The effect of the ISO dial setting on camera metering

[GeorgianBay1939]

This is a continuation of " Two functions of a camera's ISO dial." which ran out a few days ago.

First let me apologize for disappearing after starting a controversial thread. Health issues intervened and I learned some things other than the functions of an ISO dial. :-D No, I did not abandon DM's Ship of Fools.

_____________________

I decided here to separate the two functions, metering and mapping, so that we could drill down to clarify each function.

_____________________

An email (from an accomplished photographer) said, "I don't know and I don't care." So I should tell you why I started the thread in the first place:

A couple of weeks ago I heard a well-respected, experienced, ex-film, photographer answer a question from a youngster who had just added his first full frame camera to his collection of gear. The youngster (in his 50s) was a beginning raw shooter who was learning to stretch his envelope. Good science background and a bit of a perfectionist. Kids grown up and he was developing a hobby.

In answer to the question, "What does the ISO dial REALLY do when I change it?",

this is what the older photographer said,

"You are changing one of the three brightness variables so that you'll have to change either f/ or ss to get a proper exposure. The ISO dial is calibrated to give you optimum exposure in varying light conditions. Turn ISO down in bright conditions, up in dark conditions. Increasing the ISO setting increases the sensor's gain to compensate for lack of brightness in the scene."

After thinking about it, I realized that the above explanation would limit the growth of the youngster and contemplated how I would answer such a question when asked by a developing (!) photographer.

"Two functions of a camera's ISO dial" was my first attempt at forming such an answer. After reading the thread I realize that both parts, metering and mapping, of the answer needed improvement. Since the metering portion seemed to be almost ok, I would ask that you limit your comments to the metering portion of the answer in this thread. I will start a thread to deal with the (much more controversial) mapping portion later.

_________________________________________________

Here is the context:

What do changes to the ISO dial do to a camera’s inner workings?

1 It changes the set points of the camera's exposure metering system, (EMS).

2 It changes the tone values of the output image by a combination of sensor level gain adjustment and by modifying the tone mapping of exposure to output image.

Item 2 seems controversial and will be put up in a later thread.

___________________________________________________



Back to Item 1: It changes the set points of the camera's Exposure Metering System, (EMS).

1 (a) What does the camera's exposure metering system (EMS) do?

The EMS measures (normally through the lens) the average luminance of the scene, L, and based on that measurement and a factory calibration of ISO Speed, S, the EMS recommends a pair of exposure settings, N and t (f/ and ss) to give a good image result of the exposure. The factory calibration complies with the ISO 2720:1974 Standard to establish the relationship between average scene luminance, L, and recommended camera settings, N and t, for a selected ISO setting, S, for that particular camera model.

1(b) How does changing the ISO setting knob change the EMS recommendations?

Based on the measured average scene Luminance, L, and the calibration curves installed at the factory the EMS will return recommended EV settings , N and t, according to the Exposure Metering Equation , programmed into the EMS :



View: original size

For the above equation to be valid for a measured average scene luminance, L, the recommended camera settings, (N^2)/t , will change with changes in ISO speed, S.

Therefore for every setting of S, there is only one (N^2)/t (and only one pair of N, t) that will satisfy the equation.

That pair is the recommended exposure that the EMS offers to the photographer.

Therefore for a given average scene luminance, L, every time the ISO knob is changed, S is changed and the recommended exposure settings, N, t, (f/,ss) will change. Also, every time the average scene luminance, L, is changed (as measured by the meter) the recommended exposure settings will change, even if the ISO speed, S, is held constant.

The above paragraph (and its underlying equation) relates Exposure, (which only depends on L, N, t,) and ISO Speed, S, to give a "good image result". Two parts of Exposure, N and t, and ISO Speed, S, are sometimes crafted into an "Exposure Triangle", which should be read very critically as it is often leads to false conclusions.

This relationship of N ,t and S is easily seen when using the P(program) Mode in a camera: For a given L, adjusting S will change the exposure pairs. And, for a fixed L and S, changes in N will cause changes in t and vice versa.

[The definition of Exposure Value = Log_2 ([N^2)/t] is no accident as it puts all of the above in "stops", making relationships of exposure variables simple.]

The outputs of the EMS, the exposure settings, inform the various AutoExposure (AE) Modes [(P(rogram), A(perture), S(peed)] of the Mode dial. Some cameras have fully Automatic and/or Scene Modes which are also informed by the EMS.

It is important to note that the above is only valid when the photographer is using an AE Mode or is following the meter recommendation in M(anual) Mode. If the photographer disregards the meter recommendation (by using M(anual), Spot Metering with AE(lock), LiveView Histogram, or blinkies, etc) to inform his/her exposure decisions, the above equation is invalid and EV (and consequently exposure, Hv) is totally independent of S for any L.* Exposure compensation shifts the equation to provide a "compensated" recommended exposure pair.

* In this case the photographer is probably using independent camera settings (subject to artistic constraints regarding motion blur and DoF/Diffraction) which would optimize raw file outputs in terms of highlight protection, minimum noise and maximum PDR as opposed to recommended camera settings which would normally return an output image with a standard acceptable brightness.

___________________________

I would ask that you review and criticize the above. Please restrict your comments to the metering function of the ISO dial. I am sure that there will be a lot of other opportunities to add to the discussion of the other function of the ISO Dial!

Please no plagiarism or excessive noise, like this static:

http://www.dpreview.com/forums/post/57259249

http://www.dpreview.com/forums/post/57259530

Many thanks.

--
Tom
The best part of growing old is having the opportunity to do so.

Brtthome's Blog

photos, prose, prattle and possibilities

brtthome.wordpress.com

 

[Jeff]

Tom,

I like how you have posed the question ... as a matter of responding to a query from a beginning photographer seeking a deeper understanding of the ISO control. I just wonder if your proposed response might assume, so to speak, 'facts not in evidence' for a beginner asking a question like this.

My own approach would be different with an attempt to ground the issue in terms of visualization and control of the image rather than the technical aspects of how the camera operates. In this regard, I think there are four essential points the beginner would need to know ...

Point 1. The purpose of the ISO control is to calibrate the amount of light captured by the camera to the relative tones of human perception. Our eyes routinely handle 1000-fold and larger changes in light level just from going from outdoors to indoors. Changing shutter speed and f-number, alone, can't accommodate such a range. And if you also want to use those controls creatively for motion blur and depth of field adjustment, then there is no hope. The fix is an additional control to tell the camera how to translate the amount of light captured in an exposure into the tonal values of human perception. That's what the ISO control does, in a more-or-less industry standard fashion.

Point 2. The image processing chain begins in the camera and ends once you've produced the final image. By setting the ISO control on the camera, the camer has information needed to produce a final jpeg image, or, at a minimum, optimize the internal electronics to get as much information as possible for a raw file. It's also used to calibrate the auto-exposure mechanism of your camera. The details depend on the technology within the camera.

Point 3. A high ISO means that you low light levels will correspond to brighter tones in the final image. Low light levels are intrinsically noisy, no matter how good your camera's systems might be. The more light you capture, the better.

Point 4. Camera's also have upper limits on how much light they can capture.

I don't see why an explanation to a beginner would want to delve much deeper into this topics, particularly the math (unless you're talking to an engineer).

--
Jeff
http://www.flickr.com/photos/jck_photos/sets/

 

[Detail Man]

This is a continuation of " Two functions of a camera's ISO dial." which ran out a few days ago.

First let me apologize for disappearing after starting a controversial thread. Health issues intervened and I learned some things other than the functions of an ISO dial. :-D No, I did not abandon DM's .Ship of Fools.

_____________________

I decided here to separate the two functions, metering and mapping, so that we could drill down to clarify each function.

_____________________

An email (from an accomplished photographer) said, "I don't know and I don't care." So I should tell you why I started the thread in the first place:

A couple of weeks ago I heard a well-respected, experienced, ex-film, photographer answer a question from a youngster who had just added his first full frame camera to his collection of gear. The youngster (in his 50s) was a beginning raw shooter who was learning to stretch his envelope. Good science background and a bit of a perfectionist. Kids grown up and he was developing a hobby.

In answer to the question, "What does the ISO dial REALLY do when I change it?",

this is what the older photographer said,

"You are changing one of the three brightness variables so that you'll have to change either f/ or ss to get a proper exposure. The ISO dial is calibrated to give you optimum exposure in varying light conditions. Turn ISO down in bright conditions, up in dark conditions. Increasing the ISO setting increases the sensor's gain to compensate for lack of brightness in the scene."

After thinking about it, I realized that the above explanation would limit the growth of the youngster and contemplated how I would answer such a question when asked by a developing (!) photographer.

"Two functions of a camera's ISO dial" was my first attempt at forming such an answer. After reading the thread I realize that both parts, metering and mapping, of the answer needed improvement. Since the metering portion seemed to be almost ok, I would ask that you limit your comments to the metering portion of the answer in this thread.

I will start a thread to deal with the (much more controversial) mapping portion later.

That the transfer-function of a non-linear system (with RGB color-space rendered tone-curve transfer functions and any associated gamma-correction applied ... when mapping image-data ... into such RGB color-spaces) does not constitute a linear system (having one constant scale-factor) is not controversial.

All analytical assumptions drawn surrounding non-linear output image-data, as well as in-camera metering system behavior in relation to such non-linear output image-data, require attention be paid to "mappings" !

From the display device, and into the vast and mysterious pre-optic ganglia (having an overall consistency of that of tofu, or of toothpaste) we conceptually meander - but between our (perhaps pristinely) modeled hardware and the truly ineffable psycho-visual impressions and interpretations that we in some manner formulate, there may exist legions of (seemingly inscrutably) structured and arranged "bundles of wiring":

Source: http://www.detectingdesign.com/images/HumanEye/Muller Glial Cells.jpg

.

________________________________________________

Here is the context:

What do changes to the ISO dial do to a camera’s inner workings?

1 It changes the set points of the camera's exposure metering system, (EMS).

2 It changes the tone values of the output image by a combination of sensor level gain adjustment and by modifying the tone mapping of exposure to output image.

Item 2 seems controversial and will be put up in a later thread.

___________________________________________________

Back to Item 1: It changes the set points of the camera's Exposure Metering System, (EMS).

1 (a) What does the camera's exposure metering system (EMS) do?

The EMS measures (normally through the lens) the average luminance of the scene, L, and based on that measurement and a factory calibration of ISO Speed, S, the EMS recommends a pair of exposure settings, N and t (f/ and ss) to give a good image result of the exposure. The factory calibration complies with the ISO 2720:1974 Standard to establish the relationship between average scene luminance, L, and recommended camera settings, N and t, for a selected ISO setting, S, for that particular camera model.

1(b) How does changing the ISO setting knob change the EMS recommendations?

Based on the measured average scene Luminance, L, and the calibration curves installed at the factory the EMS will return recommended EV settings , N and t, according to the Exposure Metering Equation , programmed into the EMS :

View attachment 1285560

View: original size

For the above equation to be valid for a measured average scene luminance, L, the recommended camera settings, (N^2)/t , will change with changes in ISO speed, S.

Therefore for every setting of S, there is only one (N^2)/t (and only one pair of N, t) that will satisfy the equation.

That pair is the recommended exposure that the EMS offers to the photographer.

Therefore for a given average scene luminance, L, every time the ISO knob is changed, S is changed and the recommended exposure settings, N, t, (f/,ss) will change. Also, every time the average scene luminance, L, is changed (as measured by the meter) the recommended exposure settings will change, even if the ISO speed, S, is held constant.

The above paragraph (and its underlying equation) relates Exposure, (which only depends on L, N, t,) and ISO Speed, S, to give a "good image result". Two parts of Exposure, N and t, and ISO Speed, S, are sometimes crafted into an "Exposure Triangle", which should be read very critically as it is often leads to false conclusions.

This relationship of N ,t and S is easily seen when using the P(program) Mode in a camera: For a given L, adjusting S will change the exposure pairs. And, for a fixed L and S, changes in N will cause changes in t and vice versa.

[The definition of Exposure Value = Log_2 ([N^2)/t] is no accident as it puts all of the above in "stops", making relationships of exposure variables simple.]

The outputs of the EMS, the exposure settings, inform the various AutoExposure (AE) Modes [(P(rogram), A(perture), S(peed)] of the Mode dial. Some cameras have fully Automatic and/or Scene Modes which are also informed by the EMS.

It is important to note that the above is only valid when the photographer is using an AE Mode or is following the meter recommendation in M(anual) Mode. If the photographer disregards the meter recommendation (by using M(anual), Spot Metering with AE(lock), LiveView Histogram, or blinkies, etc) to inform his/her exposure decisions, the above equation is invalid and EV (and consequently exposure, Hv) is totally independent of S for any L.* Exposure compensation shifts the equation to provide a "compensated" recommended exposure pair.

* In this case the photographer is probably using independent camera settings (subject to artistic constraints regarding motion blur and DoF/Diffraction) which would optimize raw file outputs in terms of highlight protection, minimum noise and maximum PDR as opposed to recommended camera settings which would normally return an output image with a standard acceptable brightness.

___________________________

I would ask that you review and criticize the above. Please restrict your comments to the metering function of the ISO dial. I am sure that there will be a lot of other opportunities to add to the discussion of the other function of the ISO Dial!

Please no plagiarism or excessive noise, like this static:

http://www.dpreview.com/forums/post/57259249

http://www.dpreview.com/forums/post/57259530

Many thanks.

 

[xpatUSA]

1 (a) What does the camera's exposure metering system (EMS) do?

The EMS measures (normally through the lens) the average luminance of the scene, L, and based on that measurement and a factory calibration of ISO Speed, S, the EMS recommends a pair of exposure settings, N and t (f/ and ss) to give a good image result of the exposure. The factory calibration complies with the ISO 2720:1974 Standard to establish the relationship between average scene luminance, L, and recommended camera settings, N and t, for a selected ISO setting, S, for that particular camera model.

The reference is not to the Standard itself. It is to an article by Doug Kerr about photographic exposure meters, not in-camera metering.

My cameras vary from this description because I have them set to 'spot' metering meaning that only part of a scene is metered - not "the average luminance of the [whole] scene".

I think that, strictly speaking, the "EMS" measures the illuminance that is incident upon the light-measuring sensor, not the luminance of the scene or part thereof.

My cameras do not recommend "N and t" per se. Instead they show the equivalent of a needle which moves above or below zero as I wave the camera around (well, when in manual anyway). In aperture or shutter priority, the cameras only recommend one parameter, not a pair thereof.

In 'P' mode, you get what the manufacturer thinks you need - which is surely variable between mfg/models. But nobody uses 'P' mode anyway ;-)

1(b) How does changing the ISO setting knob change the EMS recommendations?

Based on the measured average scene Luminance, L, and the calibration curves installed at the factory the EMS will return recommended EV settings , N and t, according to the Exposure Metering Equation , programmed into the EMS :

View attachment 1285560

View: original size

For the above equation to be valid for a measured average scene luminance, L, the recommended camera settings, (N^2)/t , will change with changes in ISO speed, S.

Therefore for every setting of S, there is only one (N^2)/t (and only one pair of N, t) [my emphasis] that will satisfy the equation.

This reads as if there is only one unique pair for a given ISO that will "center the needle" but, if I change the settings of N and t by equal and opposite stop amounts (in manual), the needle remains centered. Before "in manual" is objected to, that is no different than messing with EC, the which action also keeps the needle centered and changes the pair values appropriately.

Example: f/8 at 1/100 sec = f/5.6 at 1/200 sec. That is to say that there is more than one pair of settings that keep the needle centered and the examples do have different values of N^2/t.

Based on the the fact that I've said nothing beyond "Exposure 101", perhaps your wording could be clarified?

In terms of readability, Tom, would it be possible to reduce the amount of bolding and underlining which some folks find distracting or which some might perceive as "shouting"?

--
Ted

 

[GeorgianBay1939]

Tom,

I like how you have posed the question ... as a matter of responding to a query from a beginning photographer seeking a deeper understanding of the ISO control. I just wonder if your proposed response might assume, so to speak, 'facts not in evidence' for a beginner asking a question like this.

But the question is simple: What does this control do?

If I am flying an airplane: What does the control column do? When I pull back on it? Ans: When I pull back on it, the associated cabling/hydraulics/electrics causes the the elevator to rotate upwards, reducing the lift of the tailplane, causing the plane to pitch upwards. Pulling back makes the kite go up.

Back to the camera: What am I doing when I adjust the "aperture" ring ? At a constant FL, I am making the entrance pupil smaller, similar to the pupil of my eye when I enter a bright room, reducing the intensity of light transiting the lens system and therefore the intensity of light illuminating the sensor. Other things being equal I am reducing the Exposure (Hv) of the camera for that shutter cycle.

My own approach would be different with an attempt to ground the issue in terms of visualization and control of the image rather than the technical aspects of how the camera operates.

For sure. You want to give advice so that the learner fully understands how to control the final image. ... so that he has something really good to work with when he loads the (image or raw) file from his camera onto his computer. The file will enable him to bring out the highlights, and minimal noise all while having no motion blur and adequate DoF with no diffraction.

In this regard, I think there are four essential points the beginner would need to know ...

Point 1. The purpose of the ISO control is to calibrate the amount of light captured by the camera to the relative tones of human perception. Our eyes routinely handle 1000-fold and larger changes in light level just from going from outdoors to indoors. Changing shutter speed and f-number, alone, can't accommodate such a range. And if you also want to use those controls creatively for motion blur and depth of field adjustment, then there is no hope. The fix is an additional control to tell the camera how to translate the amount

(intensity might be better than amount to get across the fact of area density of light energy on the sensor)

of light captured in an exposure into the tonal values of human perception. That's what the ISO control does, in a more-or-less industry standard fashion.

Point 1 is a great paragraph!!! A beauty of a summary. It explains why we have an ISO dial.

Point 2. The image processing chain begins in the camera and ends once you've produced the final image. By setting the ISO control on the camera, the camer has information needed to produce a final jpeg image, or, at a minimum, optimize the internal electronics to get as much information as possible for a raw file. It's also used to calibrate the auto-exposure mechanism of your camera. The details depend on the technology within the camera.

Also good. How do you set the ISO control to "optimize the internal electronics to get as much information as possible for a raw file." ?

More general to say exposure metering instead of auto-exposure (which is informed by exposure metering of course.)

Point 3. A high ISO means that you low light levels will correspond to brighter tones in the final image. Low light levels are intrinsically noisy, no matter how good your camera's systems might be. The more light you capture, the better.

True, but how do you capture more light? How do you set the learner up for learning the difference between brightness and exposure, how to insulate him from the falsehoods of the triangle?

Point 4. Camera's also have upper limits on how much light they can capture.

How do you convert the above statement into something that the learner can use when making exposure decisions?

I don't see why an explanation to a beginner would want to delve much deeper into this topics, particularly the math (unless you're talking to an engineer).

I agree. I only threw in the exposure metering equation because it is sooooo misunderstood ... by users of the exposure triangle who usually confuse the use of that equation.

 

[Jeff]

Tom,

I like how you have posed the question ... as a matter of responding to a query from a beginning photographer seeking a deeper understanding of the ISO control. I just wonder if your proposed response might assume, so to speak, 'facts not in evidence' for a beginner asking a question like this.

But the question is simple: What does this control do?

If I am flying an airplane: What does the control column do? When I pull back on it? Ans: When I pull back on it, the associated cabling/hydraulics/electrics causes the the elevator to rotate upwards, reducing the lift of the tailplane, causing the plane to pitch upwards. Pulling back makes the kite go up.

Back to the camera: What am I doing when I adjust the "aperture" ring ? At a constant FL, I am making the entrance pupil smaller, similar to the pupil of my eye when I enter a bright room, reducing the intensity of light transiting the lens system and therefore the intensity of light illuminating the sensor. Other things being equal I am reducing the Exposure (Hv) of the camera for that shutter cycle.

My own approach would be different with an attempt to ground the issue in terms of visualization and control of the image rather than the technical aspects of how the camera operates.

For sure. You want to give advice so that the learner fully understands how to control the final image. ... so that he has something really good to work with when he loads the (image or raw) file from his camera onto his computer. The file will enable him to bring out the highlights, and minimal noise all while having no motion blur and adequate DoF with no diffraction.

In this regard, I think there are four essential points the beginner would need to know ...

Point 1. The purpose of the ISO control is to calibrate the amount of light captured by the camera to the relative tones of human perception. Our eyes routinely handle 1000-fold and larger changes in light level just from going from outdoors to indoors. Changing shutter speed and f-number, alone, can't accommodate such a range. And if you also want to use those controls creatively for motion blur and depth of field adjustment, then there is no hope. The fix is an additional control to tell the camera how to translate the amount

(intensity might be better than amount to get across the fact of area density of light energy on the sensor)

Yes! My blunder to say amount ... intensity would definitely be more correct.

of light captured in an exposure into the tonal values of human perception. That's what the ISO control does, in a more-or-less industry standard fashion.

Point 1 is a great paragraph!!! A beauty of a summary. It explains why we have an ISO dial.

Point 2. The image processing chain begins in the camera and ends once you've produced the final image. By setting the ISO control on the camera, the camer has information needed to produce a final jpeg image, or, at a minimum, optimize the internal electronics to get as much information as possible for a raw file. It's also used to calibrate the auto-exposure mechanism of your camera. The details depend on the technology within the camera.

Also good. How do you set the ISO control to "optimize the internal electronics to get as much information as possible for a raw file." ?

More general to say exposure metering instead of auto-exposure (which is informed by exposure metering of course.)

Point 3. A high ISO means that you low light levels will correspond to brighter tones in the final image. Low light levels are intrinsically noisy, no matter how good your camera's systems might be. The more light you capture, the better.

True, but how do you capture more light? How do you set the learner up for learning the difference between brightness and exposure, how to insulate him from the falsehoods of the triangle?

Point 4. Camera's also have upper limits on how much light they can capture.

How do you convert the above statement into something that the learner can use when making exposure decisions?

I don't see why an explanation to a beginner would want to delve much deeper into this topics, particularly the math (unless you're talking to an engineer).

I agree. I only threw in the exposure metering equation because it is sooooo misunderstood ... by users of the exposure triangle who usually confuse the use of that equation.

 

[GeorgianBay1939]

This is a continuation of " Two functions of a camera's ISO dial." which ran out a few days ago.

First let me apologize for disappearing after starting a controversial thread. Health issues intervened and I learned some things other than the functions of an ISO dial. :-D No, I did not abandon DM's .Ship of Fools.

_____________________

I decided here to separate the two functions, metering and mapping, so that we could drill down to clarify each function.

_____________________

An email (from an accomplished photographer) said, "I don't know and I don't care." So I should tell you why I started the thread in the first place:

A couple of weeks ago I heard a well-respected, experienced, ex-film, photographer answer a question from a youngster who had just added his first full frame camera to his collection of gear. The youngster (in his 50s) was a beginning raw shooter who was learning to stretch his envelope. Good science background and a bit of a perfectionist. Kids grown up and he was developing a hobby.

In answer to the question, "What does the ISO dial REALLY do when I change it?",

this is what the older photographer said,

"You are changing one of the three brightness variables so that you'll have to change either f/ or ss to get a proper exposure. The ISO dial is calibrated to give you optimum exposure in varying light conditions. Turn ISO down in bright conditions, up in dark conditions. Increasing the ISO setting increases the sensor's gain to compensate for lack of brightness in the scene."

After thinking about it, I realized that the above explanation would limit the growth of the youngster and contemplated how I would answer such a question when asked by a developing (!) photographer.

"Two functions of a camera's ISO dial" was my first attempt at forming such an answer. After reading the thread I realize that both parts, metering and mapping, of the answer needed improvement. Since the metering portion seemed to be almost ok, I would ask that you limit your comments to the metering portion of the answer in this thread.

I will start a thread to deal with the (much more controversial) mapping portion later.

That the transfer-function of a non-linear system (with RGB color-space rendered tone-curve transfer functions and any associated gamma-correction applied ... when mapping image-data ... into such RGB color-spaces) does not constitute a linear system (having one constant scale-factor) is not controversial.

I agree with you. I use the word to describe a lot of the chit-chat here. A lot of folks don't fully understand "tone mapping" ... including me. And we've all been pummeled by every manufacturer calling that knob the "ISO sensitivity" knob. ( even Panasonic!)

All analytical assumptions drawn surrounding non-linear output image-data, as well as in-camera metering system behavior in relation to such non-linear output image-data, require attention be paid to "mappings" !

I thought that Jeff did a good job of summarizing all this in his point 1. You?

From the display device, and into the vast and mysterious pre-optic ganglia (having an overall consistency of that of tofu, or of toothpaste) we conceptually meander - but between our (perhaps pristinely) modeled hardware and the truly ineffable psycho-visual impressions and interpretations that we in some manner formulate, there may exist legions of (seemingly inscrutably) structured and arranged "bundles of wiring":

Source: http://www.detectingdesign.com/images/HumanEye/Muller Glial Cells.jpg

.

What I really like is the discussion of how we learn to see. Most of what we see now, in our elder years, was learned eons ago and are being recalled, stimulated by little bits of information from those magical ganglia. Magick!

That photo reminded me of epithelial tissue that has caused me some grief over the last decade, especially over the last few days. Prognosis is good, though. :-D

Did I get the metering part right?

Thanks, DM.

________________________________________________

Here is the context:

What do changes to the ISO dial do to a camera’s inner workings?

1 It changes the set points of the camera's exposure metering system, (EMS).

2 It changes the tone values of the output image by a combination of sensor level gain adjustment and by modifying the tone mapping of exposure to output image.

Item 2 seems controversial and will be put up in a later thread.

___________________________________________________

Back to Item 1: It changes the set points of the camera's Exposure Metering System, (EMS).

1 (a) What does the camera's exposure metering system (EMS) do?

The EMS measures (normally through the lens) the average luminance of the scene, L, and based on that measurement and a factory calibration of ISO Speed, S, the EMS recommends a pair of exposure settings, N and t (f/ and ss) to give a good image result of the exposure. The factory calibration complies with the ISO 2720:1974 Standard to establish the relationship between average scene luminance, L, and recommended camera settings, N and t, for a selected ISO setting, S, for that particular camera model.

1(b) How does changing the ISO setting knob change the EMS recommendations?

Based on the measured average scene Luminance, L, and the calibration curves installed at the factory the EMS will return recommended EV settings , N and t, according to the Exposure Metering Equation , programmed into the EMS :

View attachment 1285560

View: original size

For the above equation to be valid for a measured average scene luminance, L, the recommended camera settings, (N^2)/t , will change with changes in ISO speed, S.

Therefore for every setting of S, there is only one (N^2)/t (and only one pair of N, t) that will satisfy the equation.

That pair is the recommended exposure that the EMS offers to the photographer.

Therefore for a given average scene luminance, L, every time the ISO knob is changed, S is changed and the recommended exposure settings, N, t, (f/,ss) will change. Also, every time the average scene luminance, L, is changed (as measured by the meter) the recommended exposure settings will change, even if the ISO speed, S, is held constant.

The above paragraph (and its underlying equation) relates Exposure, (which only depends on L, N, t,) and ISO Speed, S, to give a "good image result". Two parts of Exposure, N and t, and ISO Speed, S, are sometimes crafted into an "Exposure Triangle", which should be read very critically as it is often leads to false conclusions.

This relationship of N ,t and S is easily seen when using the P(program) Mode in a camera: For a given L, adjusting S will change the exposure pairs. And, for a fixed L and S, changes in N will cause changes in t and vice versa.

[The definition of Exposure Value = Log_2 ([N^2)/t] is no accident as it puts all of the above in "stops", making relationships of exposure variables simple.]

The outputs of the EMS, the exposure settings, inform the various AutoExposure (AE) Modes [(P(rogram), A(perture), S(peed)] of the Mode dial. Some cameras have fully Automatic and/or Scene Modes which are also informed by the EMS.

It is important to note that the above is only valid when the photographer is using an AE Mode or is following the meter recommendation in M(anual) Mode. If the photographer disregards the meter recommendation (by using M(anual), Spot Metering with AE(lock), LiveView Histogram, or blinkies, etc) to inform his/her exposure decisions, the above equation is invalid and EV (and consequently exposure, Hv) is totally independent of S for any L.* Exposure compensation shifts the equation to provide a "compensated" recommended exposure pair.

* In this case the photographer is probably using independent camera settings (subject to artistic constraints regarding motion blur and DoF/Diffraction) which would optimize raw file outputs in terms of highlight protection, minimum noise and maximum PDR as opposed to recommended camera settings which would normally return an output image with a standard acceptable brightness.

___________________________

I would ask that you review and criticize the above. Please restrict your comments to the metering function of the ISO dial. I am sure that there will be a lot of other opportunities to add to the discussion of the other function of the ISO Dial!

Please no plagiarism or excessive noise, like this static:

http://www.dpreview.com/forums/post/57259249

http://www.dpreview.com/forums/post/57259530

Many thanks.

--
Tom
The best part of growing old is having the opportunity to do so.

Brtthome's Blog

photos, prose, prattle and possibilities

brtthome.wordpress.com

 

[GeorgianBay1939]

1 (a) What does the camera's exposure metering system (EMS) do?

The EMS measures (normally through the lens) the average luminance of the scene, L, and based on that measurement and a factory calibration of ISO Speed, S, the EMS recommends a pair of exposure settings, N and t (f/ and ss) to give a good image result of the exposure. The factory calibration complies with the ISO 2720:1974 Standard to establish the relationship between average scene luminance, L, and recommended camera settings, N and t, for a selected ISO setting, S, for that particular camera model.

The reference is not to the Standard itself. It is to an article by Doug Kerr about photographic exposure meters, not in-camera metering.

Yes, but the Background in Kerr's article is a concise explanation of the process. The best that I've come across. Can you give me a better one? Can you give me something that shows the difference between in-camera metering and reflected light metering?

My cameras vary from this description because I have them set to 'spot' metering meaning that only part of a scene is metered - not "the average luminance of the [whole] scene".

Good point. I do the same. Perhaps use "average luminance of the metered scene". Or maybe a blurb on the difference between the metering modes.

I think that, strictly speaking, the "EMS" measures the illuminance that is incident upon the light-measuring sensor, not the luminance of the scene or part thereof.

Yes, and from that measurement, EMS infers the luminance of (the part of) the scene being metered. I'm nervous to include "incident" in the context of metering for obvious reasons.

My cameras do not recommend "N and t" per se. Instead they show the equivalent of a needle which moves above or below zero as I wave the camera around (well, when in manual anyway).

I have something similar in M(anual) Mode.

In aperture or shutter priority, the cameras only recommend one parameter, not a pair thereof.

Yes, when a priority is selected, the camera recommends the other parameter.

In 'P' mode, you get what the manufacturer thinks you need - which is surely variable between mfg/models. But nobody uses 'P' mode anyway ;-)

I only use it when I am using flash, because I don't know anything about flash!

1(b) How does changing the ISO setting knob change the EMS recommendations?

Based on the measured average scene Luminance, L, and the calibration curves installed at the factory the EMS will return recommended EV settings , N and t, according to the Exposure Metering Equation , programmed into the EMS :

View attachment 1285560

View: original size

For the above equation to be valid for a measured average scene luminance, L, the recommended camera settings, (N^2)/t , will change with changes in ISO speed, S.

Therefore for every setting of S, there is only one (N^2)/t (and only one pair of N, t) [my emphasis] that will satisfy the equation.

This reads as if there is only one unique pair for a given ISO that will "center the needle" but, if I change the settings of N and t by equal and opposite stop amounts (in manual), the needle remains centered. Before "in manual" is objected to, that is no different than messing with EC, the which action also keeps the needle centered and changes the pair values appropriately.

Yeah, this is difficult because I didn't want to use Exposure Value (the properly defined way, not the way that is used in most of the "literature" where it is equated with, LV, Light Value.) as it adds another layer of complexity. I will have to mull over this.

Example: f/8 at 1/100 sec = f/5.6 at 1/200 sec. That is to say that there is more than one pair of settings that keep the needle centered and the examples do have different values of N^2/t.

Not really. When you look at detail 2^2.5 = 5.656854249492381 which, when squared is 32.000000, close enough for government work!!!

Based on the the fact that I've said nothing beyond "Exposure 101", perhaps your wording could be clarified?

Good thing to work on. Helps to clarify my thinking and to get rid of all the crapola that I've "learned" on these various forums!!! :-D :-D

In terms of readability, Tom, would it be possible to reduce the amount of bolding and underlining which some folks find distracting or which some might perceive as "shouting"?

--
Ted

For sure! Thanks a lot for your good feedback, Ted.

--
Tom
The best part of growing old is having the opportunity to do so.

Brtthome's Blog

photos, prose, prattle and possibilities

brtthome.wordpress.com

 

[GeorgianBay1939]

Tom,

I like how you have posed the question ... as a matter of responding to a query from a beginning photographer seeking a deeper understanding of the ISO control. I just wonder if your proposed response might assume, so to speak, 'facts not in evidence' for a beginner asking a question like this.

But the question is simple: What does this control do?

If I am flying an airplane: What does the control column do? When I pull back on it? Ans: When I pull back on it, the associated cabling/hydraulics/electrics causes the the elevator to rotate upwards, reducing the lift of the tailplane, causing the plane to pitch upwards. Pulling back makes the kite go up.

Back to the camera: What am I doing when I adjust the "aperture" ring ? At a constant FL, I am making the entrance pupil smaller, similar to the pupil of my eye when I enter a bright room, reducing the intensity of light transiting the lens system and therefore the intensity of light illuminating the sensor. Other things being equal I am reducing the Exposure (Hv) of the camera for that shutter cycle.

My own approach would be different with an attempt to ground the issue in terms of visualization and control of the image rather than the technical aspects of how the camera operates.

For sure. You want to give advice so that the learner fully understands how to control the final image. ... so that he has something really good to work with when he loads the (image or raw) file from his camera onto his computer. The file will enable him to bring out the highlights, and minimal noise all while having no motion blur and adequate DoF with no diffraction.

In this regard, I think there are four essential points the beginner would need to know ...

Point 1. The purpose of the ISO control is to calibrate the amount of light captured by the camera to the relative tones of human perception. Our eyes routinely handle 1000-fold and larger changes in light level just from going from outdoors to indoors. Changing shutter speed and f-number, alone, can't accommodate such a range. And if you also want to use those controls creatively for motion blur and depth of field adjustment, then there is no hope. The fix is an additional control to tell the camera how to translate the amount

(intensity might be better than amount to get across the fact of area density of light energy on the sensor)

Yes! My blunder to say amount ... intensity would definitely be more correct.

of light captured in an exposure into the tonal values of human perception. That's what the ISO control does, in a more-or-less industry standard fashion.

Point 1 is a great paragraph!!! A beauty of a summary. It explains why we have an ISO dial.

Point 2. The image processing chain begins in the camera and ends once you've produced the final image. By setting the ISO control on the camera, the camer has information needed to produce a final jpeg image, or, at a minimum, optimize the internal electronics to get as much information as possible for a raw file. It's also used to calibrate the auto-exposure mechanism of your camera. The details depend on the technology within the camera.

Also good. How do you set the ISO control to "optimize the internal electronics to get as much information as possible for a raw file." ?

More general to say exposure metering instead of auto-exposure (which is informed by exposure metering of course.)

Point 3. A high ISO means that you low light levels will correspond to brighter tones in the final image. Low light levels are intrinsically noisy, no matter how good your camera's systems might be. The more light you capture, the better.

True, but how do you capture more light? How do you set the learner up for learning the difference between brightness and exposure, how to insulate him from the falsehoods of the triangle?

Point 4. Camera's also have upper limits on how much light they can capture.

How do you convert the above statement into something that the learner can use when making exposure decisions?

I don't see why an explanation to a beginner would want to delve much deeper into this topics, particularly the math (unless you're talking to an engineer).

I agree. I only threw in the exposure metering equation because it is sooooo misunderstood ... by users of the exposure triangle who usually confuse the use of that equation.

 

[Detail Man]

This is a continuation of " Two functions of a camera's ISO dial." which ran out a few days ago.

First let me apologize for disappearing after starting a controversial thread. Health issues intervened and I learned some things other than the functions of an ISO dial. :-D No, I did not abandon DM's .Ship of Fools.

_____________________

I decided here to separate the two functions, metering and mapping, so that we could drill down to clarify each function.

_____________________

An email (from an accomplished photographer) said, "I don't know and I don't care." So I should tell you why I started the thread in the first place:

A couple of weeks ago I heard a well-respected, experienced, ex-film, photographer answer a question from a youngster who had just added his first full frame camera to his collection of gear. The youngster (in his 50s) was a beginning raw shooter who was learning to stretch his envelope. Good science background and a bit of a perfectionist. Kids grown up and he was developing a hobby.

In answer to the question, "What does the ISO dial REALLY do when I change it?",

this is what the older photographer said,

"You are changing one of the three brightness variables so that you'll have to change either f/ or ss to get a proper exposure. The ISO dial is calibrated to give you optimum exposure in varying light conditions. Turn ISO down in bright conditions, up in dark conditions. Increasing the ISO setting increases the sensor's gain to compensate for lack of brightness in the scene."

After thinking about it, I realized that the above explanation would limit the growth of the youngster and contemplated how I would answer such a question when asked by a developing (!) photographer.

"Two functions of a camera's ISO dial" was my first attempt at forming such an answer. After reading the thread I realize that both parts, metering and mapping, of the answer needed improvement. Since the metering portion seemed to be almost ok, I would ask that you limit your comments to the metering portion of the answer in this thread.

I will start a thread to deal with the (much more controversial) mapping portion later.

That the transfer-function of a non-linear system (with RGB color-space rendered tone-curve transfer functions and any associated gamma-correction applied ... when mapping image-data ... into such RGB color-spaces) does not constitute a linear system (having one constant scale-factor) is not controversial.

I agree with you. I use the word to describe a lot of the chit-chat here. A lot of folks don't fully understand "tone mapping" ... including me. And we've all been pummeled by every manufacturer calling that knob the "ISO sensitivity" knob. ( even Panasonic!)

All analytical assumptions drawn surrounding non-linear output image-data, as well as in-camera metering system behavior in relation to such non-linear output image-data, require attention be paid to "mappings" !

I thought that Jeff did a good job of summarizing all this in his point 1. You?

"His point 1" where ? Jeff seems to get the important things in due course. It's non-linear "mappings".

From the display device, and into the vast and mysterious pre-optic ganglia (having an overall consistency of that of tofu, or of toothpaste) we conceptually meander - but between our (perhaps pristinely) modeled hardware and the truly ineffable psycho-visual impressions and interpretations that we in some manner formulate, there may exist legions of (seemingly inscrutably) structured and arranged "bundles of wiring":

Source: http://www.detectingdesign.com/images/HumanEye/Muller Glial Cells.jpg

.

What I really like is the discussion of how we learn to see. Most of what we see now, in our elder years, was learned eons ago and are being recalled, stimulated by little bits of information from those magical ganglia. Magick!

That photo reminded me of epithelial tissue that has caused me some grief over the last decade, especially over the last few days. Prognosis is good, though. :-D

Did I get the metering part right?

Thanks, DM.

As I attempted to communicate in the PMs that I sent to you shortly after your original post on your previous thread on these subjects, ...

... all in-camera systems, and all RGB color-space mapped information displayed on any/all on-camera indicators [EVF/LCD, so-called "Exposure" meters (not), Histograms, Blinkies, Zebra Stripes, etc.] relate to in-camera encoded JPGs ...

... and are thus, in essence, operationally unrelated to RAW-level recording (where the image-sensor composite photosite output is essentially a linear function, and the only universal issue of (operational) interest is whether any of the (output) DNs of RAW RGB channels constitute results of non-linear clipping.

________________________________________________

Here is the context:

What do changes to the ISO dial do to a camera’s inner workings?

1 It changes the set points of the camera's exposure metering system, (EMS).

2 It changes the tone values of the output image by a combination of sensor level gain adjustment and by modifying the tone mapping of exposure to output image.

Item 2 seems controversial and will be put up in a later thread.

___________________________________________________

Back to Item 1: It changes the set points of the camera's Exposure Metering System, (EMS).

1 (a) What does the camera's exposure metering system (EMS) do?

The EMS measures (normally through the lens) the average luminance of the scene, L, and based on that measurement and a factory calibration of ISO Speed, S, the EMS recommends a pair of exposure settings, N and t (f/ and ss) to give a good image result of the exposure. The factory calibration complies with the ISO 2720:1974 Standard to establish the relationship between average scene luminance, L, and recommended camera settings, N and t, for a selected ISO setting, S, for that particular camera model.

1(b) How does changing the ISO setting knob change the EMS recommendations?

Based on the measured average scene Luminance, L, and the calibration curves installed at the factory the EMS will return recommended EV settings , N and t, according to the Exposure Metering Equation , programmed into the EMS :

View attachment 1285560

View: original size

For the above equation to be valid for a measured average scene luminance, L, the recommended camera settings, (N^2)/t , will change with changes in ISO speed, S.

Therefore for every setting of S, there is only one (N^2)/t (and only one pair of N, t) that will satisfy the equation.

That pair is the recommended exposure that the EMS offers to the photographer.

Therefore for a given average scene luminance, L, every time the ISO knob is changed, S is changed and the recommended exposure settings, N, t, (f/,ss) will change. Also, every time the average scene luminance, L, is changed (as measured by the meter) the recommended exposure settings will change, even if the ISO speed, S, is held constant.

The above paragraph (and its underlying equation) relates Exposure, (which only depends on L, N, t,) and ISO Speed, S, to give a "good image result". Two parts of Exposure, N and t, and ISO Speed, S, are sometimes crafted into an "Exposure Triangle", which should be read very critically as it is often leads to false conclusions.

This relationship of N ,t and S is easily seen when using the P(program) Mode in a camera: For a given L, adjusting S will change the exposure pairs. And, for a fixed L and S, changes in N will cause changes in t and vice versa.

[The definition of Exposure Value = Log_2 ([N^2)/t] is no accident as it puts all of the above in "stops", making relationships of exposure variables simple.]

The outputs of the EMS, the exposure settings, inform the various AutoExposure (AE) Modes [(P(rogram), A(perture), S(peed)] of the Mode dial. Some cameras have fully Automatic and/or Scene Modes which are also informed by the EMS.

It is important to note that the above is only valid when the photographer is using an AE Mode or is following the meter recommendation in M(anual) Mode. If the photographer disregards the meter recommendation (by using M(anual), Spot Metering with AE(lock), LiveView Histogram, or blinkies, etc) to inform his/her exposure decisions, the above equation is invalid and EV (and consequently exposure, Hv) is totally independent of S for any L.* Exposure compensation shifts the equation to provide a "compensated" recommended exposure pair.

* In this case the photographer is probably using independent camera settings (subject to artistic constraints regarding motion blur and DoF/Diffraction) which would optimize raw file outputs in terms of highlight protection, minimum noise and maximum PDR as opposed to recommended camera settings which would normally return an output image with a standard acceptable brightness.

___________________________

I would ask that you review and criticize the above. Please restrict your comments to the metering function of the ISO dial. I am sure that there will be a lot of other opportunities to add to the discussion of the other function of the ISO Dial!

Please no plagiarism or excessive noise, like this static:

http://www.dpreview.com/forums/post/57259249

http://www.dpreview.com/forums/post/57259530

Many thanks.

 

[xpatUSA]

GeorgianBay1939 wrote:Therefore for every setting of S, there is only one (N^2)/t (and only one pair of N, t) [my emphasis] that will satisfy the equation.

This reads as if there is only one unique pair for a given ISO that will "center the needle" but, if I change the settings of N and t by equal and opposite stop amounts (in manual), the needle remains centered. Before "in manual" is objected to, that is no different than messing with EC, the which action also keeps the needle centered and changes the pair values appropriately.

Yeah, this is difficult because I didn't want to use Exposure Value (the properly defined way, not the way that is used in most of the "literature" where it is equated with, LV, Light Value.) as it adds another layer of complexity. I will have to mull over this.

Example: f/8 at 1/100 sec = f/5.6 at 1/200 sec. That is to say that there is more than one pair of settings that keep the needle centered and the examples do have different values of N^2/t.

Not really. When you look at detail 2^2.5 = 5.656854249492381 which, when squared is 32.000000, close enough for government work!!!

Tom, I didn't understand that rebuttal at all. Are you saying that f/8 at 1/100 sec and f/5.6 at 1/200 sec are not the same exposure?

Are we talking at cross-purposes and where did "detail 2^2.5" come from?

Thanks a lot for your good feedback, Ted.

You're most welcome

 

[Jeff]

<snip>

I thought that Jeff did a good job of summarizing all this in his point 1. You?

"His point 1" where ? Jeff seems to get the important things in due course. It's non-linear "mappings".

<snip>

"in due course" .... hmmm ..... I guess that's good.

In any event, Point 1. refers to my first reply to the OP. Here's a cut and paste (with a few edits) :

Point 1. The purpose of the ISO control is to calibrate the exposure captured by the camera to the relative tones of human perception. Our eyes routinely handle 1000-fold and larger changes in light intensity just from going from outdoors to indoors. Given the state of current technology, changing a camera's shutter speed and f-number, alone, can't accommodate such a large range. If you also want to use those settings creatively for controlling motion blur and depth of field, then there is no hope. The fix is an additional control that tells the camera how the captured exposure will map to tones in the final image. That's what the ISO control does in a more-or-less industry standard fashion.

--
Jeff
http://www.flickr.com/photos/jck_photos/sets/

 

[GeorgianBay1939]

GeorgianBay1939 wrote:Therefore for every setting of S, there is only one (N^2)/t (and only one pair of N, t) [my emphasis] that will satisfy the equation.

This reads as if there is only one unique pair for a given ISO that will "center the needle" but, if I change the settings of N and t by equal and opposite stop amounts (in manual), the needle remains centered. Before "in manual" is objected to, that is no different than messing with EC, the which action also keeps the needle centered and changes the pair values appropriately.

Yeah, this is difficult because I didn't want to use Exposure Value (the properly defined way, not the way that is used in most of the "literature" where it is equated with, LV, Light Value.) as it adds another layer of complexity. I will have to mull over this.

Example: f/8 at 1/100 sec = f/5.6 at 1/200 sec. That is to say that there is more than one pair of settings that keep the needle centered and the examples do have different values of N^2/t.

You say above ... "the examples do have different values of (N^2)/t." I say that they have the same values. First example: (8^2)/t =64 x 100 =6.4K Second example: (5.6^2)/t is an approximation of (5.6569..... ^2)/t = 32 x 200 = 6.4K First example = second example.

2^0; 2^.5; 2^1; 2^1.5; 2^2; 2^2.5; 2^3; 2^3.5; 2^4; 2^4.5 <<< f/ sequence

1 ;1.414; 2 ;2.828 ; 4 ;5.6569; 8 ;11.314; 16 ; 22.63 <<< Precise

1 ;1.4 ; 2 ;2.8 ; 4 ; 5.6 ; 8 ; 11.3 ; 16 ; 22 <<< Working ("gov't work")

Not really. When you look at detail 2^2.5 = 5.656854249492381 which, when squared is 32.000000, close enough for government work!!!

Tom, I didn't understand that rebuttal at all. Are you saying that f/8 at 1/100 sec and f/5.6 at 1/200 sec are not the same exposure?

I say that they are the same.

Are we talking at cross-purposes and where did "detail 2^2.5" come from?

You'll laugh when you see above. My calculator has Log base two [argument ] and 2 ^ [arg] .... very handy when playing with stops!!

Thanks a lot for your good feedback, Ted.

You're most welcome

 

[GeorgianBay1939]

<<<< Big Snip>>>>

That the transfer-function of a non-linear system (with RGB color-space rendered tone-curve transfer functions and any associated gamma-correction applied ... when mapping image-data ... into such RGB color-spaces) does not constitute a linear system (having one constant scale-factor) is not controversial.

I agree with you. I use the word to describe a lot of the chit-chat here. A lot of folks don't fully understand "tone mapping" ... including me. And we've all been pummeled by every manufacturer calling that knob the "ISO sensitivity" knob. ( even Panasonic!)

All analytical assumptions drawn surrounding non-linear output image-data, as well as in-camera metering system behavior in relation to such non-linear output image-data, require attention be paid to "mappings" !

I thought that Jeff did a good job of summarizing all this in his point 1. You?

"His point 1" where ? Jeff seems to get the important things in due course. It's non-linear "mappings".

Here: http://www.dpreview.com/forums/post/57291814

Where he says this:

In this regard, I think there are four essential points the beginner would need to know ...

Point 1. The purpose of the ISO control is to calibrate the amount of light captured by the camera to the relative tones of human perception. Our eyes routinely handle 1000-fold and larger changes in light level just from going from outdoors to indoors. Changing shutter speed and f-number, alone, can't accommodate such a range. And if you also want to use those controls creatively for motion blur and depth of field adjustment, then there is no hope. The fix is an additional control to tell the camera how to translate the amount

(intensity might be better than amount to get across the fact of area density of light energy on the sensor)

of light captured in an exposure into the tonal values of human perception. That's what the ISO control does, in a more-or-less industry standard fashion.

<<<Big snip>>>

From the display device, and into the vast and mysterious pre-optic ganglia (having an overall consistency of that of tofu, or of toothpaste) we conceptually meander - but between our (perhaps pristinely) modeled hardware and the truly ineffable psycho-visual impressions and interpretations that we in some manner formulate, there may exist legions of (seemingly inscrutably) structured and arranged "bundles of wiring":

Source: http://www.detectingdesign.com/images/HumanEye/Muller Glial Cells.jpg

.

What I really like is the discussion of how we learn to see. Most of what we see now, in our elder years, was learned eons ago and are being recalled, stimulated by little bits of information from those magical ganglia. Magick!

That photo reminded me of epithelial tissue that has caused me some grief over the last decade, especially over the last few days. Prognosis is good, though. :-D

Did I get the metering part right?

Thanks, DM.

As I attempted to communicate in the PMs that I sent to you shortly after your original post on your previous thread on these subjects, ...

I have been somnolent most of the day. Unable to get through much new stuff. Still recuperating from a mule kick.

... all in-camera systems, and all RGB color-space mapped information displayed on any/all on-camera indicators [EVF/LCD, so-called "Exposure" meters (not), Histograms, Blinkies, Zebra Stripes, etc.] relate to in-camera encoded JPGs ...

... and are thus, in essence, operationally unrelated to RAW-level recording (where the image-sensor composite photosite output is essentially a linear function, and the only universal issue of (operational) interest is whether any of the (output) DNs of RAW RGB channels constitute results of non-linear clipping.

Very instructive. Supports the case for some sort of readout of DNs with and indicator of clipping for shooting RAW.

Clipping could come from two source, eh? Oversaturation of the pixel (too much exposure) and/or some sort of "overflow" resulting from too much gain during the analogue to digital conversion?? Right?

If right, the indicator should discriminate between the two sources of clipping so that the photog knows to either reduce exposure (N, t) or to reduce gain, (ISO dial). Right?



<<< Big Snip>>>>>

--
Tom
The best part of growing old is having the opportunity to do so.

Brtthome's Blog

photos, prose, prattle and possibilities

brtthome.wordpress.com

 

[Detail Man]

<snip>

I thought that Jeff did a good job of summarizing all this in his point 1. You?

"His point 1" where ? Jeff seems to get the important things in due course. It's non-linear "mappings".

<snip>

"in due course" .... hmmm ..... I guess that's good.

Well, from the immortal utterances of Emperor "Biggest D_____", "Wome wasn't built in a day" ...

In any event, Point 1. refers to my first reply to the OP. Here's a cut and paste (with a few edits):

Point 1. The purpose of the ISO control is to calibrate the exposure captured by the camera to the relative tones of human perception.

ISO setting-values have nothing to do with affecting (sensor-level) radiometric/photometric Exposure. (Only) if enclosed in a (metering-system or user-controlled) adjustment regimen can such be validly said.

Our eyes routinely handle 1000-fold and larger changes in light intensity just from going from outdoors to indoors. Given the state of current technology, changing a camera's shutter speed and f-number, alone, can't accommodate such a large range. If you also want to use those settings creatively for controlling motion blur and depth of field, then there is no hope. The fix is an additional control that tells the camera how the captured exposure will map to tones in the final image. That's what the ISO control does in a more-or-less industry standard fashion.

That is about user needs for (preview/review) on-camera "indicators" in recording in-camera JPGs - but has little to do with recording in RAW format (where all that matters is the practical question, "am I clipping ?").

.

None of any of the above (as post-edited, at least) mentions or relates to non-linear system "mappings".

.

Best Regards,

DM

 

[xpatUSA]

Tom, I didn't understand that rebuttal at all. Are you saying that f/8 at 1/100 sec and f/5.6 at 1/200 sec are not the same exposure?

I say that they are the same.

Rats! You spotted my deliberate mistake of dividing N^2 by e.g. 100, not 1/100 :-(

I now say they are the same without prejudice to anything else I said, yer Honor!

 

[GeorgianBay1939]

<snip>

I thought that Jeff did a good job of summarizing all this in his point 1. You?

"His point 1" where ? Jeff seems to get the important things in due course. It's non-linear "mappings".

<snip>

"in due course" .... hmmm ..... I guess that's good.

In any event, Point 1. refers to my first reply to the OP. Here's a cut and paste (with a few edits) :

Point 1. The purpose of the ISO control is to calibrate the exposure captured by the camera to the relative tones of human perception. Our eyes routinely handle 1000-fold and larger changes in light intensity just from going from outdoors to indoors. Given the state of current technology, changing a camera's shutter speed and f-number, alone, can't accommodate such a large range.

Might be useful to reinforce that the above relates to exposure, Hv.

If you also want to use those settings creatively for controlling motion blur and depth of field, then there is no hope.



The fix is an additional control that tells the camera how the captured exposure will map to tones in the final image. That's what the ISO control does in a more-or-less industry standard fashion.

Relate above to gollywops flow chart? ... this one:

Should gollywop reconsider his labels? Brightening or Mapping?

--
Jeff
http://www.flickr.com/photos/jck_photos/sets/

What makes Point 1 so attractive is that it does not refer to film and all that archaic stuff that creates false analogies for youngsters who have been brought up in a digital age.... or for oldsters (like me) who know very little about film science.

--
Tom
The best part of growing old is having the opportunity to do so.

Brtthome's Blog

photos, prose, prattle and possibilities

brtthome.wordpress.com

 

[GeorgianBay1939]

<snip>

I thought that Jeff did a good job of summarizing all this in his point 1. You?

"His point 1" where ? Jeff seems to get the important things in due course. It's non-linear "mappings".

<snip>

"in due course" .... hmmm ..... I guess that's good.

Well, from the immortal utterances of Emperor "Biggest D_____", "Wome wasn't built in a day" ...

In any event, Point 1. refers to my first reply to the OP. Here's a cut and paste (with a few edits):

Point 1. The purpose of the ISO control is to calibrate the exposure captured by the camera to the relative tones of human perception.

ISO setting-values have nothing to do with affecting (sensor-level) radiometric/photometric Exposure. (Only) if enclosed in a (metering-system or user-controlled) adjustment regimen can such be validly said.

Our eyes routinely handle 1000-fold and larger changes in light intensity just from going from outdoors to indoors. Given the state of current technology, changing a camera's shutter speed and f-number, alone, can't accommodate such a large range. If you also want to use those settings creatively for controlling motion blur and depth of field, then there is no hope. The fix is an additional control that tells the camera how the captured exposure will map to tones in the final image. That's what the ISO control does in a more-or-less industry standard fashion.

That is about user needs for (preview/review) on-camera "indicators" in recording in-camera JPGs - but has little to do with recording in RAW format (where all that matters is the practical question, "am I clipping ?").

.

None of any of the above (as post-edited, at least) mentions or relates to non-linear system "mappings".

.

Best Regards,

DM

You are a tough taskmaster DM!!!

"calibrate the exposure" is a bad phrase. I didn't read it that way, I read into it ... "calibrate the results of the exposure".

 

[xpatUSA]

<gigantic snip>
Very instructive. Supports the case for some sort of readout of DNs with and indicator of clipping for shooting RAW.

Clipping could come from two source, eh? Oversaturation of the pixel (too much exposure) and/or some sort of "overflow" resulting from too much gain during the analogue to digital conversion?? Right?

Excluding most Sigma cameras - which do not have VGAs between the sensor and the ADCs. Those few that do are unpopular with users, due to "lack of headroom".

 

[Detail Man]

<snip>

I thought that Jeff did a good job of summarizing all this in his point 1. You?

"His point 1" where ? Jeff seems to get the important things in due course. It's non-linear "mappings".

<snip>

"in due course" .... hmmm ..... I guess that's good.

Well, from the immortal utterances of Emperor "Biggest D_____", "Wome wasn't built in a day".

In any event, Point 1. refers to my first reply to the OP. Here's a cut and paste (with a few edits):

Point 1. The purpose of the ISO control is to calibrate the exposure captured by the camera to the relative tones of human perception.

ISO setting-values have nothing to do with affecting (sensor-level) radiometric/photometric Exposure. (Only) if enclosed in a (metering-system or user-controlled) adjustment regimen can such be validly said.

Our eyes routinely handle 1000-fold and larger changes in light intensity just from going from outdoors to indoors. Given the state of current technology, changing a camera's shutter speed and f-number, alone, can't accommodate such a large range. If you also want to use those settings creatively for controlling motion blur and depth of field, then there is no hope. The fix is an additional control that tells the camera how the captured exposure will map to tones in the final image. That's what the ISO control does in a more-or-less industry standard fashion.

That is about user needs for (preview/review) on-camera "indicators" in recording in-camera JPGs - but has little to do with recording in RAW format (where all that matters is the practical question, "am I clipping ?").

.

None of any of the above (as post-edited, at least) mentions or relates to non-linear system "mappings".

You are a tough taskmaster DM !!!

You vill schtep. ! Schtepping, schtepping, schtepping vith "Heinrich" ! ...

"calibrate the exposure" is a bad phrase. I didn't read it that way, I read into it ... "calibratethe resultsof the exposure".

This is relatively straightforward. Do you understand the subject of dependent (output) variables (of a system, or a symbolic mathematical identity describing a system) being a *mathematical function* of (one or more, simultaneously acting) independent (input) variable(s) of that mathematical function ?

(Linearly) recorded RAW-level RGB channel DN values are a function of (sensor-level) Exposure. RAW-level RGB channel DN values are but one of (potentially several) distinct and separate independent-variables that may (via electro-robotic-metering, or user adjustment discretion on the camera-lens system user-controls) form a (negative-signed, it turns out) functional "feedback-loop"(entirely "JPG-brightness-centric" in nature).

The first rule of feedback-loop ("servo-system") analysis is that closed feedback-loops *must be opened* in order to be able to coherently and thus meaningfully analyze the system (as opposed to chasing one's own conceptual tail, without having verifiable clues as to the nature of the system causalities involved).