My head is hurting...

[davidedric]

Is it true that for the same scene, and for the same aperture, and for the same ISO, one gets the same shutter speed for correct exposure, regardless of sensor size? (Ignoring transmission efficiency of the lens). I've been trying to work it out from first principles, and my brain has seized.

Thanks,

Dave

[DanK]

yes.

Think of it this way. Forget ISO for the moment. Assume that you set an aperture manually and hold a lens over a piece of white paper, at a distance that leaves the image circle well defined. Now, draw two rectangles of different sizes, but both completely inside the image circle. You would have the same intensity of illumination of the larger and smaller rectangles.

Now, replace those two rectangles with sensors of those sizes, both with base ISO=100. Raising ISO from the base value on a digital camera simply amplifies the signal generated by the sensor. Set both to ISO 200, and in both cases, you would double the signal strength. Since it started out equal, it would be equal when doubled.

[Geoff F]

How are you metering?

I find that when taking two shots of the same subject in Aperture Priority mode on a tripod with the only difference being a slight change in focus point, I can get some variation in shutter speed.

Particularly noticeable when doing macro shots such as moths under studio conditions. Evaluative metering can change if it has more of an area to work by, and particularly so with spot metering.

[Manfred M]

Is it true that for the same scene, and for the same aperture, and for the same ISO, one gets the same shutter speed for correct exposure, regardless of sensor size? (Ignoring transmission efficiency of the lens). I've been trying to work it out from first principles, and my brain has seized.

Thanks,

Dave

Yes, but it depends on your defininition of "scene". How's that for a clear answer?

If both shots are framed identically; i.e. for example you are using a wide open f/2.8 25mm lens on a mFT camera (crop factor 2x) and are taking the same shot with a wide open f/2.8 50mm lens on a full frame camera, then the answer is yes.

Now if you are using an incident light meter and are shooting a portrait and are metering the light falling on your subjects face. You will set your camera settings identically (ISO, shutter speed and aperture), regardless of crop factor, focal length. The face will come out looking the same, regardless of what happens with the background.

[DanK]

How are you metering?

I find that when taking two shots of the same subject in Aperture Priority mode on a tripod with the only difference being a slight change in focus point, I can get some variation in shutter speed.

Particularly noticeable when doing macro shots such as moths under studio conditions. Evaluative metering can change if it has more of an area to work by, and particularly so with spot metering.

I think this is introducing an additional variable.

The description I gave assumed uniform lighting of the surface.

Clearly, if the surface is not uniformly illuminated, changing sensor size with the same lens--which changes field of view--can change the mix of luminance picked up by the meter. Imagine, for example, that the smaller, inner rectangle is more illuminated than the area around it. Then evaluative metering would give you a different reading. This is not a function of sensor size; it is a function of the illumination of the area in the image circle.

The bottom line is that for any given FOV, you can forget about sensor size when metering. What matters is the amount of light, how it is distributed across the FOV, and the metering algorithm your camera is using.

---

I see that Manfred was posting a similar answer as I typed this.

[Shadowman]

The shutterspeed fluctuation (1/ 8, 1/8, 1/10") between 3-5 successive shots might be so slight that you won't notice the difference. Also, would you think you'd see a difference when shooting wide open as opposed to using a smaller aperture?

[davidedric]

Thanks, guy's. I thought it must be so, but I had seen a couple of posts in another place that seemed to suggest otherwise.

Manfred, yes, I used scene as a shorthand way of saying with an appropriate focal length to allow identical framing, and assuming that the light had not changed.

I was mentally using Dan's example. So I thought, let's assume that the two different sensors have the same number of pixels. Then each pixel in the large sensor is receiving more photons than each pixel in the small sensor, and that means........Then it all went hazy. Perhaps I needed to add in that it means a poorer signal to noise ratio in the smaller sensor, and hence more noise in the image. Or to put it another way, to get the same signal to noise ration in the two, I would need to turn down the gain by decreasing iso in the smaller sensor, so I would need a longer exposure. I can relax now

Happy New Year all,

Dave

[Manfred M]

How are you metering?

I find that when taking two shots of the same subject in Aperture Priority mode on a tripod with the only difference being a slight change in focus point, I can get some variation in shutter speed.

Particularly noticeable when doing macro shots such as moths under studio conditions. Evaluative metering can change if it has more of an area to work by, and particularly so with spot metering.

Geoff - you are correct here, but only because the lens's focusing mechanism changes your "scene" slightly when focusing this close to the subject.

So in fact, the metering only changes because the scene itself changes due to the lens element's movement inside the lens barrel. While the same thing happens in any shot, it is undetectable in work outside of the extreme closeup work.

[Manfred M]

Ignoring transmission efficiency of the lens

Dave - if you ever get a chance to look at lenses built for cinematography; they have aperture settings marked in t-stops (t stands for transmission), rather than the f-stops we still photographers work with.

A t-stop number incorporates the light loss due to transmission efficiency + the change in focal length, rather than the purely mathematical definition (focal length / entrance pupil diameter) used on still cameras.

[Simon Garrett]

Is it true that for the same scene, and for the same aperture, and for the same ISO, one gets the same shutter speed for correct exposure, regardless of sensor size? (Ignoring transmission efficiency of the lens). I've been trying to work it out from first principles, and my brain has seized.

Thanks,

Dave

The way I think if it is to consider film (remember that?). A 100 ISO film is the same film whether it's cut into 35mm or 2 1/4 square. To be correctly exposed, the small and large format 100 ISO film must both be getting the same light per unit area.

Same aperture (f/ number that is, not aperture size in mm), same shutter speed means same light per unit area on the sensor.

The meter doesn't care what size the sensor is when it works out exposure. If you have a stand-alone light meter, there's nowhere to set the sensor or film size on it.

[Mike Buckley]

I'm not technically oriented, but it seems to me that most times two different sensors are being used, different sets of algorithms are also being used. I assume the different sets of algorithms would change the exposure at least a small amount. My point is that from a purely practical point of view, it's probably not possible to use two different sensors and the same algos.

[xpatUSA]

In formal terms, Wikipedia defines it quite well:

The recommended f-number and exposure time for given lighting conditions and ISO speed are given by the exposure equation N^2/t = L.S/K where:

• N is the relative aperture (f-number)
• t is the exposure time (“shutter speed”) in seconds
• L is the average scene luminance
• S is the ISO arithmetic speed
• K is the reflected-light meter calibration constant

Re-arranging for t, we get:

t = N^2 x K / L / S with no mention of sensor size at all.

Also, the expression is for the image plane - not the sensor itself, which could be a piece of bog roll for all we care

So, as is often the case, we should separate in our minds that which comes out of the sensor and that which exists in the image plane, eh?

[george013]

In formal terms, Wikipedia defines it quite well:



Re-arranging for t, we get:

t = N^2 x K / L / S with no mention of sensor size at all.

Also, the expression is for the image plane - not the sensor itself, which could be a piece of bog roll for all we care

So, as is often the case, we should separate in our minds that which comes out of the sensor and that which exists in the image plane, eh?

Luminance is light intensity per unit area. So independent from the Total surface.

On dxo they also mention the t-factor for the reviewed lenses.

George

[xpatUSA]

Luminance is light intensity per unit area.

Thank you for the reminder.

Re-arranging for t, we get:

t = N^2 x K / L / S with no mention of sensor size at all.

Also, the expression is for the image plane - not the sensor itself, which could be a piece of bog roll for all we care

Of course, the ISO arithmetic value for toilet paper is quite small . . .

[Simon Garrett]

Of course, the ISO arithmetic value for toilet paper is quite small . . .

But not zero. Leave it in the sun for a few months and it darkens a bit.

Taking an approximation from the Sunny 16 rule, and assuming the paper darkens to mid grey after 10 year continuous sunlight (24/7), I estimate the ISO value at approximately 1E-12. Give or take a few orders of magnitude.