35mm distortion 2%

[lincoln_m]

Advertisement (gone after registration)

Hi,

I notice that while the 35mm f2 asph summicron has virtually no distortion the 35 f2.5 and new 35 f1.4 have about 2% distortion at 15mm from the image centre. If one is to use these distorting lenses for subjects with straight edges, buildings, trees will the 2% ( barrel distortion I think) be noticeable? On a 7500x5000 pixel slide scan 2% is about 75 pixels bend at the edge of the image, on either side. I suspect that this is just noticeable on tall Roman pillars etc

 

I think only Film photographers can really comment because M8/M9 users will be using the 6bit code thing that I assume the software will understand and correct the distortion for each lens accordingly. Am I right here? Do M8/M9 get a new software update with each new lens added to the M-system?

 

So the new 35mm f1.4 has the same distortion and similar MFT as the 35f 2.5 but is 3 times the prices and OK yes 2 stops faster. If you compare both lenses at f5.6 then only the far corners are a little softer on the summarit.

 

I think the lack of distortion in the 35mm f2 summicron asph makes it a very special lens especially as the need for f2 or even f1.4 is probably quite rare these days especially on M9 with high iso.

 

What do you think? Is the 35f2.5 therefore a bargain at around £1100?

 

Regards, Lincoln

[Advertisement]

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

[el.nino]

I think only Film photographers can really comment because M8/M9 users will be using the 6bit code thing that I assume the software will understand and correct the distortion for each lens accordingly. Am I right here?

 

No, distortion is not being corrected at all.

 

and btw: the 50mm slx also has more distortion than the 50 cron.

=> get the cron (it's always the better choice over the lux - unless you really really really need that one stop more light)

[UliWer]

Just have a look at Reid Reviews:Welcome to ReidReviews

 

Since yesterday you find a new test with the 2/50 asph the 35mm Summarit and severaly other lenses on the M9 with comparisons in any respect you may ask for. The test will be continued with more 35mm and the M9. There are also older tests with 35mm lenses on the M8. You won't find better information.

[Walt Calahan]

My copy of Photoshop CS5 arrives tomorrow, and am looking forward to learning its distortion correction features in Camera RAW control.

 

There are other software packages that can correct lens distortion, but I don't own any.

[ho_co]

There's a thread on the topic of Summarit 35 distortion at http://www.l-camera-forum.com/leica-forum/leica-m9-forum/117056-barrel-distortion-35mm-summitar-my-m9.html. Post #17 shows the example that prompted the post.

 

The Summarit's distortion can easily be corrected by your image processing program.

[lars_bergquist]

The old Summilux ASPH has about 1.4% barrel distortion. I was originally worried about that, but while I would not use it as a copy lens (and who would use a 35mm lens on a Leica M for copy work?) I found this completely unproblematic in practical use. The new version has about 1.5% barrel, and that does not worry

me either -- I'm in line for one.

 

It is when distortion reaches 2% or more that it disturbs me. But all this is very subjective, there is no absolute limit. Also, barrel distortion is less iritating than an equal amount of pincushion.

 

The old man from the Age Before Scanning

[pgk]

Advertisement (gone after registration)

Being practical about distortion means that in effect, the only time that it is likely to be noticeable (and therefore need correcting) is generaly when straight edges are in an image parallel to the image sides. In practice this means man made subject matter or obvious naturally straight edges such as the sea or a lake where this is the horizon. If these are in many of your images then yes, in my experience the Summilux asph will show noticeable distortion if you are the sort of viewer who is aware of such things. However the distortion is gradual and unlike some earlier designs does not have any waviness so is easily corrected in post processing.

 

If you shoot images which do not require a very fast lens and you do not want to spend time adjusting them in post then the 'cron is clearly the more usable lens. If on the other hand you will utilise the extra stop and are prepared to post process the few images where distortion may be obvious, then the 'lux may be a better bet. The Summarit also distorts but is considerably cheaper! I have owned all three but have settled on the fabulous current 'lux because it suits my requirements best (and mine works exactly as it should). All three are superb lenses with differing attributes.

[01af]

While a distortion of -2 % is leaning towards the high side for a prime lens with an unspectacular focal length, it still is low enough not to be a problem in most cases. However it tends to become worse at close range where it might be a problem for critical subjects. Anyway, in my Summarit-M 35 mm I didn't notice any distortion at all until I did some test shots specifically to test for distortion.

 

And it's well-behaved pure barrel distortion, no weird wave-form distortion, so it can be corrected in post-processing easily if necessary.

[batmobile]

the 24 lux has pretty severe distortion up close and is rated at 2.2% I think. The Summarit can be noticeable too. This is where Zeiss lenses really are in a league of their own. Still, if you want a 35mm 1.4 or 24 1.4 for your M, Zeiss cannot help you unfortunately.

 

I use my Ms for stuff wherte distortion matters little. The only shot I have where it bothers me is one of some children in a shall hole in a wall. Funnily enough it was taken with a lens with only modest distortion but because of the neat rows of bricks (rare in itself in Afghanistan!) it looks a bit funny.

 

Distortion with fast wides is a good reason not to shoot brick walls

[Jeff S]

Just have a look at Reid Reviews:Welcome to ReidReviews

 

Since yesterday you find a new test with the 2/50 asph the 35mm Summarit and severaly other lenses on the M9 with comparisons in any respect you may ask for. The test will be continued with more 35mm and the M9. There are also older tests with 35mm lenses on the M8. You won't find better information.

 

+1 (guess that means I agree)

 

Jeff

[lincoln_m]

Hi,

Thanks for the feedback on 2% distortion. I thought it would be noticeable or just about depending on the subject and you guys have confirmed that. You might have thought Leica would have kept it low for the top spec summilux but perhaps something worse happens with no distortion, like bad flare?

 

I guess that saves me over £3K on the summilux and points me to another 35 summicron asph in black if I need another 35mm to go with my 35 summicron asph in chrome. I've had no problems with the 35 summicron asph apart from feeling a little heavy and perhaps out of place when on my black M7 instead of on the chrome MP. Either way it takes great photos on any M.

 

The summarit may be cheaper and lighter/smaller but for my seascapes or pillars in Bath city, buildings etc the walls horizontals and verticals will need correcting. I don't have PS CS4 only Aperture 2 which does not have distortion of perspective corrections (Aperture 3 might?) so the summarit might be more annoying than paying the extra £400 for a used 35summicron asph in black mint- / exc++. I'll just keep with my silver on black look when out with the M7 for now.

 

Regards, Lincoln

[Jeff S]

If you think the 35 Summicron asph has no distortion, well, then you really could benefit from joining Sean's site.

 

Jeff

[adan]

You might have thought Leica would have kept it low for the top spec summilux but perhaps something worse happens with no distortion, like bad flare?

 

Any lens design is like herding cats. There are many variables that affect one another - distortion, flare, vignetting, chromatic aberrations, spherical aberrations, max. aperture, flatness of field, image circle (the format the lens has to cover). And that's just the optics - also throw in size and weight, or whether the lens has to clear an SLR mirror or an M9 meter, and production cost (even for Leica!)

 

The old Nocti f/1 was intentionally allowed to have very high vignetting, in order that something else (I forget what) could be improved - flatness of field or spherical aberration or something.

 

It's kind of an expanded version of the old saw - "You can have fast, you can have good, and you can have cheap - pick any two."

 

Leica's lens design software allows them to play with all those variables and come up with multiple approaches to a lens of a given spec (35mm, f/1.4, covers 24 x 36). Then they have to decide which variables are most important.

 

Leica probably could have designed a distortionless 35 f/1.4 (keeping the other variables equal) - at the size and for the price of a Nocti f/0.95. But is that the "best lens"? Presumably, for the intended function of an f/1.4 35mm lens (not something one buys for landscape or architecture, but for low-light work) - distortion was lower on the priority list than other variables.

 

"You can have f/1.4, your can have 35mm, you can have distortionless, and you can have a 46mm filter size - pick any three."

Edited May 28, 2010 by adan

[Xmas]

Any lens design is like herding cats.

 

The old Nocti f/1 was intentionally allowed to have very high vignetting, in order that something else (I forget what) could be improved - flatness of field or spherical aberration or something.

 

It's kind of an expanded version of the old saw - "You can have fast, you can have good, and you can have cheap - pick any two."

Hi

 

Vignetting allows the lens to be smaller, lighter and the off axis performance to be 'improved', the off axis humps in the MTF graphs are a typical result, when vignetting is 'optimised/matched to off axis performance'.

 

Zeiss allow their lenses to 'grow' in optimisation for performance, Leica constrain size in optimisation, adding asphs to achieve, cause lots of people like small and light, and complain loudly, CV go for small, slow & cheaper.

 

Noel

[SJP]

Any lens design is like herding cats. <snip>

 

The strategy with that is to feed them at regular & predictable intervals e.g. exactly at 20:00'00" (add more digits as required) - then herding of multiple cats is dead easy.

 

Lens design is much harder.

[rsh]

If I wanted or needed to buy a 35 Summilux ASPH, New Version or Old, I would buy it for the extra stop whether or not there was some distortion.

 

As it is, I do not need the extra stop and I am content with the 35 Summicron ASPH. Please remember, there is not a bad Leica lens, but the lens you choose should be the one that fits your needs the best. If price is an issue, then you must compromise, but in the compromise, you lose no image quality, although you may miss certain qualities a particular lens imparts to an image.

[cliffp]

Has anyone yet worked out the correct distortion correction settings for the Summarit 35mm in Photoshop?

[tobey bilek]

Leica lenses should not have distortion period. The old ones did not, so I view distortion as regressing.

 

If I were willing to accept distortion, I would use Nikon lenses. I never noticed distortion on Leicaflex lenses or RF lenses from the 1960`s to Mandler.

 

Shame on Leica

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[ho_co]

You're right, Tobey!

 

My mother wasn't distorted either. Why should I allow distortions in my life?

 

It's clear, isn't it, that Leica invented distortion as a distraction to keep us from noticing how much money they're making.

 

 

(And I love your SoL closer. )

Edited May 31, 2010 by ho_co

[pgk]

I never noticed distortion on Leicaflex lenses or RF lenses from the 1960`s to Mandler.

You are right - they had character:D!