low ISO values

[Manfred M]

As a low ISO shooter, I would definitely have a look with the one at ISO 25. The base ISO (i.e. the lowest numeric ISO) is what the sensor natively records. Anything higher than that is a result of amplifying the sensor signal, which has all kinds of negative impact (lower dynamic range, higher noise and lower colour depth), so in my view it is a bit of a kludge, rather than a "true" higher ISO.

[Adzman808]

As a low ISO shooter, I would definitely have a look with the one at ISO 25. The base ISO (i.e. the lowest numeric ISO) is what the sensor natively records. Anything higher than that is a result of amplifying the sensor signal, which has all kinds of negative impact (lower dynamic range, higher noise and lower colour depth), so in my view it is a bit of a kludge, rather than a "true" higher ISO.

Yes. That’s how very digital camera works, with either / or post and pre ADC amplification

Personally I’ve come to the conclusion that the typical photography forum member isn’t the typical camera buyer. In much the same way a car enthusiast (or petrol head) is not the typical car buyer, but rather a small market segment

[William W]

I was very fond of Kodachrome 25.

WW

[george013]

I was very fond of Kodachrome 25.

WW

I assume for the texture/quality of the negative, not for the exposure settings.
A main difference between analogue and digital is that with analogue one can change film and make use of the differences of them. In digital I can't change the sensor.

George

[xpatUSA]

I assume for the texture/quality of the [Kodachrome 25] negative, not for the exposure settings.

A main difference between analogue and digital is that with analogue one can change film and make use of the differences of them. In digital I can't change the sensor.

George

Exactly, George.

Which is why I think that the whole business of "ISO" for digital cameras is pure hogwash, just to keep the filmsters happy. Or even for commercial reasons - by introducing yet another "bigger is better" number for the Marketing Dept.

[Adzman808]

by introducing yet another "bigger is better" number for the Marketing Dept.

My point exactly, like megapixels used to be

[Manfred M]

My point exactly, like megapixels used to be

And still is. If one prints large images (I do), I look to more megapixels and a larger sensor offers significant advantages. If one only posts images on the internet, then I would agree.

[Manfred M]

I was very fond of Kodachrome 25.

WW

Me too. I never quite forgave Kodak for dropping that film, although I did shoot Kodachrome 64 for many years after Kodachrome 25 was discontinued.

[Manfred M]

I assume for the texture/quality of the negative, not for the exposure settings.
A main difference between analogue and digital is that with analogue one can change film and make use of the differences of them. In digital I can't change the sensor.

George

The Kodachrome line was a colour reversal (i.e. slide film) that had beautiful, saturated colours; this was not a negative film.

Let's also be honest that faster film mean more grain as the silver halide crystals were larger, so there was / is still a trade-off when shooting a faster (higher ISO) film.

[DanK]

A main difference between analogue and digital is that with analogue one can change film and make use of the differences of them. In digital I can't change the sensor.

George

Yes, I recall many times awkwardly trying to change film with the camera zipped into a changing bag...

I'm sure that many people back in the day selected films based on other characteristics, but with the limited skills I had back then, I selected mostly for speed: lower speed and higher quality when I didn't need the speed, and higher quality with whatever drawbacks it entailed when I needed more speed. I didn't use Tri-X or push it to 800 because I wanted more pronounced grain; I did it because I needed the extra speed.

Of course, over time, many people came to value the characteristics of specific films; for example, the grainy Tri-X appearance became a classic for casual portraits. But if one wants the characteristics of specific films, there are numerous software packages that emulate them quite well. For example, SilverEfex can emulate quite a list of black and white films.

I chanced on an interesting comparison of digital with film recently: http://www.clarkvision.com/articles/dynamicrange2/. With respect to his comments on spatial resolution at the end, bear in mind that this is not a new article, and he was using an 8.3 MP APS-H digital camera.

[george013]

The Kodachrome line was a colour reversal (i.e. slide film) that had beautiful, saturated colours; this was not a negative film.

Let's also be honest that faster film mean more grain as the silver halide crystals were larger, so there was / is still a trade-off when shooting a faster (higher ISO) film.

The point is that the photographer chooses his film for a certain purpose, and that could be a dia positive. When I remember well I used FujiColor for the colors. The film is not a part of the camera.
What I also want to mention is that the low iso is not always higher in new models. I just compared the D700 and the D750. The D700 has a base of 200, the D750,much newer, has a base of 100. Technical and software development make it possible to use that sensor to a higher iso-value with reasonable results.
D500 iso 100-51200
D5 iso 100-102400
D4 iso 200-12800
D7200 iso 100-25600.
If the base iso is to high for some reason one can use a neutral filter.

George

[xpatUSA]

The point is that the photographer chooses his film for a certain purpose, and that could be a dia positive. When I remember well I used FujiColor for the colors. The film is not a part of the camera.

D500 iso 100-51200
D5 iso 100-102400
D4 iso 200-12800
D7200 iso 100-25600.

I recall that my D50 was 200 ISO.

If the base ISO is too high for some reason, one can use a neutral filter.

Using a ND filter to "lower the ISO" - an interesting thought, George.

[William W]

I assume [that you liked Kodachrome 25] for the texture/quality of the negative, not for the exposure settings.

Yes, mostly. Let’s say 90% because of the film’s characteristics, yet also 10% because of the ASA rating (ASA 25): although Kodak recommended against long exposure times, I quite liked using Kodachrome 25 between 1 second and 30 seconds exposure, especially for night artificial light scenes. One peculiarity of the Reciprocity Effect of Kodachrome 25 was it seemed to accentuate the Red~Magenta which worked out quite acceptable for night street landscapes and similar lighting scenarios.

The (long term) archival qualities of Kodachrome 25 are legendary. The three main elements of Kodachrome 25’s characteristics which thrilled me, were:

> Extreme fine grain
> Extreme sharpness
> Precise rendition of the subtleties of hue variances

These three elements made this film excellent for any legal and forensic work.

Additionally, utilizing the Cibachrome printing processes, (I used for Portraiture), the Final Print was immaculate and these Prints, even today, have a cult following.

***

A main difference between analogue and digital is that with analogue one can change film and make use of the differences of them. In digital I can't change the sensor.

Yes. Agree.

Additionally, with film we can make use of changing the PROCESSING of the film: which is different to PRINTING the negative (or positive).

The point I am making is, additional to the point George makes - (having a choice of various films to plonk into any one camera and NOT having a choice of different sensors to plonk in any one camera): many of those ‘any one films’ could be converted into a range of DIFFERENT films by:
> rating the film differently and/or
> push/pull processing the development and/or
> using a range different developers.


There is a direct similarity between Analogue Wet Printing and Digital Post Processing. But whist the raw conversion phase of the digital file transfer is like the Development phase of Analogue Processing, the Digital raw conversion does not seem to allow the latitudes that the choices of different films and the choices of different development processes of those films gave us. Although noted that I do like playing with the range of Fuji film emulsion options on my Fuji X100s.

The aforementioned Tri-X is one of many examples of how one film could provide a range of options.

For my uses it was three films in one:
> Rated at Kodak’s suggestion ASA400 (actually there were two 320 and 400), this was a great all purpose film most suitable for reportage
> Rated at ASA100 (or sometimes ASA 50) a good studio portrait film, especially for men, producing punchy contrast negatives with good shadow detail and with a bit more ‘grit’ than using Pan-X (ASA125)
> Rated at ASA1600 (sometimes ASA 3200) excellent for night sports and indoor low light, sans flash

***

Returning directly to the Opening Post:

In predigital times camera's had ISO-settings as low as 25 or even 6, though I don't know if there were films available with such a 'slow character'.

It was common to have films rated at around ASA 25. One most commonly used was Kodachrome 25, which I have mentioned.

Not so long ago (around 2005) a popular B&W Neg Film was EFKE 25 and this was often rated at ASA 12 to ASA 10 or sometimes ASA 6; those intentional exposures were usually accommodated by Pulling the Development.

. . . In digital times many a camera's default ISO-value is 200, and values below ISO 100 can not be chosen on the majority of today's camera's - which is a limitation when we wish to record movement.

Yes, technically a ‘limitation’ – but as already mentioned a 4 Stop ND Filter ‘effectively’ will get a Digital Camera capable (only) of (lowest) ISO 100 … to ISO 6

- What could be the reason/s behind this particularity?

I agree with the reasons already given,

i.e. I expect cost of (TR&D) development, directly related to expectation of major user requirements (i.e. maximizing sales) is the reason.

The vast majority of photos are made at ISO100 to ISO xxxx. There is more the consumer passion for "high ISO" not "low ISO" - (if you like more a need for fast Shutter Speeds and not slow Shutter Speeds.

Those folk who want a really Slow Shutter speed can simply buy a ND Filter and use ISO100 (or whatever the base ISO happens to be).

It just would not provide the necessary ROI (return on investment) for the mass Digital Camera manufactures to develop a camera, with a base ISO of ISO 6 (or ISO 25).

WW

[Adzman808]

As well as the ROI equation, a significant number of modern cameras buy their sensors from Sony (I think Canon might be an exception), which means cameras are equipped with sensors that Sony are prepared to make, rather than sensors that camera OEMs might want in their cameras

[lichtloper]

Peter I'm no expert on sensor design but i think you'll find that it involves trade-offs between ISO sensitivity, Full well capacity, dynamic range and noise performance. A sensor designed with an ISO sensitivity of say 25 would probably have inferior performance at a gain setting of ISO 100 than a sensor designed to have an ISO sensitivity of 100. I think most people would do most of their shooting with a minimum ISO of 100 and thus would prefer to have the best performance possible around that setting, rather than having the opportunity to use slower shutter speeds every so often.

By the way, these so called ISO extensions or boosted minimums aren't true reductions in ISO but usually just involve over-exposure of the image.

Dave

Indeed, Dave, thanks.

I believe you're right: the sensor developer tries to find the best possible, optimal combination of ISO sensitivity, noise performance, well capacity and dynamic range quality.

I just had hoped that some technical person among us here could explain the matter on a purely technical level; I merely wish to understand.

The question arose when I was standing on a bridge over a really slowly flowing river, the Kromme Rijn (Crooked Rhine), and wondering how I could let my photograph of the river show that it was slowly flowing; it was a summer's day, with plenty of daylight. Camera: Lumix LX-5, lowest ISOsetting: 80. Used max aperture F8 and 1/4 sspeed. I didn't bring a grey filter.
That's when the question arose.

[xpatUSA]

Originally Posted by dje
Peter I'm no expert on sensor design but i think you'll find that it involves trade-offs between ISO sensitivity, Full well capacity, dynamic range and noise performance. A sensor designed with an ISO sensitivity of say 25 would probably have inferior performance at a gain setting of ISO 100 than a sensor designed to have an ISO sensitivity of 100. I think most people would do most of their shooting with a minimum ISO of 100 and thus would prefer to have the best performance possible around that setting, rather than having the opportunity to use slower shutter speeds every so often.

By the way, these so called ISO extensions or boosted minimums aren't true reductions in ISO but usually just involve over-exposure of the image.

Dave

Indeed, Dave, thanks.

I believe you're right: the sensor developer tries to find the best possible, optimal combination of ISO sensitivity, noise performance, well capacity and dynamic range quality.

I just had hoped that some technical person among us here could explain the matter on a purely technical level; I merely wish to understand.

What Dave said, plus:


As to sensor 'performance', the factors are photodiode area, bias voltage, quantum efficiency and Color Filter Array transmittance.

All of these plus what Dave said combine to form a value of illuminance in lux-sec that will saturate the sensor (Hsat). Once that is known, the ISO can be calculated by ISO's formula for 'ISO Speed' per ISO 12232.

(The formula for film is ISO = 10/Hsat per ISO 27271.)

For digital, they complicate matters by changing the constant for 'ISO Speed' such that ISO = 78/Hsat which gives about a 1/2 stop of headroom.

For most manufacturers you won't find Hsat listed for any sensor and I am only able to figure it out for one of my cameras from a sensor data-sheet (0.81 lux-sec), which gives a actual base ISO of 96 - permitting the manufacturer to call it '100 ISO'.

So, comparing film to digital, we can figure Hsat closely enough for film, remembering that a digital ISO setting does NOT affect the sensitivity of the sensor.

Simple enough to change the film formula to Hsat = 10/ISO.

So, for film ISO 6, Hsat = 10/6 = 1.67 lx.s BUT, for digital and that same ISO 6, Hsat = 78/6 = 13 lx.s !

I'm not certain that such a sensor even exists and that would explain, at least to me, why you won't find a setting of ISO 6 on any modern conventional digital camera.

By quoting some numbers, references and formulae I'm hoping that the level of the above is technical enough for you. If not, there's plenty more here:

http://www.imatest.com/docs/sensitivity_ei/

www.kilopixel.net/publications/ISSCC_ISO_BAER.pdf

HTH

[lichtloper]

What Dave said, plus:


As to sensor 'performance', the factors are photodiode area, bias voltage, quantum efficiency and Color Filter Array transmittance.

All of these plus what Dave said combine to form a value of illuminance in lux-sec that will saturate the sensor (Hsat). Once that is known, the ISO can be calculated by ISO's formula for 'ISO Speed' per ISO 12232.

(The formula for film is ISO = 10/Hsat per ISO 27271.)

For digital, they complicate matters by changing the constant for 'ISO Speed' such that ISO = 78/Hsat which gives about a 1/2 stop of headroom.

For most manufacturers you won't find Hsat listed for any sensor and I am only able to figure it out for one of my cameras from a sensor data-sheet: 0.81 lux-sec that gives a actual base ISO of 96, permitting the manufacturer to call that '100 ISO'.

So, comparing film to digital, we can figure Hsat closely enough for film, remembering that a digital ISO setting does NOT affect the sensitivity of the sensor.

Simple enough to change the film formula to Hsat = 10/ISO.

So, for film ISO 6, Hsat = 10/6 = 1.67 lx.s BUT, for digital and that same ISO 6, Hsat = 78/6 = 13 lx.s !. I'm not certain that such a sensor even exists and that would explain to me why you won't find a setting of ISO 6 on any digital camera.

By quoting some numbers, references and formulae I'm hoping that the level of the above is technical enough. If not, there's plenty more here:

http://www.imatest.com/docs/sensitivity_ei/

www.kilopixel.net/publications/ISSCC_ISO_BAER.pdf

HTH

Thank you, xpatUSA!
This is complicated stuff, but was was asking for it... .
I'm going to re-read your post and chew on it until it dawns on me, until I get
your 'conclusion': "...explains why you won't find a setting of ISO 6 on any digital camera."

Great stuff, expatUSA,
and I'm also grateful for the links you mentioned.

[xpatUSA]

Thank you, xpatUSA!
This is complicated stuff, but was was asking for it... .
I'm going to re-read your post and chew on it until it dawns on me, until I get
your 'conclusion': "...explains why you won't find a setting of ISO 6 on any digital camera."

Allow me to expand on my conclusion. I said:

"BUT, for digital and that same ISO 6, Hsat = 78/6 = 13 lx.s !. I'm not certain that such a sensor even exists"

I was talking about "most digital cameras" and the probability of finding one which actually had such an Hsat, which I believe is close to zero.

Look here for some more about that, film versus digital by R N Clark:

http://www.clarkvision.com/articles/...ital.summary1/

And more about specific digital sensors here:

http://www.clarkvision.com/articles/...mance.summary/

There's a Kodak sensor in there with a whopping 550,000e- 'well full' !

What that is in lux-sec, I have no idea . . .

[Manfred M]

There's a Kodak sensor in there with a whopping 550,000e- 'well full' !

Kodak sensor?? Kodak sold their sensor business in 2011. I know they did CCD sensors and in fact Leica used Kodak sensors when they got into the digital camera business.

[Manfred M]

As well as the ROI equation, a significant number of modern cameras buy their sensors from Sony (I think Canon might be an exception), which means cameras are equipped with sensors that Sony are prepared to make, rather than sensors that camera OEMs might want in their cameras

Not correct Adam - there are an number of companies out there making sensors, not just Sony. I remember that Nikon used Aptina sensors in the V and J series cameras. Samsung, Toshiba, Foveon (owned by Sigma and make a non-Bayer array sensor found in Sigma cameras), STMicroelectronics are some other sensor makers, in addition to Sony and Canon.