Tony Beach
Veteran Member
- Messages
- 11,967
- Solutions
- 5
- Reaction score
- 7,035
I think you meant jibe, not jive. If so, I agree.
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
FF Lens on APS-C body - confusing
- Thread starter keiththom
- Start date
Tony Beach
Veteran Member
- Messages
- 11,967
- Solutions
- 5
- Reaction score
- 7,035
Not always, not even usually.
Don't take this the wrong way, because I agree with all the stuff in your post you wrote after the above, but this sums it up nicely.
+1
A poor DX lens is no panacea for the ills you would suffer from a poor FX lens, and a good lens is a good lens regardless of which format you use it on. In fact, many FX lens issues largely go away when used on a DX sensor.
He's a technology author with over thirty books to his credit. More to the point, he actually provides the math to support his conclusions in his videos and publications which puts him a step above the other 99% of YouTubers. Where's yours.
Judging by your 8,000+ posts to this forum your services aren't in such high demand out in the real world. Now p off.
Last edited:
Now that's amusing as I posted about 5000 times when I was CEO of a group of manufacturing companies: it's just a matter of time management. Now I realize that Tony is a talented software guy and a bit of a media celebrity with an interest in photography. To his credit, in the previous thread he tried to argue his case rather than citing his authority.
altair8800
Veteran Member
- Messages
- 2,000
- Reaction score
- 443
n057
Veteran Member
altair8800
Veteran Member
- Messages
- 2,000
- Reaction score
- 443
I agree with all you said. I think the problem is that when FX lenses are tested on DX cameras, they are placed 1.5x the distance used when on FX cameras. Amazing what a 1.5x TC equivalent can do for FX resolution when compared to DX.
One needs to understand the weakness to see how pointless the idea is.
You can look up any lens test and you will see that sharpness is best in the centre of the frame and falls off towards the corners. Other problems like vignetting are also more prominent in the corners. If you put a FX lens on a DX camera, you are practically cropping away the corners of the image circle created by the lens. So in the end only the better centre of the image circle made by the lens ends up on your sensor.
On a DX lens the design of the lens matches the sensor size, so that the corners of the image created by the lens actually match your sensor size. Which means that the fall off in sharpness towards the corner actually is visible in your image.
It is also wrong that you magnify imperfections, just imagine 2 image circles trying to cover your rectangular sensor.
The DX one does cover it, but the corners of the rectangular sensor are close to the edges of the image circle.
The FX circle has a larger diameter, so the corners of the sensor have more distance to the edge of the image circle.
Set to the same focal length both lenses will create the same image on the DX sensor.
You can look up any lens test and you will see that sharpness is best in the centre of the frame and falls off towards the corners. Other problems like vignetting are also more prominent in the corners. If you put a FX lens on a DX camera, you are practically cropping away the corners of the image circle created by the lens. So in the end only the better centre of the image circle made by the lens ends up on your sensor.
On a DX lens the design of the lens matches the sensor size, so that the corners of the image created by the lens actually match your sensor size. Which means that the fall off in sharpness towards the corner actually is visible in your image.
It is also wrong that you magnify imperfections, just imagine 2 image circles trying to cover your rectangular sensor.
The DX one does cover it, but the corners of the rectangular sensor are close to the edges of the image circle.
The FX circle has a larger diameter, so the corners of the sensor have more distance to the edge of the image circle.
Set to the same focal length both lenses will create the same image on the DX sensor.
M
Michael Oersigh
Guest
Hi
Well that's interesting, does Nikon impute an optical formula that takes into account sensor size or do they just concentrate on glass quality using their optical tests? If sensor size was taken into account then we have nothing to worry about, but that would mean TN would be right. Turning the argument on its head, FX lenses would be a compromise because of the DX sensor. If they concentrated only on glass quality then Nikon would know about the magnified imperfections. Nikon have never mentioned this in any of their technical journals.
Well that's interesting, does Nikon impute an optical formula that takes into account sensor size or do they just concentrate on glass quality using their optical tests? If sensor size was taken into account then we have nothing to worry about, but that would mean TN would be right. Turning the argument on its head, FX lenses would be a compromise because of the DX sensor. If they concentrated only on glass quality then Nikon would know about the magnified imperfections. Nikon have never mentioned this in any of their technical journals.
Those details will certainly not be resolved in the D7100's sensor with a Nyquist limit of about 128 lp/mm. At best these very high resolution lenses have a measurable - but by no means linear - impact on detail bounded by the sensor's limit. Of course these very same lenses are usually free of aberrations and perhaps that is the primary reason they perform better than average lenses?
What's a typical MTF for an average lens? For example, a 100mm f/4 (25mm clear aperture) would theoretically be capable of resolving details as small as 4.56 arc seconds. In real world conditions and assuming an average MTF (I know, no such thing.), what would you ballpark guesstimate the lens's actual resolution limit to be?
I was surprised (though in hindsight I shouldn't have been) to find that it takes a really long, large aperture lens to achieve close to 1::1 equivalence in lens resolution limit to sensor pixel pitch on a 24MP FX or DX body. It makes sense in that a 24MP camera, matched with most lenses, delivers a wealth of detail while still being somewhat forgiving of mushy atmospherics and even less-than-perfect photographic technique.
If the MP race continues and full frame cameras push beyond 50 MP toward 100 MP sensors, there may be a few photogs pulling their hair out over the degradation of IQ at the pixel level.
I was surprised (though in hindsight I shouldn't have been) to find that it takes a really long, large aperture lens to achieve close to 1::1 equivalence in lens resolution limit to sensor pixel pitch on a 24MP FX or DX body. It makes sense in that a 24MP camera, matched with most lenses, delivers a wealth of detail while still being somewhat forgiving of mushy atmospherics and even less-than-perfect photographic technique.
If the MP race continues and full frame cameras push beyond 50 MP toward 100 MP sensors, there may be a few photogs pulling their hair out over the degradation of IQ at the pixel level.
Kerry Pierce
Veteran Member
Ah, yes, we do agree.
I have no idea why I have been using jive in place of jibe, but I appreciate the point and will correct it henceforth.
Kerry
altair8800
Veteran Member
- Messages
- 2,000
- Reaction score
- 443
Maybe not. The Nikon V3 has pixel density which would give 135Mp for FF. Here is an example with AF-I 400 with a TC-20E III attached:
http://www.dpreview.com/forums/post/55290461
Also, here is something to think about for those who are afraid of diffraction:
http://www.lensrentals.com/blog/2013/03/overcoming-my-fentekaphobia/
Michel F
Veteran Member
That's interesting. Back home and running the numbers, a 200MP full frame sensor would still have pixels larger than the smallest resolvable detail delivered by a 400 f/2.8. A 500mm or 600mm f/4 would resolve at very nearly a 1:1 ratio with pixel pitch. In light of this, I should probably retract that previous comment. There appears to be ample room for the megapixel wars to continue raging without fear of running up against lens resolution limits.
Tony Beach
Veteran Member
- Messages
- 11,967
- Solutions
- 5
- Reaction score
- 7,035
Lens or otherwise, there will be practical limits. I doubt there will be many examples of moiré on 50+ MP DSLRs.
--
http://imageevent.com/tonybeach/twelveimages
Last edited:
n057
Veteran Member
Lucky Sky
Veteran Member
n057
Veteran Member
Keyboard shortcuts
- f
- Forum