Digital light meter.

[Greg Miller]

But I stand on my original claim: you can see things with an analog display, that are invisible to any digital device. Whether we need to see those things, or should even care, I won't debate.

So I can't help but (politely) be curious how you would expose or develop your film differently if you had accurate exposure readings finer than .1 stop. Does the light remain that constant from the time you make your exposure reading to the time you trip the shutter? Do your shutters have the ability to set and expose with that level of precision? Is your film development process so precise as to realize any benefits? Is your printing process that precise? Is your end to end process that precise?

[Marko]

But I stand on my original claim: you can see things with an analog display, that are invisible to any digital device. Whether we need to see those things, or should even care, I won't debate.

The real question is: do the things you are seeing really exist?

Given that any light meter, regardless of the display system, is based on a silicon sensor, therefore, both a "digital" and an "analog" one essentially "see" the same amount of information.

Frank has already explained that the granularity of a digital display is a matter of engineering decision, not the precision of the instrument. IOW, they are in effect limited to displaying less information than they actually see.

A needle display, on the other hand, is essentially a solenoid with a needle attached to the moving core. Where the needle stops in between the two marks is largely determined by the characteristic dissipation incurred by the solenoid AND by the mechanical inertia of the needle.

There is no way you can seriously prove that a mechanical readout is somehow more precise than a digital one given the same signal.

[Leigh]

You may see the needle at a point between two marked lines and think that it is giving you a value which is between the two...

It is.

The accuracy and linearity of any decent d"Arsonval meter movement is excellent, typically 3% for non-mirrored movements and 2% or better for mirrored.

To put this in perspective, 0.1 stop = ~7%, while 0.01 stop = ~0.7%.

Analog meters do not suffer from quantization error that's inherent in any digital system.

ALL digital systems have an error of ±1 least significant digit, in addition to all other errors in the system. It's a consequence of the method used to convert analog to digital, and is completely irrespective of the nature of the parameter being measured.

As a practical example, for an exposure meter to accurately display aperture readings to 0.1 stop, the acquisition system and display would need a resolution of 0.01 stop.

Optical power measuring instruments are readily available with accuracies of 0.01% or better, so it's not a question of technology. It's just the fact that the photographic industry does not require that level of accuracy for exposure determination. Available lenses and shutters could never come close to that accuracy.

- Leigh

[Alan Davenport]

So I can't help but (politely) be curious how you would expose or develop your film differently if you had accurate exposure readings finer than .1 stop. Does the light remain that constant from the time you make your exposure reading to the time you trip the shutter? Do your shutters have the ability to set and expose with that level of precision? Is your film development process so precise as to realize any benefits? Is your printing process that precise? Is your end to end process that precise?

Did you actually read what I wrote?

[Alan Davenport]

Optical power measuring instruments are readily available with accuracies of 0.01% or better, so it's not a question of technology. It's just the fact that the photographic industry does not require that level of accuracy for exposure determination. Available lenses and shutters could never come close to that accuracy.

And one of the major driving forces in the industry's move from analog meters to digital: when you get rid of the delicate d'Arsonval movement and substitute a drop-in digital module, the meter is cheaper to manufacture! (Note that I did not claim it is cheaper to purchase, just to build...)

[Greg Miller]

Did you actually read what I wrote?

Yes, I did. You made a claim nd at the same time say you will not defend any practical purpose of that claim.

People generally don't make claims without having a purpose (unless they are a troll). I gave you credit for not being a troll and intentionally said I was asking politely about what your point was, and was intellectually curious as to why you would bother to make such a claim.

So at this point, I will abandon politeness and simply ask, regarding your claim, "so what"?

[Marko]

It is.

The accuracy and linearity of any decent d"Arsonval meter movement is excellent, typically 3% for non-mirrored movements and 2% or better for mirrored.

To put this in perspective, 0.1 stop = ~7%, while 0.01 stop = ~0.7%.

Analog meters do not suffer from quantization error that's inherent in any digital system.

ALL digital systems have an error of ±1 least significant digit, in addition to all other errors in the system. It's a consequence of the method used to convert analog to digital, and is completely irrespective of the nature of the parameter being measured.

As a practical example, for an exposure meter to accurately display aperture readings to 0.1 stop, the acquisition system and display would need a resolution of 0.01 stop.

Optical power measuring instruments are readily available with accuracies of 0.01% or better, so it's not a question of technology. It's just the fact that the photographic industry does not require that level of accuracy for exposure determination. Available lenses and shutters could never come close to that accuracy.

- Leigh

Even the cheapest digital multimeters, the $10 ones found at registers in Fry's , have accuracy of 0.5% or better for DC. Bench-tops are typically better than 0,01% and laboratory models reach to few ppm.

I keep referring to multimeters for a reason, because that is exactly what a lightmeter is - a device that measures low voltage DC signal generated by a photo-sensitive silicon sensor. Both analog and digital light meters are built around the same type of sensor, it's just the measuring and display systems that differ.

It's all nice and dandy as pointless discussions go , but this particular application is simply too coarse and insensitive for any real differences between the two types to matter.

In other words, it's not the hardware, it's the software that matters.

[Leigh]

... this particular application is simply too coarse and insensitive for any real differences between the two types to matter.

Exactly. This is photography, not rocket science.

In other words, it's not the hardware, it's the software that matters.

In this case, "soft" = gray matter. What's between the ears makes all the difference.

- Leigh