ISO 100, OMD-EM1II and firmware 3.0?

Wolfgang, I only ever felt the need to use the extended LOW ISO with my E-30, not with either of my E-M1s (MkI or MkII).

With the E-30, which was a real noise machine given half a chance, using ISO 100 certainly lessened the noise. Never noticed any other changes for good or ill.

The main advantage is that you are getting a longer exposure on your sensor because the (nominal) ISO is lower. If you catch twice as many photons this way, the noise will be lower.

I suspect that using extended LOW settings with any camera that offers it/them is going to help with noise, but little else.

With the E-M1 MkII, using the lower extended settings keeps DR much the same, and that's a considerable improvement on previous sensors, most of which caused the DR to drop when using the extended setting (e.g. my E-M1 MkI). See Bill Claff's site for comparisons of whatever cameras you choose. It is here:

Architeuthis said:

I have installed firmware 3.0 on my OMD-EM1II.

I wonder, whether the new ISO 100 brings better dynamic range, less noise and better colors compared to the base ISO 200, when used for raw files?

Click to expand...

As John suggests in his reply, the advantage of a lower ISO is that you get (by definition) a larger exposure, which means a larger signal to noise ratio. That should give you a better actual dynamic range, simply because the largest 'signal' you're actually capturing is larger by a factor of two, This doesn't show up in 'dynamic range' charts, since they are based on the maximum signal you could capture, which you won't get if you expose according to the ISO setting, which is what most people do. Bill Claff's charts, to which John referred, have a peculiarity in that the ISO range is nominal ISO (i.e. the number your camera will tell you, rather than the ISO given by the exposure used) so don't really tell you much about what you'll get if you let exposure follow the meter. As for raw files, raw files are no different from JPEGs, apart from there is likely more NR and other signal processing going in for raw files, except that for raw files, you have control over processing, so you don't need to follow the meter (for instance, you could 'expose to the right'). In that respect, the new ISO 100 has no advantage over the old 200.

Architeuthis said:

Already in earlier firmware versions ISO 64 was available, but as far as I understand it, this does not improve DR, noise or color, but is useful only when too much light is present, to maintain wide aperture and increase exposure time...

Click to expand...

I think your understanding a a bit awry. Nothing is different about the old ISO 64 except that the meter is calibrated for a larger exposure, and the in-camera processing is set up to render that larger exposure 'correctly'. This you get the advantage of larger exposure, as John suggested, which means less noise, more actual DR, if you expose according to the meter, and hence better colours. The risk is that you have less highlight headroom, and 64 encroaches into the JPEG standard headroom, so you can clip bright highlights. No problem if the scene is flat. Whether or not the 'better' you get is noticeable in general use is a completely different matter.

--
263, look deader.

bobn2 said:

Architeuthis said:

I have installed firmware 3.0 on my OMD-EM1II.

I wonder, whether the new ISO 100 brings better dynamic range, less noise and better colors compared to the base ISO 200, when used for raw files?

Click to expand...

As John suggests in his reply, the advantage of a lower ISO is that you get (by definition) a larger exposure, which means a larger signal to noise ratio. That should give you a better actual dynamic range, simply because the largest 'signal' you're actually capturing is larger by a factor of two, This doesn't show up in 'dynamic range' charts, since they are based on the maximum signal you could capture, which you won't get if you expose according to the ISO setting, which is what most people do. Bill Claff's charts, to which John referred, have a peculiarity in that the ISO range is nominal ISO (i.e. the number your camera will tell you, rather than the ISO given by the exposure used) so don't really tell you much about what you'll get if you let exposure follow the meter.

Click to expand...

I find it almost amusing seeing people posting link to Bill's site while talking about something completely different than what this site shows.

bobn2 said:

As for raw files, raw files are no different from JPEGs, apart from there is likely more NR and other signal processing going in for raw files, except that for raw files, you have control over processing, so you don't need to follow the meter (for instance, you could 'expose to the right'). In that respect, the new ISO 100 has no advantage over the old 200.

bobn2 said:

Already in earlier firmware versions ISO 64 was available, but as far as I understand it, this does not improve DR, noise or color, but is useful only when too much light is present, to maintain wide aperture and increase exposure time...

Click to expand...

I think your understanding a a bit awry. Nothing is different about the old ISO 64 except that the meter is calibrated for a larger exposure, and the in-camera processing is set up to render that larger exposure 'correctly'. This you get the advantage of larger exposure, as John suggested, which means less noise, more actual DR, if you expose according to the meter, and hence better colours. The risk is that you have less highlight headroom, and 64 encroaches into the JPEG standard headroom, so you can clip bright highlights. No problem if the scene is flat. Whether or not the 'better' you get is noticeable in general use is a completely different matter.

Click to expand...

I think what he means is that native ISO is not the same as the extended ISO. In other words added through software update option will only allow for slower shutter speeds, but not necessarily better rendered images. If I read it correctly.

JaKing said:

Wolfgang, I only ever felt the need to use the extended LOW ISO with my E-30, not with either of my E-M1s (MkI or MkII).

With the E-30, which was a real noise machine given half a chance, using ISO 100 certainly lessened the noise. Never noticed any other changes for good or ill.

The main advantage is that you are getting a longer exposure on your sensor because the (nominal) ISO is lower. If you catch twice as many photons this way, the noise will be lower.

I suspect that using extended LOW settings with any camera that offers it/them is going to help with noise, but little else.

With the E-M1 MkII, using the lower extended settings keeps DR much the same, and that's a considerable improvement on previous sensors, most of which caused the DR to drop when using the extended setting (e.g. my E-M1 MkI). See Bill Claff's site for comparisons of whatever cameras you choose. It is here:

http://www.photonstophotos.net/Charts/PDR.htm#Olympus OM-D E-M1,Olympus OM-D E-M1 Mark II

Click to expand...

This is very interesting, JaKing.

One must say, however, that the nominal ISOs that the camera manufacturers provide seem to be lies: DxO states that on the EM1II both nominal ISO64 as well as ISO200 correspond to real live ISO of 83. For EM1 nominal ISO100 and ISO200, the real live ISO is 122 in both cases: https://www.dxomark.com/Cameras/Com...1-versus-Olympus-OM-D-E-M1-Mark-II___909_1136

When EM1 and EM1II real ISO vs. DR is compared, the performance of both cameras is pretty the same...

So what is the difference between extended ISO (I assume also the "new" ISO100 is extended ISO) and base ISO (both mean ISO83 (EM1II) or ISO122 (EM1), respectively)? Does the "new" ISO100 mean just another ISO83 setting?

Wolfgang

Architeuthis said:

JaKing said:

Wolfgang, I only ever felt the need to use the extended LOW ISO with my E-30, not with either of my E-M1s (MkI or MkII).

With the E-30, which was a real noise machine given half a chance, using ISO 100 certainly lessened the noise. Never noticed any other changes for good or ill.

The main advantage is that you are getting a longer exposure on your sensor because the (nominal) ISO is lower. If you catch twice as many photons this way, the noise will be lower.

I suspect that using extended LOW settings with any camera that offers it/them is going to help with noise, but little else.

With the E-M1 MkII, using the lower extended settings keeps DR much the same, and that's a considerable improvement on previous sensors, most of which caused the DR to drop when using the extended setting (e.g. my E-M1 MkI). See Bill Claff's site for comparisons of whatever cameras you choose. It is here:

http://www.photonstophotos.net/Charts/PDR.htm#Olympus OM-D E-M1,Olympus OM-D E-M1 Mark II

Click to expand...

This is very interesting, JaKing.

One must say, however, that the nominal ISOs that the camera manufacturers provide seem to be lies: DxO states that on the EM1II both nominal ISO64 as well as ISO200 correspond to real live ISO of 83. For EM1 nominal ISO100 and ISO200, the real live ISO is 122 in both cases: https://www.dxomark.com/Cameras/Com...1-versus-Olympus-OM-D-E-M1-Mark-II___909_1136

Click to expand...

Beware of those figures. What DxOmark calls 'measured ISO' isn't 'ISO'. For a start, it applies to the raw files, and ISO is very clear that 'ISO' does not apply to raw files. (it was always clear to those that understood the standard, but in the latest version, they have spelled it out explicitly, for those, like DxOmark, who were abusing the standard). DoO mark's measurement is still useful, just misnamed. What it tells you is what is the lowest ISO you could set your meter to (if you had a handheld meter separate from the camera) and still get raw files where the highlights weren't clipped.

Architeuthis said:

When EM1 and EM1II real ISO vs. DR is compared, the performance of both cameras is pretty the same...

Click to expand...

No 'real ISO'. That's a misunderstanding of what ISO is.

Architeuthis said:

So what is the difference between extended ISO (I assume also the "new" ISO100 is extended ISO) and base ISO (both mean ISO83 (EM1II) or ISO122 (EM1), respectively)? Does the "new" ISO100 mean just another ISO83 setting?

Click to expand...

They are all the ISO settings which they claim to be. The ISO 100 setting sets the meter and processes for 100 ISO. It has a more sensible amount of raw headroom than the old 200 ISO setting. I suspect that the other ISOs will also have less raw headroom, as they do on the EMIX.

So just for the ignoramus, how does increasing or decreasing the ISO setting in the camera affect RAW files?

katastrofa said:

So just for the ignoramus, how does increasing or decreasing the ISO setting in the camera affect RAW files?

Click to expand...

That depends on the camera, ISO doesn't standardise it. For a few cameras, the ISO control does nothing whatsoever to the raw files. For most, each stop of ISO reduces the saturation exposure by a stop, sometimes not at the ends of the ISO range. It might also affect the level of electronic noise (called read noise), reducing up to some point on the ISO scale.

The ISO control does, in general, what ISO says it should, which is control how much exposure gives a given lightness in the processed output. Most manufacturers will also use the ISO control as an indicator of likely exposure, to optimise the electronics for that expected expose. The former effect, which is what ISO mandates, only applies JPEG files, the latter, about which ISO has nothing to say at all, and as above varies camera to camera, applies to all files.

Architeuthis said:

I have installed firmware 3.0 on my OMD-EM1II.

I wonder, whether the new ISO 100 brings better dynamic range, less noise and better colors compared to the base ISO 200, when used for raw files?

Already in earlier firmware versions ISO 64 was available, but as far as I understand it, this does not improve DR, noise or color, but is useful only when too much light is present, to maintain wide aperture and increase exposure time...

Wolfgang

Click to expand...

Base is the lowest analog gain before A/D conversion. It's the lowest noise setting.

Extended (top and bottom of the range) is digital manipulation. No SNR improvement. And only useful for shooting JPEGs.

Architeuthis said:

JaKing said:

Wolfgang, I only ever felt the need to use the extended LOW ISO with my E-30, not with either of my E-M1s (MkI or MkII).

With the E-30, which was a real noise machine given half a chance, using ISO 100 certainly lessened the noise. Never noticed any other changes for good or ill.

The main advantage is that you are getting a longer exposure on your sensor because the (nominal) ISO is lower. If you catch twice as many photons this way, the noise will be lower.

I suspect that using extended LOW settings with any camera that offers it/them is going to help with noise, but little else.

With the E-M1 MkII, using the lower extended settings keeps DR much the same, and that's a considerable improvement on previous sensors, most of which caused the DR to drop when using the extended setting (e.g. my E-M1 MkI). See Bill Claff's site for comparisons of whatever cameras you choose. It is here:

http://www.photonstophotos.net/Charts/PDR.htm#Olympus OM-D E-M1,Olympus OM-D E-M1 Mark II

Click to expand...

This is very interesting, JaKing.

One must say, however, that the nominal ISOs that the camera manufacturers provide seem to be lies: DxO states that on the EM1II both nominal ISO64 as well as ISO200 correspond to real live ISO of 83. For EM1 nominal ISO100 and ISO200, the real live ISO is 122 in both cases: https://www.dxomark.com/Cameras/Com...1-versus-Olympus-OM-D-E-M1-Mark-II___909_1136

When EM1 and EM1II real ISO vs. DR is compared, the performance of both cameras is pretty the same...

So what is the difference between extended ISO (I assume also the "new" ISO100 is extended ISO) and base ISO (both mean ISO83 (EM1II) or ISO122 (EM1), respectively)? Does the "new" ISO100 mean just another ISO83 setting?

Wolfgang

Click to expand...

Wolfgang, Bob has already summed things up quite well. Don't trust DxO!

A longer exposure (longer shutter speed/larger aperture, or both) will always give less noise (higher signal to noise ratio).

My observations over many years show that Olympus uses the tone curves differently from most other manufacturers. Gentler roll off at both ends, and generally moved to the left.

Both of these have ramifications for both raw and JPEG files. With (particularly) early cameras, it was easier to blow highlights - the meter being miscalibrated in my E-510 didn't help this! ISO 100 was more like ISO 125/160. Other ISO settings were fine ...

A good friend with a D700 alerted me further to this difference and its effect in practice. When he pushed his camera up/down to the limits of the tone curve there was very noticeable, harsh clipping.

My conclusions around this interesting thing is that (generally) exposing to the right has little (if any) benefit for most Olympus cameras.

Also, my more modern E-M1 MkI and MkII have much more highlight headroom at base ISO (200), and some of that appears to be lost at extended low ISOs. It pays to chimp ... though not generally a problem.

JaKing said:

Architeuthis said:

JaKing said:

Wolfgang, I only ever felt the need to use the extended LOW ISO with my E-30, not with either of my E-M1s (MkI or MkII).

With the E-30, which was a real noise machine given half a chance, using ISO 100 certainly lessened the noise. Never noticed any other changes for good or ill.

The main advantage is that you are getting a longer exposure on your sensor because the (nominal) ISO is lower. If you catch twice as many photons this way, the noise will be lower.

I suspect that using extended LOW settings with any camera that offers it/them is going to help with noise, but little else.

With the E-M1 MkII, using the lower extended settings keeps DR much the same, and that's a considerable improvement on previous sensors, most of which caused the DR to drop when using the extended setting (e.g. my E-M1 MkI). See Bill Claff's site for comparisons of whatever cameras you choose. It is here:

http://www.photonstophotos.net/Charts/PDR.htm#Olympus OM-D E-M1,Olympus OM-D E-M1 Mark II

Click to expand...

This is very interesting, JaKing.

One must say, however, that the nominal ISOs that the camera manufacturers provide seem to be lies: DxO states that on the EM1II both nominal ISO64 as well as ISO200 correspond to real live ISO of 83. For EM1 nominal ISO100 and ISO200, the real live ISO is 122 in both cases: https://www.dxomark.com/Cameras/Com...1-versus-Olympus-OM-D-E-M1-Mark-II___909_1136

When EM1 and EM1II real ISO vs. DR is compared, the performance of both cameras is pretty the same...

So what is the difference between extended ISO (I assume also the "new" ISO100 is extended ISO) and base ISO (both mean ISO83 (EM1II) or ISO122 (EM1), respectively)? Does the "new" ISO100 mean just another ISO83 setting?

Wolfgang

Click to expand...

Wolfgang, Bob has already summed things up quite well. Don't trust DxO!

A longer exposure (longer shutter speed/larger aperture, or both) will always give less noise (higher signal to noise ratio).

My observations over many years show that Olympus uses the tone curves differently from most other manufacturers. Gentler roll off at both ends, and generally moved to the left.

Both of these have ramifications for both raw and JPEG files. With (particularly) early cameras, it was easier to blow highlights - the meter being miscalibrated in my E-510 didn't help this! ISO 100 was more like ISO 125/160. Other ISO settings were fine ...

A good friend with a D700 alerted me further to this difference and its effect in practice. When he pushed his camera up/down to the limits of the tone curve there was very noticeable, harsh clipping.

My conclusions around this interesting thing is that (generally) exposing to the right has little (if any) benefit for most Olympus cameras.

Click to expand...

Really? So how does one shoot in snow, on the beach or in dark forest? Just do what the camera tells you to do? That's you conclusions?

JaKing said:

Also, my more modern E-M1 MkI and MkII have much more highlight headroom at base ISO (200), and some of that appears to be lost at extended low ISOs. It pays to chimp ... though not generally a problem.

Click to expand...

I thank everybody for the valuable input so far. I am still confused about the meaning of ISO in digital photography and the difference between extended and base ISO...

I will continue later, have to go to work...

Wolfgang

Thanks!

Chris Noble said:

Architeuthis said:

I have installed firmware 3.0 on my OMD-EM1II.

I wonder, whether the new ISO 100 brings better dynamic range, less noise and better colors compared to the base ISO 200, when used for raw files?

Already in earlier firmware versions ISO 64 was available, but as far as I understand it, this does not improve DR, noise or color, but is useful only when too much light is present, to maintain wide aperture and increase exposure time...

Wolfgang

Click to expand...

Base is the lowest analog gain before A/D conversion. It's the lowest noise setting.

Extended (top and bottom of the range) is digital manipulation. No SNR improvement. And only useful for shooting JPEGs.

Click to expand...

These definitions do not always work. In any case, what you're calling 'base' is not the lowest noise setting. It's generally the highest noise setting so far as electronic noise goes. So far as the shot noise goes, any increase in exposure will lower it, so the (so called 'extended') settings below will give less noise.

bobn2 said:

Chris Noble said:

Architeuthis said:

I have installed firmware 3.0 on my OMD-EM1II.

I wonder, whether the new ISO 100 brings better dynamic range, less noise and better colors compared to the base ISO 200, when used for raw files?

Already in earlier firmware versions ISO 64 was available, but as far as I understand it, this does not improve DR, noise or color, but is useful only when too much light is present, to maintain wide aperture and increase exposure time...

Wolfgang

Click to expand...

Base is the lowest analog gain before A/D conversion. It's the lowest noise setting.

Extended (top and bottom of the range) is digital manipulation. No SNR improvement. And only useful for shooting JPEGs.

Click to expand...

These definitions do not always work.

Click to expand...

What kind of work are you referring to? They are definitions.

bobn2 said:

In any case, what you're calling 'base' is not the lowest noise setting. It's generally the highest noise setting so far as electronic noise goes.

Click to expand...

Lowest gain = lowest noise introduced by the electronic processing.

bobn2 said:

So far as the shot noise goes, any increase in exposure will lower it, so the (so called 'extended') settings below will give less noise.

Click to expand...

You are confusing noise due to the particle nature of light (shot noise), which the camera electronics can't do anything about, with the noise introduced by the camera electronics, which is what base ISO refers to.

Measured in auto exposure mode with RawDigger.



Headroom is the difference between the camera's 18% grey level and maximum output (4095-255=3840 or 11.9 EV). For the normal ISO range the average headroom is 4.1 stops. For the low range it drops to around 2.4 stops. Total noise is higher in the low ISO range, but the signal level increases even more so the Signal/Noise Ratio continues to improve.

You'll lose about about 1 stop of headroom in highlights but gain 1 stop of shadow headroom. Loss in highlights may even be more (haven't calculated it of course ).

I use Low ISO only if there are no highlights that I want to preserve. At ISO64 and 100, EM1 II gives very clear pixels.

Chris Noble said:

bobn2 said:

Chris Noble said:

Architeuthis said:

I have installed firmware 3.0 on my OMD-EM1II.

I wonder, whether the new ISO 100 brings better dynamic range, less noise and better colors compared to the base ISO 200, when used for raw files?

Already in earlier firmware versions ISO 64 was available, but as far as I understand it, this does not improve DR, noise or color, but is useful only when too much light is present, to maintain wide aperture and increase exposure time...

Wolfgang

Click to expand...

Base is the lowest analog gain before A/D conversion. It's the lowest noise setting.

Extended (top and bottom of the range) is digital manipulation. No SNR improvement. And only useful for shooting JPEGs.

Click to expand...

These definitions do not always work.

Click to expand...

What kind of work are you referring to? They are definitions.

Click to expand...

They don't work when it comes to sensibly delineating 'base' or 'extended

Chris Noble said:

Chris Noble said:

In any case, what you're calling 'base' is not the lowest noise setting. It's generally the highest noise setting so far as electronic noise goes.

Click to expand...

Lowest gain = lowest noise introduced by the electronic processing.

Click to expand...

Incorrect. It confirms my impression that you're working on the very common but completely nonsense accounts of what ISO is, what it works and what is the significance and effect of 'gain'.

Chris Noble said:

Chris Noble said:

So far as the shot noise goes, any increase in exposure will lower it, so the (so called 'extended') settings below will give less noise.

Click to expand...

You are confusing noise due to the particle nature of light (shot noise), which the camera electronics can't do anything about, with the noise introduced by the camera electronics, which is what base ISO refers to.

Click to expand...

I'm confusing nothing. I'm separating the two, which you failed to do, for the sake of clarity. With respect to the noise introduced by the camera electronics (or read noise) that reduces as ISO is increased, in general.

The situation is confusing, but I did now some recherches in Internet and think I know now what extended ISO means:

Here, among other decriptions, is an interesting post about the difference between native (and base) ISO and extended ISO: https://petapixel.com/2015/06/24/native-versus-extended-the-science-and-marketing-of-iso-ranges/

Chris Noble was very close to it. The question is just, whether the raw file is affected by the extended ISO setting or only the JPEG is. In other posts/articles, not referenced here, I saw that indeed the raw file can be affected by extended settings, but not always. The way depends on the camera model and manufacturer. In the case of EM1-II people are saying that the digital information is processed towards highlight clipping, improving S/N in less bright pixels (dynamic range at the cost of S/N). So extended ISO remains useful for people that work with JPEGs out of the camera, but not the ones who postprocess raw files...

For my needs I must conclude the extended ISO 64 and 100 are useless. It would need lower ISO in circumstances where I would require lower sensor sensitivity, not leading to output clipping (e.g. having sunbursts directly in the frame) :-(

Hopefully future MFT sensors will provide lower base ISO (but I am quite happy with my EM1-II as it is)...

There is also no way that I doubt DxO benchmarks. The only way to challenge them would be to make benchmark by my own and I do not have the means, time and also not the moods to do so (prefer to go diving and make photos)...

Here the interesting description how DxO measures ISO sensitivity (and it is clear from the description and what is said above that extended ISO64 (and ISO100) and base ISO200 in the EM1II are almost the same regarding raw files): https://www.dxomark.com/dxomark-camera-sensor-testing-protocol-and-scores/

Wolfgang