imagemeister Loc: mid east Florida

Imatest, Lens Rentals, MTF charts, Photozone, and yes, even some qualified user's opinions. NONE of these alone should be relied upon for absolute accuracy !

Mark7829 Loc: Calfornia

No, they shouldn't but in addition to, they need to become knowledgeable in all aspects of a lens, beginning with elements and groups, glass that reduces vignetting, distortion, aberration, coatings, internal or external zoom, motors, materials, etc. Yet it seems that only criteria that anyone seems to look for is sharpness and that is often subject to the OP own observation.

Sharpness is overrated. I have seen some incredibly sharp images but lacking in subject, story, composition and light, become meaningless. Some of the greatest photographs I have seen are not sharp at all.

Technology is overrated. We argue over minutia and miss the whole point - compelling emotionally engaging images. You can have the greatest and best but if it is not put to proper use, it almost becomes worthless and your time wasted.

imagemeister Loc: mid east Florida

Maybe your most compelling and profound comments on UHH .....!

Gene51 Loc: Yonkers, NY, now in LSD (LowerSlowerDelaware)

:XD:

dmsM43

It depends on the sensor in your crop-sensor camera. Most, if not all of them, are designed to work with lenses made for them. To say that you are using only the central part of the image of a full frame lens on a crop sensor is oversimplified. My experience with using full frame lenses on my M43 cameras, is that they work OK for lenses from 35mm and up in focal length. Wider angle full frame lenses just don't work very well whether you are using Leica type full frame lenses or retrofocus SLR lenses. I recall owners of the Sony Nex 7, an APS-C, camera reported the same problem.
It's interesting that with my full frame Sony A7R, I can get good results with old SLR lenses down to 20mm. That's because Sony incorporated "micro lenses" in front of their full frame sensor so that it would work with full frame wide angle lenses. Leica has done the same thing with their cameras as well. However, I don't know of any manufacturer of crop sensor cameras that has done this to make their cameras more usable with full frame lenses.
As I said, you can get good results with lenses of 35mm and longer focal lengths, but you will get better results with lenses designed for that specific camera and sensor.
I did look at the Tony Northrup video, and I don't know where he gets the change in f-stop, because there isn't any, unless he's talking about depth of field which is a different can of worms. In other words, that 300mm f2.8 full frame lens mentioned above is still going to be a 300mm f2.8 when it's adapted to a M43 camera. The only difference is that the angle of coverage on the M43 camera is equivalent to the angle of view of a 600mm f2.8 on a full frame camera. And the big advantage to putting that 300mm f2.8 on your crop-sensor camera is a huge saving in weight and size. I don't think anyone even makes a 600mm f2.8, but if they did it would be huge and really heavy.

selmslie Loc: Fernandina Beach, FL, USA

This is more likely to be an issue with mirror-less cameras whose lenses have been designed so a rear element is significantly closer to the sensor than it is for an SLR or DSLR. In other words, they use a lower Back Focus distance.

A mirror-less camera expects some of the light to be arriving at the sensor surface at an angle further from the perpendicular.

An SLR/DSLR lens (or to a lesser extent a rangefinder lens designed for the Leica M system) will be illuminating the sensor from a greater Back Focus distance as is evident from the thickness of the adapter required for a mirror-less camera. This is bound to upset the best laid plans of the designers.

But since the SLR/DSLR cameras can share lenses without the need for adapters, these considerations should not visibly affect the use of full frame lenses on a crop sensor DSLR. Each manufacturer uses its own standard Back Focus assumption.

I can't really speak from experience about how ultra-wide lenses fit into the mix but, if you think about the geometry, even an ultra-wide angle lens on a DSLR needs to leave room for the mirror, so the angles involved between the rear of the lens and the sensor are not significantly affected. The retro-focus design is really accomplished at the front of the lens.

dmsM43

I suspect you may be referring to Nikon cameras and lenses, I don't think that Canon has that much interchangeability between full frame and crop sensor cameras and lenses. I don't have any direct experience with Nikon, so I can't say how well their full frame lenses would work on their crop sensor bodies. I did lend a Nikkor 28-85mm MF full frame film lens to a friend with a D7200, but he hasn't got around to trying it, so until I hear from him, I really don't know.
I did read an interesting article going the other direction, where a photographer was using e-mount crop sensor lenses on his full frame Sony A7R II body. He just used the camera's crop sensor mode and and still got 18mp images while using the smaller, lighter, and less expensive APS-C lenses.
I would guess that would also work with Nikon. However, as I said before, if you want the best image quality, you should use the lenses designed for your camera, if you can afford them.

selmslie Loc: Fernandina Beach, FL, USA

My experience is mostly with Nikon SLR/DSLR (46.5 mm flange to sensor), Leica M (27.8 mm) and Sony E (18 mm). Canon FL/FD is 42 mm. Other than the specific numbers, the same principles apply. The difference between 46.5 or 42 mm and 18 or 19 mm is huge and can upset the sensor design assumptions.

Leica M and Voigtlander were designed without the mirror in mind and there is at least on ultra-wide lens I know of that sits so far back into the body that the meter will not register correctly. These lenses were originally designed for use with film.

joer Loc: Colorado/Illinois

I don't think he is talking about any change in actual aperture.

He talking about change in DOF equivalent to a corresponding aperture.

dmsM43

That is correct. But I don't know why Tony Northrup brought it up in the first place, since the DOF issue is there no matter whether the lens is a full frame lens or a crop frame lens. And since very few modern lenses even have DOF scales on them, why bring it up. This is not unique to digital photography, if you cropped your 35mm film, you would also have the same DOF issue as you would have using a crop sensor over a full size sensor. Anyway, I think it just added to the confusion.

Mark7829 Loc: Calfornia

I like TN. It takes initiative, and courage to even undertake some of the topics he has. I don't always agree with him but I appreciate someone who tries to shed light (pun intended) on complex and dynamic topics where it seems everyone is an expert.

dmsM43

I agree with you about TN. In this case, however, I think he could have left out the DOF issue on this topic and avoided some confusion. As a former school teacher, I always subscribed to the idea of covering one topic at a time which in this case involved the pro's and cons of using full frame lenses on crop sensor bodies, and in particular, image sharpness issues.

joer Loc: Colorado/Illinois

I agree here as well as your statement about applying the crop ratio to the f-stop relative to DOF only.

My understand of the DXO numbers has become clearer. The sharpness values can be misleading.

Don't get me wrong, the sensor and camera have a lot do with IQ as do other things.

Less sharpness is a characteristic of the lens. A given lens on a D700 is lo less sharp than it is on a D810.

What the DXO numbers tell you is that it will retain its detail at a much high image magnification.

joer Loc: Colorado/Illinois

Corrected.

Gene51 Loc: Yonkers, NY, now in LSD (LowerSlowerDelaware)

Mark,

One thing is "owning" the knowledge and understanding of said complex topics, then sharing this with others as writers like Thom Hogan, The good folks at Photozone.de, Bjørn Rørslett, Nasim Mansurov, Sean McHugh, the guys over at Imaging Resource and many others who have taken the time to do the research, conduct measurements and tests along the guidelines of the scientific method and provide fact-based information, or provide anecdotal and subjective experiences - clearly identified as such - so that we can learn, make prudent decisions when purchasing gear and software and however else we may use the info provided.

Then there are others.

This is the response I received this morning from DXO - along with the questions asked - regarding their PMpix ratings on lenses and bodies, quality of FX lenses on DX bodies, and how to use their rating systems. Of particular note is their explanation that FX lenses are NOT less sharp when used on a DX camera. The person responding was Sophie Cornillet-Jeannin,, Support and Assistance on the DXO Mark team.

You can decide just how valuable and accurate the information presented in the videos actually is.

1. Are full frame lenses (like a Nikon FX) sharper on crop sensor (like a Nikon DX) body? There is a photographer out there that cites your PMpix data to make the point that a lens that has a high PMpix number on an FX sensor will likely have a smaller PMpix number on a DX sensor, which means that the lens that is perfectly sharp on FX will be not as sharp on the DX.

Lenses are not sharper or less sharp on DX sensor. Usually, pixel pitch will be smaller on DX format than on FX format. That means that for a 24 Mpix sensor, it will be easier to achieve an high PMPix score on a full frame 24 Mpix sensor than on a 24 Mpix APS-C sensor.

2. If you place an FX lens on a DX body, does the crop imply that the image quality will be more uniform from corner to corner than the same lens on an FX body because you are using the center of the lens, which tends to have better performance than the edges and corners?

Sharpness uniformity is of course better on DX format sensor for a FX lens as we will crop in the full frame image output.

Or for that matter would the FX lens be sharper than a DX lens of the same focal length on a DX camera?

It depends on the lens optical design.

3. Can the PMpix be regarded as a “crop” or “digital zoom” function? For example, if an FX lens has a PMpix of 7 on an 18mp crop camera, does this mean that you are losing 11mp as if you either cropped in post processing, and therefore losing detail?

Not exactly, it means that on this sensor surface, we could replace the 18 Mpix sensor by a perfect lens and a 7 Mpix sensor

Abot crop: it is true that if you crop you loose sensor surface, so on a theoretical point of view you loose resolution and also information capacity.

For more information please refer to our documentation here:
http://www.dxomark.com/About/Sensor-scores/Viewing-Conditions


I think TN needs to rethink and re-shoot his videos on this subject, before anyone else buys into that nonsense.