MTF 1/3" Nikon "G" Adapter

[Stefan Immler]

Hi there --

I am in need to add a telephoto lens to my JVC and am intrigued by the new MTF Nikon G Lens to 1/3" Adapter. In addition to accepting the "no iris ring" Nikon G lenses, this adapter seems to be compatible with virtually all Nikon lenses.

Nikon G to 1/3" bayonet adaptor

In particular, the Nikon 70-300mm/4.5-5.6 telephoto zoom could be a great addition due to its small price tag, size & weight, constant length and non-rotating front lens element (perfect for attaching a matte box on rails).

Has anyone experience with the new MTF G -> 1/3" adapter and that Nikon lens? (JVC, Sony, or Panasonic)

Cheers

[Steve Phillipps]

You have to be very careful using stills lenses on small chip cameras, whatever their quality, but I'd be very wary about using anything but the absolute top of the line models (like Canon L series or Nikon EDIF) or you'll likely get iffy images.
With that in mind, for a lot of people you'd be better off just getting the standard Nikon adapter and a secondhand manual focus lens - they are cheaper and they tend to have better manual focus controls, and an iris on the lens.
Steve

[Brian Standing]

Originally Posted by Steve Phillipps

You have to be very careful using stills lenses on small chip cameras, whatever their quality, but I'd be very wary about using anything but the absolute top of the line models

Really? I would have thought exactly the opposite. If you're looking at an SLR lens with a 1/3" CCD, you're cropping out the exact center portion of the lens, which in most lenses, is the sharpest, most optically pure part of the lens. I have always understood that one of the biggest difference between cheap SLR lenses and expensive ones is the edge-to-edge sharpness. If you're throwing away the edges by cropping out the center, who cares how sharp they are?

Is there something else here I'm missing?

[Steve Phillipps]

Yeah, you would think so, but it doesn't work that way. Because the pixels on these chips are so small the lenses need massive resolving power, and you'll find that the line per mm of a 1/3" Fujinon lens will be many times greater than even the best 35mm stills lens ever made.
In practice the results rarely are as bad to the eye as the science suggests they should be, but it's definitely there.
Steve

[Stefan Immler]

I don't think this is an issue with Nikon glass. Nikon makes some of the best lenses. I wouldn't use Tamron or Sigma lenses on my JVC, but original Nikon glass should give excellent results.

[Steve Phillipps]

Originally Posted by Stefan Immler

I don't think this is an issue with Nikon glass. Nikon makes some of the best lenses. I wouldn't use Tamron or Sigma lenses on my JVC, but original Nikon glass should give excellent results.

Not neccessarily, with most brands there are tiers of quality, and some are definitely better than others, and a good Sigma like the 120-300 will well-outperform the cheaper Nikons.
Steve

[Bill Ravens]

I'm using an MTF adapter to mount older(with aperture rings) Nikon 35mm glass on my HD110. I achieve image qualities with the Nikon zooms that FAR exceed the images I get from my Fujinon 17x zoom lens . No, absolutely NO CA problems, which, in my mind, is a huge step forward. Considering the 7x magnification factor of a non-optical adapter, the smaller Nikon zoom is a 36-72mm effectively yielding the equivalent of 242-504mm. Incidentally, I paid $15 US for this lens from B&H used lens department. A pretty significant cost savings over the Fujinon 18x.

[Steve Phillipps]

Originally Posted by Bill Ravens

I'm using an MTF adapter to mount older(with aperture rings) Nikon 35mm glass on my HD110. I achieve image qualities with the Nikon zooms that FAR exceed the images I get from my Fujinon 17x zoom lens . No, absolutely NO CA problems, which, in my mind, is a huge step forward. Considering the 7x magnification factor of a non-optical adapter, the smaller Nikon zoom is a 36-72mm effectively yielding the equivalent of 242-504mm. Incidentally, I paid $15 US for this lens from B&H used lens department. A pretty significant cost savings over the Fujinon 18x.

In the words of Scotty from Star Trek, that's certainly bending the laws of physics Captain! Unless the Fujinon is an absolute dog. Even the most expenive stills lenses on even 2/3" HD cameras don't look as good as video lenses, so on 1/3" they should really suffer. But as I said the results never seem to be as bad the science would suggest.
Steve

[Stefan Immler]

Originally Posted by Steve Phillipps

Even the most expenive stills lenses on even 2/3" HD cameras don't look as good as video lenses, so on 1/3" they should really suffer.

There are plenty of videos on this and other sites that prove the exact opposite. From what I have seen, high-quality Nikon lenses outperform most video lenses.

I decided to get one of the MTF Nikon -> 1/3" adapters and check it out.

[Mike Tapa]

Originally Posted by Steve Phillipps

In the words of Scotty from Star Trek, that's certainly bending the laws of physics Captain! Unless the Fujinon is an absolute dog. Even the most expenive stills lenses on even 2/3" HD cameras don't look as good as video lenses, so on 1/3" they should really suffer. But as I said the results never seem to be as bad the science would suggest.
Steve

This issue has always been a bit of a paradox.
However, over the years, I have carried out MTF measurements on many lens systems whist working at Optex and even though the benchmark for, say 2/3" HD, is around 90% at 56LP/mm in zone 1, I have found that any half decent lens can resolve 140 to 200 LP/mm.
Actual MTF at those resolutions will be well down, but the resolving power is nonetheless there.

[Bill Ravens]

Many a techie has been fooled by the obvious, thinking the equations(academic wisdom) show all, while a little empirical evidence(street smarts) show otherwise !!
I stand by my comments. Chromatic Aberration on the Fujinons(at least the ones that come with the HD110) are absolutely hideous. yet the 35mm lenses perform wonderfully. I doubt it's the coatings, since the same coatings can be put on a video lens. I suspect it's because the 1/3 inch sensors only use the center of the 35mm optic. Or, perhaps, CA is influenced by a 3D sensor(RGB being on different layers) that some lenses are more sensitive to than others. Regardless, the point is that the physics models are flawed and don't represent reality.

Resolving power, aside, Mike, I wonder how optimized the 35mm lenses are for minimum CA, using more corrective optics than video lenses?

[Steve Phillipps]

Originally Posted by Bill Ravens

Many a techie has been fooled by the obvious, thinking the equations show all!!
I stand by my comments. Chromatic Aberration on the Funinons is absolutely hideous. yet the 35mm lenses perform wonderfully. I doubt it's the coatings, since the same coatings can be put on a video lens. I suspect it's because the 1/3 inch sensors only use the center of the 35mm optic.

Resolving power, aside, Mike, I wonder how optimized the 35mm lenses are for minimum CA?

AFAIK Lateral Chromatic Abberation should be consistent across the frame, it should make no difference whether it's the centre or edge, nor should aperture make any difference - but again, we seem to going outside the laws of physics and optics!
I agree though that the CA on a lot of HD lenses is very bad, even on the Canon HJ40 it's horrible, but I believe it's to do with the extra resolving power in the design, it's a trade-off.
I'm surprised by Mike's finding as I know other engineers have certainly seen the opposite, and I've done my own tests that have also shown that stills lenses fall short of HD ones - though often fairly close. For example I tested the Canon 150-600 and 300 f2.8 against the HJ40 and even when the HJ40 had the 2x extender on it was sharper than the stills lenses! Also tried the Nikon 55 micro and 105 2.5 - two of the sharpest lenses Nikon has ever made, no question - and they were significantly less sharp than the stock lens on the EX3.

Steve

[Brian Standing]

Argh. My head is swimming. At some point, doesn't the resolving power of the medium have some influence? A 35mm lens is designed for 35mm photographic film, which is a higher resolution format than any but the highest HD video format.

Would a 1/3"-chip at 720p even notice the difference between a good and a poor 35mm SLR lens? (Let's just talk about resolution, here, instead of chromatic aberration.)

[Jason Davis]

Stefan, I own an MTF adapter, I am currently using nikon nikkor 80-200 the picture quality is fantastic and I am sure you will not be dissapointed. I am not using the "new" G adapter. The image is much better than my 16x fuji lens.

[Bill Ravens]

Brian...

I'm afraid you're wanting to simplify things to the point of reducing the discussion to a meaningless level. It is not a simple subject. Chromatic aberration and resolving power are two different parameters that partially describe optical performance. While one can keep the discussion to one or the other, it's simply burying ones head in the sand to ignore the other. And there are other parameters such as coma and astigmatism, which haven't even been brought up.

Good 35mm lenses, at the sweet spot, are designed to resolve to the level of film grain size. It has nothing to do, per se, with the format size, since film grain doesn't change size as the format size changes. When working with digital imaging, one has to consider the number of pixels and format size as the limit to resolving power. At what point do pixel sizes begin to approach film grain sizes? The z direction depth also will effect the apparent resolving power. I would guess a 2k digital image on a super 16 film format is approximately the same resolution as a fine grained 35mm film. RED cameras deliver 4k performance in a 35mm format size. At this point, the use of a 35mm optic resolving power will begin to bump up against the resolution available from the sensor.

And again, resolving power has nothing to do with Chromatic aberration.