FF Lens on APS-C body - confusing

I'm shooting with a D7100 (until my D500 gets here). I switched from the 70-300 kit lens to a 80-400 "pro" lens. I loved the difference the 80-400 made.

So I'm listening to Tony Northrup's video, "20 things photographers get wrong," and Tony argues that since my "APS-C body only covers about 45% of the the lens, it is magnifying the defects in the lens by 200%" and I'm actually getting worse images than I would from the much cheaper kit lens.

Admittedly, I'm not photography tech savvy, but "in my mind, and to my eye," my pro lens shots look better than my kit lens shots.

Am I duping myself by buying good glass? Or should I strive for cheap glass for better images as Tony argues?
 
I'm shooting with a D7100 (until my D500 gets here). I switched from the 70-300 kit lens to a 80-400 "pro" lens. I loved the difference the 80-400 made.

So I'm listening to Tony Northrup's video, "20 things photographers get wrong," and Tony argues that since my "APS-C body only covers about 45% of the the lens, it is magnifying the defects in the lens by 200%" and I'm actually getting worse images than I would from the much cheaper kit lens.

Admittedly, I'm not photography tech savvy, but "in my mind, and to my eye," my pro lens shots look better than my kit lens shots.

Am I duping myself by buying good glass? Or should I strive for cheap glass for better images as Tony argues?
 
It is pointless, because he got the physics wrong, as there is no magnification going on. The image circle that a lens needs to produce to cover the DX sensor is constant, it does not matter if the image circle is a tight fit (DX) or if it is a lot bigger (FX), what ends up on the sensor is just the same. And that is the fault of his "mathematics" exercise, he looks at the sensor size not on the image circle produced by the lens. A FX lens does not care if only 4X% of the image it produces are recorded by the sensor.
 
The following may support Tony's assertion that putting a FX lens on a DX camera is a bad idea because it magnifies the flaws.

My Tamron 150-600 was acceptably sharp all the way up to 400mm on a D7100 but noticably soft at 600mm. However on a D700 it was acceptably sharp right the way up to 600mm. Magnification of an imperfection is one possible explanation.

(Incidently I've seen otther 150-600's produce sharp images all the way up to 600mm on a 7100 which is understandale if off brand lenses do have greater quality variation.)

What makes no sense at all is: Why would lens brand name manufacturers like Nikon or Cannon grind and assemble their FX lenses or "pro glass" to a lower tollerance specification than their DX lenses?
 
The following may support Tony's assertion that putting a FX lens on a DX camera is a bad idea because it magnifies the flaws.

My Tamron 150-600 was acceptably sharp all the way up to 400mm on a D7100 but noticably soft at 600mm. However on a D700 it was acceptably sharp right the way up to 600mm. Magnification of an imperfection is one possible explanation.

(Incidently I've seen otther 150-600's produce sharp images all the way up to 600mm on a 7100 which is understandale if off brand lenses do have greater quality variation.)

What makes no sense at all is: Why would lens brand name manufacturers like Nikon or Cannon grind and assemble their FX lenses or "pro glass" to a lower tollerance specification than their DX lenses?
It makes sense

You are comparing 2 cameras with very different pixel densities. The D700 is just 5,7MP in DX the D7100 is 24 MP, that is about double the resolution. You would need to use a 54MP FX camera to compare the results without having the pixel density playing a role.

Nobody disputes that higher pixel density makes lens defects more visible, but this is true for DX and FX lenses alike.
 
It is pointless, because he got the physics wrong, as there is no magnification going on. The image circle that a lens needs to produce to cover the DX sensor is constant, it does not matter if the image circle is a tight fit (DX) or if it is a lot bigger (FX), what ends up on the sensor is just the same. And that is the fault of his "mathematics" exercise, he looks at the sensor size not on the image circle produced by the lens. A FX lens does not care if only 4X% of the image it produces are recorded by the sensor.
 
The following may support Tony's assertion that putting a FX lens on a DX camera is a bad idea because it magnifies the flaws.

My Tamron 150-600 was acceptably sharp all the way up to 400mm on a D7100 but noticably soft at 600mm. However on a D700 it was acceptably sharp right the way up to 600mm. Magnification of an imperfection is one possible explanation.

(Incidently I've seen otther 150-600's produce sharp images all the way up to 600mm on a 7100 which is understandale if off brand lenses do have greater quality variation.)

What makes no sense at all is: Why would lens brand name manufacturers like Nikon or Cannon grind and assemble their FX lenses or "pro glass" to a lower tollerance specification than their DX lenses?
It makes sense

You are comparing 2 cameras with very different pixel densities. The D700 is just 5,7MP in DX the D7100 is 24 MP, that is about double the resolution. You would need to use a 54MP FX camera to compare the results without having the pixel density playing a role.

Nobody disputes that higher pixel density makes lens defects more visible, but this is true for DX and FX lenses alike.

--
hobby aviation photographer
The pixel density thing is negated in this instance though because when I noticed this problem I trawled Flickr for the same DX lens camera combo and there were lots of images that were much sharper at 600mm than mine ever was. I specifically looked at pictures of small birds because of the detail and because generally we all have to crop around the same amount to fill the screen (due to their flighty nature).
 
Last edited:
It isn't magnification because the process is not optical, but it is spatial frequency. The way I understand it is that if I put a 400 f2.8 lens on a D500 full resolution will be 3712/15.7 = 236 l/mm. If I shoot the same scene on a D5 with 600 f4 full resolution is 3840/23.9 = 161 l/mm. In order for the lens/sensor combination to achieve MTF50 (roughly speaking what we perceive as sharp) at full resolution in both pictures the 400 f2.8 needs to be as sharp (in the center) at 236 l/mm as the 600 f4 is at 161 l/mm (across the frame).

--
Jim
 
Last edited:
The following may support Tony's assertion that putting a FX lens on a DX camera is a bad idea because it magnifies the flaws.

My Tamron 150-600 was acceptably sharp all the way up to 400mm on a D7100 but noticably soft at 600mm. However on a D700 it was acceptably sharp right the way up to 600mm. Magnification of an imperfection is one possible explanation.

(Incidently I've seen otther 150-600's produce sharp images all the way up to 600mm on a 7100 which is understandale if off brand lenses do have greater quality variation.)

What makes no sense at all is: Why would lens brand name manufacturers like Nikon or Cannon grind and assemble their FX lenses or "pro glass" to a lower tollerance specification than their DX lenses?
It makes sense

You are comparing 2 cameras with very different pixel densities. The D700 is just 5,7MP in DX the D7100 is 24 MP, that is about double the resolution. You would need to use a 54MP FX camera to compare the results without having the pixel density playing a role.

Nobody disputes that higher pixel density makes lens defects more visible, but this is true for DX and FX lenses alike.
 
The 150-600 is only made in a version designed for FX sensors. There is no DX version. So what you see are either better edits or people having a better copy of your lens. Pixel density can never be ignored.

--
hobby aviation photographer
 
Last edited:
Of course the 70-200 f2.8 VR was a 35mm lens developed specifically to complement DX sensors and 35mm film cameras. It worked well on DX due to the high center sharpness but didn't work well in the corners with the new FX sensors. I not only have this lens, but another lens - the Tamron 28-75 f2,8 "screwdriver" - that is similarly good on DX and film but not much good on FX.

The irony is that the 70-200 f2.8 VR II is better on DX than the original - because it is even sharper in the center - except for the "focus breathing" issue.
 
Of course the 70-200 f2.8 VR was a 35mm lens developed specifically to complement DX sensors and 35mm film cameras. It worked well on DX due to the high center sharpness but didn't work well in the corners with the new FX sensors. I not only have this lens, but another lens - the Tamron 28-75 f2,8 "screwdriver" - that is similarly good on DX and film but not much good on FX.

The irony is that the 70-200 f2.8 VR II is better on DX than the original - because it is even sharper in the center - except for the "focus breathing" issue.
 
Don't know if this has already been explained and I didn't see Tony's video but I don't agree with him. I think he is assuming that FF sensors will have larger pixels and lower pixel density than an equivalent MP DX sensor, which will therfore show imperfections less. That is true except many FX sensors have the similar pixel densities as a DX sensor, they just have more of them in the bigger area.
 
The 150-600 is only made in a version designed for FX sensors. There is no DX version. So what you see are either better edits or people having a better copy of your lens. Pixel density can never be ignored.
 
Last edited:
I'm shooting with a D7100 (until my D500 gets here). I switched from the 70-300 kit lens to a 80-400 "pro" lens. I loved the difference the 80-400 made.

So I'm listening to Tony Northrup's video, "20 things photographers get wrong," and Tony argues that since my "APS-C body only covers about 45% of the the lens, it is magnifying the defects in the lens by 200%" and I'm actually getting worse images than I would from the much cheaper kit lens.

Admittedly, I'm not photography tech savvy, but "in my mind, and to my eye," my pro lens shots look better than my kit lens shots.

Am I duping myself by buying good glass? Or should I strive for cheap glass for better images as Tony argues?
 
It is pointless, because he got the physics wrong, as there is no magnification going on. The image circle that a lens needs to produce to cover the DX sensor is constant, it does not matter if the image circle is a tight fit (DX) or if it is a lot bigger (FX), what ends up on the sensor is just the same. And that is the fault of his "mathematics" exercise, he looks at the sensor size not on the image circle produced by the lens. A FX lens does not care if only 4X% of the image it produces are recorded by the sensor.

--
hobby aviation photographer
There has to be magnification somewhere. We don't all imagine that the focal length if a lens appears to increase from FX to DX. It's effectively between the sensor and the screen and if there are any flaws magnification magnifies them.
The "magnification" that Tony is probably talking about is the magnification that takes place when the DX sized sensor image is brought up to the same size as the FX image, ie 1.5x.

Let's say that it takes 10 steps of magnification to print an FX image. If we only magnify the DX image the same 10 steps, it would be smaller than the FX print. You'd have to go another 1.5x to get the same sized print.

It has nothing to do with the lens magnification. But, the more the pixels are enlarged, the larger any recorded defects will be in the print.

You can substitute viewing device instead of print and it remains the same.

Kerry

--
When is it "Okay" to be mean, petty or unethical?
-
my gallery of so-so photos
http://www.pbase.com/kerrypierce/root
 
Last edited:
The following may support Tony's assertion that putting a FX lens on a DX camera is a bad idea because it magnifies the flaws.

My Tamron 150-600 was acceptably sharp all the way up to 400mm on a D7100 but noticably soft at 600mm. However on a D700 it was acceptably sharp right the way up to 600mm. Magnification of an imperfection is one possible explanation.
Or the D700 does not have enough resolution to see the finer imperfections.
(Incidently I've seen otther 150-600's produce sharp images all the way up to 600mm on a 7100 which is understandale if off brand lenses do have greater quality variation.)

What makes no sense at all is: Why would lens brand name manufacturers like Nikon or Cannon grind and assemble their FX lenses or "pro glass" to a lower tollerance specification than their DX lenses?
Compare "pro glass" MTF charts to those for DX. Do you find any DX better than "pro glass".
 
"So I'm listening to Tony Northrup's video, "20 things photographers get wrong," and Tony argues that since my "APS-C body only covers about 45% of the the lens, it is magnifying the defects in the lens by 200%"

That makes 21, he is inventing his own things to get wrong. My advice is to ignore EVERYTHING people say about the differences between APS and FF except the angle of view.

--
Andrew Skinner
 
Last edited:

Keyboard shortcuts

Back
Top