Lenses, are they infinite in resolution?

[helgetjelta]

Hi, I had a discussion with a photographer who just bouht a new phantom camera, a 2K+ camera. And he more or less laught at me of the 4K hysteria, telling me that the optics in a lense will not give you more then 2K. That for a 4K picture you needed to have almost a double opening of the lense. I told him "here we see differences in meanings...and teaching..., the 2K camp vs the 4K and 8K camp"

So my question is:

Lenses, since they are optical, do have an infinite resolution ? It is only up to the sensor, how small the "pixels" you can get out of the lights coming in, right ?

cheers Helge

 

[Karim D. Ghantous]

I am not an optics expert, but I can tell you that lenses do have limits on their resolution. I can also tell you that the person you were talking to is unintelligent. What he said can be true. But (I am simplifying, now) it depends on the lens and it depends on the width of the photosites on the sensor.

 

[Pascal Scheffers]

You cannot have infinite resolution, that physically impossible. Theoretical lenses are mathematical games. In the real world, for starters, each ray of light has a certain size which is small, but not zero. More relevant to lenses is the fact that each color as we record it is a different wavelength and therefore bends differently through the glass. You can fix this to certain degree, which is what 'low dispersion glass' or 'LD' is all about in lenses, and then you stack them (lens elements) to try to get them back together... but you always end up losing resolution. And this isn't the only contributor, heck, I'm not even sure if this is the biggest component of resolution 'loss' in optical lenses.

 

[Erich Ocean]

Lenses, since they are optical, do have an infinite resolution ? It is only up to the sensor, how small the "pixels" you can get out of the lights coming in, right ?

Lenses do not have "infinite" resolution. The image is always blurred as it passes through the lens (this is what we mean when we say a lens is "soft"). All lenses blur fine details at some point. Most lenses also can only resolve the finest detail in the center of the lens, and get blurry as you move away from the center.

Recording a blurry image with small pixels does not make it less blurry. ;-)

 

[Erich Ocean]

In the real world, for starters, each ray of light has a certain size which is small, but not zero.

Since light has no mass (being pure energy), I'm fairly certain that photons do not have a "size". They do have a wavelength, though, being electromagnetic radiation; this becomes important in certain situations where the wave-particle duality aspect of light comes into play.

Point being, I think suggesting that light (photons) have a physical size is incorrect, and I can say with certainty that physical simulations of light do not take into account the "size" of a photon, only the wavelength.

I think you're actually thinking of the circle of confusion that occurs when light passes through a lens. This may seem like a "beam", giving the appearance that "rays" have a "width", but that's not what's actually happening at the photon level. Hope this helps!

 

[David Rasberry]

From Wikipedia:

Lens resolution

The ability of a lens to resolve detail is usually determined by the quality of the lens but is ultimately limited by diffraction. Light coming from a point in the object diffracts through the lens aperture such that it forms a diffraction pattern in the image which has a central spot and surrounding bright rings, separated by dark nulls; this pattern is known as an Airy pattern, and the central bright lobe as an Airy disk. The angular radius of the Airy disk (measured from the center to the first null) is given by

\sin \theta = 1.22 \frac{\lambda}{D}

where

θ is the angular resolution,
λ is the wavelength of light,
and D is the diameter of the lens aperture.

Two adjacent points in the object give rise to two diffraction patterns. If the angular separation of the two points is significantly less than the Airy disk angular radius, then the two points cannot be resolved in the image, but if their angular separation is much greater than this, distinct images of the two points are formed and they can therefore be resolved. Rayleigh defined the somewhat arbitrary "Rayleigh criterion" that two points whose angular separation is equal to the Airy disk radius to first null can be considered to be resolved. It can be seen that the greater the diameter of the lens or its aperture, the greater the resolution. Astronomical telescopes have increasingly large lenses so they can 'see' ever finer detail in the stars.

 

[Fredrik Harreschou]

Hei, Helge!

You´ve gotten a lot of very technical explanations. The real world answer to your question is that any decent 35mm lens resolves more than 4K. Arri recommends at least a 4K scan of film. Those films are shot with 35mm glass. Same thing with DSLR, with the logic of your friend the 12/16/24/etc MP sensors of the cameras would be useless.

MVH
Fredrik

 

[David Rasberry]

Hei, Helge!

You´ve gotten a lot of very technical explanations. The real world answer to your question is that any decent 35mm lens resolves more than 4K. Arri recommends at least a 4K scan of film. Those films are shot with 35mm glass. Same thing with DSLR, with the logic of your friend the 12/16/24/etc MP sensors of the cameras would be useless.

MVH
Fredrik

You want the optics to be the limiting factor on resolution. You have problems with aliasing and moire when the sensor is the limiting factor.

Resolution is also relative to sensor size. A lens that would resolve 4k on a 35mm sensor would only resolve half that on a 16mm sized sensor.

 

[Martin Stevens]

Since light has no mass (being pure energy), I'm fairly certain that photons do not have a "size".

Since E=mc2 or E=m, photons do indeed have mass and size.

 

[Ketch Rossi]

NO, lenses are not made with Infinite resolution...

Inf act more then once was said that literally all Camera companies have their Glass Obsolete by their Sensor capacity, Including Canon "L" Series, glass,
and this is why also recent tests with EPIC show that Canon is softest when compared to Red Pro Primes or Master Primes, and that currently the new King of the Block in resolution
are the LEICA S-C Primes made form Scratch to resolve 8K and above...

But I must add is way more complicated then this...

 

[Christopher Barrett]

Most lenses you'll be using will be able to resolve for a 4k sensor. Resolving Power... that's how the sharpness of a lens can be tested. Largely, we look at how many lines a given lens can resolve within a millimeter. The sharper lenses can resolve, or distinguish, more lines per millimeter and no, all lenses are not equal in this respect, not at all.

When I started shooting stills digitally I was using Large Format lenses and a 39megapixel digital back and they seemed sharp enough. When I moved up to a 60megapixel back, the sensor was easily out-resolving my lenses and I had to move to the new higher resolution digital formula lenses from Schneider and Rodenstock.

So no, lens resolution is not infinite, but yes you will see the difference between 4k and 2k with most lenses that you could mount on a Red. That's why people love the CaNikons... besides being cheaper, small and light they are also designed to resolve and cover 5 k sensors... and the Tilt/Shift lenses cover even beyond that.

Cheers,
CB

 

[Keith Putnam]

Here's a video clip of Eric Fossum, the inventor of the CMOS sensor, explaining Diffraction Limit, which is what causes point sources of light to "smear out" once they pass through a lens (starting at 24:23).

He tells us that the smallest area on a chip that a point source of light can cover, due to diffraction limit, is about 4 microns. I believe the photosite size on the MX chip is 5.4 microns, and photosites can be significantly smaller than 4 microns on other sensors if engineers so desire.
So what happens when your photosites are smaller than the diffraction limit? As Eric Fossum puts it, "Maybe someday somebody will figure out in software how to make the images better."
Regardless, there's been plenty of testing to show that lenses resolve line pairs well beyond 2K. Remember all that flap about how the RED One "wasn't really a 4K camera"? Yeah, all those tests collected plenty of data to refute the claim that lenses can't resolve detail finer than 2K.

 

[Peter Dmitriyev]

you realize how small 4k is right? 4k is the equivalent of like 9 or 10 megapixels. i.e. the resolution of an iPhone camera. A canon 5d markii as an example shoots 24 megapixel still photos. That means any damn still lens is capable of resolving far more than 4k or 5k etc.

 

[Jason Comparetto]

Wow, there's so many different technical explanations to the same question. Even involving Einstein himself, wow! But in any of these explanations your "photographer" "friend" is on crack. Like Peter said, 4k is nothing compared to today's 21MP sensors.

 

[Pawel Achtel]

you realize how small 4k is right? 4k is the equivalent of like 9 or 10 megapixels. i.e. the resolution of an iPhone camera. A canon 5d markii as an example shoots 24 megapixel still photos. That means any damn still lens is capable of resolving far more than 4k or 5k etc.

Canon 5D II is exactly 5.7k camera and unless you want to shoot for 1950's 4:3 television set, the extra vertical pixels are wasted for cinema aspect ratio.

As to the subject matter, yes, many lenses struggle to resolve 2k on the edges. Most resolve 5k in the centre.

As to diffraction limit, it is aperture dependent.

red= Green= Blue=
0.6 0.53 0.47
micron micron micron
================================================
f/ratio diffraction spot diameter in microns
================================================
2 2.9 2.6 2.3
2.8 4.1 3.6 3.2
4 5.9 5.2 4.6
5.6 8.2 7.2 6.4
8 11.7 10.3 9.2
11 16.1 14.2 12.6
16 23.4 20.7 18.3
19 27.8 24.6 21.8
22 32.2 28.5 25.2
32 46.8 41.4 36.7
45 65.9 58.2 51.6
64 93.7 82.8 73.4
================================================

source: Roger N. Clark

You can see that shooting at apertures smaller than about f/4.0 starts imparing resolution due to diffraction limit.

Good reading here:
http://www.cambridgeincolour.com/tutorials/diffraction-photography.htm
http://www.luminous-landscape.com/tutorials/resolution.shtml
http://www.clarkvision.com/articles/does.pixel.size.matter/

 

[Dave Blackham]

So what I think your saying is shooting at the optimum f stop of a given lens is important to achieve maximum resolution but it may not be the desired f stop for practical or creative requirements.

 

[Peter Dmitriyev]

http://www.kenrockwell.com/tech/lens-sharpness.htm

 

[Ron Goodman]

Don't leave out the Leica M series...small and potent!

 

[Ron Goodman]

Food for thought. I recently read an independent lens test comparing exactly, the best of primes of the same focal length, I believe it was 24mm, from Canon, Nikon and the Leica viewfinder camera on their flagship full size 36mm wide sensors. Now the sensors had different pixel counts with the Leica being the lowest at 18Meg. The Canon was 21 plus and the Nikon 24 or so. Same scene, same light, cropped to 1 percent of frame area at various sites within the full sensor.....The winner Leica, very close second Nikon....distant third Canon. The little lenses that can and do!