Full Frame vs. Crop in terms of sharpness

[warheit12]

So i didnt no where to ask this so i am going to go ahead and ask this here.

Lets assume you have cropped sensor camera with a 50mm f1.2 lens and a full frame camera 50mm f1.2 lens. you shoot the exact same shot with the two cameras(assume its iso 100 and aperture shutter and everything is the same.) is the sharpness going to be greater on the full frame camera?

My question is, does the sensor factor in when it comes to sharpness or does it all boil down to the lens?

 

[jaomul]

Crop likely sharper as the best part of lenses usually centre and sharpness falls off at edges

 

[KmH]

However, a larger FF sensor generally has better inherent resolution by dint of it's larger pixels when compared to a crop sensor camera.
Plus there are some fast lenses that are sharp all the way out to the edges.

The area of a FF sensor is 864 sq mm. The area of a Nikon APS-C sensor is about 368 sq mm (Canon APS-C is smaller, about 332 sq mm).

There are in fact several image sensor factors that contribute to image sharpness. Things like pixel size, pixels pitch, and how aggressive the low-pass filter (anti-aliasing) in front of the image sensor is.

Nikon's D800 claim to fame is the cameras new low-pass filter design, though the design is such that moiré will more often be a problem for D800 users.

 

[ronlane]

However, a larger FF sensor generally has better inherent resolution by dint of it's larger pixels when compared to a crop sensor camera.
Plus there are some fast lenses that are sharp all the way out to the edges.

The area of a FF sensor is 864 sq mm. The area of a Nikon APS-C sensor is about 368 sq mm (Canon APS-C is smaller, about 332 sq mm).

There are in fact several image sensor factors that contribute to image sharpness. Things like pixel size, pixels pitch, and how aggressiveness the low-pass filter (anti-aliasing) in front of the image sensor is.

Nikon's D800 claim to fame is the cameras new low-pass filter design, though the design is such that moiré will more often be a problem for D800 users.

WOW, what a great answer. I don't understand much of it.

 

[sekhar]

You can't compare the two because they produce different images: the FF would cover a broader area than the crop. To match the image from the crop camera, you'd need to take the FF image, crop it, and expand it (assuming same image resolution for both the cameras). If you do that, of course the re-sized image from the FF camera is going to be softer.

 

[fjrabon]

First, sharpness isn't an actual real thing. There's no way to measure sharpness. There is only apparent softness. As keith noted, FF sensors have an advantage in that they can produce the same resolution while having larger pixels. This allows for less bleed into adjacent pixels, which increases sharpness. Further, because of the increased pixel density on many crop frame sensors, to achieve the same resolution crop frames have more of a problem with diffraction and noise. diffraction and noise both make images seem less sharp. Noise especially hits sharpness on two fronts, in that it both directly detracts from sharpness, and further, if you try to remove it, the only way to get rid of it is to in some way soften the image.

 

[unpopular]

First, sharpness isn't an actual real thing. There's no way to measure sharpness. There is only apparent softness.

Yeah, bro. Unless you can explain yourself, this is complete nonsense.

 

[unpopular]

Here's a simple experiment that you can do yourself! Here's what you'll need:

A coherent light source
A razor blade
An APS-C camera
A Full frame camera
(no lenses required!)

Align the blade face to be parallel with the sensor, preferably rotated at a 45° angle.

Align the light source to be as perpendicular to the knife and sensor, illuminating the sensor such that the blade casts a shadow over a portion of the sensor, while leaving the other portion illuminated.

In a darkened room, without a lens attached, expose the shadow of the knife edge. You want the white portion to be at maximum exposure without bloom.

Process the file normally without any unique adjustment.

Use the Plot Profile function in ImageJ and sample perpendicularly the edge profile. Whichever plot has a steeper slope has greater edge definition, i.e. sharpness.

This can be easily quantified using a slope function, provided that the slope is calculated from corresponding regions.

 

[KmH]

I link to this group of tutorials several times a day.

Apparently, few actually take the time to read any of them - Tutorials – Sharpness

 

[unpopular]

You could also measure optical resolution by placing a specifically designed test pattern directly onto the sensor.

But I wouldn't recommend it

----

ETA: actually, you could use a pulsing laser line that travels over the sensor durring exposure.

 

[Robin_Usagani]

You can't compare the two because they produce different images: the FF would cover a broader area than the crop. To match the image from the crop camera, you'd need to take the FF image, crop it, and expand it (assuming same image resolution for both the cameras). If you do that, of course the re-sized image from the FF camera is going to be softer.

To a point. If the cropped file of FF camera can still produce the resolution of the media, I would say the FF is sharper. I agree with you though. I mean if after you crop it you print 4x6 and you have enough pixels, it is most likely sharper.

 

[KmH]

Many of today's FF image sensors have resolving power that exceeds the resolution limit of most lenses.

 

[unpopular]

^^ in theory, but 'megapixels' don't correspond directly to optical resolution due to interpolation.

As far as I've known, optical resolution of the camera itself has never been adequately tested. They might very well, still; but simply because a lens can resolve the equivalent of 14mp doesn't necessarily mean that a 24mp camera can out resolve the lens.

 

[digital flower]

I haven't noticed any difference between the two as far as sharpness goes.