Summilux-SL 50 MM F/1,4 ASPH

[Chaemono]

Advertisement (gone after registration)

Because of the assumption that because there are in camera corrections for this lens, apparently baked into the DNG, Leica's claim that it is a reference lens is an empty piece of marketing for an overpriced lens.

I'll wait till I have my copy to make a judgement whether there are actually in camera corrections for this lens. When I open SL DNGs created with the following M lenses 28 Lux, 35 Cron, Noctilux, 50 Apo Cron, and 90 Apo Cron LR tells me that the integrated lens profiles are applied. But no digital corrections are actually applied until profile corrections are enabled. With DNGs from the native 24-90 SL lens the baked in corrections by Leica are automatically applied in LR and cannot be undone there as we know.

 

If the 50 Lux-SL is to be a reference lens, I would expect it to be maximally optically corrected. Opening the files and getting the message that the integrated lens profile (or the Leica baked-in corrections) has (have) been applied by LR doesn't mean anything. That's the way the DNGs with the above mentioned M lenses are shown. Now, Adam seems to have been able to separate the Leica baked-in digital corrections and shows some barrel distortion. Again, I'll wait for my copy to see for myself.

[IkarusJohn]

I'll wait till I have my copy to make a judgement whether there are actually in camera corrections for this lens. When I open SL DNGs created with the following M lenses 28 Lux, 35 Cron, Noctilux, 50 Apo Cron, and 90 Apo Cron LR tells me that the integrated lens profiles are applied. But no digital corrections are actually applied until profile corrections are enabled. With DNGs from native 24-90 SL lens the baked in corrections by Leica are automatically applied in LR and cannot be undone there as we know.

 

...

 

 

I'm unlikely to get this lens (despite my preference for 50mm), primarily because it is large and expensive.  I already have very good 50mm lenses which work as well if not better on the SL than the M.  And I really like the SL sensor.

 

As to corrections, I think there are two different things happening.  When you mount a coded lens onto an M or SL, the firmware in the camera makes corrections to the DNG file, in the same way that the firmware applies white balance and a number of other corrections.  These are "baked" into the DNG file, and are already applied when you open the file in LightRoom - this is more than vignetting, distortion, while balance and sharpening.  You can remove those corrections as Adam has done, though I don't see the point.  You can also apply the lens profile or your own profiles in LightRoom, using the LightRoom adjustments.

 

The "digital corrections" everyone is worried about here are the corrections made in the firmware in the camera, not the lens profiles in LightRoom - quite why we have both, I'm not sure.  I don't really understand why LightRoom doesn't pick up the lens coding in the EXIF file and apply the lens profile automatically.  That said, my version of LightRoom doesn't seem to have all the camera and lens profiles (God knows why - probably some incompetence on my part), so I ignore them.  

 

Cheers

John

[zlatkob]

You are still making the error of separating the digital corrections from the optical ones. The optical corrections have been made with the digital corrections in mind. One cannot exist without the other. Lens corrections are not a user control and certainly not a basic one.

 

Sorry, but you are just making up rules as you go.  You say, "The optical corrections have been made with the digital corrections in mind. One cannot exist without the other."  But of course they can.  There is no rule or law that says the digital corrections must be baked into the DNG.  You've simply made up a rule and presented it as a priori.  Nevermind the evidence of professional cameras that leave final decisions about digital lens corrections to the user; let's just pretend they don't exist, shall we?

[ZHNL]

Sorry I'm later then I thought I would be with this.. And sorry it's not a more in-depth review. I'm really in a rush to get everything done before I fly off tomorrow. And I landed one unexpected job that took way more time then it was worth TBH...

 

Anyway for what it's worth.. And I'll try to answer specific questions if you have any once I've settled down after flying off.

 

http://www.jupitersnake.com/review/50-summilux-sl-f1-4-review-first-look/

Adam, thanks for the reviews. Very helpful. I can see it is a job in hurry but I have to say I still can get a lot from what I see here.

 

I think if CA performance is true,(means no correction baked in DNG) it is quite respectful, it is better than 90-APO_m and better than most 50mm prime lens save for a few real APO design.

 

Distortion is insane IMHO without correction, I don't remember have seen any 50mm lens behave this bad other than some zoom at this focal. I will not judge sharpness at this time and will not even take your word for it since yours is only copy show up so far. From early samples I saw in this thread, I know it is great in center and good enough for the lens intend to be used. It is not a big concern for me. but say it is not sharp in corner till f5.6 sounds scary. I am sure it must be better.

 

Now, Rendering, This is big to me. The out of focus rendering instantly remind me one lens: Zeiss 35mm ZM distagon f1.4. I am not a fan of that lens personally. (There are many people like it for different reason.) TBH, I feel most modern BEST lenses with good correction fall into this kind of rendering. The few exceptions I have seen so far are: RX1, 85mm GM, Nikon 105f1.4. (of course a few S glass and 50APO)

 

For SL, there may not be better option. But optically, I see this lens will facing strong competition from a few AF lens introduced in past few years. Name a few: sigma Art, Zony 50mm ZA, 55mm Sonnar etc..

 

I don't see it in the OTUS league, at least not so far but it is a AF.

 

Edit; as for rendering, I know this is very subjective. And there are few lens can satisfy everybody especially shooting WO, it is better judge it for more cases especially slow down a stop as most lens I mentioned are f2 lenses.

[Chaemono]

The "digital corrections" everyone is worried about here are the corrections made in the firmware in the camera, not the lens profiles in LightRoom - quite why we have both, I'm not sure.  

 

Thanks for clarifying John. I downloaded the OPCODE Remover to remove those corrections made in the firmware in the camera and see for myself.

[IkarusJohn]

Thanks for clarifying John. I downloaded the OPCODE Remover to remove those corrections made in the firmware in the camera and see for myself.

 

 

I guess it will be interesting to see the extent of the corrections, but to be honest of marginal passing interest only.  I fall into the camp for whom digital correction is not the crime of the century, and it is just another tool in the designer's bag of tricks.  I have no doubt that Leica can produce a lens as optically correct as any other lens manufacturer, perhaps more so; this lens is supposed to be "without compromise" (I'm paraphrasing), but there are limitations.  The mount, the register distance, AF and size (to a degree).  I take on board that with the increased mount size and less constraint on lens size, there should be no reason for this lens to be worse than the APO-Summicron-M 50.

 

But, if Leica can make gains by letting some "imperfections" through which can easily be corrected digitally, what do I care?  It's all part of the package.  This lens isn't designed to be used with Sonys or anything else - unlike Zeiss, Leica still makes cameras, and its lenses are designed for use with them.  I should add that even if this lens can be adapted for use with other cameras (why?), I can think of no reason why the "baked in" corrections cannot be replicated with careful post processing.

 

That rather assumes that there is some advantage for Leica in letting this apparent optical imperfections slip through to be corrected in the firmware.  Who knows?  The proof of this pudding is in mounting the lens on the SL and examining the output from the camera.  Removing the digital corrections is surely, ultimately pointless.  Of marginal interest, but pointless.

 

I should add that the fruit and veg man (who many people here rate) seems to be withholding judgment until the final lens and firmware are released.

 

Cheers

John

[Chaemono]

Advertisement (gone after registration)

Removing the digital corrections is surely, ultimately pointless.  Of marginal interest, but pointless.

 

Thanks John. Your comments are much appreciated. I respect your point above but let me explain. As you say, there is no reason for the 50 Lux-SL to be worse than the 50 Apo Cron. And Leica claims that it is the best 50mm lens ever, a reference lens. But if the distortions are as insane as Adam shows and digital corrections are applied at the expense of resolution it cannot possibly be the best 50mm lens ever made.

[IkarusJohn]

I understand the point you make, but ultimately for me, if I bought this lens, any optical imperfections would be irrelevant.

 

I don't agree that the distortions are "insane" at all.  The barrel distortion is there, and easily corrected, but not extraordinary.  Anyway, I guess we'll agree to differ - I am not really interested in theoretical issues - I put the lens I want on the camera, and I take pictures.  I then process what comes out of the camera.  I certainly don't go to the effort of second guessing what Leica does in camera - I tried that with my M Edition 60, and I couldn't even get close to what the camera produces.

 

That said, I don't use the Jpegs from my cameras (the M Edition 60 and Monochrom don't even produce anything beyond the thumbnails).

 

All the best.

John

[zlatkob]

I guess it will be interesting to see the extent of the corrections, but to be honest of marginal passing interest only.  I fall into the camp for whom digital correction is not the crime of the century, and it is just another tool in the designer's bag of tricks.  I have no doubt that Leica can produce a lens as optically correct as any other lens manufacturer, perhaps more so; this lens is supposed to be "without compromise" (I'm paraphrasing), but there are limitations.  The mount, the register distance, AF and size (to a degree).  I take on board that with the increased mount size and less constraint on lens size, there should be no reason for this lens to be worse than the APO-Summicron-M 50.

 

But, if Leica can make gains by letting some "imperfections" through which can easily be corrected digitally, what do I care?  It's all part of the package.  This lens isn't designed to be used with Sonys or anything else - unlike Zeiss, Leica still makes cameras, and its lenses are designed for use with them.  I should add that even if this lens can be adapted for use with other cameras (why?), I can think of no reason why the "baked in" corrections cannot be replicated with careful post processing.

 

That rather assumes that there is some advantage for Leica in letting this apparent optical imperfections slip through to be corrected in the firmware.  Who knows?  The proof of this pudding is in mounting the lens on the SL and examining the output from the camera.  Removing the digital corrections is surely, ultimately pointless.  Of marginal interest, but pointless.

 

I should add that the fruit and veg man (who many people here rate) seems to be withholding judgment until the final lens and firmware are released.

 

Cheers

John

 

I prefer to have the digital corrections in my bag of tricks, not exclusively in the designer's bag of tricks.  Removing digital corrections isn't pointless to me.  When photography is one's work, details like this get to the heart of whether a camera is a professional tool or not.

[LocalHero1953]

If this lens is not to be released into the wild till January, I cannot believe that this is anywhere near final f/w. Reviews now are interesting and gratefully received, but I look forward to reviews of the final version.

 

And choosing to use a lens without the digital corrections that are designed integrally with the optical design seems to me a bit like choosing to use a lens without an apochromatic correction element, because it's somehow more pure. Sure it's simpler, but you're likely to get a cr4p image.

[zlatkob]

If this lens is not to be released into the wild till January, I cannot believe that this is anywhere near final f/w. Reviews now are interesting and gratefully received, but I look forward to reviews of the final version.

 

And choosing to use a lens without the digital corrections that are designed integrally with the optical design seems to me a bit like choosing to use a lens without an apochromatic correction element, because it's somehow more pure. Sure it's simpler, but you're likely to get a cr4p image.

 

It has nothing to do with "more pure".  It has to do with expressing and crafting your image.  It has to do with the photographer being the decision maker.  If you've never seen how digital corrections can (sometimes) make an image worse, then it is hard to explain.  Maybe it just sounds theoretical, rather than practical.  The concept of manual control is consistent with the idea of Leica as professional tool or an artist's tool.  The concept of "We built-in automation, whether you like it or not" is more consistent with another kind of camera.

[jaapv]

If this lens is not to be released into the wild till January, I cannot believe that this is anywhere near final f/w. Reviews now are interesting and gratefully received, but I look forward to reviews of the final version.

 

And choosing to use a lens without the digital corrections that are designed integrally with the optical design seems to me a bit like choosing to use a lens without an apochromatic correction element, because it's somehow more pure. Sure it's simpler, but you're likely to get a cr4p image.

Ach, Paul, what is a spot of CA and Coma between friends...

BTW, despite several heads close to the corners on Sean's sample images, I see no egghead distortion, making the whole discussion slightly absurd. The idea of creative control by the photographer by removing some of the lens corrections appears rather curious to me.

[zlatkob]

Ach, Paul, what is a spot of CA and Coma between friends...

BTW, despite several heads close to the corners on Sean's sample images, I see no egghead distortion, making the whole discussion slightly absurd. The idea of creative control by the photographer by removing some of the lens corrections appears rather curious to me.

 

It's something you can easily see and likely noticed if you ever compared before / after correction for barrel distortion in Lightroom.  Even with a 50mm lens, correction causes pixels near the edge of the frame to move more than pixels that are nearer the center of the frame.  So a round object is not stretched uniformly, but rather is elongated more on the side that is nearest to the edge of the frame.  In other words, after correction, it bulges out a little on one side.  If you've ever photographed a person with rectilinear wide angle lens and placed them near the edge of the frame, then you know the unnatural stretched look.  The same thing happens with a fully corrected 50mm lens, albeit in a more subtle way.  We're talking about a small but noticeable change in the shape of the object.  You can say that it doesn't matter for many photos, but you can't deny the phenomenon itself.

SaveSave

[adamdewilde]

The digital corrections IN THIS CASE definitely have a negative effect on people photos. Since that is what I shoot, for me the digital corrections are a negative.

This comes at a shock to me, as I've owned plenty of 50mm lenses that are AF/Smaller/Better distortion wise. 

 

I could only come to the conclusion which I mentioned earlier in this thread that the lens may have been designed for a larger sensor simply to allow for Leica to use the sweet spot. As either a rush job or a cost saving measure. I'm not a lens designer, I don't even know if my theory makes sense.. But being a company who prides itself on optical excellences I don't understand why this lens isn't optically excellent for the price/size (which from the start was and still is my complaint). 

 

The end result is acceptable in 90% of the cases.. But that percentage drops if you shoot people. I actually much prefer the rendering of this lens to the Sigma ART and a lot of other 50mm lenses. I just wish it didn't make portraits look wonky like it does.

 

 

IF I could sell my SL+Zoom, I wouldn't have regrets when this lens hits the market. BUT since I don't think I'll get a good price on my kit (nowhere near what I paid for it). I will likely end up buying this lens, or pending delivery date of the 75APO, I might hold out for that. Since I've always favoured shooting portraits with a 75mm over a 85mm or 90mm

I'm currently using my remaining S-lens (100S) on the SL, and frankly it has nowhere near the charm it does when mounted on the S-006. I still think the best S-lens for the SL is the 70S. And I regret selling mine cheap, as it would be a good stop-gap option. 

 

 

End of the day, I still don't think anything out there beats an M240 + 50APO (simple, quick to focus, as weather proof as I need it and a pure joy to use). Some of you may disagree and that's cool with me, but really when I say this, I'm not exaggerating.. The 50APO has basically ruined me for any other lens. And to me, that's the mark of a true reference lens. 

[adamdewilde]

It's something you can easily see and likely noticed if you ever compared before / after correction for barrel distortion in Lightroom.  Even with a 50mm lens, correction causes pixels near the edge of the frame to move more than pixels that are nearer the center of the frame.  So a round object is not stretched uniformly, but rather is elongated more on the side that is nearest to the edge of the frame.  In other words, after correction, it bulges out a little on one side.  If you've ever photographed a person with rectilinear wide angle lens and placed them near the edge of the frame, then you know the unnatural stretched look.  The same thing happens with a fully corrected 50mm lens, albeit in a more subtle way.  We're talking about a small but noticeable change in the shape of the object.  You can say that it doesn't matter for many photos, but you can't deny the phenomenon itself.

SaveSave

 

This is true... I have some shots of my wife (the woman who was in white w/gold purse) that she won't allow me to post because the distortion was really wonky for lack of a better word. And even so, I think she wanted to adjust the images of herself that made it into the post because something looked off.. And I routinely shoot her with 50mm lenses so it's not a focal length bias (though her favourite setup is S-006+120S).

[pop]

....Even with a 50mm lens, correction causes pixels near the edge of the frame to move more than pixels that are nearer the center of the frame.  So a round object is not stretched uniformly, but rather is elongated more on the side that is nearest to the edge of the frame.  In other words, after correction, it bulges out a little on one side. ...

 

That is the nature of the distortion and its correction. The distortion is created by a lens with a magnification factor that varies with the distance from the center of the image. So is, of course, the correction.

 

You'd just as soon state that everything seems darker when you wear sunglasses.

[wattsy]

Sorry, but you are just making up rules as you go.  You say, "The optical corrections have been made with the digital corrections in mind. One cannot exist without the other."  But of course they can.  There is no rule or law that says the digital corrections must be baked into the DNG.  You've simply made up a rule and presented it as a priori.  Nevermind the evidence of professional cameras that leave final decisions about digital lens corrections to the user; let's just pretend they don't exist, shall we?

 

 

I think you have to make a distinction between digital corrections that are meant to improve an already optically corrected (as far as possible) lens that is not camera body specific and which is designed to be primarily used via an optical viewing system (for example, a modern Nikon AF-S lens that, whilst typically used on a Nikon DSLR, will also be perfectly fine on a film body like the F5) and a lens that is designed to be solely used in conjunction with an EVF based digital body that implements the necessary digital corrections. The latter lens is likely to be deliberately less optimal from an optical perspective (i.e. as a standalone optical unit) because, by not correcting (optically) for those distortions or aberrations which are easily corrected digitally, the lens designer can avoid the introduction of other more serious (less easily corrected digitally) side effects. The overall balance of optical and digital correction results in a "better" overall image and removing those digital corrections is therefore akin to removing the aspherical lens from a conventional optically corrected lens.

[jaapv]

Thank you Ian, you managed to put it more concisely than I was able to.

 

Another aspect of digital correction is that a lens designer will not have to correct aberrations that are aperture-specific in the optical part. On a purely optically corrected lens he will go for a compromise-optimum correction, in an hybrid corrected lens he will shift them into the digital part of the lens corrections.

[Vieri]

My understanding about lens design is that it is, ultimately, a compromise. Certain optical formulas have certain qualities: i.e. no distortion, no vignetting, sharpness all over the frame, no coma, APO, speed, etc., depending of course on focal length as well. However, no optical formula has them all, at least not at reasonable size / weight / price, otherwise everyone would be using such a magic formula. Thus, when designing a lens, the designer is bound to make choices favouring one or the other aspects. With digital, you have the possibility of fixing in FW what you couldn't fix optically: in this case, good lens designers would solve optically what is best solved optically, and digitally what is best solved digitally. I guess we all agree so far.

 

What we don't seem to agree is on whether it should be possible to turn digital correction on and off. In this case, apparently, distortion is corrected digitally; the only reason we are discussing this is because distortion is something that somebody might want to be off at times. I am pretty sure we wouldn't be having this discussion if we were talking about CA, or coma, or any other design parameter. So, the solution is theoretically easy: those who prefer to be able to turn off distortion should ask Leica to give them such an option, and see what happens. Perhaps, just for distortion it might be a good feature to have, even if it goes a bit against Leica's philosophy of simplicity and perfection and even if it probably would be at the very bottom of a list of FW changes.

 

To me, it is not very far from asking a car manufacturer to give me the possibility of switching ABS off, which seems a good example. So far, no car manufacturer did it, but if you want to be in control of your breaking, you might ask...

 

Best,

 

Vieri 

[Advertisement]

Hallo,

Look here: Leica L-Mount

[jaapv]

The problem here is that nobody of us knows what other lens corrections except distortion may be made  digitally.

IIRC the  digital lens corrections of the Q can be avoided by using Raw Therapee as a raw converter. Might that not be the solution in this case?