Home

Please support mflenses.com if you need any graphic related work order it from us, click on above banner to order!

SearchSearch MemberlistMemberlist RegisterRegister ProfileProfile Log in to check your private messagesLog in to check your private messages Log inLog in

Wide-angle lenses and chromatic aberration
View previous topic :: View next topic  


PostPosted: Sat Nov 26, 2011 7:36 pm    Post subject: Wide-angle lenses and chromatic aberration Reply with quote

It has been observed that MF wide-angle lenses are prone to various sorts of color fringing (chromatic aberration) that are visible in digital photographs, but not so much in film photography because of different ranges of tolerance. I would like to know your personal experiences, opinions and expertise perhaps with respect to the following questions:

(1) Is there any appreciable difference in CA when you attach the same lens to a FF rather than a crop DSLR?

(2) Are there any differences in registering CA between the different makes of DSLRs, for example Canon, Nikon, Sony, Pentax, etc.?

(3) Is there any clever trick to reduce CA, for example by using an UV filter, stopping down the lens or other?

(4) Do you know of any wide-angle lenses between 28 and 19mm that show little CA, preferably on the cheaper side (M42) that you would recommend?

Thanks for your input.


PostPosted: Sat Nov 26, 2011 7:40 pm    Post subject: Reply with quote

Tokina RMC 3.5/17 on NEX-3 has so little CA I consider it to be not worth considering, same with AT-X version on EOS.


PostPosted: Mon Nov 28, 2011 10:38 am    Post subject: Reply with quote

There are "fringe killer" filters but they affect the color reproduction
Of course by cutting off some viopet/blue and deep red. Only
Available in smaller size.


PostPosted: Mon Nov 28, 2011 11:24 am    Post subject: Re: Wide-angle lenses and chromatic aberration Reply with quote

Fujinonuser wrote:
It has been observed that MF wide-angle lenses are prone to various sorts of color fringing (chromatic aberration) that are visible in digital photographs, but not so much in film photography because of different ranges of tolerance.


No, that effect is not CA, nor a matter of tolerances, but the visible effect of sensor pit shadowing in tangential light. Film is inherently two-dimensional, hence it does not matter what direction its light comes from, and symmetric film age lenses sometimes have a rear node few millimetres off the film surface. Sensors have a three-dimensional lattice structure, and prefer their light to be as perpendicular as possible. Accordingly, CA free high end rangefinder wides from Leitz and Zeiss fringe like mad onm a digital camera, while any cheap retrofocal wide from a long register SLR system won't.

On the other hand, as it is a sensor artefact, surprisingly large amounts can be removed computationally without impact on the image, and any raw converter today will do that - so it isn't really a issue that should have too much of an impact on lens choice, unless you want to use the lens on critical subjects that use up all the colour space (like sunsets).

As to 1: The same focal length lens is narrower angle on a crop sensor, hence the incident angle on the sensor is not as oblique as on a FF, and the effect is less pronounced (or even invisible).

2: Longer registers have more telecentric lens designs - it does not make that much of a difference within different SLR systems where the registers are within 25% or so, but non-SLR lenses with their much shorter register are far more affected.

3. Stopping down may make matters better or worse, depending on whether lattice shadows or cross-filter pollution is the more severe problem. Usually it won't help much as it will increase the one as much as it reduces the other...

4. The longer register the better - Nikon mount lenses are a good choice when it comes to that. And beware, if you buy too cheap you may end up with lenses with genuine (also visible on film) CA...


PostPosted: Mon Nov 28, 2011 11:56 am    Post subject: Re: Wide-angle lenses and chromatic aberration Reply with quote

Fujinonuser wrote:

(1) Is there any appreciable difference in CA when you attach the same lens to a FF rather than a crop DSLR?


Yes. All lenses are weaker at the edges compared to their centre. All optical aberrations, including chromatic, will get worse near the edges.

Quote:
(2) Are there any differences in registering CA between the different makes of DSLRs, for example Canon, Nikon, Sony, Pentax, etc.?


The difference is not determined by the brand, rather by the density of the sensor cells and their angle. The denser the cells, and the straighter their angle at the edges, the higher the probability of purple fringing becoming more visible. Note that the digital sensor do not create aberrations per se, but can -and will- magnify the inherent aberrations of a lens. I have both the 5D and the 5D Mark II. It is my direct experience that purple fringing that looks bad on the 5DMarkII looks better on the 5D, and quite often it simply disappears on film.

Quote:
(3) Is there any clever trick to reduce CA, for example by using an UV filter, stopping down the lens or other?


Stopping down the lens usually reduces the purple fringing to a minimum until it becomes virtually invisible. This because by stopping down you are excluding the edges of the glass from contributing to the image. As already noted in point 1, the edges of a glass are always it's weaker point. This is the reason why most lenses today use aspherical glass in their most crucial elements, especially for APO (apochromatic) lenses.
UV has nothing to do with purple fringing.

Quote:
(4) Do you know of any wide-angle lenses between 28 and 19mm that show little CA, preferably on the cheaper side (M42) that you would recommend?


Like I always say, cheap and good don't go well together, and this is especially true at the extremes of the focal lenghts range, i.e. for super wides and long teles.
Having that said, I am happy with my Contax Distagons. But they can not be classified as cheap, unfortunately.
It is possible that Olympus wides do perform well and with little aberrations, but you should better ask Olympus experts.
One lens that is reasonably affordable and performs very well is Yashica's ML 24mm f/2.8 lens for C/Y mount.


PostPosted: Thu Dec 01, 2011 9:53 pm    Post subject: Re: Wide-angle lenses and chromatic aberration Reply with quote

Thank you very much for your helpful clarification.

Quote:
No, that effect is not CA, nor a matter of tolerances, but the visible effect of sensor pit shadowing in tangential light.


This would imply that not only lens characteristics (such as the lens design), but also the sensor resolution and differences in lattice structure among the various camera models and manufacturers will contribute to the effect, as Orio suggests in his reply. Could we say that a Nikon or Pentax DSLR will produce less fringing than say a Canon DSLR?

Quote:
On the other hand, as it is a sensor artefact, surprisingly large amounts can be removed computationally without impact on the image, and any raw converter today will do that


I must admit that I haven't delved into raw converters so far. Are there any tutorials for the Canon system how to remove color fringing that you might recommend? Or handy one-button-solutions?


PostPosted: Thu Dec 01, 2011 10:15 pm    Post subject: Re: Wide-angle lenses and chromatic aberration Reply with quote

Quote:
Stopping down the lens usually reduces the purple fringing to a minimum until it becomes virtually invisible.


Of course, sharpness will then also be reduced due to diffraction and vagabonding light within the mirror compartment, especially in a crop camera.

Quote:
Like I always say, cheap and good don't go well together, and this is especially true at the extremes of the focal lenghts range, i.e. for super wides and long teles.


I was quite surprised by the performance of an old M42 Soligor 28mm lens probably manufactured by Tokyo Optical (Tokina), but those are lucky finds.

Thank you for your information.


PostPosted: Fri Dec 02, 2011 6:23 pm    Post subject: Re: Wide-angle lenses and chromatic aberration Reply with quote

Orio wrote:
Fujinonuser wrote:

(1) Is there any appreciable difference in CA when you attach the same lens to a FF rather than a crop DSLR?


Yes. All lenses are weaker at the edges compared to their centre. All optical aberrations, including chromatic, will get worse near the edges.


Orio, you expressed ideas in the quoted post that aren't consistent with what I think I know about what controls aberrations. Please take a look at http://toothwalker.org/optics.html


PostPosted: Sat Dec 03, 2011 12:34 pm    Post subject: Re: Wide-angle lenses and chromatic aberration Reply with quote

Fujinonuser wrote:
It has been observed that MF wide-angle lenses are prone to various sorts of color fringing (chromatic aberration) that are visible in digital photographs, but not so much in film photography because of different ranges of tolerance. I would like to know your personal experiences, opinions and expertise perhaps with respect to the following questions: ....

My own experience and understanding are limited. However if you are interested in numbers (and I am) then lots of information is available. Listed below are 6 summaries of Chromatic Aberration measurements carried out by Photozone. Make of them what you will. The summaries are divided into 3 groups in order to provide sensor-sensor and manufacturer-manufacturer comparisons. I am not able to judge the quality of the Photozone procedures nor their accuracy or precision. Note that Photozone report CA in units of pixels. For my own purposes I have converted these into distances. Other sites such as Lenstip report CA differently. It might be interesting to compare the Photozone results with the Lenstip measurements – and with those of other sites such as dpreview.
Quote:
(1) Is there any appreciable difference in CA when you attach the same lens to a FF rather than a crop DSLR?

Photozone looked at the Samyang 14mm f/2.8 IF ED AS on both the Canon 5D (Full Format, 13.2 Mpix, 8.04 micron pixel spacing) and the Canon 50D (APS-C, 15.2 Mpix, 4.68 micron spacing).
For the 5D, Chromatic Aberration (CA) at the borders increased from 0.15 pixels (1.2 microns) at f/2.8 to 0.26 (2.1) at f/8.
For the 50D, CA decreased from 0.88 (4.1) at f/2.8 to 0.54 (2.5).
So for the 5D CA increases on stopping down and for the 50D the opposite is true. For all apertures the CA is smaller for the larger sensor and remember that measurements are taken at the borders which will be further away from the centre of the image for the large sensor than for the small sensor.

The Nikon D3x (Full format, 24.6 Mpix, 5.9 micron spacing) and the Nikon D200 (APS-C, 10.2 Mpix, 6.05 micron spacing) have very similar pixel densities.
Photozone tested the Nikon 20 mm f/2.8 AFD on each of these cameras. In both cases border CA increased on stopping down from f/2.8 to f/8, from 11.3 to 13.7 microns for the D3x, and from 11.2 to 12.3 microns for the D200.
So here the larger sensor does show more border CA. I think the difference is small.

Quote:
(2) Are there any differences in registering CA between the different makes of DSLRs, for example Canon, Nikon, Sony, Pentax, etc.?

Photozone looked at the Sigma 8-16 mm f/4.5-5.6 DC USM using the Canon 50D (APS-C, ~15 Mpix, 4.73 micron pixel spacing) and the Nikon D7000 (APS-C, ~16 Mpix, 4.68 micron). The pixel densities are similar; the Nikon sensor is slightly larger than the Canon sensor. Border CA is larger for the Nikon than for the Canon (e.g. at 12mm FL, f/8, CA=3.7 microns for the Canon and 4.3 microns for the Nikon).


PostPosted: Sat Dec 03, 2011 6:33 pm    Post subject: Re: Wide-angle lenses and chromatic aberration Reply with quote

Fujinonuser wrote:
It has been observed that MF wide-angle lenses are prone to various sorts of color fringing (chromatic aberration) that are visible in digital photographs, but not so much in film photography because of different ranges of tolerance.


Modern wide angle lenses are prone to lateral chromatic aberrations issues as well. Older MF lenses (wide angle or not) tend to have a weakness for purple fringing - this is almost always the case with Japanese lenses, but I found Zeiss and Russian lenses to do much better in this area for some reason.


PostPosted: Sat Dec 03, 2011 6:48 pm    Post subject: Reply with quote

Of your four questions, I can only address #3. I have had very good luck removing all CA from my images using Paint Shop Pro, both v9.01 and X2 Ultimate. There is a specific menu selection for this and it can be used for removing various CA colors. The good thing about PSP is the above are older versions of this program and they can usually be found on places like eBay for very cheap. I also use PSP X2 and to a lesser extent v9.01 for most of my image editing needs. There are things I can do with PSP that I haven't been able to get done in Photo Shop.


PostPosted: Wed Dec 07, 2011 12:00 am    Post subject: Reply with quote

cooltouch wrote:
I have had very good luck removing all CA from my images using Paint Shop Pro, both v9.01 and X2 Ultimate.

Michael, I know PSP Version 9.00, but I cannot find the menue item you're speaking of. Where should I look for it? I agree with you that Photoshop is not the only option.


PostPosted: Wed Dec 07, 2011 12:08 am    Post subject: Re: Wide-angle lenses and chromatic aberration Reply with quote

John, those are very interesting findings. If my interpretation is correct, then the sensor size and layout does have an effect on the visibility of color fringing, as also the intensity and spectral composition of the available light. I noticed some lenses to be bad performers in sunlight, while they convinced in interior rooms.

And of course, it might be possible for camera manufacturers to reduce or eliminate the effects of color fringing by some additional software algorithm. But then we are again within the reach of RAW converters.


PostPosted: Wed Dec 07, 2011 1:14 am    Post subject: Reply with quote

Fujinonuser wrote:
cooltouch wrote:
I have had very good luck removing all CA from my images using Paint Shop Pro, both v9.01 and X2 Ultimate.

Michael, I know PSP Version 9.00, but I cannot find the menue item you're speaking of. Where should I look for it? I agree with you that Photoshop is not the only option.


Go to Adjust > Photo Fix and click on Chromatic Aberration . . . I also have mine set up so there is a hot button for this function in a tool bar. I have the v9.01 patch installed, which as far as I know is just a bug fix. Should be the same as v9.0.