Strange aperture behavior

[BernardC]

Advertisement (gone after registration)

2 hours ago, wizard said:

if there is tons of light, and you have set the camera to boost the sensor signal (by setting ISO at 800), resulting in a considerably higher signal value than if you had used base ISO, then why does the camera stop down the aperture when the actual signal value is not boosted 

That's not how current sensors work, although some sensors may have worked this way in the past.

Changing the ISO within that range doesn't boost anything, it just tells the raw converter where to put middle grey. Here's an Arri image that shows it visually:

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

The whole page is very interesting: https://www.arri.com/en/learn-help/technology/image-processing

The sensor readout is the same at all of these ISO values, it's just the location of middle grey that changes. That's why people complain of "blown highlights" at low ISO, as well as a great ability to "lift the shadows." In the example above, there are 5.6 stops of highlight range at ISO 200, and 8.9 stops of shadow range. Leica's numbers will be slightly different, because they use different sensors, but the principle is the same.

I don't know if this will help or not, but try to visualize the same graph with the bars aligned on their top (instead of being aligned on middle grey). The six bars would be the same height, and the middle grey line would slope down.

 

Note: many sensors now have "dual native ISO". This is implemented by using two different A/D paths. It's not a boost per-se, just a different voltage range being sampled. Think of it like "line vs. mic" for audio inputs. This isn't relevant here, because the camera will always choose the less-sensitive path if you tell it that the scene is brightly lit. I'm not even sure that dual-ISO is used for EVF display, although I guess you could test that.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[wizard]

vor 14 Stunden schrieb BernardC:

Changing the ISO within that range doesn't boost anything, it just tells the raw converter where to put middle grey. Here's an Arri image that shows it visually:

Thanks so much, that does help indeed. I did not know that it is just the position of middle grey that will be changed when changing ISO values. 

[SrMi]

5 hours ago, wizard said:

Thanks so much, that does help indeed. I did not know that it is just the position of middle grey that will be changed when changing ISO values. 

Bernard's example is valid for Arri cameras, not for still cameras like Leica. With still cameras, you do not have more highlight dynamic range when you increase ISO (easy to test by yourself). The dynamic range of still cameras decreases as you raise ISO.

 

[wizard]

vor 3 Minuten schrieb SrMi:

Bernard's example is valid for Arri cameras, not for still cameras like Leica. With still cameras, you do not have more highlight dynamic range when you increase ISO (easy to test by yourself). The dynamic range of still cameras decreases as you raise ISO.

Well then, do still cameras amplify, as I had assumed, the sensor's output (instead of shifting the middle grey point) when ISO is increased?

[SrMi]

6 minutes ago, wizard said:

Well then, do still cameras amplify, as I had assumed, the sensor's output (instead of shifting the middle grey point) when ISO is increased?

Yes, they do amplify. 

Here is a correct diagram:

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

The article discussing ISO is here:

...photoelectrons (e-) collected by the photosites are put through various stages of linear amplification, then converted to digital numbers by an Analog to Digital Converter and finally stored in a file as raw values (referred to as DN or ADU, see figure 1 above).

P.S.: Sometimes, at high ISOs, digital scaling is used instead of amplification.

Edited December 17, 2021 by SrMi

[SrMi]

This article discusses various ISO implementations. Bernard's post describes EI (exposure index) in ARRI. It is discussed at the end of the article.

Ins and outs of ISO: where ISO gets complex

 

[BernardC]

Advertisement (gone after registration)

I clicked on the link, and it took me directly to this: "ISO is not just analog amplification, and doesn't dictate what happens at the Raw level", and "Before 2006 most cameras did increase lightening by increasing analog amplification with each ISO step". Isn't that what I wrote?

It goes on, about post-2006 cameras: "If you ignore the exposure implications and give ISO 100 and ISO 200 modes the same exposure, you end up with identical values in the Raw files."

I like the Arri article better, because it's written for cinematographers and it was proof-read by Arri's sensor engineers. In the end, the two articles are just different ways of explaining the same concepts.

[SrMi]

1 hour ago, BernardC said:

I clicked on the link, and it took me directly to this: "ISO is not just analog amplification, and doesn't dictate what happens at the Raw level", and "Before 2006 most cameras did increase lightening by increasing analog amplification with each ISO step". Isn't that what I wrote?

Further in the article:

This provides plenty of freedom for manufacturers to use different combinations of analog amplification and tone curve, so long as a given exposure results in the expected lightness.

Also:

For now, we’ve not seen a stills/video camera explicitly use an EI approach to exposure.

1 hour ago, BernardC said:

It goes on, about post-2006 cameras: "If you ignore the exposure implications and give ISO 100 and ISO 200 modes the same exposure, you end up with identical values in the Raw files."

That is correct because, for raws, the exposure is more relevant than the ISO setting, especially if the difference in ISO is a few stops. The image lightness will differ, though.

The main point that I am trying to convey is that still cameras include some signal amplification when increasing ISO. Arri and other recent video cameras operate at base amplification (ISO 800 for Arri). With Arri, as shown by their graph, increasing ISO increases highlight headroom. That is not the case with still cameras.
You can simulate Arri's behavior with still cameras by keeping your ISO fixed and changing exposure. In that case, the lightness must be adjusted in the post, and you get the same effect on highlight headroom when increasing exposure (decreasing EI) or decreasing exposure (increasing EI).

[BernardC]

15 hours ago, SrMi said:

The main point that I am trying to convey is that still cameras include some signal amplification when increasing ISO. Arri and other recent video cameras operate at base amplification

Sure, but that's not really what the article says. It says that adding "analogue amplification" is a deprecated strategy, but some cameras might still use a tiny amount of it, even though it reduces dynamic range.

Frankly, I'm not impressed by the article. It has all the hallmarks of "a little bit of knowledge is a bad thing." It says one thing and provides an example of the opposite, it makes a claim and provides no example, it starts talking about one thing, drops it, comes back to it, drops it again, and mentions it in passing later. In contrast, the Arri article was proof-read by people who fully understand the topic, and it shows.

One example: the article says that the GFX "stops adding amplification at higher ISO," but provides no evidence that it adds any "amplification" at any ISO. What is the article trying to say? Is the GFX an example of how modern cameras don't provide "analogue amplification?" Because that's what the examples show, and what the other reference to the GFX says, but not what the caption says! The caption implies that it does amplify, but it doesn't say when this might happens, only when it definitely doesn't happen, which is incidentally when you would expect it to happen (at higher ISO). If a student handed-in this paper, I would fail it.

[Photojoejoe21]

On 12/14/2021 at 8:26 AM, BernardC said:

That's a ISO/speed/aperture combination corresponds to a scene 16x brighter than direct mid-day summer sunlight (1/ISO @ f:16). Do you see the issue at exposure combinations that are closer to what one would expect?

Just to reiterate what's going on here, the settings I used anyone is likely to encounter. Basically somewhere close to :1/60/-1/125, ISO 100-200, f/16 shooting in bright sun.

I should also note that I have returned my Q2 due to this erratic aperture behavior for an Fujifilm X-Pro 3, with which I am not able to replicate the same aperture behavior. Neither am I able to replicate the aperture abnormality I experienced on my Q2 with either my Sony A9 or my Sony A1. 

Super bummed because I really wanted to make the Q2 work. I still love that camera. But, despite the ways the Q2 is superior to the X-Pro 3, the Fuji does not suffer from that aperture behavior and for me, that reliability makes it worth the tradeoffs, as many of them as there are.

Should also mention that it's been over two weeks since I first started contacting Leica customer service about this issue and still have not heard back from them either via phone or e-mail.

[SrMi]

On 12/19/2021 at 11:48 AM, Photojoejoe21 said:

Just to reiterate what's going on here, the settings I used anyone is likely to encounter. Basically somewhere close to :1/60/-1/125, ISO 100-200, f/16 shooting in bright sun.

I should also note that I have returned my Q2 due to this erratic aperture behavior for an Fujifilm X-Pro 3, with which I am not able to replicate the same aperture behavior. Neither am I able to replicate the aperture abnormality I experienced on my Q2 with either my Sony A9 or my Sony A1. 

Super bummed because I really wanted to make the Q2 work. I still love that camera. But, despite the ways the Q2 is superior to the X-Pro 3, the Fuji does not suffer from that aperture behavior and for me, that reliability makes it worth the tradeoffs, as many of them as there are.

Should also mention that it's been over two weeks since I first started contacting Leica customer service about this issue and still have not heard back from them either via phone or e-mail.

Sorry to hear that Q2 did not work out for you.

We got a bit of sun now, so I tried out your settings but I could not reproduce it (automatic exposure, 1/60/-1/125, ISO 100-200, f/16). 

Good luck!

[NickT]

Have you tried "playing" with "Enhanced live view" which was introduced with firmware 3? It is found on menu page 3 "Live view settings.

[SrMi]

4 hours ago, NickT said:

Have you tried "playing" with "Enhanced live view" which was introduced with firmware 3? It is found on menu page 3 "Live view settings.

Good suggestion. With "Enhanced Live View" on, the aperture stays wide open. So there is a solution for SL2 cameras if such an issue occurs in real life. Unfortunately, there is no such option in Q2.

[beewee]

Enhanced Live View will only kick in (automagically) in dark scenes. Under normal scenes, even when enabled, you’ll get the normal EVF mode. One drawback of Enhanced Live View is that focus peaking is disabled when it kicks in.