Why most cameras as default, forcing you to press a button for changing ISO via dial?

mahidoes

Veteran Member
Messages
3,897
Solutions
2
Reaction score
2,940
Location
Jaffna, LK
Most people and even camera manufactures might jump and answer it like below
So you won't accidently change it

f5277c6421bd4dd78ccec6a90f1b251b.jpg


But you could change shutter speed as well as aperture also accidently which will mess the exposure in same way. Why camera manufactures allow to change them directly but not ISO.

I shoot 98% of the time manual mode now because I use EVF not OVF. I decide what shutter and What aperture i want for the session. Because other than exposure these two can mostly affect the feel of the image as they also adjust DOF and control motion. But ISO only affect the noise along with light level.

So when light go down or up I want to adjust ISO first. But by default most camera ask you to press a button first. Even if you could assign it to a dial still I'll have an unused button most of the time i can't assign to some thing else.
I really like the ISO dial in Fuji.

Fuji ISO dial

Fuji ISO dial

But even Fuji for low end camera like X-T20 I use they require you to press a button before hand. The newer X-S10 they have a dedicated ISO button. Doh.

I think the idea of changing ISO less often come from Film Era in my opinion. You put a film then you only have Shutter and Aperture to play with.

Now considering we have EVF (so manual shooting coming very often even with beginners) and we no longer insert films in digital cameras. I think changing ISO should be given same priority as Shutter and Aperture.
I would even love a mode dial position for Auto ISO. Because that would be the best mode to shoot with. But sadly in all cameras that is burred under menus.


--
If you like my pictures follow me on Insta
https://www.instagram.com/mahidoes/
 
Last edited:
What about the D7200? More or less flat as well, but lower - is lower 'better' (and higher 'worse'?
Yes but you need to adjust for the number of pixels. The D7200 has ~24 MP while the D7000 has 16 MP. Therefore, for each pixel of the D7000, the read noise from 1.5 pixels of the D7200 is added in quadrature, so to compare them, you would multiply the numbers from the D7200 by sqrt(1.5), i.e. approximately 1.22.

This is what that page means by:
These raw values are not appropriate for comparing camera models because they are not adjusted for area.
The adjustment is not too difficult to do, but you still need to do it.
 
Last edited:
What about the D7200? More or less flat as well, but lower - is lower 'better' (and higher 'worse'?
Yes but you need to adjust for the number of pixels. The D7200 has ~24 MP while the D7000 has 16 MP. Therefore, for each pixel of the D7000, the read noise from 1.5 pixels of the D7200 are added in quadrature, so to compare them, you would multiply the numbers from the D7200 by sqrt(1.5), i.e. approximately 1.22.

This is what that page means by:
These raw values are not appropriate for comparing camera models because they are not adjusted for area.
The adjustment is not too difficult to do, but you still need to do it.
Thank you!
 
PS: how/where could I find out which of the current generation of camera sensors can be seen/treated as ISO-less?
You can look at the input-referred read noise page from photonstophotos.net and see at which setting the chart for a given camera starts to be more or less flat. The camera is then considered to be “ISO-less” / “ISO-invariant” beyond that setting.
Thank you!

However, when I look up the D7000 that bobn2 referred to back in 2010 as ISO-less (see also the post just above mine), I don't see the flatness you mention below ISO 1000 ?
I would say that being “ISO-less” is not really a binary thing, there is some tolerance involved. Here in particular, note that the curve expands to use the available space, and that the actual range is just from 3.1 from 2.4 e⁻, which is not that much.

Compare to a 60D from the same year for example: https://www.photonstophotos.net/Charts/RN_e.htm#Canon EOS 60D_14,Nikon D7000_14
The thread I linked was right at the inception of ISOlessness, and we've learned a lot more since then, as have the camera manufacturers. There hasn't been a subsequent camera as ISOIess as the D7000, and mostly the read change voltage gain tweaks have become really really rather subtle (another reason what 'ISO' isn't 'gain') and dual conversion gain has been added.
https://www.photonstophotos.net/Charts/RN_e.htm#Nikon%20D7000_14,Nikon%20D7200_14,Nikon%20Z%205_14,Olympus%20OM-D%20E-M1%20Mark%20II_12,Panasonic%20Lumix%20DC-G9_12


What about the D7200? More or less flat as well, but lower - is lower 'better' (and higher 'worse'?
As some have noted, you need to take the pixel count into account to compare, and even then it isn't simple. Whilst a hight pixel count camera might deliver a bot more read noise pro-rata, it also provides better data for NR to work, so you can apply a bit more NR before it becomes noticeable.

In any case, you have to remember that read noise at these levels is only visible in the very deepest shadows. Until you get to quite high ISO settings the black level is likely above the read noise level, which renders any difference moot.

--
Is it always wrong
for one to have the hots for
Comrade Kim Yo Jong?
 
PS: how/where could I find out which of the current generation of camera sensors can be seen/treated as ISO-less?
You can look at the input-referred read noise page from photonstophotos.net and see at which setting the chart for a given camera starts to be more or less flat. The camera is then considered to be “ISO-less” / “ISO-invariant” beyond that setting.
Thank you!

However, when I look up the D7000 that bobn2 referred to back in 2010 as ISO-less (see also the post just above mine), I don't see the flatness you mention below ISO 1000 ?
I would say that being “ISO-less” is not really a binary thing, there is some tolerance involved. Here in particular, note that the curve expands to use the available space, and that the actual range is just from 3.1 from 2.4 e⁻, which is not that much.

Compare to a 60D from the same year for example: https://www.photonstophotos.net/Charts/RN_e.htm#Canon EOS 60D_14,Nikon D7000_14
The thread I linked was right at the inception of ISOlessness, and we've learned a lot more since then, as have the camera manufacturers. There hasn't been a subsequent camera as ISOIess as the D7000, and mostly the read change voltage gain tweaks have become really really rather subtle (another reason what 'ISO' isn't 'gain') and dual conversion gain has been added.
https://www.photonstophotos.net/Charts/RN_e.htm#Nikon%20D7000_14,Nikon%20D7200_14,Nikon%20Z%205_14,Olympus%20OM-D%20E-M1%20Mark%20II_12,Panasonic%20Lumix%20DC-G9_12


What about the D7200? More or less flat as well, but lower - is lower 'better' (and higher 'worse'?
As some have noted, you need to take the pixel count into account to compare, and even then it isn't simple. Whilst a hight pixel count camera might deliver a bot more read noise pro-rata, it also provides better data for NR to work, so you can apply a bit more NR before it becomes noticeable.

In any case, you have to remember that read noise at these levels is only visible in the very deepest shadows. Until you get to quite high ISO settings the black level is likely above the read noise level, which renders any difference moot.
Thank you for this additional note - much appreciated
 

Keyboard shortcuts

Back
Top