Playing with my Sekonic L-398A calculator dial

[Ted Cousins]

The dial pointer is set to about 0.8 of the way between 40 fc and 80 fc. So that is about 2^(0.8) x 40 ≅ 70 fc. 70 fc x 10.764 ≅ 753 lx.

[edit]From the literature, N²/t=E.S/340 = 221.612 and log₂(221.612) = 7.792 Ev ... not too far off, eh ...

 

[Doug Kerr]

Hi, Ted,

View attachment 14396

The dial pointer is set to about 0.8 of the way between 40 fc and 80 fc. So that is about 2^(0.8) x 40 ≅ 70 fc. 70 fc x 10.764 ≅ 753 lx.

From the literature, N²/t=L.S/340 = 221.612 and log₂(221.612) = 7.792 Ev ... not too far off, eh ...

The Ev scale in the picture seems to read about 7.55 (an eyeball estimate).

But on the main dials, the f-number corresponding to 1 sec for t seems to be about f/12.9 (from another eyeball estimate), which would make the Ev about 7.38.

So the "circular slide rule" scales are perhaps not precise.

Thank you for that analysis.

Best regards,

Doug

 

[Ted Cousins]

Hi, Ted,

And the Ev scale seems to read about 7.6.

Thank you for that.

Best regards,

Doug

Oops fixed my deliberate mistake E.S/340 not L.S/340

 

[Doug Kerr]

Hi, Ted,

Oops fixed my deliberate mistake E.S/340 not L.S/340

I knew what you meant.

Best regards,

Doug

 

[Doug Kerr]

Evaluation of the exposure calculator of the Sekonic L-398A here suggests that the value of C baked into the calculator is about 328.

That is based on the following observations:

ISO=100
t=1 sec
N= f/8 (the calculator dial was set for a "precise" alignment of the 1 second and f/8 marks)
E=210 lux (that being from an eyeball of the resulting illuminance input scale setting as being 19.5 ft-c).

That is based on the notion that the meter movement reading is intended to be accurate as to illuminance (or pseudo-illuminance, given the "Norwood" collector).

Best regards,

Doug

 

[Doug Kerr]

Hi, Ted,

I note that in your picture it looks as if you do not have S set to exactly ISO 100:

Best regards,

Doug

 

[Doug Kerr]

Silly though this might seem, a bit more scrupulous photogrammetry gave the following data for a certain setting of the L-298A exposure calculator:

ISO=100 (at the detent point)
t=1 sec
N= f/8 (the calculator dial was set for a "precise" alignment of the 1 second and f/8 marks)
E=201 lux (by photogrammetry of the illuminance input scale, seeming to show an interpolated value of 18.7 footcandles)

That would say that the "baked-in" value of C was about 314.

Maybe.

Best regards,

Doug

 

[Doug Kerr]

Hi, Ted

Hi, Ted,

I note that in your picture it looks as if you do not have S set to exactly ISO 100:

Might have been just a bit of parallax - I ran into that when doing my bush-league photogrammetry!

Best regards,

Doug

 

[Doug Kerr]

Hi, Ted,

The manual for the Sekonic L-398A exposure meter tells that with the hemispherical collector in place C is 340, while with the flat collector in place C is 250.

This can certainly be true.

But if it is, then it is not possible that in both configuration modes the meter movement indication of illuminance (quasi-illuminance, in the case of the hemispherical collector) is actually correct.

We find by reverse engineering of the exposure calculator that, if the input value (from the meter movement) is actually the illuminance (or quasi-illuminance), then C would have to be something like 314 (and the exposure calculator has no idea which collector is in place).

This is a conundrum I have wrestled with in the past.

The manual only suggests that with the flat collector in place (so the meter response is proportional to actual illuminance) is the "illuminance" indication of the meter movement correct.

If it is, then for the mode with the hemispherical collector in place the meter movement indication is essentially an arbitrary number for transfer of the meter indication into the exposure calculator. (As you know, in many exposure meters the meter movement is only marked with an arbitrary scale, to that end.)

If that is the situation, then we cannot conclude what the value of C is merely by reverse-engineering the exposure calculator. That would have to be done with a calibrated illuminance source.

Best regards,

Doug

 

[Ted Cousins]

Meanwhile, earlier today I did a calc. for a dial setting of 20 fc and it was much closer to the dial EV:

L-398A

N^2/t dome = E.S/340

where E = illuminance lux

let E customary = 20fc ergo 215.278 lux

so N^2/t = 215.278x100/340 = 63.3171 = 5.985 Ev

Almost exactly same as L-398A dial calculator, 6 EV at 100 ISO unless my 'rithmatic is in error.

In this thread, I'm ignoring what's screwed into the head, as indeed just said "the exposure calculator has no idea which collector is in place" ... I'm just playing with the dial. If I use the formula for the Lumidisk, of course I get 0.4 EV less exposure for the same fc dial setting (not really meter fc reading).

 

[Ted Cousins]

Hi, Ted,

I note that in your picture it looks as if you do not have S set to exactly ISO 100:

View attachment 14397

Best regards,

Doug

Yep, the image was found on-line ...
... mine aren't that good ...

 

[Ted Cousins]

Hi, Ted,

The manual for the Sekonic L-398A exposure meter tells that with the hemispherical collector in place C is 340, while with the flat collector in place C is 250.

This can certainly be true.

But if it is, then it is not possible that in both configuration modes the meter movement indication of illuminance (quasi-illuminance, in the case of the hemispherical collector) is actually correct.

We find by reverse engineering of the exposure calculator that, if the input value (from the meter movement) is actually the illuminance (or quasi-illuminance), then C would have to be something like 314 (and the exposure calculator has no idea which collector is in place).

This is a conundrum I have wrestled with in the past.

The manual only suggests that with the flat collector in place (so the meter response is proportional to actual illuminance) is the "illuminance" indication of the meter movement correct.

If it is, then for the mode with the hemispherical collector in place the meter movement indication is essentially an arbitrary number for transfer of the meter indication into the exposure calculator. (As you know, in many exposure meters the meter movement is only marked with an arbitrary scale, to that end.)

If that is the situation, then we cannot conclude what the value of C is merely by reverse-engineering the exposure calculator. That would have to be done with a calibrated illuminance source.

Best regards,

Doug

 

[Ted Cousins]

The manual only suggests that with the flat collector in place (so the meter response is proportional to actual illuminance) is the "illuminance" indication of the meter movement correct.

If it is, then for the mode with the hemispherical collector in place the meter movement indication is essentially an arbitrary number for transfer of the meter indication into the exposure calculator. (As you know, in many exposure meters the meter movement is only marked with an arbitrary scale, to that end.)

If that is the situation, then we cannot conclude what the value of C is merely by reverse-engineering the exposure calculator. That would have to be done with a calibrated illuminance source.

No numbers at all on the meter face of my Sekonic L-158 'Auto-Lumi'!

https://www.manuallib.com/download/pdf1/SEKONIC-L-158-MANUAL.PDF

No 'K' given ... do the scales imply one?

 

[Doug Kerr]

Hi, Ted,

Meanwhile, earlier today I did a calc. for a dial setting of 20 fc and it was much closer to the dial EV:

L-398A

N^2/t dome = E.S/340

where E = illuminance lux

let E customary = 20fc ergo 215.278 lux

so N^2/t = 215.278x100/340 = 63.3171 = 5.985 Ev

Almost exactly same as L-398A dial calculator, 6 EV at 100 ISO unless my 'rithmatic is in error.

In this thread, I'm ignoring what's screwed into the head, as indeed just said "the exposure calculator has no idea which collector is in place" ... I'm just playing with the dial. If I use the formula for the Lumidisk, of course I get 0.4 EV less exposure for the same fc dial setting (not really meter fc reading).

On my L-298A, if I set the input to 20 fc (as closely as I can, given the nature of the dial scale markings), and ISO 100 (on the detent point, one photographic exposure recommendation is almost exactly 1 sec and f/8. That would imply a C of about 336 (again assuming that the input dial is actually set to a meter reading that is actually the quasi-luminance observed by the meter).

Best regards,

Doug

 

[Ted Cousins]

Hi, Ted,


On my L-398A, if I set the input to 20 fc (as closely as I can, given the nature of the dial scale markings), and ISO 100 (on the detent point, one photographic exposure recommendation is almost exactly 1 sec and f/8. That would imply a C of about 336 (again assuming that the input dial is actually set to a meter reading that is actually the quasi-luminance observed by the meter).

Best regards,

Doug

Yes, I got the same implied value of C earlier today. Maybe Sekonic rounded it up.

best,

Ted.

 

[Doug Kerr]

The fact that the rated value of C for the Sekonic L-398A exposure meter is 340 with the dome collector in place and 250 with the flat collector in place is difficult to understand.

The purpose of the dome collector is to essentially practice the "Norwood" concept of single-measurement exposure metering in such cases as the use of Key-fill" lighting of a human face.

I note that before the emergence of the Norwood technique incident light exposure meters generally were intended to respond to the illuminance on the plane of the meter's detector.

I also note that for an instrument to respond to the illuminance on some plane its directivity pattern (its relative sensitivity to light arriving from different angles) ideally would be a cosine pattern. That pattern would theoretically be given by a flat light detector (or a flat translucent "collector" over the detector proper).

Donald Norwood pioneered an incident light measurement technique based on an instrument whose directivity pattern was significantly different from the cosine pattern. It turned out that a meter with hemispherical ("dome") collector in theory gave a "cardiod" directivity pattern, which was a close approximation to the pattern Norwood has at first said was ideal.

If we follow Norwood]s reasoning (along different lines) in his various papers, I conclude that his technique, as he visualized a meter with essentially a cardiod directivity pattern , calibrated so that for a single light source located on the meter axis it would recommend the same photographic exposure as would an exposure meter that responded to true illuminance.

In terms of the incident light metering calibration factor, C, that would suggest that a "Norwood scheme" meter should have the same value of C as a "true illuminance based" meter.

Now the definition of C for a meter with other than a cosine pattern is problematical. The defining equation (in ISO 2720) revolves around the illuminance observed by the meter. But a "cardiod" meter does not respond to illuminance.

A likely answer (for which I have no proof) is that for a meter with a cardiod pattern, C is determined in terms of the illuminance of a single light source on the meter axis.

Bu then, if we go to the implications of Norwoods work, we would expect the value C for a cardioid directivity (dome collector) configuration of a meter to be the same as the value of C for a cosine pattern (flat collector) meter. Yet we see that Sekonic has departed from that.

There are many paradoxes if we simulate a "Norwood scheme" meter with various hypothetical illumination setups.

It may well be that Sekonic has detained that the best situation will be attained if the two modes had different values of C.

Best regards,

Doug

 

[Ted Cousins]

The fact that the rated value of C for the Sekonic L-398A exposure meter is 340 with the dome collector in place and 250 with the flat collector in place is difficult to understand.

The purpose of the dome collector is to essentially practice the "Norwood" concept of single-measurement exposure metering in such cases as the use of Key-fill" lighting of a human face.

I note that before the emergence of the Norwood technique incident light exposure meters generally were intended to respond to the illuminance on the plane of the meter's detector.

I also note that for an instrument to respond to the illuminance on some plane its directivity pattern (its relative sensitivity to light arriving from different angles) ideally would be a cosine pattern. That pattern would theoretically be given by a flat light detector (or a flat translucent "collector" over the detector proper).

Donald Norwood pioneered an incident light measurement technique based on an instrument whose directivity pattern was significantly different from the cosine pattern. It turned out that a meter with hemispherical ("dome") collector in theory gave a "cardiod" directivity pattern, which was a close approximation to the pattern Norwood has at first said was ideal.

If we follow Norwood]s reasoning (along different lines) in his various papers, I conclude that his technique, as he visualized a meter with essentially a cardiod directivity pattern , calibrated so that for a single light source located on the meter axis it would recommend the same photographic exposure as would an exposure meter that responded to true illuminance.

In terms of the incident light metering calibration factor, C, that would suggest that a "Norwood scheme" meter should have the same value of C as a "true illuminance based" meter.

Now the definition of C for a meter with other than a cosine pattern is problematical. The defining equation (in ISO 2720) revolves around the illuminance observed by the meter. But a "cardiod" meter does not respond to illuminance.

A likely answer (for which I have no proof) is that for a meter with a cardiod pattern, C is determined in terms of the illuminance of a single light source on the meter axis.

But then, if we go to the implications of Norwoods work, we would expect the value C for a cardioid directivity (dome collector) configuration of a meter to be the same as the value of C for a cosine pattern (flat collector) meter. Yet we see that Sekonic has departed from that.

There are many paradoxes if we simulate a "Norwood scheme" meter with various hypothetical illumination setups.

It may well be that Sekonic has detained that the best situation will be attained if the two modes had different values of C.

Best regards,

Doug

 

[Doug Kerr]

It is difficult to be precise when "reverse engineering" the Sekonic L-398A exposure meter's exposure calculator, especially because it seems that the result will vary a bit depending upon the part of the scales we consider.

But it is probably reasonable to consider that the value of C implied by the calculator (with the dome collector in place) is nearly the 340 stated by the manufacturer (this contingent on the quasi-illuminance indicated by the meter movement being accurate).

That having been said, then for the meter to exhibit a C of 250 with the "flat" collector in place (as stated by the manufacturer), then with the flat collector in place, it would be required that the actual illuminance be 250/340 times the indication of the meter movement. Said another way, the indication of the meter movement would have to be 1.36 times the actual luminance.

That could well be true, a result of differing "transmissivities" of the dome and flat collectors.

But of course it means that using the instrument as an illuminometer (with the flat collector in place) [and such usage is in fact described in the manual] would produce results that were about 36% too high.

If, alternatively, the meter movement indication with the flat collector in place was accurate as to illuminance, then the value of C stated for that configuration would be entirely wrong (it would in fact have to be about 340).

So go figger!

Best regards,

Doug

 

[Ted Cousins]

A likely answer (for which I have no proof) is that for a meter with a cardiod pattern, C is determined in terms of the illuminance of a single light source on the meter axis.

But then, if we go to the implications of Norwoods work, we would expect the value C for a cardioid directivity (dome collector) configuration of a meter to be the same as the value of C for a cosine pattern (flat collector) meter. Yet we see that Sekonic has departed from that.

I just did the test with my A19 LED desk lamp on-axis 15" from the LumiDome and from the LumiDisk. Bearing in mind the cardioid and cosine polar diagrams on your website, I was not surprised that the meter readings were the same: 48 fc. I expect that off-axis readings would be different depending how far off-axis and depending on the radiation pattern of the incident light.

 

[Doug Kerr]

Hi, Ted,

I just did the test with my A19 LED desk lamp on-axis 15" from the LumiDome and from the LumiDisk. Bearing in mind the cardioid and cosine polar diagrams on your website, I was not surprised that the meter readings were the same. I expect that off-axis readings would be different depending how far off-axis and depending on the radiation pattern of the incident light.

Interesting result. Thanks.

Note that the directivity patterns I show are all relative, and do not attempt to say how the "0°" sensitivity of two meters (or the same meter in two different configurations) might differ.

I any case, based on what I said in a post just above, I am surprised that with what we can reasonably consider to be the same luminous flux density incidence on the meter, the indicated "illuminance" is essentially the same for both collector configurations.

Given that the exposure calculator does not have, for example, different "pointers" to be used to set the meter movement indication for the two collector configurations, this would tell us that the value of C is the same for both collector configurations.

There are many things we know about this that I can't fit together!

Best regards,

Doug

 

[Doug Kerr]

The manual I have for the Sekonic L-398A exposure meter says this about measuring illuminance:

But this is in error. Illuminance pertains to the impact of the light on a (possibly hypothetical) surface of some given orientation. To measure the illuminance that a certain light source gives to such a surface, we need to orient the meter so that its collector is parallel to that surface, not "parallel with the light source", which I think we must interpret as meaning, "With the a line perpendicular to the Lumidisc pointing at the light source".

The measurement described above is actually of the luminous flux density of the light source at the location of the measurement, a property of the light beam, not dependent on an assumed target of some orientation. Illuminance and luminous flux density are often confused, and it is no help that the official scientific symbol for both is the same, E.


Best regards,

Doug

 

[Doug Kerr]

The manuals for many exposure meters with an incident light mode describe the use of the meter to determine the "ratio" between two (or more) light sources.

It is generally recommended that this be done with the "flat" collector in place.

The quantity of interest, and which is measured, is actually the luminous flux density of the respective light source (although we almost never see that said). This can be measured with an instrument intended to measure illuminance merely by orienting it so that a line perpendicular to its "collector" points at the respective light source.

Since it is the ratio between luminous flux density of the two (for more)sources that is of interest, it does not matter if the "illuminance" scale of the meter movement is in error by a fixed ratio.

Best regards,

Doug

 

[Ted Cousins]

Hi, Ted,

I note that in your picture it looks as if you do not have S set to exactly ISO 100:

View attachment 14397

Best regards,

Doug

Tsk ... a whopping 0.152 EV off ...

Not my meter anyway, stole the picture off the web.

 

[Doug Kerr]

In an earlier post in this thread, I stated that, for the Sekonic L-398A espsure meter, given that the manufacturer states that the value of C with the "dome" light collector in place is 340 while with the flat" collector in place the value of C is 250, then if the meter movement "illuminance" indication in the "dome" mode is correct (it is actually quasi-illuminance in that case), then the meter movement indication in the "flat" configuration would be 340/250 of the actual illuminance.

My presentation was not very complete. Here is a more complete presentation:

******
This presentation uses specific values for the Sekonic L-398A exposure meter in its two incident light modes.

Again, the manufacturer states that the value of C with the meter in the dome configuration is 340, and the value of C with the meter in its flat collector configuration is 250.

I also note that in this meter the transfer of the meter movement indication into the exposure calculator uses exactly the same scale for either meter configuration.

I now refer to the incident light metering formula given in ISO 2720: (I write it here in a slightly different arrangement)

where C is the incident light metering calibration constant; t is the exposure time (shutter speed), in seconds; N is the aperture, as an f-number; E is the illuminance (or quasi-illuminance) observed by the meter; and S is the film/camera sensitivity, as an ISO speed.

Now when we go from the dome case to the flat case, C by definition decreases in the ratio 250/340.

Assume that for both cases, we have the same input "illuminance" to the exposure calculator. That is, the meter dials will be in the same position, and so t/N^2 will be the same for both cases.

Thus, considering the formula above, as we go from the dome case to the flat case, the value of E must also decrease in the ratio 250/340.

The input to the exposure calculator is taken directly from the indication of the meter movement, so that must be the same for both cases. But the value of E the meter sees will be less in the flat case than in the dome case by the ratio 250/340.

Thus the "sensitivity" of the photodetector and meter movement system for the flat case must be 340/250 times what it is for the dome case.

Accordingly, if (as is suggested by other work here), the quasi-luminance indicated by the meter movement in the dome case is accurate, then the luminance indicated by the meter movement in the flat case must be 340/250 times the actual value.

Quod erat demonstrandum.

Best regards,

Doug

 

[Doug Kerr]

About "quasi-illuminance"

In several places above I have used the term "quasi-illuminance" without explaining it. I will do that here.

For an instrument to actually determine the illuminance upon some surface from the existing illumination, it must "weight" the density of luminous flux arriving from different directions by the cosine of the angle each component's direction of arrival makes with a line perpendicular to the instrument's "receptor" (which we assume is oriented parallel to the receiving surface of interest).

This comes from the basic definition of "illuminance". The cosine gets into the picture not in some esoteric way but merely from a basic geometric consideration (which I will not discuss here).

That is, the directivity pattern of the instrument's receptor (the sensitivity of the receptor to light versus the angle of arrival of the light) must be a cosine pattern.

In an incident light exposure meter with a dome light collector (or the equivalent), the directivity pattern is intentionally not a cosine pattern but rather is something like a cardiod pattern. This is to make the instrument follow the "Norwood" scheme of exposure metering.

Since the meter in its dome configuration does not have a cosine directivity pattern, its response to a certain illumination will not (usually) actually be the illuminance of that illumination. In the interest of rigor, I speak of the quantity that is determined by the meter in its "dome" configuration as "quasi-illuminance".

Best regards,

Doug

 

[Doug Kerr]

Sekonic L-398A exposure meter–values of C

1. Reverse engineering of the exposure calculator suggests that, for the actual value of C the meter exhibits in its "dome collector" configuration to be the stated value of 340, the "footcandle" reading of the meter movement would very nearly have to be the actual quasi-illuminance observed by the meter.

Note that the information in the manual makes no suggestion that, in the dome configuration, the footcandle reading of the meter movement would be the actual quasi-illuminance the meter observed. But it seems that it nearly is.

2. From that. for the actual value of C the meter exhibits in its "disk collector" configuration to be the stated value of 250, the "footcandle" reading of the meter movement would very nearly have to be 1.36 times the the actual illuminance observed by the meter.

Note that the information in the manual directly suggests that, in the disk configuration, the footcandle reading of the meter movement would be the actual illuminance the meter observed. But it seems that it isn't.

Best regards,

Doug

 

[Doug Kerr]

Sekonic L-398A exposure meter–exposure calculator inconsistency

One (certainly small) source of inconsistency of the exposure calculator with different input settings is this:

All scales on the exposure calculator have their major markings at the same intervals throughout their length. That interval is intended to be one stop (2:1 in exposure implication).

But on the time (shutter speed) scale the values those marks are labeled as meaning include (for a series long enough to show the point here), in seconds:

1/2, 1/4, 1/8, 1/15, 1/30, 1/60

Note that these are all from the preferred series of camera shutter speeds. Thus they are values expected to be found on a camera's shutter speed setting dial (or menu).

But in fact 1/15 s is about 0.91 stop less than 1/8 s, whereas the layout of the calculator dial is based on a 1-stop difference.

Again, this is a minor discrepancy, of no consequence in practical work. But it can contribute to some of the small numerical conundrums encountered when "reverse engineering" the L-398A exposure calculator.

There is a somewhat similar situation on the aperture (f-number) scale, but it is harder to see. But the preferred "labeling" of aperture f-numbers is based on rounded values, while it is expected that the calculator will work with the cosponsoring "exact" values. In this light, there is no place in the set of markings of the aperture scale where there is what we consider to be an "inexact" interval.

Best regards,

Doug

 

[Doug Kerr]

Incident light exposure meter–determination of "C"

ISO 2720 defines the value of the meter calibration constant for an incident light exposure meter, C, by way of a formula that involves the "illuminance", E, observed by the meter.

But a meter with a "cardioid" sirectivity pattern (one of two incidnet light meter flavors explicitly recognized in ISO 2720) does not repond to illuminance but rather to a property I call "quasi-illuuminance".

A reaonable way to work around this would be to test the meter response for each configuration with a single light source coming along the "axis" of the meter ("head on"). (The two measures–illuminance and quasi-iiluminance– would be the same for such a light source.)

And in fact, ISO 2720 prescribes just that.

Best regards,

Doug

 

[Doug Kerr]

Sekonic L-398A exposure meter–value of C implied by the expsure calcualtor

Quite careful reverse engineering of the expsure calculator on our Sekonic L-398A expsure meter, including phtogrammetric meaurement of the input setting to prduce an integral Ev expsure reommendation, tells that, if indeed the input value (as read from the meter movement) is the actual quasi-illuminace observed by the meter , the value of C would be 314.5. (This is at an input value of 18.7 footcandles. Hosever, given that the scales of the calculator are quite uniform, we wouold expect the same result for greater values.)

Note that, subject to the conditions stated above, this is independent of the kind of collector fitted. (The way that the value of C becomes different for the two types of collector is that the condition, "the input value to the calculator is the actual illuminance (or quasi-illuminance) observed by the meter", would not be true for both collector types.)

Best regards,

Doug