ISO standard output sensitivity and recommended exposure index in the 1D Mk III

In its 2006 edition, the ISO standard for determining the "ISO speed" of digital still cameras took on two new alternative measures of camera sensitivity, the standard output sensitivity (SOS) and the recommended exposure index (REI). As I understand it, these were included at the behest of CIPA, the Japanese camera technical organization, which had developed these measures and earlier promulgated them in one their own standards.

Before I go further, let me clarify two terms:

ISO speed - a property of a particular type of film or a particular digital camera (in any given "ISO" setting mode). It tells us (in a certain sense) the sensitivity of the medium to exposure. It is determined in a specified way under specified conditions.

Exposure index - what we tell an exposure meter the ISO speed of the medium is. Most commonly, we set the exposure index to the applicable ISO speed, but we might not.

The new measures are:

ISO standard output sensitivity (SOS) - Going right to the bottom line (skipping the definitions and algebra), this is always a number that is 0.71 times the (saturation-based) ISO speed for the camera. It can be presented this way:

"ISO 400 (SOS)" or "ISO 400 (standard output sensitivity)".


ISO recommended exposure index (REI) - This is what the manufacturer of a particular camera recommends we set as the exposure index on our meter. It is the value the manufacturer feels will give good results in a lot of situations. It is not (necessarily) determined under any standardized procedure or based on any standardized defintion. It can be presented this way:

"ISO 400 (REI)" or "ISO 400 (recommended exposure index)"


These terms appear in the specifications for the Canon EOS 1D Mk III (perhaps for the first time in the Canon line - I'm not sure). They are basically presented this way:

"ISO 200, 400, 800 . . . (standard output sensitivity, recommended exposure index)"

My first reaction was, "well, which is it - SOS or REI?" But I think this is what they mean: the ratings that are presented are ISO SOS values (presumably determined under the ISO standard), which by implication are what we should (under ordinary circumstances) set as the exposure index in a meter. They are also what Canon recommends we set as the exposure index in a meter (REI). A little redundant, but . . .

As a matter of fact, a while ago I determined what exposure index values were used in the Canon EOS automatic exposure systems, based on a "kitchen table" test to verify the proper operation of EOS automatic exposure systems passed on by Chuck Westfall. I determined that the exposure index values used by Canon we perhaps 65% of the value that would be determined under the ISO procedures for (saturation-based) ISO speed. I didn't realize then that Canon was ahead of their time - that these "ISO speeds" (as they were called then) were actually (at least very nearly) SOS values (not yet at the time defined by the ISO). (Canon is of course a participant in the CIPA group responsible for this work.)

Hi, Doug:

Here is some additional information on this topic:

1. The CIPA document you are referring to is Publication DC-004, "Sensitivity in Digital Cameras" and an English translation of it can be downloaded at this URL:

http://www.cipa.jp/english/hyoujunka/kikaku/pdf/DC-004_EN.pdf

2. This document was published on April 20, 2006, and it went into effect for all digital cameras announced by Japanese manufacturers after October 20, 2006. The EOS-1D Mark III is the first Canon Digital SLR to be affected by this new standard.

3. According to CIPA DC-004, Section 3: "Characterisitics to be specified as 'Sensitivity' are the assigned values of Standard Output Sensitivity (symbol: SOS), and/or Recommended Exposure Index (symbol: REI). The use of both or either shall be acceptable. When both are listed, Standard Output Sensitivity shall come first, followed by Recommended Exposure Index. It is also acceptable to use the terms 'Standard Output Sensitivity' and/or 'Recommended Exposure Index' as notational terms of each value instead of using the word 'Sensitivity'.

4. The specifications of the EOS-1D Mark III, as listed in the Instructions on page 198, list "ISO Speed (Recommended Exposure Index)."

Doug, can you refer me to the documentation you found that lists both terms for the EOS-1D Mark III? Thanks.

Best Regards,

Chuck Westfall
Director/Media & Customer Relationship
Camera Marketing Group/Canon U.S.A., Inc.

Hi, CHuck,

Chuck Westfall(Canon USA) said:

Hi, Doug:

Here is some additional information on this topic:

1. The CIPA document you are referring to is Publication DC-004, "Sensitivity in Digital Cameras" and an English translation of it can be downloaded at this URL:

lhttp://www.cipa.jp/english/hyoujunka/kikaku/pdf/DC-004_EN.pdf

Click to expand...

Thanks. I already had a draft. It was part of my research on this topic. I'll download the real thing now.

2. This document was published on April 20, 2006, and it went into effect for all digital cameras announced by Japanese manufacturers after October 20, 2006. The EOS-1D Mark III is the first Canon Digital SLR to be affected by this new standard.

Click to expand...

I thought perhaps so. Thanks.

3. According to CIPA DC-004, Section 3: "Characterisitics to be specified as 'Sensitivity' are the assigned values of Standard Output Sensitivity (symbol: SOS), and/or Recommended Exposure Index (symbol: REI). The use of both or either shall be acceptable. When both are listed, Standard Output Sensitivity shall come first, followed by Recommended Exposure Index. It is also acceptable to use the terms 'Standard Output Sensitivity' and/or 'Recommended Exposure Index' as notational terms of each value instead of using the word 'Sensitivity'.

Click to expand...

Indeed. I had sort of assumed that this meant a manufacturer could specify different values of SOS and REI, but certainly stating the same value for both should be legitimate. Still, I hardly see the point. Either the single value you put forth is objectively determined (SOS), or it is empirically chosen (REI). But of course I don't know the intent of the standard drafters.

4. The specifications of the EOS-1D Mark III, as listed in the Instructions on page 198, list "ISO Speed (Recommended Exposure Index)."

Click to expand...

Ah, I don't have that.

Doug, can you refer me to the documentation you found that lists both terms for the EOS-1D Mark III? Thanks.

Click to expand...

In most particular, it shows up in the White Paper for the EOS-1D Mark III (at least in the version I have).

It also shows up in a number of Web discussions of the camera by retailers evidently taken from some Canon document. Here is a random example:

http://www.pictureline.com/products/14178/Canon_EOS_1D_Mark_III_Digital_Camera/

It also shows up in the specifications for the cameras on the Canon USA site, here:

http://www.usa.canon.com/consumer/controller?act=ModelInfoAct&fcategoryid=139&modelid=14999

Do I take it from your comment above regarding the specifications in the camera instructions that Canon's current intent is to state these values as REI only and not as SOS?

Thanks so much for your help.

Best regards,

Doug

>>Do I take it from your comment above regarding the specifications in the camera instructions that Canon's current intent is to state these values as REI only and not as SOS?<<

That is correct, and furthermore, I anticipate that Canon will stick to REI for its new digital SLRs going forward from here.

Best Regards,

Chuck Westfall
Director/Media & Customer Relationship
Camera Marketing Group/Canon U.S.A., Inc.

Hi, Chuck,

Chuck Westfall(Canon USA) said:

>>Do I take it from your comment above regarding the specifications in the camera instructions that Canon's current intent is to state these values as REI only and not as SOS?<<

That is correct, and furthermore, I anticipate that Canon will stick to REI for its new digital SLRs going forward from here.

Click to expand...

Thanks for clarifying that.

This makes sense. Notwithstanding the thousands of man hours of excruciating sophistry, and hundreds of pages of ill-explained algebra, in the work over the years on ASA speeds for film and their ISO successors, the "basic" type of reflected light metering that is the major user of the numerical results of this work remains largely an empirical process, where there isn't even a single criterion by which we can judge "success". The REI definition is a candid address of this situation.

It is ironic that we have now made essentially a full circle from the time of the intitial introduction of the Weston exposure meter, before there were any standard definitions of film "speed", when Weston was kind enough to suggest, for each of the the then-popular film types, an exposure index value they thought would give "good results in many cases" [my phrase, not theirs].

Thanks again for your help.

Best regards,

Doug