Jump to content

24-90mm Focus Shift (Diglloyd)


agencal

Recommended Posts

Advertisement (gone after registration)

I just returned from a trip and have skimmed through some of this thread concerning the distortion, particularly at the wide end, of the 24-90 SL lens. Much of the conversation concerned the digital corrections applied in Lightroom to deal with distortion, vignetting, and C A.

I found claims of up to 5 degrees of distortion correction necessary at 24mm. That is a lot.

Because of my concern with architectural photography, from the beginning I have looked for distortion and found none. This is why I have rarely used a zoom lens over the years. I know it's there somewhere.

First of all I looked for it through the excellent viewfinder, checking along a vertical edge of a window, or the horizontal edge of a roofline. But nothing was noticeable . This is all before the image comes face to face with Lightroom. By comparison, the distortion of the Zeiss 21 f2.8 Contax Yachica, that we all lusted after a few years ago, was clearly visible in the viewfinder. 

So if this visible distortion exists, it must then be somehow corrected in the camera before it is seen in the viewfinder. But this does not seem to be the case. It is the profile corrections in Lightroom that gets blamed.

After reading this thread I downloaded some recent files into an older Lightroom 4.4 which knew absolutely nothing about the SL and the 24-90 lens. No profiles to be found there. 

The 24mm files with horizontal or vertical architectural details along the top, bottom or sides, showed no visible distortion in the older Lightroom program either. In other words the images looked exactly like they do with the currant Lightroom program.

This all leads me to believe that the lens is of exceptional design in almost every way.

So what am I missing here? 

Edited by douglas ball
  • Like 1
Link to post
Share on other sites

It's a superb lens, but what some fret about is that they are not getting the pixels recorded by the camera. Instead they get a mathematically smeared version of them. Whether you can see any adverse effect of the smearing -- and there must be some -- is what preoccupies some. 

Link to post
Share on other sites

 

After reading this thread I downloaded some recent files into an older Lightroom 4.4 which knew absolutely nothing about the SL and the 24-90 lens. No profiles to be found there. 

The 24mm files with horizontal or vertical architectural details along the top, bottom or sides, showed no visible distortion in the older Lightroom program either. In other words the images looked exactly like they do with the currant Lightroom program.

This all leads me to believe that the lens is of exceptional design in almost every way.

So what am I missing here? 

 

The DNG files contain "op-codes" which dictate the corrections for barrel distortion at the focal length used.  It is quite noticeable at 24mm focal length with this lens.  The DNG spec, which is probably understood by LR even as far back as 4.4, tells any raw file development program what to do to correct, and LR in particular does this without making any comments about it.  Other tools, such as Capture One, do make the correction explicit, and it can be overridden or ignored in cases where there are no lines that need to be straightened. The best way to think about the correction is that it inverts the distortion by giving you an excellent approximation to the larger image that would have been captured by a non-distorting lens in an array with the same pixel spacing, and this larger image is then trimmed to give a true 24 MPx image at the focal length intended.  Yes, there is interpolation involved, but it maps each collected ("actual") pixel of R, G, and B values into the four pixels in the "virtual" image that was really there.  Not more tricky than Bayer de-mosaicing.

 

scott 

Edited by scott kirkpatrick
  • Like 2
Link to post
Share on other sites

If focus shift is visible @ f 11 it must be massive not to disappear in DOF. On such a lens? And without exciting comment from the users on this forum, who can be a pretty critical bunch?

Digiloyd is not too reliable, but Ming Thein must be taken seriously.

 

Not sure why you would make that last comment. They both give user experiences, so I don't find either more credible than the other, other than Digiloyd makes louder noises when he finds fault with a pieces of equipment.

Link to post
Share on other sites

Advertisement (gone after registration)

I understand everything you say, but it still feels as if you're objecting to digital corrections as a matter of principle rather than because you are getting (or seeing) inferior photographs .

 

In practice virtually everyone seems to be saying what wonderful photos they're getting from it. So I'm still not clear about what the problem is: if it's not about the quality of the photos, what else is it about? Is it a matter of principle? Do you feel the lens is worth a lot less than the asking price, regardless of the quality of photos it can produce?

 

I would agree. I find digital corrections have less negative impact on 24mp files than they did on 12mp files, it's really a non-issue and literally everyone is doing it now. The lens on the Sony RX1/r has very, very strong distortion and vignetting which is software corrected, and the lens on that camera is universally praised, so...

 

In the Leica world, the X-Vario also has a lot of software corrections applied and I suspect the Q does as well, though I've been so happy with my Q images that I haven't really investigated.

 

Another example would be my $2000 Nikon 24-70 f/2.8 which has pretty crazy distortions at different focal lengths and often need correction in Lightroom, the only difference being that I have to actually tell Lightroom to do it, instead of Lightroom doing it manually. 

Link to post
Share on other sites

Thank You Scott. A very good explanation. Still, when I look carefully through the viewfinder or the rear display, should I not see the distortion if it is great enough to cause smearing. Or are the corrections simply being made to the jpeg image I am seeing in the viewer?

Yes, the Jpeg is corrected as well.  In fact with cameras that use proprietary raw formats, the clue to distortion is perfect jpegs and distorted raw files during the early period when the software from COne and other third non-Adobe parties are struggling to catch up.  You see this at the moment with Fuji, for example.  And it was the case with Olympus a year ago.

 

scott

Link to post
Share on other sites

And Ming Thein makes better pictures, plus holding to a high technical standard.

 

scott

 

Matter of opinion, I think. Not to disparage Ming's photography, but I feel he's far more concerned with the technical quality of his photos, than the artistic quality. Most of his work leaves me cold. But again, matter of opinion and everyone's tastes differ. 

Link to post
Share on other sites

Here is a comparison with the Noctilux. http://scenictraverse.com/blog/2016/3/8/50v50 Not sure whether the observed differences are due to focus shift, field curvature or operator error. 


  • The best possible bokeh comes from the Noctilux at f/0.95 (if this surprises anyone, we have a problem).
  • At f/4, the better bokeh actually belongs to the Vario-Elmarit. Shooting the Noctilux at f/2 actually renders bokeh very similar to the Vario-Elmarit at f/4
  • Even at f/8, the Vario-Elmarit maintains some soft bokeh compared to the Noctilux
  • Sharpness and color rendition is very similar between the two lenses. Any differences could easily be the result of Lightroom adjustments (and could be resolved in Lightroom).
Link to post
Share on other sites

 

Here is a comparison with the Noctilux. http://scenictraverse.com/blog/2016/3/8/50v50 Not sure whether the observed differences are due to focus shift, field curvature or operator error. 

  • The best possible bokeh comes from the Noctilux at f/0.95 (if this surprises anyone, we have a problem).
  • At f/4, the better bokeh actually belongs to the Vario-Elmarit. Shooting the Noctilux at f/2 actually renders bokeh very similar to the Vario-Elmarit at f/4
  • Even at f/8, the Vario-Elmarit maintains some soft bokeh compared to the Noctilux
  • Sharpness and color rendition is very similar between the two lenses. Any differences could easily be the result of Lightroom adjustments (and could be resolved in Lightroom).

 

I find this article a bit suspicious. The f8 picture was user error and the writer did not even realize it. in his picture, 24-90@f8 showed nothing in focus where as with Noctilux, pretty much everything is still in reasonable focus. If both picture was focus to exactly the same spot from exactly the same distance, I doubt that dof will be all that different. Also calling Noctilux, the best lens in the world is very questionable. True it is the most expensive 50mm in the world (as far as I know, MSRP wise) but I would say that it is a specialty lens that is well suited to certain condition only. Being most expensive is not synonymous with the best.  Also trying to evaluate bokeh where you only have featureless white wall for background and trying to determine bokeh on out of focus area of the subject alone is a rather strange in my opinion.

Just my 2 cents :)

  • Like 2
Link to post
Share on other sites

  • 3 months later...

And Ming Thein makes better pictures, plus holding to a high technical standard.

 

scott

 

 

I have been a subscriber to Lloyd's site on and off for years.  I like Ming's site and his reviews.

 

I don't believe Lloyd has ever liked a Leica camera that I am aware of.  

Link to post
Share on other sites

I have been a subscriber to Lloyd's site on and off for years.  I like Ming's site and his reviews.

 

I don't believe Lloyd has ever liked a Leica camera that I am aware of.  

 

He liked the M , but he doesn't want to pay only for the name Leica

Link to post
Share on other sites

As I recall, Godfrey has tested and tested for this focus shift, and has come to the conclusion that it was an AF issue resolved by firmware 2.0.  I see no reason to doubt his findings;; conversely, much as I like Ming, I don't think he has re-run his tests post firmware 2.0 and more critically, he just doesn't like this camera one bit; DigiLloyd is a friend of Ming's and as I recall, they did their testing of the camera at about the same time?

 

On this issue, I haven't noticed a problem and I believe that both Ming and Lloyd have it wrong.

  • Like 1
Link to post
Share on other sites

Yes, I tested the SL24-90 for focus shift extensively, at all marked focal length settings, and found none. 

 

One issue with this (now ancient and mostly forgotten) report from Digilloyd is that the alleged problem was misnamed. Focus shift is an inherent optical property of a lens—my tests proved (at least to me) that no such issue existed. What he's actually reported was a problem with the AF system: the plane of critical focus chosen by the SL with the SL24-90 lens with certain zoom settings tended be closer to the near-side edge of the DoF zone than with other cameras when the subject was at a large distance. 

 

This might have been true, although I personally never noticed it. I don't shoot at/near infinity focus very often, and the critical focus plane has always been fine for my photos. I did some testing once upon a day and found it always well within the DoF zone with fw 1.2. With firmware 2.0, Leica tuned the AF system's focusing—I've not seen much difference myself, and it seems only once in a while when someone comes across DL's site that is there anything posted on the subject. 

 

With the thousands of cameras shipped since January and the number of people not seeing any problems with the AF system, I wouldn't worry about it. 

  • Like 1
Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...