availability

[commtrd]

My question has to do with availability of Nikon lenses. If some lenses are usually not available for sale (most

vendors are constantly out of stock) does that imply that certain lens is soon going to be offered in an upgraded

model? For example the 70-200 zoom has been very difficult to find for some time at the mainstream dealers (B & H,

etc.) and was in stock in seemingly very small quantities at smaller retailers for substantially more money, usually a

few hundred dollars more until even they sold out.

 

So why does Nikon not make lenses available for sale? It would seem to be in their best interest to do so.

 

On a side note, is there a possibility that sensor resolution could increase to a point that currently available lenses

could not perform adequately with those sensors? How would that manifest itself in real-life photography? Would that

be something that only pixel-peepers could discern? Is there actually a point where pixel density becomes self-

defeating?

[Rene11664880918]

It seems that the lenses are produce by batch. They are not in constant production. Now is the 70-200, a couple of months

ago was the 300 f4 and so on. We can't predict any upgrades when it comes to lenses.

 

It was also discussed before, to make a camera with a bigger sensor they will need to increase the size of the body since

the distance between the lens and the sensor would have to increase. It would be a different format so I don't think there is

nothing to worry about it. Rene'

[leonard_forte1]

The 300mm f/4 also has not been available at B&H for a couple of months! I know because I've been waiting for it

[josephwalsh]

Rene, I believe Keith is talking not about increasing sensor size but rather increased pixel density on DX and/or FX.

 

Good questions, Keith. I don't know the answers but hope someone who does will chime in. In particular, your last sentence gives one pause.

[jay_poel]

I think we're close to the sensor out performing the lens and showing the lens faults. I'll try to find the article I was reading that discussed it.

[josephwalsh]

Come to think of it, I was on a shoot where the photographer was using the top of the line Canon (would that have been a 1DsII ?) Anyway, the highest Canon pixel count which is around 20.

 

This is a very competent guy...big time NYC advertising photographer...He was complaining about how dissatisfied he was with several of his Canon lenses, both primes and zooms, since getting that body.

[richard_driscoll]

"is there a possibility that sensor resolution could increase to a point that currently available lenses could

not perform adequately with those sensors?"

 

Yes there certainly is and there is evidence that we are already partially there.

 

Of the current models, compare a D40, D80/D60 and D300 which have 6, 10 and 12 MP sensors. The maximum

theoretical resolution of the D40 is about 60 cycles/mm, the D80 is about 80 cycles/mm and the D300 about 90

cycles/mm. In practice the anti-aliasing filters will tend to reduce these figures somewhat.

 

If you look at the MTF curves that Nikon publish you can get a measure of the performances of various lenses at

full aperture and you'll see that they aren't actually quoted for anything over 30 cycles/mm. The 30 cycles/mm

curves show a marked loss of contrast especially towards the edges of the field. Note that curves for 60

cycles/mm and beyond aren't even published!

 

This is telling you that at full aperture the 10 MP sensor is already being compromised over the 6 MP sensor and

the 12 MP sensor even more so.

 

Of course when the lens is stopped down things get better up to perhaps f/8-11 at which point diffraction tends

to diminish the differences again.

 

All this is at least partially supported by writers to this forum who say that in practice they don't actually

observe all that much difference between the high and low resolution sensors.

 

To see the MTF curves go to:

 

http://imaging.nikon.com/products/imaging/lineup/lens/af/index.htm

 

.... and choose a lens. Scroll down to see the MTF curves. The 30 cycles/mm ones are in blue. Remember that

these curves represent very roughly half the resolution of a D40! Note that they label the curves lines/mm. I'm

pretty sure this is an error in translation since 30 lines/mm would only be 15 cycles/mm and I can't believe the

lenses are that poor!

 

Note too that the DX lenses aren't quoted at any higher cycles/mm than the FX ones whereas in fact they should be

quoted at about 45 cycles/mm to be equivalent.

[bob_sunley]

Yes digital is finally catching up to film in resolution in 35mm format. Fine grain films had better resolution than lenses 40 years ago. Most Nikkor primes were easily capable of resolving 50 line pairs per millimeter back in the 60's. That translates to 8 mpixel, tho with a bayer filtered sensor, you have to at least triple that pixel count, and have a sensor with around 25 mpixel to avoid interpolation artifacts. Current lenses can have a resolution greater than 70 lpm, so the pixel count has to increase a lot, to somewhere around 36 mpixel in a 24x36mm sensor.

[richard_driscoll]

Bob,

 

When you say "Current lenses can have a resolution greater than 70 lpm" is that lines pairs per millimetre? What is the contrast at that resolution? Please can you give us a reference. Isn't diffraction a problem at such high resolutions?

 

Doesn't the Bayer interpolation result in chrominance errors rather than loss of luminance resolution? Surely the eye is has very low chrominance resolution doesn't it?