90 Apo Purple Fringing

[lct]

Advertisement (gone after registration)

If the 90/2 apo were a "purple machine", all lenses including apo's would deserve this rumor in case of major overexposure. If you want to see what CA can do at f/2 take a Summicron 90/2 v3 but the apo is much better from this standpoint. The Summicron 90/2 v2 is even better though but it is a bulkier lens. No experience with 90/2 v1. 

[onasj]

I just don’t see this purple fringing issue with my copy of the 90 APO. See the attached 100% crop, shot at a highly angled focus calibration device. The level of APO correction is nearly as good as the 50 APO, which is the best I’ve ever seen. It’s actually better than the Zeus’s Otus 55 shot on my D810 tested in the same way, and about as good as the Sigma 85/1.4 Art on the D810.

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!

[Advertisement]

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

[adan]

Well, (see below) white-balanced, and with brightness normalized and a bit of saturation added to show what's really going on - the fringing is there.

 

Really obvious is the green-purple divergence in the foreground and background blurs, even barely out of focus (top and bottom of the "0" focus target). To the point that the white paper background itself becomes tinted completely purple or green at the top and bottom of this crop (around the numbered squares). That's a lot of "aberration" (from the Latin aberrare: "wander from the true path, stray, err, make a mistake") of the purple vs. green light rays.

 

But as I said previously, this is how even APO lenses render blurs, and has nothing to do with the definition of APO.

 

Less obvious, but not non-existent, is a bit of purple edge fringing on a black-white border even in the plane of focus - see little square showing the color my eyedropper sampled from one pixel right on the edge of the "0" focus target bar. And a higher contrast range (black leaf silhouetted against sky, for example) will bring out that bit of purple even more.

 

Now, measuring and "seeing" are two different things. And some people will be pickier about what they can see (or measure ). And of course the "remove fringes" button in post-processing can easily delete a 1-pixel fringe. And it is relative - one lens can be better than another, without being perfect.

 

 

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!

[lct]

Well, (see below) white-balanced, and with brightness normalized and a bit of saturation added to show what's really going on [...]

 

A lot of saturation added i would say. More than 50% with my old PS3. At this count the 50/2 apo must show the same aberrations or worse. Not a realistic test IMHO .

[NB23]

I just don’t see this purple fringing issue with my copy of the 90 APO. See the attached 100% crop, shot at a highly angled focus calibration device. The level of APO correction is nearly as good as the 50 APO, which is the best I’ve ever seen. It’s actually better than the Zeus’s Otus 55 shot on my D810 tested in the same way, and about as good as the Sigma 85/1.4 Art on the D810.

This is the internet: "perfect lens,, the best APO correction ever, almost as good as the Masterful Sigma art (Sigma? Really?)... and so on". And yet, there are loads of Green and purple going on.

[adan]

lct - more or less saturation doesn't matter.

 

If the purple wasn't there in the first place, no amount of saturation would create it. It's just turning up the volume on something that's already there.

[onasj]

Advertisement (gone after registration)

No one said the lens was “perfect” (expect NB23). All lenses, I’ve tested including the 55 and 85 Otus have some CA visible in this test. My point is that straight out of the camera the 90 APO is well above average. And in real portraiture settings the CA is sufficiently controlled to not limit shot quality or usability in my opinion.

 

NB23: if you don’t realize that the Sigma Art line currently offers some of the best 35 mm format lenses in existence then you are really missing out.

[lct]

There is no such thing as lenses sans CA. At least i have never seen that so far. Apo lenses have simply less CA than others, which is the case of lenses like 50/2 apo and 90/2 apo compared to their otherwise excellent predecessors. 

[NB23]

No one said the lens was “perfect” (expect NB23). All lenses, I’ve tested including the 55 and 85 Otus have some CA visible in this test. My point is that straight out of the camera the 90 APO is well above average. And in real portraiture settings the CA is sufficiently controlled to not limit shot quality or usability in my opinion.

 

NB23: if you don’t realize that the Sigma Art line currently offers some of the best 35 mm format lenses in existence then you are really missing out.

I find it amazing that the shot you showed actually showcased the very opposite of what you were trying to show.

 

Your post is the classic example of why people shouldn't rely blindly on other people's internet posts.

One man's high standards is another man's substandards.

 

Too much misinfirmation passed on as authoritative proof.

[NB23]

There is no such thing as lenses sans CA. At least i have never seen that so far. Apo lenses have simply less CA than others, which is the case of lenses like 50/2 apo and 90/2 apo compared to their otherwise excellent predecessors.

If so is the case, we are facing the fact that APO is either a lie or a marketing gimmick.

 

They can't sell a non-alcoholic beverage "with some, or a full doze" of alcohol in it, can they? That would be a crime.

 

But they can market a APO lens wich is totally uncorrected and tell us that it's still apo to some extent.

 

Don't forget, we live in an era of Accepted BS.

[adan]

Let's be clear about what an APO lens is - it is a lens that will focus three wavelengths/colors in the same focus plane. Period. That does not mean it will focus ALL colors in the same plane.

 

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!

 

(Diagram from Wiki Commons, original by cmglee; reproduced under GNU license)

 

Lens designers can pick and choose - to some extent - which three colors will focus correctly. Through choice of glass and exact optical layout. In many APO lenses, the "third color" is in the invisible infrared part of the spectrum (useful for astronomical or other scientific work). Without access to Leica or Zeiss lens software, or rigorous lab-grade scientific testing, we can't know exactly which colors any particular lens is optimized for.

 

That's the operating definition of "APO" - which existed in the scientific world long before "marketing" was an issue. The first "consumer" APO lens was the Leitz APO-Telyt-R 180mm f/3.4 (once permission was granted from the U.S. Navy, who commissioned it, to release it to the public in 1975.) The idea of an apochromatic lens dates back to at least 1881 (Ernst Abbe, first APO lens ever made) or even to 1817 (Joseph von Fraunhofer, inventor of the spectroscope and diffraction grating, and maker of achromatic telescopes).

 

An APO lens is not even the pinnacle of color correction (sorry, APO owners ) - for that, you need a Superachromat. For normal photography, the only such lenses that I know of are the Zeiss 250 or 350mm Superachromats for the Hasselblad. Which still won't actually focus all colors in one plane, but by focusing four, reduce (but do not eliminate) the remaining error in the others.

 

As a side-effect of correcting the longitudinal chromatic aberrations, APO/SuperA lens usually also improve spherical aberration (in 2 or 3 colors). So they may be "better" lenses beyond what they do for color-fringing. E.G. the 90 APO is less "dreamy" than the 90 non-APO Summicron @ f/2, purely due to reducing the "soft, glowy" secondary spherical aberration images overlaid from the edges of the elements @ f/2.

Edited October 17, 2017 by adan

[onasj]

The world isn't so black and white, NB23.  Like everything else, there are different degrees of apochromatic correction.  In my opinion, the 90 APO's correction is quite good.  When I subject the Nikon 85/1.4G, which is considered a classic portrait lens but suffers from a high degree of CA (and is not designed as an APO lens), to the same FoCal test I get highly color fringed results that look like Adan's super color-enhanced version... but without any color enhancement!  

 

And yet I used the Nikon 85/1.4G to shoot the attached portrait (low quality JPEG to fit into 500 kb), despite its rather pronounced CA, which is much worse than that of the 90 APO.  My main point I'm trying to convey is that the 90 APO's CA is well within what's needed to shot decent portraits.

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!

[Michael Geschlecht]

Hello Everybody,

 

I think that the traditional colors that are all in focus at the same focal plane with an apochromat are red, green & blue.

 

And to be an apochromat the lens only has to bring the 3 colors to the same focal plane at the center of the image.

 

A number of companies making apochromatic lenses go beyond the above specifications when designing & building their own lenses which they call apochromats.

 

Best Regards,

 

Michael

Edited October 18, 2017 by Michael Geschlecht

[NB23]

I love my 90 apo and Nikkor 85 1.4D, which I use a lot. And I am not a measurebator although I may sound like one.

It's just that I don't think that the 90APO deserves the apo tag if I judge its massive CA wide open.

Edited October 18, 2017 by NB23

[Paul J]

Apo isn't really a destination, it's a design objective on a sliding scale with some lenses being better corrected than others. I guess you could say Perfect APO is a destination but it doesn't really exist.

 

If you compare the Summicron to the APO-Summicron, particularly when it came out so long ago, I think most will agree that relatively speaking it's undeniably apo. By standards we are able to judge today it's probably less so.

 

I found the Summicrons CA hindering and distracting on digital but the APO-Summicron I eventually swapped it for is a different lens in that regard. In every way I found it a much better lens and not even that long ago it was considered by many to be a sort of bench mark across all brands. 

Edited October 18, 2017 by Paul J