Jump to content

Super-Elmar 21 distortion and lens coding


ph.
 Share

Recommended Posts

Advertisement (gone after registration)

Am 1.7.2018 um 16:55 schrieb Raid Amin:

Don recently coded my Zeiss Hologon after converting its G mount to M.

Apart from giving exif data, I presume  coding serves  to correct for optical & sensor shortcomings so that lenses can be produced more cheaply.

However, if you use the "verlauffilter there would not be any need for countering vignetting and given its lack of distortion , not much need for correction. Back when I had mine using Kodachrome, the corners were unacceptable, so I got rid of it. I presume such shortcomings cannot be corrected by any electronic trick.

p

Link to post
Share on other sites

You presume wrongly. It has nothing to do with lens price. The only thing these codes do is correct the (cyan) vignetting and Italian flag shift that some lenses not designed with the limitations of digital sensors in mind in the past exhibit , and insert EXIF data.

You also have a completely wrong idea about modern hybrid lens design.

Link to post
Share on other sites

38 minutes ago, jaapv said:

The only thing these codes do is correct the (cyan) vignetting and Italian flag shift that some lenses not designed with the limitations of digital sensors in mind in the past exhibit , and insert EXIF data.

Are you sure? I seem to notice distortion corrections as well.

Link to post
Share on other sites

12 minutes ago, jaapv said:

Possibly - but on lenses that were not designed with these corrections in mind  - as Leica M lenses are. They are optically optimized designs, not hybrid ones.

I meant on M lenses, not "optimized" ones. 

Link to post
Share on other sites

Advertisement (gone after registration)

16 minutes ago, jaapv said:

All lenses are optimized, some only optically, like the Leica M lenses and some in a hybrid design both optically and electronically in symbiosis, like virtually all new mirrorless lens designs.

So i've been using "optimized" lenses for 30+ years really? I'm learning every day. Now you stated that the only effect of 6-bit coding, besides exif data, is to correct for vignetting and Italian flag on M lenses if i read you well. So i asked if you are sure of that because i seem to have noticed that 6-bit coding does distortion correction on M lenses as well. Would you have any info about that? Just curious.

Link to post
Share on other sites

No I have not. I certainly did not notice it, but it may be the case. However, normally distortion correction is applied at a later stage, or even in postprocessing, e.g. lens profiles in Lightroom.

Of course you have been using optimized lenses, except for pinhole lenses ;)

Link to post
Share on other sites

Just now, jaapv said:

No I have not. I certainly did not notice it, but it may be the case. However, normally distortion correction is applied at a later stage, e.g. the lens profiles in Lightroom.

I wonder if distortion correction is applied by the camera firmware not only on jpeg but also dng files. I seem to have noticed this on two 6-bit M lenses so far, namely SE 21/3.4 asph and, to a lesser extent, 7art 35/2, but i did not check on other lenses so i may be wrong.

Link to post
Share on other sites

At any rate it is clear that the colour shift corrections are made at an early stage and EXIF is written into the DNG file. The 21/3.4 is a notoriously low distortion lens, 1.3% at the most, so I doubt whether there is much need for correction.

Link to post
Share on other sites

1 minute ago, jaapv said:

At any rate it is clear that the colour shift corrections are made at an early stage and EXIF is written into the DNG file. The 21/3.4 is a notoriously low distortion lens, 1.3% at the most, so I doubt whether there is much need for correction.

Question is how this distortion measure was taken. Before or after 6-bit correction?

Link to post
Share on other sites

14 minutes ago, jaapv said:

Certainly before. What distortion are you talking about BTW? The pin-cushion/barrel/mustache linear distortion or perspective distortion (falling lines and eggheads in the corners)?

I meant optical distortion (barrel...) but i'm not sure to share your certainty with respect. Would you have any link about the moment distortion measures are taken by Leica, before or after 6-bit coding? Just curious but i seem to have noticed significant firmware distortion corrections out of the SE 21/3.4 asph on M240 and/or digital CL. I will display them here if i can retrieve the files.

Link to post
Share on other sites

Yes, that is a good idea. The distortion by this lens is mustache-type, which is devilishly difficult to correct in postprocessing.

 

The distortion charts are on the Leica website:

That is always without six-bit coding corrections, as Leica sells M lenses for use on film cameras as well.

 

https://us.leica-camera.com/content/download/102625/906177/version/2/file/Super-Elmar-M21-TechnicalData.pdf

Link to post
Share on other sites

I'm quite sure the Leica M coding affects only vignetting and edge colors. Sean Reid has pointed out that distortion correction always reduces resolution, and has not mentioned comparing M lenses with and without corrections, as he does with other types.

However, there may be another layer, as Lightroom can also do distortion correction from the DNG file, so if the lens model is identified in the DNG, Lightroom could be adding correction even though Leica doesn't do it in-camera.

Vignetting and edge color correction won't affect resolution, but may add a bit of noise, as it basically amplifies, like using higher ISO selectively on colors. That's why M lens edge corrections are more effective at lower ISO settings, as there is more "headroom" for adjustment.

 

Link to post
Share on other sites

41 minutes ago, TomB_tx said:

I'm quite sure the Leica M coding affects only vignetting and edge colors. Sean Reid has pointed out that distortion correction always reduces resolution, and has not mentioned comparing M lenses with and without corrections, as he does with other types. [...]

Too bad he has not as such a comparo would have been interesting IMHO. 
I dont use LR and i've no vegetables on hand ;) but here's what i found in my garden. 
SE 21/3.4 asph on Digital CL. Lens profile "Super-Elmar-M 1:3.4/21 ASPH".
Same raw file converted with and w/o lens profile through Iridient Developer.
(6MB files)

Lens profile off:
https://photos.smugmug.com/photos/i-TFJ9xsM/0/ab5ffda3/O/i-TFJ9xsM.jpg

Lens profile on:
https://photos.smugmug.com/photos/i-t9MpWtm/0/0bad8edc/O/i-t9MpWtm.jpg

Link to post
Share on other sites

Important not to confuse what the camera does to the image, and what post-processing software does to the image.

The 6-bit code and Leica cameras do not, by default, correct the distortion of a lens. Only the vignetting and color stains.

If one chooses to turn on automated corrections in software afterwards, any lens corrections can be applied then (by reading the 6-bit EXIF lens ID and the software choosing the appropriate corrections).

But it is not clear to me whether the correcting profile is included in the camera file data and thus built by Leica ("To correct the distortion of this lens, apply these settings") - or whether it comes in the software (e.g. Iridient) instead, from their own lens tests.

I'm not aware of Leica making any M lens that depends on digital computing to correct distortion in the taking of the picture - because Leica still respects the thousands of film users who want to, for example, use a 21 Super-Elmar on their M3/4/5/6/7, and print it via a lab or darkroom, not via LR or Iridient or C1, etc.

 

Link to post
Share on other sites

Hard to believe that Iridient can be the culprit given that the file suffering from distortion is the one with no profile above. Anyway, i got similar results when removing the opcodes from the raw file through DNG Cleaner below. Basic raw conversion through Apple Preview. Any explanation? 

Opcodes off:
https://photos.smugmug.com/photos/i-fVxVdBX/0/e8de07bd/O/i-fVxVdBX.jpg

Opcodes on:
https://photos.smugmug.com/photos/i-HVFgCdg/0/7be59dce/O/i-HVFgCdg.jpg

Link to post
Share on other sites

Same test with a CV 15/4.5 coded as TE 16-18-21 (WATE) on digital CL. Opcodes removed from the raw file through DNG Cleaner. Basic raw conversion through Apple Preview again. Pretty obvious that 6-bit coding affects distortion as i suspected. 

Opcodes off:
https://photos.smugmug.com/photos/i-gwJts7n/0/cf424a17/O/i-gwJts7n.jpg

Opcodes on:
https://photos.smugmug.com/photos/i-J5dWQXF/0/df5cd323/O/i-J5dWQXF.jpg

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.

Loading...
 Share

×
×
  • Create New...