Best Prime lens for APS-C

[helged]

Advertisement (gone after registration)

Unfortunately, the S lenses are not compatible wuth the CL and TL bodies. Only Leica knows whether this is possible (it should, given that the S lenses wirk on the SL), or if this is low on the priority list. I have added this wish on the 'CL firmwate update' thread. Additional support may strengthen the request.

[SaW]

Thanks helged !

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[grahamhoey]

Not sure if this has been asked already but am thinking of either the TL 23 f/2 or the Elmirat M 28 f/2.8. Other than the obvious focal length difference and speed any comments or thoughts on preference. They both seem really nice small travel lenses.

[lct]

The TL 23/2 is not that compact a lens with its hood on. My M 28/2.8 asph v1 (with 39mm filters and accessory hood) looks tiny by comparison.

[grahamhoey]

Thanks, how do you think the 2 lenses compare in image quality. I always think the M glass is better but am I wrong there.

[lct]

I have no experience with the TL 23/2 but it has a very good reputation aside from its fly by wire focusing, lack of aperture ring and electronically corrected distortion. Sean Reid did an interesting review of this lens a couple of years ago if memory serves.

[jaapv]

Advertisement (gone after registration)

Digital correction is a plus. It allows for better image quality than optical corrections alone whilst reducing cost and size.

[grahamhoey]

Thank you, I just read Sean Reid report from 2014. He seems to suggest that the digital correction reduces some of the lens resolution as opposed to optical correction. The other issue that has been more inclined to the M glass is ability to pre zone focus which doesn’t seem to be a very practical option on the TL

[scott kirkpatrick]

I'm a long term subscriber to Sean Reid's reviews, and appreciate the care that he devotes to making his work reproducible over the years.  But I disagree with some of his conclusions, in particular his vigorous opposition to software corrections.  I don't think his evidence in all cases supports the problems he claims to see, and he provides a confused explanation of what software correction involves.  Besides straightening lines at the edge of a picture, the corrections in Leica's DNGs correct some residual lateral chromatic aberration -- providing a sort of software APO enhancement.  Even Sean seems to be coming around to the idea that the manufacturer's recommendation might be the right thing to do.  

 

I think the most interesting comparison of TL prime to Leica M prime would be the TL Summilux 35 to the 35/1.3 FLE M lens.  The latter works really well on my CL, but I have seen a few very nice shots with the TL equivalent, which might be the perfect CL downsizing of the latest SL 50 autofocus lens.  Zone focus is no problem with an M lens on the CL, as both aperture and distance scales are external.  You can set Auto ISO for the camera to use in figuring out the rest, and bang away.

 

Another interesting comparison is the TL zooms against the TL "pancake" 18 and 23 primes.  I spent some time over the weekend pixel-peeping with the 11-23 TL and the 23/2.0.  No firm conclusions yet, but at least for close work (<1m and f/5.6-8), the zoom seems to be the stronger lens.  Both are sharp and contrasty across the whole frame at normal distances and f-stops.  I have posted some 23/2.0 shots in the image thread.  Also waiting for a closer look is the 24/2.8 M.

Edited December 31, 2017 by scott kirkpatrick

[jaapv]

Thank you, I just read Sean Reid report from 2014. He seems to suggest that the digital correction reduces some of the lens resolution as opposed to optical correction. The other issue that has been more inclined to the M glass is ability to pre zone focus which doesn’t seem to be a very practical option on the TL

That is the crux. Digital correction is not opposed to optical correction. It is NOT a way to correct poor design. It is an extra degree of freedom for the designer to correct lens aberrations.

 

It works like this: lens correction is about shifting aberrations. If you correct one, it will create several others, which in turn must be corrected elsewhere. That means a good designer will tweak each element to make the sum of corrections produce as small a drawing error of the lens as possible. But there will always be residual aberrations.

 

A modern lens design will correct a lens optimally by optical means, shifting the residual aberrations mainly into distortion, leaving a higher level of correction for other aspects. Distortion -and lateral chromatic aberration- can be corrected digitally, leaving us with a better lens than one that would have been designed with only optical means. And yes, the digital corrections will involve a slight loss of resolution, but it is a small price to pay for a better quality overall. The minimum resolution needed for impeccable image rendering is 80 lp/mm (figures by Zeiss) and even digitally corrected lenses will outresolve this value in the corners.

 

Strangely enough, some people seem to think that digital correction is a substitute for poor design.  I greatly respect Sean Reid (now that his name has been dropped, not singling him out) and like him a lot as a person, but this is one of his few blind spots. The lens designers of the top brands, be it Leica, Canon, Zeiss, Nikon, Fuji etc. will have nothing to do with sloppy work, and they have all embraced this way to design better lenses within the parameters (size, lens speed, cost, mechanical restrictions, etc.) that they have been set.

 

As for zone focus, it is better written as zone misfocus. An image is focussed sharply in one plane only. Zone focus determines the area which falls within the gradient of acceptable unsharpness.

[scott kirkpatrick]

Jaap, I think we agree, and you have nailed the big picture -- design for digital optimizes the final image by taking advantage of both optical and digital corrections.  To be "shocked, shocked!" by the digital corrections is to ignore all the terrible tradeoffs that have always been a part of optical design, leading to lenses with great "character," but not always the most faithful imaging. 

 

P.S.  One of these days, someone is going to incorporate a mild radially-dependent sharpening to go along with distortion correction, using the same DNG information, and then we will no longer look for soft grains of sand in the corners of our pictures.

Edited December 31, 2017 by scott kirkpatrick

[lct]

[...] Strangely enough, some people seem to think that digital correction is a substitute for poor design. [...]

 

It is actually. I like much my little Panasonic LX100 (clone of the Leica D-Lux 109) but it crops my images to correct for some huge distortion. If i knew that any of my "serious" cameras did that sort of thing i would return it immediately. YMMV.

[jaapv]

MMV indeed. A purely optically designed lens of the same level would create a smaller angle of view, giving you the same image as the crop, at a lower state of correction. The only difference is that the user is able to make the final digital correction visible. by showing the unused -inferior- field. What sense does that make?

 

More and more "serious" cameras do this, should you be returning you Q, X, X Vario, TL2 (+native lens) -just to remain with Leica; any other brand will have such a list; you will be stuck in the past, I fear. I suspect that even the superior SL lenses have some digital corrections built in.

 

An example from the past is Olympus OM. Some of the Zuiko lenses reach an amazingly high image quality, for instance the 75-150 Zuiko zoom is arguably the best  compact mid-range zoom lens of its time - but the designers could only reach this level of correcttion by introducing very pronounced barrel- to pincushion distortion when zooming, making the lens unsuitable for subjects with straight lines except at about 110 mm.

If only they would have had digital corrections available...

[lct]

I don't mind to those computerized tweakings on a €600 camera lens included but i won't spend little fortunes for the best lenses in the world if they have a lousy design and i'm pretty sure you won't either.

Edited December 31, 2017 by lct

[jaapv]

That is a supposition that is not the case.

Haven't you read my post? The design of a hybrid lens is BETTER overall, as it shifts the unavoidable residual optical aberrations to be corrected digitally, making it the superior lens.

[lct]

I could not disagree more sorry. A lens with too much distortion, lens curvature or chromatic aberration is nothing but an inferior lens and no computer in the word can give it any optical superiority. YMMV again of course but this is a point we will never agree about i'm afraid.

[jaapv]

All lenses that are only corrected optically have residual aberrations that could benefit from digital correction. A lens that has the residual aberrations shifted to the next degree of freedom by integral design,  i.e. digital correction will always be the better lens.

The optical design can be better when some aberrations are corrected by other means.

Disagreeing will not change reality, I fear.

[lct]

This is what the "superior" or "better" lens of my otherwise excellent LX100 does before computerized tweakings. No problem for 600 euros or even a bit more for a compact camera like this but don't ask me to spend more on any M or TL lens designed this way, it will never happen. 

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

 

 

[scott kirkpatrick]

 

I suspect that even the superior SL lenses have some digital corrections built in.

 

They do.  Both SL and S lenses ask (in their DNG files) for digital corrections.  Even when there is no geometric distortion there are chromatic shifts that are best handled there.

 

Incidentally, if you look at the distortion curves that Leica and Zeiss print for their finest lenses of the <=1990s, you will see that designers used to cause distortion to be largest at middle radii, reducing it to a small value at the edges of the frame.  This solved a first order problem, but caused a second order problem -- mustache distortion!  Now the SL lenses don't include distortion curves as this is considered to be a solved problem, but I suspect they no longer build them the old way.  The last few curves that I have seen show monotonically increasing distortion, which can be corrected without side effects.  And in the course of correction (e.g. with Capture One, but not with Adobe, who hides it) you can indeed get the extra pixels at the edges.

[Advertisement]

Hallo,

Look here: Leica L-Mount

[scott kirkpatrick]

The heat generated by this discussion (LCT and Sean Reid are by no means the only people that believe that only optical correction deserves respect) shows why Mr. Karbe seldom gives interviews that go into technical detail.  Which is a shame, as I would love to know more about how lens design is conceived of today.