[KiloHotelphoto]

Correct, but some people's minds are closed to the concept, so the arguments continue. This article by DPR in 2014 did help more people to understand: What is equivalence and why should I care?: Digital Photography Review

That was a good read, sounds a lot like what Northrup said.

[Digitalis]

The lens doesn't perform any differently as a result of the crop.

This is what I find so offensive about the thought that using a full frame lens on APS-C automatically reduces the lenses performance. By using a smaller sensor all you are doing is cropping, the lens will perform pretty much the same - though optical issues such as vignetting and corner softness will be diminished compared to full frame, these optical shortcomings are still present: however, they are not recorded due to the smaller size of the format.

I use 8X10 lenses on 5X7 and 4X5 format because you can get away with some rather ridiculous camera movements that wouldn't be possible with standard 5X7 or 4X5 lenses - the image quality from the 8X10 lenses I work with is so good you could crop a 24X36mm portion from a full 8X10 negative and get a superb 16"X20" print out of it*.

* Bear in mind the lenses I work with are diffraction limited and I also use Kodak Technical pan film, but I have even printed 11"X14" prints from T-max 100 with the same degree of cropping and the results were excellent.

[cyberjunkie]

AFAIK, the only disadvantage to using FF lenses on crop cameras is that the FF lenses admit more non image forming light into the camera which can cause flare. However, by using a sufficiently long lens hood, this problem is mitigated.

And of course the big advantage is the crop sensor gets the best image quality from the centre of the lens.

I'm quite satisfied with what I get from my FF lenses--I only own one DA lens.

Finally I found a post that highlights the same points I had in mind. This will spare me cramps, typing on the virtual keyboard of my cellphone

Just two more details:
1) most FF optics available in PK mount are film-era lenses; their design, not their image circle, is the problem (especially true for zooms and super wides)
2) a good multicoating on the back glass and appropriate baffles on both camera and lens bodies should tame the non-image-forming light rays coming from an FF optic. Of course a suitable lens hood is the first step. Though I've seen vintage film lenses with poor blackening, and adapters that are a total disaster in this regard.
A typical example would be a macro lens for Pentax 6x7, Pentax 645 or Mamiya 645. All of them are still sharp enough (the 6x7 one a little less, IMHO), but the adapter is the weak link. Chinese rings are generally good and cheap, but sometimes need to be closely inspected for light leaks and reflections (shiny, non blackened material).
With some M42 lenses any adapter ring, even the original Pentax, lets plenty of light in. My remedy is an O-ring of appropriate size, or a black hair band.
Many times I found that a supposedly poor lens was plagued by light leaks or inter-reflections. Once the problem was solved the same lens proved much better!

The area of the APS-C sensor doesn't differ so much from the FF/135mm, I don't think it would justify going for different levels of sharpness, at the design level. I don't think modern day FF lenses, of any maker, are designed to be less sharp than APS-C optics. Though the larger the format they are designed for, the more light enters the camera. Proper construction and appropriate lens hoods are paramount.

EDIT:
The last message, posted by @Digitalis, reminded me that sometimes 8x10" optics could give a certain level of veiling flare, if used on 4x5" or film backs. Even a black pleated bellows can cause inter-reflections, if the material is not fully matte.
I'm sure modern time Sinar or Linhof cameras have addressed the problem (and recent LF lenses are multicoated!), but I own a few vintage cameras that have the bellows made of a kind of shiny black material, and some areas that should be blackened, unfortunately are not. On such cameras it's better to use lenses with appropriate coverage, and it's always a good idea to use the kind of adaptable lens shades used in cinema, instead of a conventional screw-in hood.

Cheers

Paolo

[Digitalis]

sometimes 8x10" optics could give a certain level of veiling flare, if used on 4x5" or film backs. Even a black pleated bellows can cause inter-reflections, if the material is not fully matte.

yes, older lens and camera and lens combinations can have problems with veiling flare. The multi-coatings used on current lenses from Schneider and Rodenstock et al have significantly reduced this problem. However older Nikkor,Schneider, Fuji, Horseman, Rodenstock, and Goerz* can suffer quite badly from veiling flare, this can be remedied with the use of pleated non-reflective bellows and pleated bellows lens hood. I never use screw in hoods.

*and lets not forget the famous Goerz hypergon.

---

Last edited by Digitalis; 11-07-2017 at 01:30 AM.

[cyberjunkie]

*and lets not forget the famous Goerz hypergon.

And the Dagor, no funky whirlwind thingy... but it dominated an era, original or cloned, along with the Zeiss Tessar.
Still today, in a controlled studio environment, I'm sure that a first generation Dagor-type Symmar could positively surprise. Of course its MC grandchildren are better, but with appropriate lighting the difference shouldn't be so huge

Cheers

Paolo

[Rondec]

MY suspicion is the DA 70 while test chart sharp doesn't have the same 3D portrait spec the 77 does. But I've never had a copy of either lens. My memory is guys wo appreciate the 70, don't appreciate the 77, and vice versa. Apparently there's two kind of people in this world to go along with the two kins who either don't like either or like them both.

Did i leave anyone out?

I own the FA 77 and owned the DA 70 prior (before the K-1's release so I didn't try it on full frame). The DA 70 on APS-C has less fringing and feels like it has more consistent center/edge performance. The FA 77 has stellar center sharpness, even wide open although if you want the borders sharp you do need to stop down a little (this is usually fine as it is a portrait lens).

My wife does a lot of portrait/wedding work and she preferred the shallower depth of field you could get with the FA 77 on crop cameras for subject isolation. She used the FA 77 and 50-135 a lot for those situations. Otherwise we were pretty happy with the DA 70.

To the whole other subject, good lenses from the film era tend to perform well now while weak lenses from that era still are weak now. Shooting them on APS-C does crop away the borders and that could help lenses that had decent center sharpness, but poor border performance, but certainly poor lenses from thirty years ago are probably still poor lenses now -- regardless of the size sensor you are shooting them on.

[BigMackCam]

To the whole other subject, good lenses from the film era tend to perform well now while weak lenses from that era still are weak now. Shooting them on APS-C does crop away the borders and that could help lenses that had decent center sharpness, but poor border performance, but certainly poor lenses from thirty years ago are probably still poor lenses now -- regardless of the size sensor you are shooting them on.

Absolutely.

A few related observations (straying off the thread's purpose, I accept):

- Lab tests and test charts might measure optical performance, but they don't measure character (witness the FA43)
- Even so-called weak lenses can have their charms; sometimes, it's those very weaknesses that give them their unique character and creative appeal (my Helios 40-2, for example)
- There are, IMHO, relatively-few truly weak lenses that offer nothing creatively regardless of the sensor format they're used on

[normhead]

The one example that really stood out is for a Olympus Stylus 1 6-64mm f2.8, they advertise it as 28-300 f2.8 equivalent but it would fit in your pocket. Anyone who has seen a 300mm 2.8 FF lens knows how big they are Tony said the real equivalent for the Olympus would be 28-300 f13, it sound right to me.

Once again Tony doesn't understand photography. He thinks aperture is a measure of depth of field, in optics it's related, in photography It isn't,

Try the Wiki explanation.

In Photography
The aperture stop of a photographic lens can be adjusted to control the amount of light reaching the film or image sensor. In combination with variation of shutter speed, the aperture size will regulate the film's or image sensor's degree of exposure to light. Typically, a fast shutter will require a larger aperture to ensure sufficient light exposure, and a slow shutter will require a smaller aperture to avoid excessive exposure.

Notice how in photography, (not optics) an aperture is used to control exposure. This is photographically the correct use of the term.

If I have a scene in front of me. ƒ2.8 100 ISO and ƒ8 will work on any camera. Because ƒ2.8 means exactly the same thing on any camera.

When Tony says it would be ƒ13, he's confusing optics, with photography. They use the term slightly differently as you can see from the Wiki article.

In optics aperture basically determines Depth of field and out of focus areas.

The aperture determines how collimated the admitted rays are, which is of great importance for the appearance at the image plane.

From an optics point of view 2.8 ff may in fact be the same as ƒ4 APS-c. The term "aperture" doesn't mean the same thing in photography. The article explains it really clearly.IN photography ƒ-stops prove a consistent system of exposure across all formats. Tony i trying to make it into a consistent system of depth of field across all formats. He needs to find another term to use, in photography, Aperture" is already taken.

So why does this matter. I look at a scene which because of lighting conditions I'm going to shoot at ƒ4, 400 ISO and 1/30s. It doesn't matter if I shoot my K-1, my K-3 or my XG-1. That will be my exposure. But on my XG-1 I will have much more DoF if that is desirable. The K-1 will have a much different look for the same exposure. What using the FF lens on the smaller sensor does is change the look of the picture, not the exposure. And that's how Tony works his slight of hand. He ignores that an image is different and he's actually comparing different looking pictures. If you adjust the larger format image so the images are the same, his math falls apart.

So my advice is don't be confused by Tony's confusion. Use the optics term if you are discussing optics, use it in the correct photographic context if you are discussing photography and lenses used for photographic purposes.

I've explained this a million times with mathematical examples. At this point I'm ready to just say, if you don't get it, believe what you want. No sweat off my back But when you manually set your bridge camera to ƒ13 instead of f 2.8 for your Tony comparison and you don't have enough light to get an exposure, don't blame me. I did my best. But of course you won't do that and neither will Tony. Somewhere in his mind he knows its all BS.

[Rondec]

You have an exposure triad -- shutter speed, aperture, and film speed/iso. In the days of film, you could technically change your film in the middle of a shoot if you need something faster, but it certainly wasn't as easy as it is now. Shutter speed usually has some floor that you don't want to go below. Aperture is something we can adjust when we want to maintain a certain shutter speed, but it has a side effect -- open it up enough and your depth of field may be too shallow for an image to be successful. These days, though, I would argue that people fiddle with their aperture more to get a certain depth of field than because they need to do so to maintain a certain exposure. Stop down for landscapes, open up for portraits, etc.

As far as the exposure triad, it is true that it works across cameras. So if you shoot f2.8 iso 1600 and 1/100 second shutter speed you will get the same exposure on a Q, K3 and K-1. But the images aren't going to actually look the same. In fact, even if you are shooting with the same angle of view, they still will look quite a bit different. I think this is part of what Tony is trying to touch on -- he just doesn't do a very good job of explaining it.

[audiobomber]

A 100mm f4 lens has a maximum aperture (iris opening) of 100mm / f4 = 25mm. This is a characteristic of the lens, irrespective of which camera it is mounted on.

> Olympus Stylus 1 6-64mm f2.8, they advertise it as 28-300 f2.8 equivalent

It is a 6-64mm f2.8 lens, and it is 28-300mm FF equivalent. It is not f2.8 FF equivalent, because the iris opening is tiny. To show the physical size of the aperture, we use the actual lens specifications:

• 6mm / f2.8 = 2.1mm
• 64mm / f2.8 = 22.9mm

If you want to use FF equivalent focal lengths,we must use the actual diameter of the iris opening to determine equivalent aperture:

• 28mm / 2.1 = f13
• 300mm / 22.9 = f13

The crop factor of the Stylus sensor is 28mm / 6mm = 4.67

If you were to mount the 100mm f4 lens above on the Stylus, the FF equivalent would be

• 4.67 x 100 = 467mm
• 4.67 x f4 = f19

The iris opening remains unchanged:467mm / f19 = 24.6mm (There is a small rounding error because I used f19 instead of f18.68)

Exposure remains as stated by the f-stop. Equivalence theory will allow you to compare DOF, noise performance, diffraction and DR between formats, given technologically similar sensors. There's no problem doing this for the current incarnation of mainstream cameras, but noise and DR do not hold for a flyer like Sigma Foveon. Here is some supporting evidence: Depth of field?...can we close this argument once and for all - Page 2 - PentaxForums.com

[normhead]

A 100mm f4 lens has a maximum aperture (iris opening) of 100mm / f4 = 25mm. This is a characteristic of the lens, irrespective of which camera it is mounted on.

> Olympus Stylus 1 6-64mm f2.8, they advertise it as 28-300 f2.8 equivalent

It is a 6-64mm f2.8 lens, and it is 28-300mm FF equivalent. It is not f2.8 FF equivalent, because the iris opening is tiny. To show the physical size of the aperture, we use the actual lens specifications:

• 6mm / f2.8 = 2.1mm
• 64mm / f2.8 = 22.9mm

If you want to use FF equivalent focal lengths,we must use the actual diameter of the iris opening to determine equivalent aperture:

• 28mm / 2.1 = f13
• 300mm / 22.9 = f13

The crop factor of the Stylus sensor is 28mm / 6mm = 4.67

If you were to mount the 100mm f4 lens above on the Stylus, the FF equivalent would be

• 4.67 x 100 = 467mm
• 4.67 x f4 = f19

The iris opening remains unchanged:467mm / f19 = 24.6mm (There is a small rounding error because I used f19 instead of f18.68)

Exposure remains as stated by the f-stop. Equivalence theory will allow you to compare DOF, noise performance, diffraction and DR between formats, given technologically similar sensors. There's no problem doing this for the current incarnation of mainstream cameras, but noise and DR do not hold for a flyer like Sigma Foveon. Here is some supporting evidence: Depth of field?...can we close this argument once and for all - Page 2 - PentaxForums.com

There's some nice work in that post audiobomber... but it become problematic when you start talking about ISO levels that fall below the threshold of visible noise, which to my mind is big mistake. In photography we tend to work with the thresholds of human perspective. Focus is an issue where you can see it. The photographic terms is "acceptable focus." Noise tends to be evaluated as a 1 or a 0. An image is too noisey or it isn't. I look at your numbers where you've equalized noise, and if you set your criteria for acceptable noise at say 50 DB (just picking a number) then all images below that number are equal. They are 1's, and all above are zeros.

At that point discussing the number of DBs of noise become irrelevant. Sure one is higher or lower, but it's not a relevant statistic. That is made even more moot by the fact that reducing images size reduces visible noise, so for most of us we will have noise levels much below the published results.

Here's a K-3 image taken at 5.6 and 640 ISO


And an XG-1 image taken at ƒ6.6 and 400 ISO


I'm sure the XG-1 image has a lot more noise, theoretically but I'm also sure it's still acceptable when I look at the images. The raw science needs to be tempered with a bit of practical experience of what is acceptable and what isn't. I tend to look at this in a practical way by setting limits for each camera, I won't shoot the XG-1 over 400 ISO, I won't shoot the k-3 over 640, and I won't shoot the K-1 over 3200 ISO. In photography this is the most simple way of dealing with noise. Clearly, the backlit XG-1 sensor provides less noise than your math would suggest it should. Just as a K-P has different noise characteristics than a K-3. Practically, not scientifically. So while your math is excellent. it needs to be tempered with your experience with the sensors. Practical photography is way more complicated than just cranking out numbers and using them to make decisions. You need empiracle evidence so you know what you're talking about.

[cyberjunkie]

I don't have a degree in optics, but I know that the laws of optics are always valid, in astronomy as well as in photography.
Aperture is a number that expresses the ratio between the focal length of a lens and the diameter of the entrance pupil of the same lens. It affects both exposure and depth of field, as we all know.
While the concept of "equivalent focal" can make sense, "equivalent aperture" is in my opinion the wrong way to explain something real, and we create a lot of confusion if we mix the two things together, as if they were the same.

Cheers

Paolo

[UncleVanya]

The equivalence of aperture is meant to show that fast lenses on small sensors aren't as fast as they seem when taking ISO limitations of the smaller densely packed sensors into account as well as to show the delta on depth of field. The problem is that sensor density isn't a fixed property nor is ISO performance strictly based on sensor size. For example the K-3 and the K-5 are the same size sensor with very different noise characteristics due to pixel density and technology.

As a heuristic it's a bit of a poor way to compare.

My LX-7 has an f1.4 - 2.3 lens. Technically since I have to shoot at fairly low ISO to get good results this is not as impressive as it sounds - but in practice after trying it against some other cameras it is a bit more practical than the "equivalence" might suggest. On the whole it is better just to understand the ISO performance and DOF of any given platform and take it from there.

[Na Horuk]

a) FF lenses project a larger image circle than the crop sensor records. This can in theory mean lower contrast due to excess light in the mirror box, but with modern lens coatings and lens hood, this is not really a problem. Especially if you get a crop focal equivalent hood. This can be a problem if you use some really old low contrast, high flare lenses; but you can't expect high contrast and saturated colours from those even on FF.
b) FF lenses can be bigger and heavier or more expensive. So it might not make sense to pay a lot of extra money for a really heavy lens, if you can find an equally good lens at a smaller size and lower cost made for APSC. This does not hold true for all lens designs, not anymore. You can find very light 50mm and very heavy 50mm lenses, for example. This is only a problem with some zoom lenses, for example DFA 70-200mm is huge, and it might make more sense to get a more affordable, smaller APSC lens like DA* 50-135mm instead. But there are too many other variables for this argument (of FF lenses being bigger and expensive) to hold up. The DA 50-135mm does not have some features that the 70-200mm has. And on APSC, the 70-200mm still has some extra reach over 50-135mm. If those factors are important to you, get the right lens
c) The aperture argument about APSC having lower aperture is something you can safely ignore.

The main reasons for using FF glass on APSC camera are:
1) You are planning on buying FF camera soon and want to be ready for it (or, you already have FF or film camera)
2) There is no APSC-only equivalent. For example, you might really like the rendering of the FA 31mm limited, but there is no APSC-only lens with same FoV, aperture, and rendering.
3) The FF lens is cheaper and gets the job done. This mostly applies to lenses from the film era.

Tl;dr: Ignore equivalence, use any lens that pleases you. If you are worried with using FF lens on APSC camera, you can buy a tighter lens hood for total ease of mind.

[audiobomber]

There's some nice work in that post audiobomber... but it become problematic when you start talking about ISO levels that fall below the threshold of visible noise, which to my mind is big mistake. In photography we tend to work with the thresholds of human perspective. Focus is an issue where you can see it. The photographic terms is "acceptable focus." Noise tends to be evaluated as a 1 or a 0. An image is too noisey or it isn't. I look at your numbers where you've equalized noise, and if you set your criteria for acceptable noise at say 50 DB (just picking a number) then all images below that number are equal. They are 1's, and all above are zeros.

My numbers are correct and objective. SNR can be measured, regardless of format or ISO level. Deciding whether the noise in any particular image is acceptable is totally subjective. We all have different standards. Lots of Q owners mount a long tele lens on a Q and love it. I tried with several good tele lenses and generally considered the results horrible, even in good light.

I'm sure the XG-1 image has a lot more noise, theoretically but I'm also sure it's still acceptable when I look at the images. The raw science needs to be tempered with a bit of practical experience of what is acceptable and what isn't. I tend to look at this in a practical way by setting limits for each camera, I won't shoot the XG-1 over 400 ISO, I won't shoot the k-3 over 640, and I won't shoot the K-1 over 3200 ISO. In photography this is the most simple way of dealing with noise. Clearly, the backlit XG-1 sensor provides less noise than your math would suggest it should. Just as a K-P has different noise characteristics than a K-3. Practically, not scientifically. So while your math is excellent. it needs to be tempered with your experience with the sensors. Practical photography is way more complicated than just cranking out numbers and using them to make decisions. You need empiracle evidence so you know what you're talking about.

I am mystified that you limit the K-3 to ISO 640. As I've said a couple of times, you need better NR.

The duck picture is subjectively pleasing, like a watercolour painting. It's nice artwork, but technically flawed. The colours are pastel and the feathers are blurry, even at 2mp resolution. These are the result of the small aperture, elevated ISO and lens diffraction.

BSI technology provides no more than a 1/3 stop improvement. I've researched this extensively in the past on DXOMark and have no interest in doing so again.