@snibgo
It sounds as if OP’s camera behaves like Fuji’s X-T4:
Select a bracketing amount (±1/3, ±2/3, or ±1). Each time the shutter is released, the camera will take a picture at the current sensitivity and process it to create two additional copies, one with sensitivity raised and the other with sensitivity lowered by the selected amount.
Have fun!
Claes in Lund, Sweden
So it appears there is no difference in current at the sensors between the ISOs because it is only one capture from the camera.
OP: I think you will get different results if you manually take 3 images at each of the iso, keeping S and A constant (and trying to keep the scene lighting constant).
The DateTimeOriginal tag of all three images is exactly the same, 2023:02:01 13:00:29. What gives?
Oh, never mind, you bracketed ISO…
Still, doesn’t make sense. Why even have the dial on the camera if it doesn’t do what it’s supposed to… ???
Sure, that’s the limiting best case. I meant: real world amplifier electronics is part of the signal chain. And that is the part that has improved in the last 10 to 15 years so much so that you can almost do photon counting with individual pixels on consumer Si-CMOS. The analog circuits after photoconversion apply gain to the signal AND contribute noise. So, not looking at those amplifier stages seemed like an omission. That was the reason I wrote what I wrote. But yes of course, photon counting has \sqrt n noise.
Err, it’s a little bit more complicated than that. Please refer to the Emil Martinec web site above (all of it). You need quite a bit of signal for the pure photon shot noise approximation to be valid.
The Olympus E-M5 Mark III Instruction Manual at Download Olympus E-M5 Mark III Instruction Manual | ManualsLib , page 133, says:
ISO BKT (ISO bracketing)
The camera varies the sensitivity over three shots while keeping
the shutter speed and aperture fixed. You can select the
bracketing increment from 0.3 EV, 0.7 EV, and 1.0 EV. Each
time the shutter button is pressed, the camera shoots three
frames with the set sensitivity (or if auto sensitivity is selected,
the optimal sensitivity setting) on the first shot, negative
modification on the second shot, and positive modification on
the third shot.
Their words “shots” and “shoots” implies to me that the shutter opens three times. But the camera was hand-held, so I suggest the manual is wrong or misleading, and the shutter opened only once, and saved three files.
I suggest the experiment is repeated. Take a photo at a certain ISO, aperture and shutter speed. Alter the ISO and take another photo. Alter the ISO again and take third photo. Develop these in a very simple way. I think the three resulting images will show different lightnesses.
@g-man is quite correct. If I take a series of photos keeping the aperture and shutter speed constant but manually changing the ISO by hand, then the raw histogram changes as I change the ISO. I did a series of 200, 400, 800, 1600, 3200, 6400, and 12800 and all the histograms were different with it moving to the right as I increased ISO. So, it appears that this whole issue is a non-issue as a result of Olympus “cheating” with how it implements ISO bracketing.
Thank you to all who replied. I wouldn’t have reached this resolution without your comments and questions.
Question is, do you hear 3 clicks or one click, when you take a bracketed set? It seems it would be one - it would be funny if the camera actually took three shots but only used one.
It also occurs to me that if that camera is ISO-invariant, it’s actually a good way of doing bracketing… unless you’re already at the lowest ISO in which you’d loose headroom I suppose.
One click.
I own an Olympus m10 Ii. There should be a menu option that takes 3 consecutive images at different iso. There are too many menu options with Oly.
Going back to your original post. You are already at a fix shuttle and aperture, so all you can do is optimize the iso. Olympis Auto iso can be an option in this scenario, keeping the camera in manual mode.
As I had written above, I view this from my experience with astronomical CCD cameras. And there, if everything is done correctly, we do reach the photon noise limit.
Yes, the analog signal from the CCD is amplified before fed into the ADC. But this does not increase the sensitivity of the CCD in terms of detected photons.
Hermann-Josef
Yup. ISO-invariant means that you CAN choose to use a much lower ISO with no penalty (and, in fact, improved dynamic range) because your read noise is fixed regardless of ISO setting, so shooting a lower ISO means you can accumulate more photons before clipping. It does not guaranteed that amplification does not change - most cameras DO change amplification.
Ther ARE cameras on the market that “bake in” this idea and do not change amplification settings, instead recording the ISO value as an offset in metadata. IIRC some of the Fuji GFX medium format bodies do this?
In this case, it COULD be that the Olympus, under the hood, decides that when ISO bracketing, it will only take one raw and then alter the JPEGs, leveraging the concept of ISO invariance. This is a bit strange and unusual. That said, I’ve never bracketed ISO due to the concept of ISO invariance - on the majority of modern cameras it has little to no benefit.
A question would be - when this bracket “series” was taken, was only one shutter actuation heard? (Edit: I saw later that it was confirmed only one click was heard. That’s a dead giveaway Olympus is doing something a bit unusual in ISO bracketing mode.)
My thinking in choosing ISO bracketing was to optimize the full range of the ADC while controlling blowouts. That is, to avoid any quantization effects at low values – in other words, to use the full 12 bits of the ADC.
I think Olympus uses lightly modified Sony sensors under the hood which are typically what is considered “ISO invariant”, whereas Canon sensors for a long time (I don’t know if they could advance their readout in recent years) had different analog gain settings for full-stop increments of changing ISO.
When you plot the dynamic range of sensors, iso-invariant ones have a monotonic decrease vs. iso increase, iso-variant ones “gain back” a little dynamic range everytime they switch to a higher gain setting.
For example in this comparison here:
https://www.photonstophotos.net/Charts/PDR.htm#Canon%20EOS%2040D,FujiFilm%20X-T1
The sony derived X-T1 sensor is iso-invariant, the Canon 40D is anything but iso-invariant.
Newer sony sensors seem to have one, or sometimes two additional gain setting at some ISO, so they do the same trick as Canon sensors.
https://www.photonstophotos.net/Charts/PDR.htm#Sony%20ILCE-7RM5
Newer Canon sensors seem to have improved their dynamic range a bit, but still seem to have more fine grained gain steps.
https://www.photonstophotos.net/Charts/PDR.htm#Canon%20EOS%20R5,Canon%20EOS%20R7
And just for clarity the DR chart for the Olympus EM5mk3
https://www.photonstophotos.net/Charts/PDR.htm#Olympus%20OM-D%20E-M5%20Mark%20III
If you just want to see the improvements in the shadows that you can gain back, photons to photons has that as well. For the EM5mk3 it looks like this:
https://www.photonstophotos.net/Charts/PDR_Shadow.htm#Olympus%20OM-D%20E-M1%20Mark%20III
From that you can see that iso bracketing will bring you up-to half a stop of improvement in noise performance in the shadows. Play around with the tool, you will see that for truly ISO-invariant sensors, like the X-T1, there will be no improvement.
Using the full DR of the sensor is recommended, bracketing might improve things, but that depends on the scene-DR and how you allocate your sensors DR with respect to its gain stages. That means that iso-bracketing with sony derived sensors requires one to know at which ISOs the switch(es) happen and then you only need to switch between those two or three ISOs to see any improvement (which then might be smallish). On old canon sensors, ISO-bracketing was really beneficial.
Case in point (40D, X-T1, EM5mk3) on the same “improvement”-chart
Yup. As an example for my past two cameras:
https://www.photonstophotos.net/Charts/PDR_Shadow.htm#Sony%20ILCE-7M3,Sony%20ILCE-7M4
The only real benefit to ISO bracketing is if you shot one image at 100 and one image at the (very obvious) dualgain cutin point. Note that according to Jim Kasson over on DPReview, sometimes shooting at the lowest possible ISO can lead to black point compensation challenges, so he frequently shoots 3-4 stops down, but does not go all the way down to the floor of the dualgain cutin point.
ADC quantization hasn’t really been an issue for any recent camera. The last time I saw ADC quantization being theorized as relevant was the A7R3 vs A7R2, due to some ADC improvements at low ISO. Although Shadow Improvement of Photographic Dynamic Range versus ISO Setting
is inconsistent with that, and with Photographic Dynamic Range versus ISO Setting