A few words on Equivalence and comparing systems

Started Apr 2, 2013 | Discussions thread
Macx Senior Member • Posts: 1,433
Re: A few words on Equivalence and comparing systems

Anders W wrote:

Macx wrote:

Anders W wrote:

Macx wrote:

texinwien wrote:

olliess wrote:

"Didn't want that light anyway."

It's the sensor not the lens, and JACS is talking about measures of lens performance w/r/t equivalence.

But I'd wager you knew this and were just trying to be witty. Unfortunately, you just ended up sounding obtuse.


Keep the sensor out of it. The sensor size is interesting because the FOV of a particular lens is dependant on it, but for the same FOV and the same physical aperture the systems are equivalent and they get the same amount of light. If we had lenses with the same FOV and the same physical aperture for micro four-thirds as we have for 135, there would be no innate shot noise advantage of the larger sensor.

In other words, it's the lenses, not the sensor, and if we had a 12-35/2 our exposure envelope or gamut would be equal to having a 24-70/4 for our 135. The loss of a stop of total light when going from a 24-70/4 on a 135 to a 12-35/2.8 on M43 is dependant on the latter being f/2.8 and not f/2.


It seems that your point (FoV instead of FL, absolute size of the aperture rather than relative size) is a pedagogical one. Your idea, if I understand you right, is that it simplifies things to use FoV and absolute aperture size because you can state equivalence in terms of equalities.

Yes, and because these are optical qualities that define what we want in a lens. It's field of view and it's depth of field and/or diffraction properties. We've learned to make the mental arithmetic needed to translate focal length and focal ratio into these values, but why the need for such translation when we could use these "proper" values as the basis for discussing how a lens behaves?

Personally, I have no trouble going between FL and FoV. And I actually find it easier to define DoF and diffraction properties on the basis of f-stops than absolute aperture. Could you perhaps tell me how to convert the rule of thumb that MFT lenses shouldn't be stopped down further than f/8 if you want to avoid significant diffraction effects into your conceptual scheme?

If I really had my way, I would use a logarithmic scale for the stops. So I would say that Av shouldn't be more than 6. That might be taking it too far though. As I wrote, I'm not asking for us to forget or make focal length and relative aperture secret. It's not either-or. Relative aperture is also useful for calculating depth of field using the same framing, for example.

I see that point but I am not sure I agree that this is the more pedagogical solution all things considered. For example, FL and max relative aperture (max f-stop) are known properties of a lens and independent of the sensor to which the lens is fitted. FoV and max absolute aperture are properties that have to be derived (calculated) from those already known and the first of them (FoV) varies depending on the size of the sensor the lens is mated with.

The known properties are the focal length and the maximum absolute (virtual) aperture. The maximum f-stop is derived from these, not the other way around.

When it comes to aperture, it works perfectly well either way. Start with the relative number and define the absolute number if you want it (I rarely do).

It does indeed. No special advantage either way.

It is true that to calculate the FoV we need to know what sensor the lens is used for. Even so, at least we would know it, instead of just having an intuitive understanding about the field of view of a 19 mm lens on a 135, for example.

I know that anyway.

Well, it was an easy one. What is the Field of View of a 3 mm focal length phone camera? NO PEEKING AT WHAT THE CROP FACTOR IS! I suppose we're blessed that we can use such simple arithmetic to convert between 135 and 4:3 values. It would be harder to do on the fly if it was an irrational number.

I don't propose that we keep focal length or the f-stops secret. But these values aren't what is interesting about how our lens behaves on our camera, and especially when discussing lens properties across systems it seems to me to be much more pertinent to concentrate on these values instead of having to work with "equivalent focal lengths" and other thought-experiments.

You say that f-stops aren't what we are interested in. So tell me, how would you convert the "sunny 16" rule into your conceptual machinery?

I wouldn't really. Is there a need to when talking about the optical qualities of a lens on different systems?

Another thing here that I am curious about. Equivalence isn't only about equality but also about how far apart things are. Say we are comparing a 12/2 on MFT with a 24/2.8 on FF. Clearly, these two lenses are not equivalent. The 12/2 is equivalent to a 24/4, not a 24/2.8. But how would you, using your conceptual machinery, express the one f-stop difference between the 12/2 and the 24/2.8? And of what practical help would that difference, as expressed in your language, be?

If the absolute aperture is half the area I would say that it has one stop less light. But I'd probably just use the focal ratio here for ease of calculation. Again, I would have changed it to a logarithmic scale if I had my way, (If it wasn't for those meddling kids!), but since I have memorised the f-stop progression (and the 1/3 points and the 1/2 points) I can hobble along using focal ratios.

And from a performance standpoint, isn't the performance of a lens at a certain absolute aperture more interesting that what it is at a relative aperture? I think so, at least.

So if you are comparing a 25 mm and 50 mm lens on MFT, which comparison would you find more interesting. A comparison at the same f-stop or a comparison at the same absolute aperture?

A 25 mm on 4:3 and a 50 mm on 135, absolute aperture here, definitely. I wouldn't really have much use comparing specific f-stops on a 25 and 50 mm to each other, I would be much more interested in how they performed individually across the range. I don't think I would often be in doubt whether to use a 25 or a 50, and use their performance at specific stops to decide. I suppose I could crop the 25 to get the same FOV as the 50?

Both ways of thinking about the matter are of course factually correct. It's just a matter of conceptualization. But in the end, I am not sure that a shift of attention away from FL and relative aperture to FoV and absolute aperture makes things any better from a pedagogical point of view.

Maybe not. The lens manufacturers isn't making it easy for us, that's for sure.

It seems I am making it difficult for you as well. It's a hard life.

Yes, but I hope I eventually learn something from all the beat downs, so don't show mercy. As long as you're polite about it, I can take it, I think.

Post (hide subjects) Posted by
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow