to achieve ETTR you are photographing with too slow a shutter speed an/or too shallow a dof)
Really it is simple - there are 2 variables available to you to put light on the sensor. (Aperture and shutter)
Give the sensor as much light as you can within the limits of your shooting environment while bearing in mind the added noise from starving your sensor.
It is all a balancing act.

[swanlefitte]

I have been quoted 3 times but 2 about my point. This quote is my point.
"It is only relevant at base ISO and only then when your scene dynamic range sits within your cameras dynamic range"
Yes. That was my point. That is the point where technique trumps tech.

[GUB]

I have been quoted 3 times but 2 about my point. This quote is my point.
"It is only relevant at base ISO and only then when your scene dynamic range sits within your cameras dynamic range"
Yes. That was my point. That is the point where technique trumps tech.

Do we have differing views of ETTR definition? To me it is exposing to the right but not clipping the bright end of the histo.

[TonyW]

I have been quoted 3 times but 2 about my point. This quote is my point.
"It is only relevant at base ISO and only then when your scene dynamic range sits within your cameras dynamic range"
Yes. That was my point. That is the point where technique trumps tech.

Ahh, the confusing part for me was that it looked like you were suggesting that ETTR was causing blurring and ETTL good detail and poor colour. Pleased to see it wasn’t

[GUB]

and only then when your scene dynamic range sits within your cameras dynamic range"

This is the bit I don't understand?

[TonyW]

and only then when your scene dynamic range sits within your cameras dynamic range

This is the bit I don't understand?

Not trying to teach granny the proverbial but ...
ETTR is all about optimising the S/N ratio and this applies only at base ISO and requires as you know that the exposure is pushed as far to the right as possible without clipping any important highlight detail. Ideally it should be based on a raw histogram but unfortunately cameras with this facility are in short supply - the JPEG histogram will lead to an underexposed raw file. As will relying on most raw editing programs as they invariably add a lot under the hood leading to a false sense of exposure!

Imagine you have a scene where you have measured the dynamic range that you need to capture and that turns out to be 11 EV, but your camera is only capable of capturing 9EV of range. You have a couple of options capture as much of the highlights as possible and let the shadows fall where they may or vice versa or turn to multi exposure (HDR)

Now imagine a scene with a dynamic range of only 8 stops sitting comfortably within the cameras DR. Making sure to optimise exposure by pushing to the right stopping short of highlight clipping will ensure that you have optimised S/N ratio leading to lower shadow noise and also maintained maximum image DR

[GUB]

Not trying to teach granny the proverbial but ...
ETTR is all about optimising the S/N ratio and this applies only at base ISO and requires as you know that the exposure is pushed as far to the right as possible without clipping any important highlight detail. Ideally it should be based on a raw histogram but unfortunately cameras with this facility are in short supply - the JPEG histogram will lead to an underexposed raw file. As will relying on most raw editing programs as they invariably add a lot under the hood leading to a false sense of exposure!

Imagine you have a scene where you have measured the dynamic range that you need to capture and that turns out to be 11 EV, but your camera is only capable of capturing 9EV of range. You have a couple of options capture as much of the highlights as possible and let the shadows fall where they may or vice versa or turn to multi exposure (HDR)

Now imagine a scene with a dynamic range of only 8 stops sitting comfortably within the cameras DR. Making sure to optimise exposure by pushing to the right stopping short of highlight clipping will ensure that you have optimised S/N ratio leading to lower shadow noise and also maintained maximum image DR

Your perception of ETTR matches mine.
And yes in high dynamic range situations you need to override it (and clip the highlights a bit) if you value the noise quality of the dark regions over the tonal range of the highlight regions.
But apart from that decision ETTR is still best practice in contrasty images isn't it?
I don't understand how that is relevant to the cardinal bird scenario.

I find the "histo is of the jpg" thing a bit overblown. But then I am always in Raw (all manual) and I leave my jpg variables alone and have highlight and shadow correction turned off. This means the relationship between the jpg histo and the raw remain fairly constant. Mostly the amount of "spare" in the histo at the clipping end is a nice margin of error if you are pedantic about highlight clipping (like I am!!)

[BarryE]

I've found that comparing an image's histogram (the jpg version with clipping info) with the RAW histogram on screen, for several types of images, has helped me "calibrate" where I can push the camera's histogram and still not blow the highlights in post. The clipping warnings and the histogram are conservative and there's often room to push the image a little further to the right and still keep in the camera's DR. However, I've also learnt that I often find I prefer a slightly underexposed image and my earlier obsession with ETTR was not helping. A slightly underexposed image, to me, has better saturation and often more mystery/interest. Less is more, perhaps ...

[TonyW]

I think one of the biggest mental hurdles to overcome is the fact that an optimally exposed raw is really likely to appear light and washed out in your raw editor with a histogram that suggests overexposure. Often you will hear comment about just how good x raw editor is in recovery of highlights.

The truth may be that your raw editor has applied various under the hood corrections including Baseline exposure corrections and what you thought of as over is actually underexposure up to the tune of +2EV. It should be a very easy task to bring the image down in exposure and up in saturation to reach image nirvana.