[kh1234567890]

I am not so sure.

Having scanned my old 35mm neg and slide collection (a variety of film, both colour and B&W) at 2700 dpi - roughly 100 pixels per mm - I'd agree with that. I have not come across any where the grain wasn't visible at that resolution. Admittedly, I did not have fancy lenses in those days, but the actual image resolution always appears to be less than the film grain.

[normhead]

This whole concept is pretty foreign to me. Anyone have an example of a picture where the resolution of a lens isn't good enough for a sensor? I can understand the concept of a lens that doesn't focus well. But the notion that the defining resolution of a lens can in some way be linked the sensor density of a photo receptor and in some way be found inadequate in other than the traditional way of determining lens quality, is a bit of a stretch. I keep hearing it. It keeps not computing. Show me some pictures.

Where's my Ryerson optical physics professor when I need him?

[Rondec]

This whole concept is pretty foreign to me. Anyone have an example of a picture where the resolution of a lens isn't good enough for a sensor? I can understand the concept of a lens that doesn't focus well. But the notion that the defining resolution of a lens can in some way be linked the sensor density of a photo receptor and in some way be found inadequate in other than the traditional way of determining lens quality, is a bit of a stretch. I keep hearing it. It keeps not computing. Show me some pictures.

Where's my Ryerson optical physics professor when I need him?

Well, I think we can show examples where a lens that seemed adequately sharp on a 6 megapixel camera, seemed soft when looking at 100 percent views on a 14 megapixel camera. I had (still have actually) an old FA J 28-80 zoom that showed exactly this. Decent sharpness on my K100, but flaws really showed up on the K10. Same with an old Sigma 70-300 zoom I had. I can't show you photos, because I have deleted them and bought new lenses that were up to the task of shooting on higher pixel density sensors.

A lot does depend if you are shooting wide open or stopped down a little bit.

[Tranzistors]

Taken with 6MP camera http://i28.tinypic.com/14n3apy.png
It clearly shows that 6MP camera outresolves Sigma zoom.

[SteveM]

There are many qualities to take into consideration when deciding if a lens is right for you and the K5. If you check many of the lens comparison threads here on the forum, you'll see that what might be the best lens for someone else is not necessarily the best one for you. Ironically, the sharpest lens is commonly not the favorite of the people posting. In addition to that, the sharpest lenses are not always the latest ones available.

I've gone in the other direction deciding to sell my newer glass to pick up some of the older legends. I can go out to buy a 35 ltd any old day but finding a 28/2.0 in new condition is no easy task. I am finding that the 28/2.0 is a little too sharp at times where it looks to be over processed straight out of the K-5.

As far as longer glass, you are not going to find much "new" over 300mm, and they will tend to be slower. The high ISO from the K-5 is perfect for these longer lenses. The detail out of ISO 3200 is amazing.

Once the K-5 sensor hype is finished, I see the market for older lenses increasing.

[atiratha]

16 mpx is not that much more than 14 mpix of K20 or K7 - I'd say as a rule of thumb - if it's good enough for K20, it's good enough for K5. If you print photos of the same size as before, the detail stays the same or can even be greater. If you are bothered by the lens performance at 100%, then consider if it's such an important criteria after all.

[normhead]

So, do sensors outresolve lenses? It depends on the lens you use, the properties of the light, the aperture and the format. Small format sensors may have surpassed the limit, this is, in most cases they are lens-limited in terms of resolution. It is easier to correct aberrations for a smaller light circle though, so you can approach diffraction-limited resolutions for lower f-numbers. The signal-to-noise ratio, however, imposes an inflexible limit to the effective resolution of the whole system, mostly due to photon shot noise.

waffle waffle waffle... no answer
Small format sensors may have surpassed the limit, or on the other hand, maybe they haven't.

Sensors for larger formats are approaching the diffraction limit of real lenses, and it is more difficult to get high levels of aberration suppression for them. The point is that you cannot fully exploit the resolution potential of high-resolution sensors with regular mass-produced lenses, particularly for larger formats.

"Sensors for larger formats. No indication this is at all relevant for a K-5 or any small format sensor.

You cannot compare the limits of two different photographic systems looking at a print because the variables that determine the subjective perception come into play. Different systems can provide comparable results on paper under certain conditions (the circle of confusion reasoning explains how that is possible), but the limit of a system must be evaluated considering the pixel as the minimum circle of confusion.

In other words.. "I know nuttink."

For all the gory details...

Do Sensors “Outresolve” Lenses?

[SteveM]

"So, do sensors outresolve lenses? It depends on the lens you use, the properties of the light, the aperture and the format"

That is a very acedemic article. It doesn't really prove anything about newer vs older lenses.

[normhead]

I do remember one of my lecturers at Ryerson stating that in some circumstances a cheap camera could out perform an expensive one because they had plastic parabolic lenses that gave them better DoF and less aberrations. I notice many new lenses are using at least one molded element, ( not ground glass) presumably to do with glass, what in the old days was only done with plastic. Following through on that one would have to go back and find out how long this has been going on. I'm guessing the higher resolution of the molded glass would be the factor that might make a newer lens more desirable. It would also explain the change in thinking about shooting with small f stops. Without a molded element shooting right through the center of the lens would give you the best resolution. With a non-spherical surface cleaning up distortion, etc. apparently now the diffraction after f11 is more of a factor, than avoiding lens aberrations. New for me, but hey.

[peterh337]

That Luminous Landscape article is good in covering the diffraction limit, which - for all types of lenses - will ensure that at a high F number you will get poor resolution (basically because the light is squeezing through a small hole and the wavelength of light becomes a limiting factor.

But at lower F number the resolution will be determined primarily by how smooth the glass surface is.

Obviously, the more elements in a lens, the more potential for mucking up the resolution. So a zoom will have a lower resolution than a simple lens.

[normhead]

Obviously, the more elements in a lens, the more potential for mucking up the resolution. So a zoom will have a lower resolution than a simple lens.

Check out the resolutions on photozone, the DA 60-250 might change your mind on that assumption.

DA 200 F2.8

DA 60-250 f4

At 200mm the 60-250 out resolves the 200 prime at every tested f-stop, sometimes by as much as 200 points. What you get from the prime is f2.8 not better resolution. You can check other Pentax primes against the 60-250 and you'll find pretty much the same thing.

[whojammyflip]

just to put the cat amongst the pigeons, the mtf of a system is not equal to the minimum of the various limiting factors like lens, sensor, scanner, printer etc, but actually a convolution of all of them, ie the square root of the sum of the errors squared, type of thinking

this means that you should use the best lens possible in order to reveal the power of the sensor

as a corollary to my comment about the 50lp/mm of the sensor, this is further reduced by the Bayer layout of the RGB pixels, ie most primes will outperform this sensor by a long way

there are some good Zeiss documents on understanding MTF curves freely available on the web. Basically, my concern here would be converting the idea of a single MTF limit into an understanding that MTF is a function...contrast vs frequency of information

all lenses can resolve 20lp/mm, but some do it better than others...which is why something like a Fuji S5 is popular even with its 6mp sensor...equally, its why compact cameras offering huge numbers of pixels are still awful...contrast at the frequency suggested as being possible is terrible.

[peterh337]

10MP+ compacts deliver the resolution OK but the result is awful because they have tiny lenses (which presumably means they are at the diffraction limit), tiny sensor pixels (loads of sensor noise), and use heavy jpeg compression (loads of image artefacts).

[Ira]

What about old FF lenses using only the sharpest center areas of the lens compared to APS-C designed lenses?

Has glass itself been improved over the past 50 years?

[peterh337]

My suspicion is that more expensive lenses are made of better glass and more accurately machined / polished.

Historically, Zeiss became well known for using pricey high refractive index glasses which result in thinner lenses, and the less glass the better.

But if you use just the small middle bit of a lens, you compromise sharpness through diffraction.