Reading mode: Light Dark

New Panasonic bodies and ISO in general

Go to last post
J

jrq23

Well-known member
Messages
102
Reaction score
56
Location
US
So, some of the specs for the new Panasonic GF7 body have been leaked, which show a low ISO limit of 200. Now I've owned a GF1 and GX1. The GF1 had an ISO limit of 100, and the GX1 has an ISO limit of 160. I'll swear that the GX1 looks fractionally softer than the GF1. I have a professional photographer friend, who's high-end backs are at ISO 50, and look super-super-sharp.

So what's the deal with ISO in digital cameras. Is it meaningful? Are these newer micro 4/3 bodies doing a disservice by bumping up the lower limit from 100 to 160 to 200?
 
Last edited by a moderator:
i dont think so.. its more important at the higher end.. i doubt i would go lower than 200 even if i had the option.. the sensors are good enough at iso 200 not to need to go lower if the desire is only image quality improvement..

trog
 
trog100 said:
i dont think so.. its more important at the higher end.. i doubt i would go lower than 200 even if i had the option.. the sensors are good enough at iso 200 not to need to go lower if the desire is only image quality improvement..

trog
Click to expand...
So wrong.

Lower ISO allows you to achieve shallow DOF in sunlight and bright conditions, without having to use an ND filter. Also useful for video.

I think the manufacturers are more interested in the whole system package, which for most people will include a kit zoom, so low ISO isn't particularly useful there.
 
hindesite said:
trog100 said:
i dont think so.. its more important at the higher end.. i doubt i would go lower than 200 even if i had the option.. the sensors are good enough at iso 200 not to need to go lower if the desire is only image quality improvement..

trog
Click to expand...
So wrong.

Lower ISO allows you to achieve shallow DOF in sunlight and bright conditions, without having to use an ND filter. Also useful for video.

I think the manufacturers are more interested in the whole system package, which for most people will include a kit zoom, so low ISO isn't particularly useful there.
Click to expand...
 
yes i wasnt thinking of it from DOF point of view.. just the sensor noise point of view..

trog
 
jrq23 said:
So, some of the specs for the new Panasonic GF7 body have been leaked, which show a low ISO limit of 200. Now I've owned a GF1 and GX1. The GF1 had an ISO limit of 100, and the GX1 has an ISO limit of 160. I'll swear that the GX1 looks fractionally softer than the GF1. I have a professional photographer friend, who's high-end backs are at ISO 50, and look super-super-sharp.

So what's the deal with ISO in digital cameras. Is it meaningful? Are these newer micro 4/3 bodies doing a disservice by bumping up the lower limit from 100 to 160 to 200?
Click to expand...
If you shoot RAW, the real base ISO limit is the so-called "measured ISO" reported by DxOMark, which tells you how much light the sensor can take before it clips. Indexed that way, the GF1 has a base ISO of 143, the GX1 146, and the GF7 probably something very close to that value too (like the GX7, GM1, and GM5).

Recent Panys all allow you to use so-called "extended ISOs" below 200 but the DxOMark "measured ISOs" for those are questionable since DxO seems to have missed that at these ISOs the sensor effectively clips before it reaches the max ADU-value obtainable at normal ISOs.
 
But it becomes less of an issue with the faster shutter-speeds capable in the newer bodies. If certainly had problems with my old E-PL1 shooting wide open with it's limit of 1/2000 - not so much of an issue with a 1/8000 shutter.

Cheers,
Paul
 
The base ISO doesn't really mean anything, it's generally just whatever the minimum sensor gain is with a tone curve applied that makes it look like the ISO given.

You could make a tone curve that would make the minimum sensor gain look like ISO 100, or a tone curve that would make it look like ISO 200. However if you make it 200 then your camera will meter a stop less, and so have a stop more highlight range.

Given how harshly m4/3 sensors clip it's probably more important to not clip highlights then to have a stop less shadow noise.

Your friend's high-end back looks sharp because it's a high end medium format back, not because it goes down to ISO 50. You can shoot RAW with m4/3 at ISO 200 and +2 EV then reduce the exposure by -2 EV in post and essentially shoot at ISO 50, but it'll never look like a medium format back.
 
Andy Crowe said:
The base ISO doesn't really mean anything, it's generally just whatever the minimum sensor gain is with a tone curve applied that makes it look like the ISO given.

You could make a tone curve that would make the minimum sensor gain look like ISO 100, or a tone curve that would make it look like ISO 200. However if you make it 200 then your camera will meter a stop less, and so have a stop more highlight range.

Given how harshly m4/3 sensors clip it's probably more important to not clip highlights then to have a stop less shadow noise.

Your friend's high-end back looks sharp because it's a high end medium format back, not because it goes down to ISO 50. You can shoot RAW with m4/3 at ISO 200 and +2 EV then reduce the exposure by -2 EV in post and essentially shoot at ISO 50, but it'll never look like a medium format back.
Click to expand...
Andy , you are talking about extended ISO settings not a genuine low base ISO. A true base ISO of 50 or even 25 { I shoot a lot of landscape :-) and in the good old days worked with Kodachrome 25 , Ektar 25 and my favorite B&W film Agfapan 25 so the low ISO setting would not be a problem for me. The BIF and sport guys might not be so keen :-)
 
amipal said:
But it becomes less of an issue with the faster shutter-speeds capable in the newer bodies. If certainly had problems with my old E-PL1 shooting wide open with it's limit of 1/2000 - not so much of an issue with a 1/8000 shutter.
Click to expand...
True to a point, but most cameras have 1/4000 shutter and it is a problem then (though of course it depends where you live and what you shoot - it is pretty bright outdoors down here).

There is still the issue of video, where you may want to avoid high shutter speeds.
 
jim stirling said:
Andy , you are talking about extended ISO settings not a genuine low base ISO. A true base ISO of 50 or even 25 { I shoot a lot of landscape :-) and in the good old days worked with Kodachrome 25 , Ektar 25 and my favorite B&W film Agfapan 25 so the low ISO setting would not be a problem for me. The BIF and sport guys might not be so keen :-)
Click to expand...
You can't arbitrarily make a genuine low base ISO sensor, it depends on how much light the sensor can physically handle before the cells fill (and thus clip) then basing the base ISO value on the best combination of shadow/highlight dynamic range.

As smaller cells fill faster than large cells it makes sense that high resolution sensors naturally have a faster base ISO, and as noise improves you'd want to make the base ISO even faster to preserve more highlights without increasing shadow noise.
 
Andy Crowe said:
Andy Crowe said:
Andy , you are talking about extended ISO settings not a genuine low base ISO. A true base ISO of 50 or even 25 { I shoot a lot of landscape :-) and in the good old days worked with Kodachrome 25 , Ektar 25 and my favorite B&W film Agfapan 25 so the low ISO setting would not be a problem for me. The BIF and sport guys might not be so keen :-)
Click to expand...
You can't arbitrarily make a genuine low base ISO sensor, it depends on how much light the sensor can physically handle before the cells fill (and thus clip)
Click to expand...
Right. Although in the future (hopefully not a very distant one), sensors will be able to handle arbitrarily low ISOs using technologies like this:

Andy Crowe said:
then basing the base ISO value on the best combination of shadow/highlight dynamic range.
Click to expand...
In reality, there are no separate DRs for shadows and highlights. There is just dynamic range, period. As long as there are no downsides in terms of sensor efficiency, the lower the base ISO, the better it is, since a low base ISO makes it possible to give the sensor more light, thereby improving the signal-noise ratio whenever the base ISO can be used.
Andy Crowe said:
As smaller cells fill faster than large cells it makes sense that high resolution sensors naturally have a faster base ISO,
Click to expand...
Not really. A high-resolution sensor with relatively small pixel may have less storage capacity. But it will also accumulate less light for any given exposure due to the smaller pixel area.
Andy Crowe said:
and as noise improves you'd want to make the base ISO even faster to preserve more highlights without increasing shadow noise.
Click to expand...
Not really. As I pointed out above, a low base ISO is just a good thing as long as it does not reduce sensor efficiency (via factors like quantum efficiency, read noise, and PRNU).
 
Anders W said:
Anders W said:
then basing the base ISO value on the best combination of shadow/highlight dynamic range.
Click to expand...
In reality, there are no separate DRs for shadows and highlights. There is just dynamic range, period. As long as there are no downsides in terms of sensor efficiency, the lower the base ISO, the better it is, since a low base ISO makes it possible to give the sensor more light, thereby improving the signal-noise ratio whenever the base ISO can be used.
Click to expand...
There is in terms of the exposure settings the camera applies though; If the base ISO was 50 then all exposures would be +2 ev brighter at any given sensor gain, and so highlights would lose 2 stops of DR and shadows would gain 2 stops of DR.

Given that highlights clip harshly while shadows do not, it makes more sense to shift the exposure to give more DR at the highlight end.
Anders W said:
Anders W said:
As smaller cells fill faster than large cells it makes sense that high resolution sensors naturally have a faster base ISO,
Click to expand...
Not really. A high-resolution sensor with relatively small pixel may have less storage capacity. But it will also accumulate less light for any given exposure due to the smaller pixel area.
Click to expand...
The storage area doesn't increase/decrease linearly though because of the extra circuitry around the pixels takes a fixed space (something BSI could potentially change though)
Anders W said:
Anders W said:
and as noise improves you'd want to make the base ISO even faster to preserve more highlights without increasing shadow noise.
Click to expand...
Not really. As I pointed out above, a low base ISO is just a good thing as long as it does not reduce sensor efficiency (via factors like quantum efficiency, read noise, and PRNU).
Click to expand...
No it's not, as above every stop lower base ISO you use you'll get 1 stop less highlight range because you're forcing the camera to increase the exposure time by 1 stop.

If you're happy with 1 stop less highlights of course then you can expose at +1 ev and get the same result (ETTR).
 
New sensors have about 2 EV better dynamic range. That allows you to overexpose your image at max. shutter speed and take it back in post process. Your ISO100 is here. The extended low ISO is the same for jpeg shooters.

Anyway, regarding to GF1 and GX1... Some denies it, but many reported and I also experienced that GX1 suffers from serious shutter shock with many lenses. That's why its pictures are less sharp than the GF1's (GF2 in my case). It has nothing to do with base ISO.

trog100 said:
yes i wasnt thinking of it from DOF point of view.. just the sensor noise point of view..

trog
Click to expand...
 
Andy Crowe said:
Andy Crowe said:
Andy Crowe said:
then basing the base ISO value on the best combination of shadow/highlight dynamic range.
Click to expand...
In reality, there are no separate DRs for shadows and highlights. There is just dynamic range, period. As long as there are no downsides in terms of sensor efficiency, the lower the base ISO, the better it is, since a low base ISO makes it possible to give the sensor more light, thereby improving the signal-noise ratio whenever the base ISO can be used.
Click to expand...
There is in terms of the exposure settings the camera applies though; If the base ISO was 50 then all exposures would be +2 ev brighter at any given sensor gain, and so highlights would lose 2 stops of DR and shadows would gain 2 stops of DR.

Given that highlights clip harshly while shadows do not, it makes more sense to shift the exposure to give more DR at the highlight end.
Click to expand...
We are talking past each other here. I am talking about lowering the ISO in the sense of increasing sensor storage capacity or other technologies aimed at increasing the amount of light the sensor can take before it clips. You are talking about how to set metering and nominal camera ISO relative to the clipping point of the sensor.

If a low nominal camera ISO is achieved simply by telling the meter to suggest more exposure while keeping the clipping point of the sensor unchanged, then you will certainly gain performance in the shadows (since they'll get more exposure) and lose in the highlights (due to increased risk of clipping), if you simply expose expose according to the meter without taking the actual clipping point of the sensor into account. But I would refer to this as changing the headroom rather than changing the dynamic range. Dynamic range is defined as S/N (often expressed as the base-2 logarithm of S/N) where S is the max signal and N one or other definition of the "noise floor", and this quantity remains the same regardless of the nominal camera ISO and the way the camera meters.

Furthermore, many RAW shooters (including myself), take the relationship between nominal camera ISO and the clipping point of the sensor into account when we set exposure. I have done so with all my digital ILCs. With the Oly E-M1 and E-M5 I allow more exposure relative to the meter reading than I did with my Pentax K100D and Pany G1 because the the former have the nominal camera ISO set higher relative to the clipping point of the sensor than the latter.
Andy Crowe said:
Andy Crowe said:
Andy Crowe said:
As smaller cells fill faster than large cells it makes sense that high resolution sensors naturally have a faster base ISO,
Click to expand...
Not really. A high-resolution sensor with relatively small pixel may have less storage capacity. But it will also accumulate less light for any given exposure due to the smaller pixel area.
Click to expand...
The storage area doesn't increase/decrease linearly though because of the extra circuitry around the pixels takes a fixed space (something BSI could potentially change though)
Click to expand...
Not really. If that were the case, smaller sensors would have lower quantum efficiency than larger, BSI aside, and that's not the case. The extra circuitry shrinks with the pixel size.
Andy Crowe said:
Andy Crowe said:
Andy Crowe said:
and as noise improves you'd want to make the base ISO even faster to preserve more highlights without increasing shadow noise.
Click to expand...
Not really. As I pointed out above, a low base ISO is just a good thing as long as it does not reduce sensor efficiency (via factors like quantum efficiency, read noise, and PRNU).
Click to expand...
No it's not, as above every stop lower base ISO you use you'll get 1 stop less highlight range because you're forcing the camera to increase the exposure time by 1 stop.

If you're happy with 1 stop less highlights of course then you can expose at +1 ev and get the same result (ETTR).
Click to expand...
See above. We are talking past each other here in the same way.
 
It's a case of:

(i) How efficiently you convert Photons to Electrons.

(ii) How deep your pixels are (i.e. how many electrons they capture compared to surface area - although "surface area" is a complicated thing, relating to angle of incidence and other crap).

(iii) How much highlight recovery space you want (i.e. how much your target exposure for pure white is below the pixel capacity).

(iv) Marketing.

Basically if you can convert a lot of Photons to Electrons (which is gradually increasing over time and running from 40-60% at present) then your base ISO is down to how many you can capture in a pixel (so if you can't make very deep Pixels you might overflow at exposures of ISO 100 and 160 and end up at ISO 200 base).
 
Last edited:
Andy Crowe said:
Andy Crowe said:
Andy , you are talking about extended ISO settings not a genuine low base ISO. A true base ISO of 50 or even 25 { I shoot a lot of landscape :-) and in the good old days worked with Kodachrome 25 , Ektar 25 and my favorite B&W film Agfapan 25 so the low ISO setting would not be a problem for me. The BIF and sport guys might not be so keen :-)
Click to expand...
You can't arbitrarily make a genuine low base ISO sensor, it depends on how much light the sensor can physically handle before the cells fill (and thus clip) then basing the base ISO value on the best combination of shadow/highlight dynamic range.

As smaller cells fill faster than large cells it makes sense that high resolution sensors naturally have a faster base ISO, and as noise improves you'd want to make the base ISO even faster to preserve more highlights without increasing shadow noise.
Click to expand...
I am talking about in the future as sensor tech invariably evolves { hopefully :-) } . Though it is not necessarily the case I would be happy to sacrifice all high ISO's above 1600 or even 800 for better base ISO performance.
 
You must log in or register to reply here.

Keyboard shortcuts

f
Forum
m
My threads
Mobile site
About
Editorial content
Cameras & Lenses
Community
All content, design, and layout are Copyright © 1998–2025 Digital Photography Review All Rights Reserved.
Reproduction in whole or part in any form or medium without specific written permission is prohibited.
When you use DPReview links to buy products, the site may earn a commission.
©GPS Media - Guides, Products, Services.
Back
Top