ETTR - the "Experts" have spoken...

[RichDesmond]

I have been a serious photographer for 40 years, and one using digital cameras since 2003. I had no idea what ETTR meant - just Googled it to find out. I thought it must be a metering mode like TTL. I have no idea how I have ever been able to take a decent picture without having this knowledge.

You're obviously one of those losers who goes out and takes pictures instead of doing what serious photographers really do, which is argue arcane points on the internet with strangers...

 

[PakkyT]

I have been a serious photographer for 40 years, and one using digital cameras since 2003. I had no idea what ETTR meant - just Googled it to find out. I thought it must be a metering mode like TTL. I have no idea how I have ever been able to take a decent picture without having this knowledge.

Ya I think it is one of those techniques that was found and used when our sensors had more noise and a less dynamic range that modern ones. I am not convinced that as a technique it is very useful anymore except to the extreme pixel peepers who are often more interested in each pixel than the photo they actually took. I tried it a little when I first heard about it back in my 4/3rds days but then decided it was a lot of work for little gain considering how & what I normally enjoyed shooting. I find it much easier to simply expose for what it is I want to capture, even if that means crushing the blacks or blowing the highlights in the things I don't feel important to the photo.

 

[Robert Watcher]

You're obviously one of those losers who goes out and takes pictures instead of doing what serious photographers really do, which is argue arcane points on the internet with strangers...

obviously. I had no idea.

 

[BushmanOrig]

obviously. I had no idea.

Love your Instagram account - great work ?

 

[dirtdevil]

Love your Instagram account - great work ?

Ahhh he's the guy of the infamous "debit/credit card payment on a hockey stick" photo.

 

[wyk]

Sure when those compared images all shot at the same ISO. As has been explained to "Einstein", no one is disputing that if you add more light to the sensor through shutter or aperture you will get better SNR and better details in your image. The way you are shooting is how a lot of us will shoot simply to make sure we have some choices to work with when we get home. However he is talking specifically about using ETTR as the image quality tool which is a pain in the arse method of trying to get the absolute highest detail you can out of your camera and it used by someone who is setting up for that one perfect shot, typically a landscape, and has the time and the patience to really work the controls to optimize that shot. No bracketing, no underexposing or proper exposing, but one shot pushed to the right on purpose.

Your (not done for the ETTR technique specifically) example illustrates the fundamental flaw if you were using this specific tool on purpose at higher ISOs. Assuming your +0.7 EV photo was done with shutter and/or aperture and not ISO, if you are able to exposure 0.7EV higher why not simply shoot at that aperture and shutter at -0.7 ISO? As an example, you are at ISO 1600, f5.6, and 1/200s and when you bracket say your shutter changes ('cause you are in A mode). So your +0.7EV shot ends up 1600, f5.6, 1/125s. On the computer you find that if you roll back the exposure to 0.0 EV it looks great, so your image is effectively ISO 1000 with the same f5.6 1/125s settings, right? Works great for those of us who hedged their bets and bracketed and we ended up with a shot we like. However if you purposely set out to use the ETTR technique and went to all that trouble, my first question to you when you showed me the shot as proof of using the technique would be "Great, but why didn't you just shoot it at ISO 1000 to begin with? That probably would have resulted an even cleaner image."

TL ; DR - Basically if trying to use ETTR when you are able to bump up exposure through shutter and aperture, you will always get an even better image by also dropping the ISO to get back to 0EV and not having to do it after on the computer.

Oh, I'm not bothering with that ETTR nonsense.
I just take photos.

 

[BushmanOrig]

Interesting reactions to a simple technique that will benefit every MFT photographer's results positively.

Instead, I see a lot of talking but no effort to personally take a camera, produce examples, and show everyone what all the talking is about... Those who seem to be really in the deep, get silly instead of simply asking?

I guess if I wrote something like full-frame sensor cameras are better than MFT cameras, they would all give it a like and continue standing small...?

Love my Olympus ?

 

[Danny.]

Interesting reactions to a simple technique that will benefit every MFT photographer's results positively.

Instead, I see a lot of talking but no effort to personally take a camera, produce examples, and show everyone what all the talking is about... Those who seem to be really in the deep, get silly instead of simply asking?

I guess if I wrote something like full-frame sensor cameras are better than MFT cameras, they would all give it a like and continue standing small...?

Love my Olympus ?

The biggest problem you have goes by the name Robin Wong. Now theres a clever guy for sure.

Danny.

 

[Paulb123]

My views are very similar to Wyk. ETTR is simply a technique too many that would take away some love of taking pictures through the OVF, knowing what you see is what you get and especially as the OVF has highlighting for exposure (so keeps me on the straight and narrow)
It just feels counter tuitive - especially as most of my photography problems can be tweaked out post processing even in JPEG.
Indeed I only really deliberately over expose when taking photos of birds in flight over a bright sky (to kill off the sky but try to get more colour in the birds).
Each to their own.

 

[BushmanOrig]

The biggest problem you have goes by the name Robin Wong. Now theres a clever guy for sure.

Danny.

Really is that all? I realize people accept many things they told without thinking or questioning... but being this sensitive about the fact that I questioned someone?

So Robin takes a night scene, places an object, and lit it up to simulate "daylight" conditions. Although it looked like a night scene he basically simulated a daylight situation. Nothing wrong about that except it is really difficult to explain a high ISO, low-light situation the way he did.

The example I showed is closer to a real-life low-light situation where it is impossible to take handheld images at ISO200. How often are you inside a building, in a city at night or during a lovely blue hour and you have to up the ISO. You also like to keep noise to a minimum.

Robin did not give a solution to a real-life example... What he probably tried was simulating a landscape onto a small object? When demonstrating ETTR at base ISO it's really important to use something like a real landscape scene where one can point out image detail versus targetting image noise (low-light conditions).

In other words in terms of base ISO, the focus will be more on image details or recording maximum image data. In low light conditions the focus will typically be on reducing noise. That said, ETTR always benefits both quality & noise.

Why? If you study my detailed article on this subject plus the references I included you will see the biggest noise generator is the sensor. The biggest image quality breakdown also takes place at the sensor level. No other camera component has as much a unique impact as the sensor. How do we optimize sensor performance? The answer is exposure...

My comment about Robin - scroll down a little

 

[BushmanOrig]

Here are some extreme examples with a static scene shot with the same aperture and shutter speed, just changing ISO. This was a very boring experiment to do and I’d much rather be shooting landscapes but I seem to have some time on my hands at the moment with no access to vast landscapes for some reason. I’ll let the collective IQ of the forum draw their own conclusion.


Extreme underexposure
ISO 200 underexposed by 3 stops, corrected back to zero by adding +3EV compensation in post:
View attachment 824862

View attachment 824863

Extreme overexposure
ISO 6400 overexposed by 2-stops corrected back to zero by reducing -2EV compensation in post (so ETTR):
View attachment 824865

View attachment 824864

Thanks for doing the effort.

I am not familiar with the MKIII. I read somewhere it is ISO compliant and will react differently to my MKII. I read only 2 Olympus models are currently compliant, the rest are not.

Please do me a favor. It looks as if the reds are breaking in your examples. Study the shadow.

It will be nice to see the 0EV image at ISO6400

 

[Bob T]

...all this noise about noise....

Back to taking photos....

 

[John King]

Thanks for doing the effort.

I am not familiar with the MKIII. I read somewhere it is ISO compliant and will react differently to my MKII. I read only 2 Olympus models are currently compliant, the rest are not.

Please do me a favor. It looks as if the reds are breaking in your examples. Study the shadow.

It will be nice to see the 0EV image at ISO6400

Here's one, Siegfried ...

Subscribe to see EXIF info for this image (if available)

Want to see some more?

 

[BushmanOrig]

Here's one, Siegfried ...

Want to see some more?

Great image John - I love the 12-50mm, probably one of Olympus best...

You do not say but it looks like the image was taken in Aperture Mode and you did not alter the exposure. If the case you could have dropped the background noise with 10 - 15% and increased the details between 5 - 8% by increasing the exposure with +/- half a stop at ISO6400.

It does seem you have no issue correcting the exposure in raw?

 

[PakkyT]

The ISO 6400 shot with 0EV compensation in post:

...and for completeness, the ISO 200 shot with 0EV compensation in post:

Funny I was just thinking about this yesterday. Since the camera captures light via aperture & shutter and then ISO amplifies that after, in theory you could shoot everything at ISO 200 like your second example, then turn it up to "6400" (+5EV) and get the same results as the SOC 6400 shot. I would be interested if you turned EV up on that second shot to +5EV, preferably in Oly Workspace, to compare to the first one (assuming both shots were shot identally other than ISO).

 

[BushmanOrig]

To be fair there was a very small AWB change between shots which I have now corrected above to make it a completely fair comparison.

My friend the reason I am posting and doing the effort to write articles is I have a passion for photography, the Olympus brand, and I love to be of value to other Olympus photographers.

I get the feeling you purposely posted incomplete information with your first lego blocks images and the second group. Fortunately, I see it and I refuse to take your bait... What disappoints me is this game is super selfish and helps nobody, its a complete waste of energy and time...

For those who do not know what is happening go look for the Peter Forsgard video on ISO compliance...

For now I am done with this discussion - this has NO value for anyone...

 

[RichDesmond]

Funny I was just thinking about this yesterday. Since the camera captures light via aperture & shutter and then ISO amplifies that after, in theory you could shoot everything at ISO 200 like your second example, then turn it up to "6400" (+5EV) and get the same results as the SOC 6400 shot. I would be interested if you turned EV up on that second shot to +5EV, preferably in Oly Workspace, to compare to the first one (assuming both shots were shot identally other than ISO).

Go back and read the LuLa thread I linked to earlier

 

[BushmanOrig]

I have no idea why that comment, of all the comments here, has offended you.

Thanks to you noticing differences in the reds I realised that I had taken my shots with Auto White Balance (AWB) switched on. The WB was 4800K +10 tint on the ISO 200 shot and 4500K +8 tint and in the ISO 6400 shot. Since noticing this, as a result of your correct observation about colour differences, I have removed that variable by levelling the WB of the former to match the later.

There is no game here. Please, take a copy of the RAW files and see for yourself.

raw files

Accepted - I will have a look at the raw files... Pls go look, you posted more than the 2 raw files??

 

[Machi]

Funny I was just thinking about this yesterday. Since the camera captures light via aperture & shutter and then ISO amplifies that after, in theory you could shoot everything at ISO 200 like your second example, then turn it up to "6400" (+5EV) and get the same results as the SOC 6400 shot. I would be interested if you turned EV up on that second shot to +5EV, preferably in Oly Workspace, to compare to the first one (assuming both shots were shot identally other than ISO).

M43 cameras aren't isoless so there is a real gain in IQ with higher ISO.
That also means that ETTR could be in theory used at higher than base ISO.
But truth is that RAW shooter can limit its M43 camera at range ISO200-400 and will loose almost nothing in the IQ area.
It can be seen here at Dpreview site where images taken at the same physical exposure (aperture/shutter speed) but at different ISO setting can be compared.

 

[PakkyT]

Go back and read the LuLa thread I linked to earlier

Took me a minute to figure out what "LuLa" is.

Interesting read. Is there a page 2 and 3 to that? When I click them I don't go to a new page but instead get an "index.php" file download.