Kristian Dowling - 28mm Summilux 1.4 Review

[IkarusJohn]

Advertisement (gone after registration)

LCA (and its effect off the plane of focus) is present in modern high contrast lenses.  It just goes with the territory.  Choose an older design if it bothers you.  

 

 

As most of my lenses are fast and modern, I do find this trying at times.  It has meant that I avoid those high contrast transitions (hot spots or where there is a brightly lit background).  Is it possible that this flaw will be designed out of modern lenses?

 

I ask from a pure position of ignorance about lens design ...

[CheshireCat]

LCA (and its effect off the plane of focus) is present in modern high contrast lenses.  It just goes with the territory.  Choose an older design if it bothers you.  

 

The latest Zeiss lenses have very little CA, still they are modern high contrast lenses. Yes they are expensive, but Leica's 28/1.4 is much more.

[Advertisement]

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

[Rick]

Hi John,

 

It seems that these high contrast lenses produce this phenomena.  I agree that the best option is to shoot around it.  I'm not sure that it is a design flaw.  It is really just a side effect of great lens design.  

 

Rick

[Rick]

The latest Zeiss lenses have very little CA, still they are modern high contrast lenses. Yes they are expensive, but Leica's 28/1.4 is much more.

 

CA is very low on Leica lenses as well.

[hiepphotog]

As most of my lenses are fast and modern, I do find this trying at times.  It has meant that I avoid those high contrast transitions (hot spots or where there is a brightly lit background).  Is it possible that this flaw will be designed out of modern lenses?

 

I ask from a pure position of ignorance about lens design ...

 

For anything 50 or above, you want lens with APO (i.e. the like of Otii). APO design doesn't work as well with WA so you need more elements and more ASPH elemetns to correct it. That would mean bigger and heavier lens. You can correct for PF by scarifying other aberrations (like more distortion, more astigmatism, more field curvature, etc.) But I think that would result in an even weaker offering.

[gpwhite]

Hi John,

 

It seems that these high contrast lenses produce this phenomena.  I agree that the best option is to shoot around it.  I'm not sure that it is a design flaw.  It is really just a side effect of great lens design.  

 

Rick

 

It is true that the 28 Summilux produces substantial CA at edges with very high contrast even at f/5.6 if you work to make it happen.

 

BUT, as we learned from the Quasilux 28mm f/1.7 and from Jaap's comments on this point several weeks ago, there is a trade-off in design of high-speed WA between producing CA and correcting for spatial distortion. You get one, or the other, take your pick.

 

Cases in point: the Quasilux shows virtually no CA, and we have learned from uncorrected DNG and evaluating the periphery of images what sort of large trade-off the Q makes to achieve this look with lots of baked-in processing;

the 28 Summilux renders images with virtually no spatial distortion (you might think they are from a 50 SX), but one does need to add the extra step in converting RAW to remove the CA.

[Rick]

Advertisement (gone after registration)

The images Kristian posted are LoCA and not CA.  I didn't see any CA in his shots.  

 

If you point the lens at a bright background with dark objects in the foreground and shoot at 1.4 you are going to get purple fringing, LoCA.

[hiepphotog]

The images Kristian posted are LoCA and not CA.  I didn't see any CA in his shots.  

 

If you point the lens at a bright background with dark objects in the foreground and shoot at 1.4 you are going to get purple fringing, LoCA.

 

By CA you mean Lateral CA, typically appeared around the edges and corners? Or you mean bokeh fringing, out of focus LoCA? I am not sure about the bokeh fringing, but I remember Sean Reid infinity test shots of the Lux 28, it does have a bit of LCA. LCA does increase as you stop the lens down. I think the term CA is too general that we might have a misunderstanding here.

[lct]

Ah yes another discussion about lateral CA, axial CA and sensor blooming please.

Edited August 3, 2015 by lct

[IWC Doppel]

A nice review. I quite fancy one now 

[Rick]

By CA you mean Lateral CA, typically appeared around the edges and corners? Or you mean bokeh fringing, out of focus LoCA? I am not sure about the bokeh fringing, but I remember Sean Reid infinity test shots of the Lux 28, it does have a bit of LCA. LCA does increase as you stop the lens down. I think the term CA is too general that we might have a misunderstanding here.

 

LoCA.  Stop down if you don't want to experience it on high contrast, fast lenses like the 28 SX.  CA this lens does not suffer from.

 

Here we go again...

[CheshireCat]

LoCA.  Stop down if you don't want to experience it on high contrast, fast lenses like the 28 SX.  CA this lens does not suffer from.

 

Here we go again...

 

Like hiepphotog just said, the problem is that "CA" is a generic term, and also includes "LoCA", of course.

So you should say: "this lens only suffers from longitudinal CA, not lateral" (which is to be demonstrated).

[gpwhite]

Like hiepphotog just said, the problem is that "CA" is a generic term, and also includes "LoCA", of course.

So you should say: "this lens only suffers from longitudinal CA, not lateral" (which is to be demonstrated).

 

Well, I am quite confused....  even after shooting the 28 Summilux a fair amount as I came to choose it over a Summicron a couple of months ago. Having nothing to do with CA, I am quite happy with my choice (both lenses produce fringing tons).

 

My lens, which is excellent by all other measures I probed, produces substantial purplish fringing at high contrast edges wherever such juxtapositions of density occur in the image. The phenomena is reduced, but not eliminated, by stopping down to f/5.6. On some images, there are also edges with green fringing in less contrasty areas, as though the green glow masks a bit of shadow.

ACR removes it all by applying the different CA sliders, whether the fringing appears in center image or periphery.

 

When I read about lateral and longitudinal CA on line, the manifestations of these different optical challenges seem to be the same in the image recorded. Rick, how is it that you can tell apart lateral from longitudinal CA in images from your 28 Summilux?

 

My view of CA is that when spotted, get ACR to remove it! I am just trying to learn about why a lens is so surely showing this rather than that, or that rather than this  .

[hiepphotog]

Well, I am quite confused....  even after shooting the 28 Summilux a fair amount as I came to choose it over a Summicron a couple of months ago. Having nothing to do with CA, I am quite happy with my choice (both lenses produce fringing tons).

 

My lens, which is excellent by all other measures I probed, produces substantial purplish fringing at high contrast edges wherever such juxtapositions of density occur in the image. The phenomena is reduced, but not eliminated, by stopping down to f/5.6. On some images, there are also edges with green fringing in less contrasty areas, as though the green glow masks a bit of shadow.

ACR removes it all by applying the different CA sliders, whether the fringing appears in center image or periphery.

 

When I read about lateral and longitudinal CA on line, the manifestations of these different optical challenges seem to be the same in the image recorded. Rick, how is it that you can tell apart lateral from longitudinal CA in images from your 28 Summilux?

 

My view of CA is that when spotted, get ACR to remove it! I am just trying to learn about why a lens is so surely showing this rather than that, or that rather than this  .

 

Lateral CA does not occur in the center and becomes more prominent as you approach the edges and corners. In the plane of focus, it appears on the edges of the object with magenta and cyan fringes opposite of each other. You can't get rid of it by stopping down; most often time, it actually gets worse. Luckily, LR/ACR can fix this problem easily. In-focus LoCA, or purple fringing, is harder to fix in LR because they just desaturate purple that might interfere with other parts of the picture. You can stop down to minimize/rid of purple fringing. PF can be accentuated on digital sensor via what is known as "sensor blooming." Out-of-focus LoCA, bokeh fringing, is the green fringing of the background bokeh and purple fringing of the foreground bokeh. This is the most difficult one to fix. Most APO lenses, Leica and Zeiss included, can minimize most other CAs, but this bokeh fringing is still quite prominent. Like the in-focus LoCA, LR/ACR can partially fix this problem but it might interfere with whole picture.

[gpwhite]

Thank you, hiepphotog, for the detailed explanation!

 

With this information in hand, I can report that the 28mm Summilux produces *all* of the chromatic aberrations that hiepphotog described, obviously depending upon the circumstances. Indeed, now that I know what is LoCA and what is LaCA, I understand why efforts to clean out some instances of either green or violet fringing wind up diluting that image's color saturation in those channels, so I get kind of stuck.

 

However aberrant its recordings are, the 28 Summilux remains a wonderful lens that I very highly recommend.

[lct]

Would you have a pic to show that surprising mix of aberrations? Just curious.

Edited August 4, 2015 by lct

[gpwhite]

I can look for one....  after all, you are asking for an unedited .jpg conversion of the DNG I have in mind.

 

Did not mean to convey that all of the aberrations are in one single image, BTW. What I intended to convey is that, depending upon the circumstances, the 28 Summilux can produce images that evidence one or more of the CA symptoms that hiepphotog explained. Is that more clear?

 

There is a particular scene, shot at f/2.8, that I remember giving me lots of trouble to make a good print. I will look tonight or tmw to see if I can oblige. Then you will need to peek through all of the pixels to find the rainbow 

[Leica Frog]

Thank you so much for a nice review, I'd ever met Lawrence at AV camera, Bangrak district, he's very nice and smart guy, you're so lucky in photographer & friend. From your review, I found you'd compared this lens in size with 35 Summilux FLE, so I'd like to see its quality comparision version if you can, I think it'll be more useful for many Leica users, thank you in advance.

[CheshireCat]

I will look tonight or tmw to see if I can oblige. Then you will need to peek through all of the pixels to find the rainbow 

 

Any news ?

[Advertisement]

Hallo,

Look here: Leica M

[gpwhite]

Any news ?

 

Yes. I posted some examples yesterday under a thread about the 28 Summilux and chromatic aberrations.