Either or...

[Shadowman]

Two very recent threads inspired me to start this thread, I think both Cambridge threads somewhat relate to my query.

Technical versus compositional excellence.
Focusing issue

The query is: FX lens on a DX body, is it the better option for performance? I've included two views found through an online search.

https://www.youtube.com/watch?v=YDbUIfB5YUc
https://www.youtube.com/watch?v=N5FvEHC4kCQ

Personally, I've used FX lenses on a DX body for years and I love the output and I guess that's all that matters, however after viewing Tony's video I wonder if I'm short changing myself.

[Manfred M]

John - every time I see one of Tony Northrup's videos I wonder if he is misleading people on purpose or if he simply knows less about photography than he claims he does. Personally I think it is a bit of both.

First of all, you do not recalculate the "effective f-stop" on a lens for FX or DX. The f-stop is simply a purely mathematical calculation of the input aperture of the lenses diameter divided by the focal length. Neither of these two basic components of the lens change, regardless of which body you put the lens on, so the f-stop cannot change. The changes in the DoF is, on the other hand, correct. That is due to the effective magnification difference between the sensor sizes.

If you read up what DxO Mark writes about how their P-MPix is calculated, it quickly becomes apparent that once more we are being mislead just a little. Most of the difference in scores is related to the difference in sensor size and sensor resolution, rather than any actual change in lens performance. Nicely said if you want the same size of image from a DX image as an FX image, you have to magnify it a lot more, so you will definitely see a reduction in resolution.

Every one of his videos I've seen contains at least one significant factual error / omission like that, so I look at all of his conclusions with a great deal of suspicion.

[xpatUSA]

Two very recent threads inspired me to start this thread, I think both Cambridge threads somewhat relate to my query.

Technical versus compositional excellence.
Focusing issue

The query is: FX lens on a DX body, is it the better option for performance? I've included two views found through an online search.

https://www.youtube.com/watch?v=YDbUIfB5YUc
https://www.youtube.com/watch?v=N5FvEHC4kCQ

Personally, I've used FX lenses on a DX body for years and I love the output and I guess that's all that matters, however after viewing Tony's video I wonder if I'm short changing myself.

Sometimes I wonder about that too, with respect to DG lenses on DC bodies of course.

Especially when Sigma true macro lenses are all DG with no DC versions offered.

Therefore, shooting macro with my SD10 or SD14 (1.7 crop) means that much of the captured light is "wasted" on lighting up the interior of the camera, some might say. Others might say different- this is, after all, a photographic forum.

On the other hand, using only the center part of a DG lens should have less abberation than a DC lens of equal "quality" but there may be a quality issue in that lenses that are good enough for 135-format cameras are less likely to be "good enough" for smaller sensors with their (generally) higher spatial resolution (pixel pitch).

Just to muddy the waters, Falk Lumo says:

"It is a myth that APS-C cameras "crop the sweet spot" of the image field of lenses which cover a larger image circle or are made for a full registration distance."

http://www.falklumo.com/lumolabs/art...nce/index.html Section 4.2.2

But doesn't clearly say why.

Ho Hum . . .

[Shadowman]

John - every time I see one of Tony Northrup's videos I wonder if he is misleading people on purpose or if he simply knows less about photography than he claims he does. Personally I think it is a bit of both.

First of all, you do not recalculate the "effective f-stop" on a lens for FX or DX. The f-stop is simply a purely mathematical calculation of the input aperture of the lenses diameter divided by the focal length. Neither of these two basic components of the lens change, regardless of which body you put the lens on, so the f-stop cannot change. The changes in the DoF is, on the other hand, correct. That is due to the effective magnification difference between the sensor sizes.

If you read up what DxO Mark writes about how their P-MPix is calculated, it quickly becomes apparent that once more we are being mislead just a little. Most of the difference in scores is related to the difference in sensor size and sensor resolution, rather than any actual change in lens performance. Nicely said if you want the same size of image from a DX image as an FX image, you have to magnify it a lot more, so you will definitely see a reduction in resolution.

Every one of his videos I've seen contains at least one significant factual error / omission like that, so I look at all of his conclusions with a great deal of suspicion.

I checked out his video which contained additional reasoning for his (Tony's) comments, but even with this explanation there are also other considerations to take into account. For instance, if comparing shots taken with an FX lens 70-300mm lens at 200mm and a DX 55-200mm at 200mm; is the image quality of the lens the same at those focal lengths or would the DX lens be better tested at its optimal focal length and then adjust the FX lens to the same.

https://youtu.be/f5zN6NVx-hY

[Manfred M]

I checked out his video which contained additional reasoning for his (Tony's) comments, but even with this explanation there are also other considerations to take into account. For instance, if comparing shots taken with an FX lens 70-300mm lens at 200mm and a DX 55-200mm at 200mm; is the image quality of the lens the same at those focal lengths or would the DX lens be better tested at its optimal focal length and then adjust the FX lens to the same.

As for image quality for a specific lens, versus another lens at a specific focal length. That is an extremely difficult question to answer. The design itself (usually driven by the selling price) means that the lens designer will have to work with different trade-offs. This will range from mechanical robustness, distortion, resolution, and all those other operating characteristics, and these will vary both by focal length and aperture (with the worst performance generally occurring at maximum aperture. Improving resolution could mean introducing more aberrations or distortion, so the design team will determine which tradeoffs they are willing to live with. Nicely said, there simply is no answer simple answer to your question.

Most of what he says in this video is correct, although he contradicts himself a couple of time and is unclear about certain concepts. I suspect some of this incorrectness might be "dumbing things down for his audience".

The biggest mistake he makes is when he compares image quality between various sensors, as the determining factor on image quality / noise is the actual pixel pitch of the sensor. The complication here is that this is generally true only for cameras that are more or less the same generation (i.e. similar technology is being used) AND the sensors have an very similar pixel pitch (i.e. the sensel light receptors are the same diameter).

[xpatUSA]

. . . AND the sensors have an very similar pixel pitch (i.e. the sensel light receptors are the same diameter).

Hate to be pedantic but most "sensel light receptors" are square or slightly rectangular - not round.

[Manfred M]

Somehow the second part of my answer didn't get loaded - weird. That concerns trying to compare different lenses.

Most lens tests tend to try to boil things down to fairly simple and easy to understand comparisons, but in truth, these results are not particularly meaningful or useful. Think about a car-buyers guide that tries to rank all cars with one or just a few numbers. That would be totally meaningless and fortunately car descriptions are handled in a more mature manner.

I find that a lot of camera and lens reviewers test parameters that are easy to test, rather than being meaningful. If you look at the DxO Mark website, they explain why they have come up the concept P-MPix because they feel that people would not know how to read a MTF chart. That in some ways is unfortunate, as some of their audience has not issues with that, but we are not given a choice.

The bottom line is that the lens designer has had to make many tradeoffs when the specific lens was designed. With very few exceptions, the number one parameter the designer has to meet is a price point (which implies a manufacturing cost + markups + distributor margins++++). Mechanical robustness, optical performance, lens speed, distortion and aberrations are all taken into account and these usually vary by shooting aperture and focal length.

This generally means trying to directly compare one lens to another is really not particularly easy.

[rpcrowe]

If I expose an image correctly using f/16 @ 1/100 second on a full frame camera, that correct exposure is f/16 @ 1/100 second using the same lens on a crop camera; as long as I am shooting the same subject under the same light conditions...

Therefore the difference between crop and full frame apertures, as far as exposure goes, is null...

[Manfred M]

Hate to be pedantic but most "sensel light receptors" are square or slightly rectangular - not round.

I definitely understand that Ted. Trying to keep the explanation short and understandable. I spent too many years having to explain technical issues to non-techies so being pedantic (except where the point was absolutely critical) did not make sense.

[Shadowman]

I decided to compare one of my lenses with the other options and my choice wasn't so bad based on the DXO comparison.
I have the f/1.8G. I know I've read other articles comparing these lenses and the 1.4G usually gets the higher rating. I only included this comparison as I was planning to compare one of my other lenses against my kit lens and wanted to see what characteristics were being analyzed. The one characteristic not listed in DXO and is included in other rating systems is print output based on aperture setting and ISO performance.

http://www.dxomark.com/Lenses/Compar...52_850_823_850