Jump to content

The Leica Forum uses cookies. Read the privacy statement for more info. To remove this message, please click the button to the right:    OK, understood.

Photo

lens correction for S lenses on the SL


  • Please log in to reply
22 replies to this topic

#1 stephan_w

stephan_w

    Erfahrener Benutzer

  • Members
  • 1,427 posts
  • City / Ort:Nord-West-Schweiz,Südbaden

Posted 14 November 2017 - 19:11

Advertisement (Gone after free registration)

just a question: As the lenses of the S don't cover the frame equally on the SL as they do on the S, wouldn't it be necessary to have distinctive lens profiles for the lenses on the SL?



#2 scott kirkpatrick

scott kirkpatrick

    Erfahrener Benutzer

  • Members
  • 4,144 posts
  • LocationJerusalem, Israel
  • City / Ort:Jerusalem

Posted 14 November 2017 - 21:30

I can't imagine that the S lenses need much of a profile, but if they do, Leica can just take the central portion of whatever vignetting correction was derived for the S and use it on the SL.  Color correction would be more work, as the chips are different, and they might be likely to leave that alone.  But I doubt that it is needed.


Edited by scott kirkpatrick, 14 November 2017 - 21:30.

___________________
https://www.flickr.c...9392@N05/albums
www.pbase.com/skirkp www.cs.huji.ac.il/~kirk

#3 Vieri

Vieri

    Leica Ambassador

  • Members
  • 1,409 posts
  • LocationSan Ginesio, Italy
  • City / Ort:San Ginesio, Italy

Posted 15 November 2017 - 08:34

Stephan, S lenses on the SL are already automatically profiled-corrected in camera exactly as SL lenses. Best regards,

 

Vieri


Vieri Bottazzini

Leica Ambassador | Formatt-Hitech Featured Artist | ABIPP EP

VIERI BOTTAZZINI WORKSHOPS | VIERI BOTTAZZINI FINE ART


#4 stephan_w

stephan_w

    Erfahrener Benutzer

  • Members
  • 1,427 posts
  • City / Ort:Nord-West-Schweiz,Südbaden

Posted 15 November 2017 - 09:19

Stephan, S lenses on the SL are already automatically profiled-corrected in camera exactly as SL lenses. Best regards,

 

Vieri

 

hm. this would mean that an additional correction in LR would not be needed, or make things worse?



#5 scott kirkpatrick

scott kirkpatrick

    Erfahrener Benutzer

  • Members
  • 4,144 posts
  • LocationJerusalem, Israel
  • City / Ort:Jerusalem

Posted 15 November 2017 - 10:48

hm. this would mean that an additional correction in LR would not be needed, or make things worse?

Whatever you like, but I wouldn't bother.


___________________
https://www.flickr.c...9392@N05/albums
www.pbase.com/skirkp www.cs.huji.ac.il/~kirk

#6 Vieri

Vieri

    Leica Ambassador

  • Members
  • 1,409 posts
  • LocationSan Ginesio, Italy
  • City / Ort:San Ginesio, Italy

Posted 15 November 2017 - 11:07

Basically, Camera Raw tells you that the lens' profile has been applied already (I don't use Lightroom, but since the engine is the same I assume it will behave in the same way). In my experience, such profile corrects CA / vignette perfectly, and distortion almost perfectly for the 24mm and 35mm I used. The 70mm I also used was corrected perfectly for everything. The 30-90mm was corrected perfectly for CA and vignette, and about 90-95% for distortion (according to the different focal lengths). If you have a clear horizon, and feel that you need to dial in some more correction, just go ahead - your eyes will tell you where to stop. Other than that, I wouldn't bother messing with what the camera gives you.

 

Hope this helps, best regards

 

Vieri


Vieri Bottazzini

Leica Ambassador | Formatt-Hitech Featured Artist | ABIPP EP

VIERI BOTTAZZINI WORKSHOPS | VIERI BOTTAZZINI FINE ART


#7 Jared

Jared

    Benutzer

  • Members
  • 758 posts

Posted 17 November 2017 - 03:21

Glad that they have already built in the profiles for the S lenses, but I'd be surprised if they actually have much content.  Since the lenses are meant to cover a larger imaging circle than the SL possesses, I would imagine distortion, lateral chromatic aberration, and in particular vignetting would all be trivial when these lenses are used on a full frame camera.



#8 Vieri

Vieri

    Leica Ambassador

  • Members
  • 1,409 posts
  • LocationSan Ginesio, Italy
  • City / Ort:San Ginesio, Italy

Posted 17 November 2017 - 08:51

Glad that they have already built in the profiles for the S lenses, but I'd be surprised if they actually have much content.  Since the lenses are meant to cover a larger imaging circle than the SL possesses, I would imagine distortion, lateral chromatic aberration, and in particular vignetting would all be trivial when these lenses are used on a full frame camera.

 

It depends on the lens Jared. The 24mm S definitely has distortion visible even on the SL; just look into an S viewfinder and you'll see it with your naked eye. The 30-90mm is a complex lens, and would distort at the wider focal length visibly on the SL as well. The 70mm is fine. The 35mm needs a very slight distortion correction on the SL.

 

Vignetting is much less of an issue, of course, again it is difficult to judge with the naked eye but I think you can safely assume it is non-existent using S lenses on the SL.

 

LCA is difficult to see with the naked eye, but when is there is there, it's not much dependent on where in the frame you look.

 

Best regards.

 

Vieri


Vieri Bottazzini

Leica Ambassador | Formatt-Hitech Featured Artist | ABIPP EP

VIERI BOTTAZZINI WORKSHOPS | VIERI BOTTAZZINI FINE ART


#9 scott kirkpatrick

scott kirkpatrick

    Erfahrener Benutzer

  • Members
  • 4,144 posts
  • LocationJerusalem, Israel
  • City / Ort:Jerusalem

Posted 17 November 2017 - 09:16

It depends on the lens Jared. The 24mm S definitely has distortion visible even on the SL; just look into an S viewfinder and you'll see it with your naked eye. The 30-90mm is a complex lens, and would distort at the wider focal length visibly on the SL as well. The 70mm is fine. The 35mm needs a very slight distortion correction on the SL.

 

Vignetting is much less of an issue, of course, again it is difficult to judge with the naked eye but I think you can safely assume it is non-existent using S lenses on the SL.

 

LCA is difficult to see with the naked eye, but when is there is there, it's not much dependent on where in the frame you look.

 

Best regards.

 

Vieri

One of my gripes as we have gone through multiple firmware upgrade cycles on the SL is that only the worst problems with M and R lens profiles (and now S) seem to rise to the top of the queue and get resolved.  The clearly damaged profiles for five of the R telephotos got fixed long ago, but there are no distortion corrections in the profiles for the 35-70 R zoom, and several others.  It was easy to check this for M and R lenses, since these can be manually set if you use any M or R lens and a dumb adapter.  I could check SL profiles in the DNG (with some Adobe tools) if I have links to DNGs taken using the Leica S lenses at the affected focal lengths on the Leica S to SL adapter so that the lens ID is known to the camera.  There are places where well-posed requests get collected for things to be fixed.

 

scott


___________________
https://www.flickr.c...9392@N05/albums
www.pbase.com/skirkp www.cs.huji.ac.il/~kirk

#10 Vieri

Vieri

    Leica Ambassador

  • Members
  • 1,409 posts
  • LocationSan Ginesio, Italy
  • City / Ort:San Ginesio, Italy

Posted 17 November 2017 - 12:45

One of my gripes as we have gone through multiple firmware upgrade cycles on the SL is that only the worst problems with M and R lens profiles (and now S) seem to rise to the top of the queue and get resolved.  The clearly damaged profiles for five of the R telephotos got fixed long ago, but there are no distortion corrections in the profiles for the 35-70 R zoom, and several others.  It was easy to check this for M and R lenses, since these can be manually set if you use any M or R lens and a dumb adapter.  I could check SL profiles in the DNG (with some Adobe tools) if I have links to DNGs taken using the Leica S lenses at the affected focal lengths on the Leica S to SL adapter so that the lens ID is known to the camera.  There are places where well-posed requests get collected for things to be fixed.

 

scott

 

Scott,

 

I wonder if R zooms do transmit focal length information to the camera. If they don't, it is clearly impossible to add correction profiles for them.

 

Best regards,

 

Vieri


Vieri Bottazzini

Leica Ambassador | Formatt-Hitech Featured Artist | ABIPP EP

VIERI BOTTAZZINI WORKSHOPS | VIERI BOTTAZZINI FINE ART


#11 gvaliquette

gvaliquette

    Erfahrener Benutzer

  • Members
  • 725 posts

Posted 17 November 2017 - 19:14

Scott,

 

I wonder if R zooms do transmit focal length information to the camera. If they don't, it is clearly impossible to add correction profiles for them.

 

Best regards,

 

Vieri

 

They do, if ROM

 

Guy



#12 Vieri

Vieri

    Leica Ambassador

  • Members
  • 1,409 posts
  • LocationSan Ginesio, Italy
  • City / Ort:San Ginesio, Italy

Posted 17 November 2017 - 19:41

They do, if ROM

 

Guy

 

Thank you Guy, I suspected as much but wasn't sure. If Scott's lenses are not ROM, that would explain why he doesn't see corrections. Best regards,

 

Vieri


Vieri Bottazzini

Leica Ambassador | Formatt-Hitech Featured Artist | ABIPP EP

VIERI BOTTAZZINI WORKSHOPS | VIERI BOTTAZZINI FINE ART


#13 scott kirkpatrick

scott kirkpatrick

    Erfahrener Benutzer

  • Members
  • 4,144 posts
  • LocationJerusalem, Israel
  • City / Ort:Jerusalem

Posted 17 November 2017 - 20:29

Thank you Guy, I suspected as much but wasn't sure. If Scott's lenses are not ROM, that would explain why he doesn't see corrections. Best regards,

 

Vieri

Cmon, Vieri, I'm sometimes careless, but I'm not that dumb!  I have and use the 15/2.8, 16-35, 35-70/4.0, and 80 Summilux, 90 APO Summicron, 100 APO Elmarit macro, 180 APO Elmarit, and 280 Telyt 4.0 R lenses, all of them purchased with ROM in order to get relatively recent copies.  They report their lens ID and focal length if they are fixed focal length, and the two zooms that I have report their current focal length (although you tend to see it only in the review of a shot).  I was the person who reported the damaged lens profiles to Leica when the SL was new, and they got fixed.  I have complained to the appropriate people about the lack of distortion corrections for the wide angle ends of R Vario lenses, and the lack of corrections for the 15's, but those didn't get enough priority to be added to SL firmware.  The fact that several of these lenses were not in-house designs may have been a factor.  

 

Anyway, I don't see distortion corrections because Leica has not provided them for certain legacy R lenses that do need them. 


Edited by scott kirkpatrick, 17 November 2017 - 20:30.

___________________
https://www.flickr.c...9392@N05/albums
www.pbase.com/skirkp www.cs.huji.ac.il/~kirk

#14 Vieri

Vieri

    Leica Ambassador

  • Members
  • 1,409 posts
  • LocationSan Ginesio, Italy
  • City / Ort:San Ginesio, Italy

Posted 17 November 2017 - 20:34

Cmon, Vieri, I'm sometimes careless, but I'm not that dumb!  I have and use the 15/2.8, 16-35, 35-70/4.0, and 80 Summilux, 90 APO Summicron, 100 APO Elmarit macro, 180 APO Elmarit, and 280 Telyt 4.0 R lenses, all of them purchased with ROM in order to get relatively recent copies.  They report their lens ID and focal length if they are fixed focal length, and the two zooms that I have report their current focal length (although you tend to see it only in the review of a shot).  I was the person who reported the damaged lens profiles to Leica when the SL was new, and they got fixed.  I have complained to the appropriate people about the lack of distortion corrections for the wide angle ends of R Vario lenses, and the lack of corrections for the 15's, but those didn't get enough priority to be added to SL firmware.  The fact that several of these lenses were not in-house designs may have been a factor.  

 

Anyway, I don't see distortion corrections because Leica has not provided them for certain legacy R lenses that do need them. 

 

Chill Scott. I am sorry if I didn't know your full history, if I didn't know whether your lenses are ROM or not, and tried to help. That will not happen again, don't worry. Best regards,

 

Vieri


  • LAHilltop said thank you to this

Vieri Bottazzini

Leica Ambassador | Formatt-Hitech Featured Artist | ABIPP EP

VIERI BOTTAZZINI WORKSHOPS | VIERI BOTTAZZINI FINE ART


#15 scott kirkpatrick

scott kirkpatrick

    Erfahrener Benutzer

  • Members
  • 4,144 posts
  • LocationJerusalem, Israel
  • City / Ort:Jerusalem

Posted 18 November 2017 - 08:16

Vieri, sorry for the aggrieved tone of my response.  The post in which I had just answered that question appeared in a different thread (https://www.l-camera...able/?p=3396428 ).  So we had been speaking in different rooms.


___________________
https://www.flickr.c...9392@N05/albums
www.pbase.com/skirkp www.cs.huji.ac.il/~kirk

#16 Jared

Jared

    Benutzer

  • Members
  • 758 posts

Posted 19 November 2017 - 08:09

It depends on the lens Jared. The 24mm S definitely has distortion visible even on the SL; just look into an S viewfinder and you'll see it with your naked eye. The 30-90mm is a complex lens, and would distort at the wider focal length visibly on the SL as well. The 70mm is fine. The 35mm needs a very slight distortion correction on the SL.
 
Vignetting is much less of an issue, of course, again it is difficult to judge with the naked eye but I think you can safely assume it is non-existent using S lenses on the SL.
 
LCA is difficult to see with the naked eye, but when is there is there, it's not much dependent on where in the frame you look.
 
Best regards.
 
Vieri


I suppose you are right about the distortion. I had been thinking in my head, “big, expensive lenses must have distortion correction built into the lens itself.” Of course, that is not necessarily the case.

With regard to LCA, though, isn’t that by definition worse the farther you are from the center of the frame? So you should see less in an SL. Or perhaps a couple of the S lenses have detectable on-axis chromatic aberration as well? Just curious. I don’t have an S camera, so no S lenses. Just M, SL, and Q.

#17 scott kirkpatrick

scott kirkpatrick

    Erfahrener Benutzer

  • Members
  • 4,144 posts
  • LocationJerusalem, Israel
  • City / Ort:Jerusalem

Posted 19 November 2017 - 08:50

Leica uses the WarpRectilinear correction provided in the DNG specification (by Adobe, going back to the 1990s) to manage distortion in the raw files from SL and , I presume, S.  For JPEGs and the EVF, they do something similar in-camera, but an optical viewfinder in a DSLR gives the game away.  The way the DNG corrections work is that, after de-mosaicing, when there are three color signals available at every pixel site, the pixel data for each color plane is mapped into the locations that the light actually came from.  If your raw file development tools allow you to see the whole corrected image, barrel distortion will give a corrected image with the corners stretched out.  This image has more pixels than you started with (often as much as 20-40 on each side), so the convention is to crop to the original dimensions.  The mapping uses only even powers of the distance from the center so the effect starts slowly at first, but is indeed seen across the whole image.  Some people, Sean Reid for example, get very upset by this, and recommend turning software corrections off to prevent it, but the information that is affected is all at spatial frequencies that cannot be reliably rendered in an array of fixed dimensions.  And the color corrections that can be accomplished by shifting the images different amounts seem valuable.  Capture One allows me to accept the corrections and use the extra pixels in the corrected image.

 

scott


___________________
https://www.flickr.c...9392@N05/albums
www.pbase.com/skirkp www.cs.huji.ac.il/~kirk

#18 Vieri

Vieri

    Leica Ambassador

  • Members
  • 1,409 posts
  • LocationSan Ginesio, Italy
  • City / Ort:San Ginesio, Italy

Posted 19 November 2017 - 09:43

Leica uses the WarpRectilinear correction provided in the DNG specification (by Adobe, going back to the 1990s) to manage distortion in the raw files from SL and , I presume, S.  For JPEGs and the EVF, they do something similar in-camera, but an optical viewfinder in a DSLR gives the game away.  ...

 

scott

 

Yes indeed, they use correction for the S as well, and using a lens such as the 24mm you can see it very clearly in the viewfinder and disappear in the DNG or Jpg. You can still see some with the naked eye in the 35mm, and pretty much none in the 70mm. Best regards,

 

Vieri


  • scott kirkpatrick said thank you to this

Vieri Bottazzini

Leica Ambassador | Formatt-Hitech Featured Artist | ABIPP EP

VIERI BOTTAZZINI WORKSHOPS | VIERI BOTTAZZINI FINE ART


#19 scott kirkpatrick

scott kirkpatrick

    Erfahrener Benutzer

  • Members
  • 4,144 posts
  • LocationJerusalem, Israel
  • City / Ort:Jerusalem

Posted 19 November 2017 - 10:25

Yes indeed, they use correction for the S as well, and using a lens such as the 24mm you can see it very clearly in the viewfinder and disappear in the DNG or Jpg. You can still see some with the naked eye in the 35mm, and pretty much none in the 70mm. Best regards,

 

Vieri

So this underscores my complaint -- there are calculated corrections for the S lenses which have left this to be dealt with in software.  Although they are identified on the SL, the SL firmware doesn't contain these corrections.  It can't be that hard to translate them.  The WarpRectilinear coefficients are scaled to the size of the pixel array, so that correction has to be made.


Edited by scott kirkpatrick, 19 November 2017 - 10:27.

___________________
https://www.flickr.c...9392@N05/albums
www.pbase.com/skirkp www.cs.huji.ac.il/~kirk

#20 Vieri

Vieri

    Leica Ambassador

  • Members
  • 1,409 posts
  • LocationSan Ginesio, Italy
  • City / Ort:San Ginesio, Italy

Posted 19 November 2017 - 11:41

So this underscores my complaint -- there are calculated corrections for the S lenses which have left this to be dealt with in software.  Although they are identified on the SL, the SL firmware doesn't contain these corrections.  It can't be that hard to translate them.  The WarpRectilinear coefficients are scaled to the size of the pixel array, so that correction has to be made.

 

Scott, S lenses ARE corrected when used with the SL. Best regards,

 

Vieri


Vieri Bottazzini

Leica Ambassador | Formatt-Hitech Featured Artist | ABIPP EP

VIERI BOTTAZZINI WORKSHOPS | VIERI BOTTAZZINI FINE ART



0 user(s) are reading this topic