[Kunzite]

It's actually required when you need to convert between any two formats and doesn't need to have anything to do with FF at all.

For example, say you have a P&S with a built-in zoom, and you want to move up to an M43 MILC, but don't want to lose that P&S zoom range, which you really like - how do you select an m43 lens that has the same FL range? You need to use equivalence.

Most people stop there at the FL conversion and don't worry about the F-stop conversion, either because they don't care or they don't know about it. But as I was trying to explain to kunzite, just because you personally don't care or don't know about it doesn't mean it's not real

Make up your minds; either "equivalence" MUST (as in RFC 2119) be used in its complete form, or one's allowed to pick what to compare.

Kunzite - I challenge you - try to answer the above question without using equivalence.

(if past is any indication I suspect you'll decline, and say something like "don't worry about it, just look through the veiwfinder and accept what you get, like I did, to learn my system.." or some variation.)

That past exists in your imagination; it's me who's still waiting for an answer to the "what should they write on lenses used on multiple formats?". Looking through the viewfinder happened when I couldn't compare between formats because I didn't had the "equivalent" FF camera+lens combination; not understanding this is probably intentional.

Your challenge is an obvious trap. I would obviously use the "old ways" (those rejected in the video as being "arbitrary and meaningless) yet you'll claim that I'm using "equivalence". Even if focal length will stay focal length, aperture won't change its meaning and I won't even try to find an "equivalent" ISO. But here it is:
1. Getting the same angle of view is done using a simple geometric relationship between AoV and FL:
- get the largest rectangle with the same aspect ratio as the "source" frame, that can be fitted into the "destination" frame.
- compute the linear ratio R between this "destination" rectangle and the "source" frame. You may call it "crop factor", though it could differ from the widely accepted crop factor (e.g. it would be ~2.12:1 for 4/3 vs. 135).
- choose a focal length which is equal to the "source" lens' focal length multiplied with the ratio R.

2. Getting the same DoF.
Yes, you will use the ratio R from before; you know how. Just open or close the aperture as required.

3. Getting the motion blur.
This is done by using the same shutter speed.

4. Getting the same exposure
Because the shutter speed is constant, light assumed constant - I will have to change ISO to compensate for the change in aperture. Anything else means a different exposure, i.e. a different result.

5. Getting the same detail.
Can't control it. Well, I suppose I could add blur to match a less detailed "source" image...

6. Getting the same noise.
No control here either; if I'm lucky, I'll get similar noise levels to that from the original picture. If not, it can't be helped.

And this could continue depending on how similar the images should be.

I'm not able to see why this concept is so alien, upsetting and disruptive to you. 'equivalence' just describes the relationship between any two formats. It's USED to 'compare specific things' between systems, that's what it's used for. It's fantastically useful for making purchase decisions if you started with one format and are moving to (or adding) another and have some preferences or settings standards you want to maintain. And it can really help you save money.

Please, if you haven't already, take the time to read through this, or jump straight to this (system comparison.) You'll notice that nowhere in there is he giving preferential treatment to one format.
.

It's definitely not alien, as I completely understood the first time I read about it. It's upsetting when people are insisting that screwing with basic optical notions is a good idea; and because it's too rigid, and don't consider photographer's needs. It's disruptive because I met people completely confused by it, e.g. believing the focal length is changing with the format then expecting different AoV between APS-C and FF 50mm lenses.

That's one of its weaknesses: that it forces you to compare in a certain way, one that might not be relevant to you. One would usually move between formats to get something different, something better, right?
Besides, detail have to be observed, noise have to be measured and observed, color reproduction, then you have size and weight, ergonomics, several performance indicators... calling it "equivalence" when it ignores so many things is not exactly appropriate IMHO.

It's too late (past midnight) to check if those pages are correct in their claims. My apologies, but I'll pass.

[amoringello]

Well, you might want to watch the video, because I do explain the physics. If you disagree with my math, just let me know where my math is wrong and I'll happily fix it.

I really liked the video as so many people re-iterate "the Xmm lens is equivalent to Ymm lens on a full frame body", and other false claims...
It irritates me to no end when professionals and the so called "rock stars" of photography perpetuate inaccurate and false info.

I guess the only part of your videos I did not agree with, and maybe I missed something, is that you seemed to make the full frame a standard by which everything should be based. The math seemed good to me, but there is never a give on one side without taking away on the other.

You stated (and maybe got a bit hyperbolic) that you feel bad that a person getting an f/2.8 lens for lets say a 4:3 camera is only really getting a f/5.6, as if they had been mislead.
Firstly, "aperture" is based on the physical properties of the lens.
It is not affected by the size of the surface on which the light from the lens is cast.

One could "feel bad" for the full frame body users that were mis-lead to thinking they had the same aperture as used on medium format cameras.
Additionally, you cannot say that the poor crop sensor cameras have it bad without also stating the benefits that it has above larger frame cameras...
Lets say I want a really deep depth of field, using f/22 on a 4:3 camera. (dunno if they go that far, but you probably get my point)
I believe the full frame would need f/44 plus require two stops more exposure (longer shutter or higher ISO).
To me, I "feel bad" that the full frame users have been mis-lead into thinking they had an f/22 lens, but max aperture might not be able to get f/44 on their lenses.

I feel like this was a part 2 of a multi-part video that missed the issue of "equivalencies". Without the reasoning behind I think it all just "sounds" flawed.

[Rondec]

It's actually required when you need to convert between any two formats and doesn't need to have anything to do with FF at all.

For example, say you have a P&S with a built-in zoom, and you want to move up to an M43 MILC, but don't want to lose that P&S zoom range, which you really like - how do you select an m43 lens that has the same FL range? You need to use equivalence.

Most people stop there at the FL conversion and don't worry about the F-stop conversion, either because they don't care or they don't know about it. But as I was trying to explain to kunzite, just because you personally don't care or don't know about it doesn't mean it's not real


.

The only format I shoot is APS-C. I don't think I own a point and shoot and I don't own any film cameras now. If I see a full frame shot, I do a mental calculation to see what sort of lens it would take to take a similar shot on my K-01 or K3. But I don't find those calculations to be particularly interesting or useful and it certainly doesn't help me when I am shooting, where it is more about composition/framing and less about matching some other format.

[TRex1123]

You stated (and maybe got a bit hyperbolic) that you feel bad that a person getting an f/2.8 lens for lets say a 4:3 camera is only really getting a f/5.6, as if they had been mislead.
Firstly, "aperture" is based on the physical properties of the lens.
It is not affected by the size of the surface on which the light from the lens is cast.

The whole thing is STUPID confusing. The video did do a nice job describing a ton of things that I have never really considered before. What the author is communicating is that the manufactures are "changing" the focal length but not the aperture. In the DOF (depth of field) equation that does not work, like he says you have to change both sides of the equation. When a manufacture of a Micro 4/3's lens says, this lens is a 35mm f1.8 lens, it is. However, when they say that this is effectively a 70mm f1.8 lens, that is false; when you are trying to make things equal....

When you are comparing the various formats you simply have to make things equal to compare them, especially for consumers. I don't believe the author said 35mm was the be all end all format BUT camera and lens manufactures seem to like to compare lenses to those on FF cameras.

[virgilr]

In terms of crop factor and magnification, I would like run this by everyone:

If two sensors capture the same pixel dimensions yet are of two different sizes, one with a larger surface (light gathering) area and one with a smaller surface area, would the smaller sensor magnify the image thus changing the sharpness of the image captured, or would it simply be a capture of a narrower field of view than the larger?

[Yassarian]

In terms of crop factor and magnification, I would like run this by everyone:

If two sensors capture the same pixel dimensions yet are of two different sizes, one with a larger surface (light gathering) area and one with a smaller surface area, would the smaller sensor magnify the image thus changing the sharpness of the image captured, or would it simply be a capture of a narrower field of view than the larger?

If all else being equal the smaller sensor effectively 'magnifies' the image by capturing a narrower field of view than the larger sensor. It doesn't change the sharpness - which is an entirely different thing. Would you say a photo taken by a 75mm lens is inherently sharper than a photo taken by a 50mm lens of the same resolution? No - sharpness has nothing to do with fov but it depends on the quality of the lens, etc etc...

[virgilr]

I received this response earlier on when asking the same sort of question- this time I decided to be clear by stating that the pixel dimensions capture were identical between the two sensors.

Let's assume we are printing for A3+, so I have to equalize for final output.

A3+ is 158,907mm^2
D800E sensor size is 864mm^2
K-5IIs sensor size is 370mm^2

To print at A3+ the image from a D800E has to be enlarged 183 times.
To print at A3+ the image from a K-5IIs has to be enlarged 430 times.

The FF sensor has 2.3 x more surface area to collect light.

As you magnify the image you change the appearance of the out of focus areas and the DoF. Ever looked at a small image and think it is in focus, but when you go larger to edit it you see that is is actually not in focus? This part of the compression effect that you get from magnification.

My point here is that the pixel dimensions are the same between the two sensors - because the surface area of the sensor (if the pixel dimensions are the same) would now not have any affect on the the amount of magnification required to print at any given size. This type of comparison couldn't be done with film - the surface area of the negative was the "pixel dimensions" - won't get into grain structure, etc (although that did play a role, I wouldn't ignore that) - magnification was an optic phenomenon of the development process, just as the lens is part of the capturing process.

With a digital sensor, the pixel dimensions ARE final resolution. There is no more "magnification" that can be done without intentionally degrading the image. Reduction is the only manipulation of the final capture that can be done.

I'm not negating the magnification effect of the narrower field of view with a smaller sensor- ,I was basically questioning the concept that the OoF rendering in an image would change depending on the sensor size. If I so choose, I can move my camera furher away from the subject to emulate the same field of view, but I am then going to absolutely change the rendering of OoF areas, as well as the depth of field represented in the image. I'm not disputing crop factor (FoV limiting is how I think about it in my head) - I'm disputing that, with the same captured pixel dimensions a smaller sensor would render differently than the larger sensor.

---

Last edited by virgilr; 05-21-2014 at 07:42 PM.

[jsherman999]

Make up your minds; either "equivalence" MUST (as in RFC 2119) be used in its complete form, or one's allowed to pick what to compare.

Didn't you just accuse equivalence of being 'too rigid' a few posts up? And didn't you yourself state that you might be interested in comparing only some attributes of the systems involved, not all? Are you contradicting yourself for effect, or did you just forget you said that?

That past exists in your imagination; it's me who's still waiting for an answer to the "what should they write on lenses used on multiple formats?".

Actually I did answer that - if you personally really want to write something on your lenses, you should write the FOV and aperture that corresponds to the format you're most comfortable with. In your case I assume that's aps-c. Equivalency doesn't require a 'base' or 'standard', but if you want it to have one in your mind you're free to choose whatever format you want for that standard. Many people think in terms of aps-c, and do the conversion 'in reverse', because they never shot film in the first place and the aps-c FOV for the focal length is more 'native' for them. Equivalency does not argue against that.

Remember - I'm talking about equivalence, and I'm not really an advocate for Tony's assertion that equivalent apertures should be written on the lens barrel, because F-stop really is a constant property that affects exposure in the equation (exposure = intensity, not talking about total light) and that info needs to be preserved as well.

You seem to be arguing against Tony still and mistakenly using equivalence as your whipping boy in that argument. I think it's time for you to carefully read the links I provided.

.

Looking through the viewfinder happened when I couldn't compare between formats because I didn't had the "equivalent" FF camera+lens combination;

That's fine, that's what I did as well. I started out with aps-c DSLR, I never shot film, I never needed to use any form of equivalence. I just bought what was available, affordable and what got my subjects in the frame from my usual shooting distances. When I started to shoot more than one format is when I benefited from knowing about it, and I basically had it thrust apon me by physics when I started using the same lenses on both formats.

Your challenge is an obvious trap. I would obviously use the "old ways" (those rejected in the video as being "arbitrary and meaningless)

Forget about the video when you talk to me. I'm talking about equivalence, have been for the past four posts or so, only referring to the video obliquely.

So, the question was:

Given: format m43: 25mm f/1.4. Q: What What kind of lens on a FF camera would you need to have, to be able to roughly take the same shot standing same distance away?

And the challenge was: Answer this correctly without using equivalence, something you've referred to in this thread as an 'abomination'.

yet you'll claim that I'm using "equivalence".

If you want to get the answer right, you'll have to. Even if you really, really, really don't want to call it 'equivalence'.

Even if focal length will stay focal length, aperture won't change its meaning and I won't even try to find an "equivalent" ISO. But here it is:
1. Getting the same angle of view is done using a simple geometric relationship between AoV and FL:
- get the largest rectangle with the same aspect ratio as the "source" frame, that can be fitted into the "destination" frame.
- compute the linear ratio R between this "destination" rectangle and the "source" frame. You may call it "crop factor", though it could differ from the widely accepted crop factor (e.g. it would be ~2.12:1 for 4/3 vs. 135).
- choose a focal length which is equal to the "source" lens' focal length multiplied with the ratio R.

And if you did that arithmetic right, you'll find the equivalent FL to 25mm on m43 would be 50mm on FF - that is, the FL that would give you the same AOV from the same distance. Let's continue.

2. Getting the same DoF.
Yes, you will use the ratio R from before; you know how. Just open or close the aperture as required.

You skipped the math in your reply, but if you did it right you'd see that the equivalent F-stop to get the same DOF (if you wanted the same DOF) would be f/2.8 on FF.

3. Getting the motion blur.
This is done by using the same shutter speed.

Yes. Which would be achieved here even if you were using a smaller f-stop if you allowed image brightening via higher ISO.

4. Getting the same exposure
Because the shutter speed is constant, light assumed constant - I will have to change ISO to compensate for the change in aperture. Anything else means a different exposure, i.e. a different result.

Yes. You are still moving down the equivalence railway here, good job. If the ISO on the m43 shot was 400, the FF shot would need ISO 1600 to get the same image brightness (notice it's brightness, not exposure - an aside a lot of folks have probs with. Strictly speaking 'exposure' is only determined by shutter speed and aperture, not ISO.)

5. Getting the same detail.
Can't control it. Well, I suppose I could add blur to match a less detailed "source" image...

There are some advantages to larger formats, for example for the same pixel count over different sensor sizes the lenses on the smaller format would need to be sharper to get the same lp/ph. But this in my opinion is the least likely thing to hold up to a real-life (vs theoretical) equivalence framework since sharpness varies so much between different lenses and even between the same model lens due to sample variation. Basically the number of pixels in the frame and the quality of lens is so much more significant that the larger-sensor advantage (ie smaller sensors needing sharper lenses) that the larger-sensor advantage probably falls almost entirely into the theoretical. Probably. I would like to be proven wrong. Add to that the fact that smaller-format lenses seem to be being made sharper now (part of the reason for their high cost) that you can't always count on it. See this page for an interesting lens-scaling example.

There's a thread on dpreview where some nice theory was posted - here's what would happen if you shot lenses of exactly equal quality on mft and FF with the same pixel count (y-axis is lp/ph) :





The area between those curves would vary depending on the lenses chosen and of course this is theoretical and depends on lenses with the exact same quality, and that's not a real-world-likely situation so take it for what it's worth.

6. Getting the same noise.
No control here either; if I'm lucky, I'll get similar noise levels to that from the original picture. If not, it can't be helped.

Actually if you kept the parameters as described above, and the sensors had roughly same-gen efficiency - the noise would be the same as well.

Now, if you didn't care about keeping the DOF the same, just the AOV and distance, you could shoot both at the same F-stop (f/1.4) and the FF image would have about two stops better noise performance and about two stops less DOF.

You just walked yourself through an equivalence discussion. You've been using it all along, partially, as I and others have tried to point out. Congrats, welcome to the gang.

It's too late (past midnight) to check if those pages are correct in their claims. My apologies, but I'll pass.

Please read them sometime.

.

[jppp]

As entertaining as reading this dispute is, it's kind of pointless. Forcing one's opinion on others is inconsiderate to say the least, and when both parties "know" they're right, futile. It's only photography, give it a rest. Go and take a picture or two.

[jsherman999]

As entertaining as reading this dispute is, it's kind of pointless. Forcing one's opinion on others is inconsiderate to say the least, and when both parties "know" they're right, futile. It's only photography, give it a rest. Go and take a picture or two.

Agreed **. Here's one I snapped yesterday, it should bring a cheerier mood than an equivalence discussion



** although this thread and subforum are for discussions like this, pictures happen in many other parts of the forum.. Some folks find it fun to learn about the underlying stuff by reading these discussions. I know I always did/do.

.

[Rondec]

If companies are honest about iso and if depth of field is not particularly important, then you will get similar exposures shooting with a 135mm f2.8 at 1/200second and iso 200 on APS-C and 200mm f2.8 at 1/200 second and iso 200 on full frame. Yes, there maybe a tiny bit more noise on the APS-C camera, but at lower iso settings (for me, 3200 and below), the difference is not noticeable. On the other hand, the real point of shooting is to get a good exposure, not to match some other format, which doesn't even exist at this point for Pentaxians.

(FA 77 shot at f2).

[northcoastgreg]

When Pentax calls their lens a DA 35mm f2.8 Limited Macro, they are not lying. If they were to call it a 53mm f2.8 equivalent, that would be lying.

Only in a very narrow, pedantic sense would it be lying. If Pentax told the "truth" about the "equivalent" aperture of the DA 35 f2.8, it would confuse far more people than it would enlighten. Nor would the truth that is concealed by calling the lens a 53mm f2.8 equivalent be of much use to the majority of photographers, but would, on the contrary, lead to another kind of deception. Many, perhaps even most photographers are interested in low apertures, not for attaining narrow DOF, but for shooting hand-held in low light. Such photographers will not only be confused, but even deceived by bringing in and insisting upon aperture equivalency.

FOV equivalency has a certain type of usefulness that many photographers can relate to. And it's the most important metric when comparing how lenses of a given focal length perform on sensors of varying sizes, as it is useful for just about any type of photography. But even FOV equivalency confuses the heck out of some photographers.

[audiobomber]

Only in a very narrow, pedantic sense would it be lying. If Pentax told the "truth" about the "equivalent" aperture of the DA 35 f2.8, it would confuse far more people than it would enlighten. Nor would the truth that is concealed by calling the lens a 53mm f2.8 equivalent be of much use to the majority of photographers, but would, on the contrary, lead to another kind of deception. Many, perhaps even most photographers are interested in low apertures, not for attaining narrow DOF, but for shooting hand-held in low light. Such photographers will not only be confused, but even deceived by bringing in and insisting upon aperture equivalency.

I disagree. It is a 35mm f2.8 lens, that is clear. It is also a 53mm f4.2 equivalent lens. FOV, DOF, perspective and even noise on a Pentax K-5 will be equivalent to a 53mm f4.2 lens on a D800. That is useful info, and seems highly understandable to me.

[Yassarian]

Well I'm assuming that at least some of us have used 35mm film cameras right? For me, I have only owned an apsc dSLR but have played with film cameras while in college. For better or for worse I'm used to that as my reference point. So at least for me, looking at the DA 35 f2.8 I automatically convert it to ~50mm f4.2 because that's how I "visualize" what I can achieve with this lens - since my reference point is the old 35mm format.

Now for some people this isn't an issue - if you have always shoot apsc then obviously why would you care about any other "equivalent"? But I'm used to the 35mm format so it's easier for me to visualize such. I suspect a lot of ppl are the same way, hence the debate.

Here's an example - from Sony's website:
http://store.sony.com/sel50f18-s-zid27-SEL50F18//cat-27-catid-All-Alpha-NEX-Lenses

Optics/Lens
Mount Material : Metal
Direct Manual Focus : Yes
Low Noise During Movie Capture : Excellent
Minimum Focus Distance : 15.3" (0.39m)
AF for Movie Capture : Yes
Lens Groups-Elements : 8 groups, 9 elements
Filter Diameter : 49mm
Lens Weight : 7.25 oz (202g)
Internal Motor : Yes (Stepping motor)
Dimensions (Max. Diameter x Length) : 2-1/2 x 2- 1/2" (62x62mm)
Aperture (Max.) : f/1.8
Aperture (Min.) : f/22
Maximum Magnification : 0.16x (APS-C)
Focal Length (35mm equivalent) : 75mm
Aperture Blade : 7 blades (Circular aperture)
Exterior Finish : Metal
Angle of View : 32°

Notice how they gave the 35mm equivalent of Focal Length, but they failed to give the proper 35mm equivalent of max aperture? If you gonna do 35mm equivalent then why not give BOTH the focal length *AND* the aperture?

I suspect it's because f1.8 sounds a bit more impressive than f2.8... ESPECIALLY if you are just browsing and you look at it and goes, wow a 75mm f1.8 lens - that's pretty good for $300. Well, it's simply not true... Sony isn't *LYING* to you it's just not telling you the whole picture to make their product look a little bit better than they really are.

Only in a very narrow, pedantic sense would it be lying. If Pentax told the "truth" about the "equivalent" aperture of the DA 35 f2.8, it would confuse far more people than it would enlighten. Nor would the truth that is concealed by calling the lens a 53mm f2.8 equivalent be of much use to the majority of photographers, but would, on the contrary, lead to another kind of deception. Many, perhaps even most photographers are interested in low apertures, not for attaining narrow DOF, but for shooting hand-held in low light. Such photographers will not only be confused, but even deceived by bringing in and insisting upon aperture equivalency.

FOV equivalency has a certain type of usefulness that many photographers can relate to. And it's the most important metric when comparing how lenses of a given focal length perform on sensors of varying sizes, as it is useful for just about any type of photography. But even FOV equivalency confuses the heck out of some photographers.

[Kunzite]

Didn't you just accuse equivalence of being 'too rigid' a few posts up? And didn't you yourself state that you might be interested in comparing only some attributes of the systems involved, not all? Are you contradicting yourself for effect, or did you just forget you said that?

I am consistently arguing against "equivalence" as presented in the video. This thread is about Tony's video.
If you're arguing for another kind of equivalence, even if you'd somehow win the argument, that can't be used as proof for the "equivalence" from the video. In particular, you cannot expand the basic idea of comparing two objects to support "equivalence".

Actually I did answer that - if you personally really want to write something on your lenses, you should write the FOV and aperture that corresponds to the format you're most comfortable with. In your case I assume that's aps-c. Equivalency doesn't require a 'base' or 'standard', but if you want it to have one in your mind you're free to choose whatever format you want for that standard. Many people think in terms of aps-c, and do the conversion 'in reverse', because they never shot film in the first place and the aps-c FOV for the focal length is more 'native' for them. Equivalency does not argue against that.

You are just refusing to admit that it makes no sense to write anything else than the real focal length and aperture on an interchangeable lens.

Remember - I'm talking about equivalence, and I'm not really an advocate for Tony's assertion that equivalent apertures should be written on the lens barrel, because F-stop really is a constant property that affects exposure in the equation (exposure = intensity, not talking about total light) and that info needs to be preserved as well.

You seem to be arguing against Tony still and mistakenly using equivalence as your whipping boy in that argument. I think it's time for you to carefully read the links I provided.

You claimed before that "You just can't stop at FL and not include F-stop if you're interested in not kidding yourself.". Again, which kind of equivalence are you talking about? Can it stay constant during our conversation, please?

Regarding the so-called challenge:
I did not use the "equivalence" I'm arguing against. I didn't use anything that wasn't available many decades ago. I didn't add "meaning" to well known notions: focal length still is a measure of converging light (no mention of projecting an image onto something), aperture is still the ratio between focal length and physical aperture, ISO works just like for film. All 3 were called "kind of arbitrary and meaningless" in the video.

I will make you remember the claims from the video (also regurgitated by other "equivalence" supporters):

Multiply focal length by crop factor
Multiply aperture by crop factor
Multiply ISO by (crop factor)2

I didn't do any of that.

- Focal length: I did not use "equivalence". I merely claimed that FXmm on format FmtX and FYmm on format FmtY would get you the same angle of view; that's an important distinction IMHO.
- Aperture: I did not use "equivalence". Instead of computing "equivalent aperture", I prefered to compute how many stops you would open/close the aperture. Alternatively, computing from scratch the aperture opening to get a given DoF is also possible but more complicated.
- ISO: definitely there's no "equivalent" ISO in my answer; I clearly explained that it can't be (or the exposure relation would break). ISO and noise aren't guaranteed directly proportional, thus it would be a mistake to assume that I would get the same amount of noise; this has to be measured and observed.

Another thing, I did not use the "crop factor", which is imprecise. The answer is not 50mm on FF, but ~46mm; I did the arithmetic right, you didn't. Oops!

You just walked yourself through an equivalence discussion. You've been using it all along, partially, as I and others have tried to point out. Congrats, welcome to the gang.

If you're claiming this, you don't understand "equivalence". But you're lucky, I know a pretty good video about it...