35 APO Acceptable Fringing?

[JPizzzle]

Advertisement (gone after registration)

Hi Everyone,

Would appreciate some thoughts on a photo I took with my 35 apo. I received the lens a few weeks back brand new from a Leica store. It hasn't gotten a ton of use really due to the holidays and scheduled. In looking over some photos from a recent walk, I noticed some slight purple fringing. Now, this isn't something that I think would bother someone as the fringing isn't "extreme," but it has me wondering if it should even be present for an APO lens. Figured I'd post on here to get the opinion of those more knowledgeable than myself.

 

Photos are at 100% and shot raw. The first was processed in LrC with remove CA on in lens correction and the second had it off. Shot wide open on an m11p.

 

Edit: Also, just want to add that I'm not an expert in lens design and not saying anything is "defective." This could totally be normal for this lens. I was just curious as to what one should expect for an APO lens. 

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!

Edited January 3 by JPizzzle

[adan]

If you blow highlights, those can fringe with almost any lens. The "excess photons" have a mind of their own, as they pass through the sensor cover glass and UV/IR filter. It is a "worst-case" scenario.

And of course 60Mpixels (M11P) will reveal 1-2-pixel fringes that simply go undetected by sensors with less resolution.

Finally, "APO" is a relative performance benchmark, not an absolute one. Especially the way Leica uses it these days - "our best-of-the-best lenses."

I don't have the APO 35, but that looks pretty darn good on a universal scale (and on M11).

In the M11 forum, you might look up the color fringing noted with the M11 using, for example, the 28mm f/1.4 Summilux ASPH (but not APO).

Beyond that I leave it to the real APO users.

[Advertisement]

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

[rustyrus]

You have to pixel peep like crazy to see any fringing- It hurts that its a boring subject and you can pick where it would fringe but stop worrying about it- Use it

 

 

[JPizzzle]

10 minutes ago, rustyrus said:

You have to pixel peep like crazy to see any fringing- It hurts that its a boring subject and you can pick where it would fringe but stop worrying about it- Use it

 

Yeah, I didn't purposely go looking for fringing. It was just the area in focus on my image in LR. Attached is the original photo. Not a super interesting shot by any means.

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!

Edited January 3 by JPizzzle

[JPizzzle]

23 minutes ago, adan said:

If you blow highlights, those can fringe with almost any lens. The "excess photons" have a mind of their own, as they pass through the sensor cover glass and UV/IR filter. It is a "worst-case" scenario.

And of course 60Mpixels (M11P) will reveal 1-2-pixel fringes that simply go undetected by sensors with less resolution.

Finally, "APO" is a relative performance benchmark, not an absolute one. Especially the way Leica uses it these days - "our best-of-the-best lenses."

I don't have the APO 35, but that looks pretty darn good on a universal scale (and on M11).

In the M11 forum, you might look up the color fringing noted with the M11 using, for example, the 28mm f/1.4 Summilux ASPH (but not APO).

Beyond that I leave it to the real APO users.

Thank you for the explanation. I thought that this was likely "normal" with any lens and was hesitant to even make this post. Just for the money and quality, I wanted to make sure mine is up to the standard it should be at if that makes sense.

[LocalHero1953]

Aside from whether a lens is performing apochromatically or not (all wavelengths refracted by the same amount), in any scenario where a fully blown highlight (all wavelengths blown) adjoins a non-blown highlight you can get this effect. Along the border zone some colour pixels (eg green) will receive just more photons than they can record (blown) while others (eg blue and red) will just be able to record the correct number. This will cause colour distortions along that border. This is not a lens effect, but rather a bayer filter and sensor effect. Raw converters try to rebuild the correct colour of such blown areas using the pixels which aren’t blown, but don’t always get it right. 

[JPizzzle]

Advertisement (gone after registration)

10 minutes ago, LocalHero1953 said:

Aside from whether a lens is performing apochromatically or not (all wavelengths refracted by the same amount), in any scenario where a fully blown highlight (all wavelengths blown) adjoins a non-blown highlight you can get this effect. Along the border zone some colour pixels (eg green) will receive just more photons than they can record (blown) while others (eg blue and red) will just be able to record the correct number. This will cause colour distortions along that border. This is not a lens effect, but rather a bayer filter and sensor effect. Raw converters try to rebuild the correct colour of such blown areas using the pixels which aren’t blown, but don’t always get it right. 

Thanks you for the explanation and re-assurance!

Edited January 3 by JPizzzle

[rustyrus]

The 35 APO is amazing- The handling, the IQ etc- 

Just go enjoy it- The lens will behave better than most in almost all scenarios. Its always a pleasure to have it on my camera. 

[don daniel]

What really bothers me about this picture is not the fringing of the APO lens, but the magenta cast of the M11. But as far as fringing is concerned: I also see it in harsh contrasts against an overexposed background with the Apo-Summicron 50mm.

[Jeff S]

20 hours ago, JPizzzle said:

   I was just curious as to what one should expect for an APO lens. 

 

 

Typically an extra few thousand $.

Jeff

[Lesslemming]

There are several posts asking the same thing about the APO 35mm (all are welcome and valid in my opinion). What you are seeing, at least this appears to be the consensus, is an artefact of the Bayer array. This phenomenon is not caused by and cannot be remedied by the optics. It is distinctly different from the more common chromatic aberration typically seen as cyan/magenta pair (or other color pair for more exotic or cinema optics) which indeed would be the fault of the lens. Here is a relevant quote from imatest:

 

Quote

Purple fringing  is not chromatic aberration, though it’s sometimes mistaken for it. 

It’s a saturation phenomenon in the sensor, also known as “blooming,” caused by overflow of electrons from highly saturated pixel sites to nearby unsaturated sites. It tends to be worst in cameras with tiny pixels (< 2μm). It has everything to do with the sensor and little to do with the lens.

The reason it’s purple is that most sensors are most sensitive to green, so that uncorrected (raw) output has a strong greenish cast (the compliment of purple). Color-correction makes the image neutral by boosting the red and blue = purple components. Hence the overflow from saturated pixels takes on a purple cast. This effect is shown in the two crops below of the area indicated by the red rectangle of the image on the right. The image on the left is a crop of the JPEG, straight out of the camera. The image on the right is a crop of the image converted from raw without color correction. The sky is still saturated, but purple fringing is not visible.

Source: https://www.imatest.com/docs/sfr_chromatic/#:~:text=Purple fringing is not chromatic,tiny pixels (< 2μm).

Note that this is an effect that is accentuated by higher mp sensors, which explains the increasing reports of M11 and this phenomenon. (As a monochrome shooter I am not worried about this)

Where does this leave you? Enjoy your lens. Or even better: sell it to me

 

Edit: may I ask a favour? Could you provide me with a shot of bokeh balls at f/2, f/2.8 and stopped down a little? I would like to see the shape produced by the aperture blades of the APO 35mm 

Edited January 4 by Lesslemming

[adan]

Up to a point, that is true - and worth remembering

But, given the test camera they mention (Fuji S7000 - introduced 2003) - that particular article is about 20 years old. And there have been a lot of changes in sensors since then (the arrival of wide-spread CMOS use, among them).

The Fuji S7000 used a CCD. And CCD sensors, with a charge ± being transported through the rows and columns of other pixels to be read, had far more "blooming" than todays' CMOS sensors (which transport voltage, not charge, and through a separate 'wiring" system, not the pixel-wells themselves.)

[pgk]

On 1/3/2024 at 1:34 AM, LocalHero1953 said:

This is not a lens effect, but rather a bayer filter and sensor effect. 

That said, depending on subject, lighting and other factors, it may be marinally more distinct with high resolving lenses capable of reproducing high contrast edge detail extremely well. Older lenses with a degree of spherical aberration may not reproduce such edges quite as abruptly which will lessen this effect. In my experience the later, aspherical lenses tend to be slightly more prone to this than prior, non-aspheric generations. But this depends on other factors too so its not an absolute.

[Lesslemming]

3 hours ago, pgk said:

That said, depending on subject, lighting and other factors, it may be marinally more distinct with high resolving lenses capable of reproducing high contrast edge detail extremely well. Older lenses with a degree of spherical aberration may not reproduce such edges quite as abruptly which will lessen this effect. In my experience the later, aspherical lenses tend to be slightly more prone to this than prior, non-aspheric generations. But this depends on other factors too so its not an absolute.

I appreciate your comment, but the idea of sharpness/resolving power/ability to sharply render abrupt edges increasing the likelihood of purple fringing always made me wonder why this effect does not get worse when a lens is stopped down (which arguably always maximises the resolving power).

 

In any case, I think what is visible here is not the dreaded chromatic aberration that some lenses exhibit. My proposal for testing the theory would be to take an image of a white distance scale, make sure the out of focus areas are completely not overexposed and then check for purple/cyan fringing in the out of focus areas.

 

Conversely, if the effect is indeed caused by overexposure and oversaturation of photo sites, if the image of the leaf against the sky above would be severely under-exposed to the point of preventing oversaturation of the sensor and then recovered in post, the image should be noisy and ugly but free of those purple artefacts, no?

Edited January 5 by Lesslemming

[TheGodParticle/Hari]

On 1/3/2024 at 6:04 AM, JPizzzle said:

Hi Everyone,

Would appreciate some thoughts on a photo I took with my 35 apo. I received the lens a few weeks back brand new from a Leica store. It hasn't gotten a ton of use really due to the holidays and scheduled. In looking over some photos from a recent walk, I noticed some slight purple fringing. Now, this isn't something that I think would bother someone as the fringing isn't "extreme," but it has me wondering if it should even be present for an APO lens. Figured I'd post on here to get the opinion of those more knowledgeable than myself.

 

Photos are at 100% and shot raw. The first was processed in LrC with remove CA on in lens correction and the second had it off. Shot wide open on an m11p.

 

Edit: Also, just want to add that I'm not an expert in lens design and not saying anything is "defective." This could totally be normal for this lens. I was just curious as to what one should expect for an APO lens. 

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!

I made a few torture tests and did not see any purple fringing, here is the article. Also comparing the APO35 with the 35/1.4FLE1 and 35/2 ASPH v1 

https://www.harisubramanyam.com/reviews/LeicaM-APO35/2

Edited January 5 by TheGodParticle/Hari

[pgk]

2 hours ago, Lesslemming said:

I appreciate your comment, but the idea of sharpness/resolving power/ability to sharply render abrupt edges increasing the likelihood of purple fringing always made me wonder why this effect does not get worse when a lens is stopped down (which arguably always maximises the resolving power).

In any case, I think what is visible here is not the dreaded chromatic aberration that some lenses exhibit. My proposal for testing the theory would be to take an image of a white distance scale, make sure the out of focus areas are completely not overexposed and then check for purple/cyan fringing in the out of focus areas.

Conversely, if the effect is indeed caused by overexposure and oversaturation of photo sites, if the image of the leaf against the sky above would be severely under-exposed to the point of preventing oversaturation of the sensor and then recovered in post, the image should be noisy and ugly but free of those purple artefacts, no?

Modern lenses usually have a fairly fast optimum aperture which is where abrupt edges are at their maximum - just look at Leica's MTF graphs of their current lenses. Stopping down will not usually increase resolving power but will slightly decrease it especially with high MPixel cameras and diffraction effects. Its not chromatic aberration and this high contrast edge fringing is difficult to predict because the edges, lighting and exposure which cause it do not appear to be simply inter-related. Reducing exposure may still result in some fringing due to the extreme contrast across the edges where it occurs with extreme lighting. Reproducing it consistently is tricky in my experience.