MyLeicaWorld Posted June 21, 2020 Share #1 Posted June 21, 2020 Advertisement (gone after registration) I got the uncoded Summicron 35 asph version 1 lens. I want to learn the differences between the coded and uncoded 35 summicron lenses. Is it just the auto recognition or more than that? Thanks.. Link to post Share on other sites More sharing options...
Advertisement Posted June 21, 2020 Posted June 21, 2020 Hi MyLeicaWorld, Take a look here Coded and uncoded M lenses. I'm sure you'll find what you were looking for!
farnz Posted June 21, 2020 Share #2 Posted June 21, 2020 The M digital cameras have offset sensels at the edges because the M register distance between the rear of the lens and the sensor cause rays from the edge of wide angle lenses to strike the sensor at oblique angles. This means that the sensel 'buckets' don't get enough light to fill them and this causes vignetting and colour shifts. The coding tells the camera which lenses is on the camera and the camera's software compensates for the vignetting and colour shifts so they are less visible. The problems above typically start from wider angle lenses than 35 mm so you'd be unlikely to see a difference between pictures from coded and uncoded 35 mm lenses. Coding the lens means that the lens's details are captured in the EXIF file. Pete. 3 Link to post Share on other sites More sharing options...
adan Posted June 21, 2020 Share #3 Posted June 21, 2020 As farnz says, the primary benefit of the lens coding is correcting color and vignetting artifacts that occur with legacy wide-angle rangefinder lenses on digital sensors. It is something that is unique to these often compact lenses that sit only 28mm or less from the image sensor - and were designed in the film era, since film is silver-impregnated gelatin, not a complex silicon "texture" of pixels. It is/was rarely a problem with SLR lenses that mount 40mm or more from the sensor (due to their moving viewing mirrors), or lenses designed in the digital era for mirrorless cameras (e.g. Fuji, Sony, Micro-4/3rds) There is no optical difference between a 35mm 'cron ASPH v.1 made without coding (pre-2006) and one built with factory coding after 2006 - in fact, Leica can add the coding to an older lens simply by installing a new lens-mount bayonet flange. There are, of course, historical optical and imaging differences between a 35mm Summicron from 1960, or 1970, or 1980, or 1997 (the ASPH). But all ASPH v.1s are identical optically and mechanically. However, when adding the coding to an older ASPH v.1, Leica does check the lens and make little adjustments if needed, to collimation and correct focusing and such. Simply to make sure the lens you get back performs at least as well as it did when it was originally made. Leica has worked on this problem from both ends, with coding on the lenses, but also by making improvements in their own M sensors over the years since the M8 arrived. The M10 sensor is more tolerant of uncoded wide-angles, with less vignetting or color artifacts near the edges, especially ones that are only "slightly" wide (35mms and some 28mms). I have used several different uncoded lenses of those focal lengths, without coding, but on the M10, without seeing the same edge effects that were present on an M9, for example. 3 1 Link to post Share on other sites More sharing options...
Mikep996 Posted June 21, 2020 Share #4 Posted June 21, 2020 (edited) Being new to this myself, I did some personal testing using a white wall as a subject and shooting wide open with my various Leica lenses. As has been stated, at 35 and longer focal lengths, I can't see any difference. With my MATE at 28mm (F4) there is a very slight bit of vignetting that could almost fall into standard deviation. With the 28mm F2.8 it's visible and with the 21mm f3.4 it's very obvious. Shooting 'normal' pictures, with the MATE, I can't see any difference. With the 28 f2.8 it depends on what the photo consists of. If it's large expanse of blue sky, I can see some darkening at the edges if I look for it. If it's street-type stuff, I don't really notice anything. The 21, OTOH, needs either a coded lens or to be manually selected in the camera menu. Otherwise the vignetting, etc is very apparent though you could use it for an effect if you wanted to. Whether the EXIF data is important is a personal decision. I didn't care about it but, OTOH, it's easy to code a lens so, IMO, it's worth coding them to have the data and not worry about vignetting/color shift at all and just leave lens selection on 'auto.' Thanks to info presented on this site, I purchased a lens flange on Ebay (China) for $15.50. It arrived in a week. Took about 15 minutes to: dab black paint with a toothpick into the appropriate indents, remove the original flange from the lens, and install the new flange. I'm going to buy another flange and code my remaining non-coded prime lens. Edited June 21, 2020 by Mikep996 Link to post Share on other sites More sharing options...
250swb Posted June 21, 2020 Share #5 Posted June 21, 2020 (edited) 4 hours ago, double_0_se7en said: I got the uncoded Summicron 35 asph version 1 lens. I want to learn the differences between the coded and uncoded 35 summicron lenses. Is it just the auto recognition or more than that? Thanks.. You can check this for yourself by choosing the lens from the Lens Detection menu in your camera, this turns it into the equivalent of a coded lens. Take pictures with the camera set at Auto Lens Detection in which your lens won't be recognised because it hasn't the correct mount, and then take some pictures in Manual mode, then try to spot the difference. Edited June 21, 2020 by 250swb Link to post Share on other sites More sharing options...
fatihayoglu Posted June 21, 2020 Share #6 Posted June 21, 2020 If this is your only uncoded lens, you can choose this lens from the menu and the camera remembers that everytime if you put this lens on and switches back if it detects any 6bit coding. Otherwise Leica Wetzlar can do it for 350 EUR or so 1 Link to post Share on other sites More sharing options...
AndreasG Posted June 21, 2020 Share #7 Posted June 21, 2020 Advertisement (gone after registration) vor 4 Stunden schrieb farnz: The coding tells the camera which lenses is on the camera and the camera's software compensates for the vignetting and colour shifts so they are less visible. Just for my basic understanding, is such compensation written into the DNG file, the JPG file or in both? Link to post Share on other sites More sharing options...
jdlaing Posted June 21, 2020 Share #8 Posted June 21, 2020 6 minutes ago, AndreasG said: Just for my basic understanding, is such compensation written into the DNG file, the JPG file or in both? In some cases both but mainly jpeg due to deeper processing. Link to post Share on other sites More sharing options...
farnz Posted June 21, 2020 Share #9 Posted June 21, 2020 As far as I'm aware it happens before demosaicing, in which case it will be cooked into the 'raw' data that's written to the DNG, the jpeg, file and the thumbnail jpeg that's created for displaying the image on the LCD, Andreas. Pete. 2 Link to post Share on other sites More sharing options...
AndreasG Posted June 21, 2020 Share #10 Posted June 21, 2020 (edited) So it means for the majority of photos which are recorded as DNG the lens coding (except the EXIF information) is useless? PS: Wrote this before I got Pete's comment. Edited June 21, 2020 by AndreasG Link to post Share on other sites More sharing options...
Dennis Posted June 21, 2020 Share #11 Posted June 21, 2020 Please help me to understand this. Sometimes RF system could be complicated. But once we master it, oh yeah ... I have two lenses at the moment: 50 Summicron V and the 2.8/35 Biogon. I'm trying these days different settings to better understand the concept of the lack of the 6-bit coding on some lenses. I created a user profile for each lens. The Cron is set to Auto, cause is 6-bit coded. The Zeiss is set to Manual (the 35 f/2 pre-Asph). And everything works fine when I manually change the profiles. But you're saying that there is no sense (or at least no significant changes) for my 35 Zeiss, because the problem is for 28 FL or wider, right? Is it always better to set to Auto and manual when is needed? Or is it better to save time and not do it? With only two lenses, let's say I create only one user profile called "ZZZ," and I set the lens detection to AUTO. When I put the 50 Cron, is correctly recognized by the camera and my user profile remain ZZZ. Awesome. When I switch to the 35 Zeiss, the Auto should be showing as 2/35, because I set it like this before, correct? But the ZZZ profile name disappears, and it shows the small gray user icon. Nothing changed in the camera settings except the lens, right? It's still ZZZ but modified? But once I put again the 50 Cron, the lens detection shows AUto, but the user profile icon remains gray. Is this the best practice? I would love to have the Exif and all corrections needed. What about I buy the CV 21 3.5 Asph soon... Can I still use ZZZ (only one user profile), or once I get a third lens, this modus operandi doesn't work anymore? Link to post Share on other sites More sharing options...
Mikep996 Posted June 21, 2020 Share #12 Posted June 21, 2020 (edited) Logically, (?) the lens correction must be applied by the processor before the data is converted to any sort of visible format. The way you have it set up will work properly - set in auto the camera will remember that the last uncoded lens was the 35 so when it sees no code it will shift to the 35. When it sees the (coded) 50, it will automatically switch to that. As far as whether it's really necessary on 35/50mm...well, seems that it's not but why not leave it in auto and be sure you are getting the best image possible (and EXIF data? Edited June 21, 2020 by Mikep996 1 Link to post Share on other sites More sharing options...
NigelG Posted June 21, 2020 Share #13 Posted June 21, 2020 Dennis To further “muddy the waters” for you - after questions from users re codings for their m-mount lenses - Zeiss issued recommendations for how best to set lens coding manually for M cameras....as users were finding that sometimes the best “match” between a Zeiss lens and the Leica software corrections was NOT the obvious closest match based on focal length+aperture. From the original table Zeiss produced the best profile for the 35 f2.8 Biogon was determined to be that for either the Leica 28 f2.8 model 11804 or the similar 11809! Both of these were rated by Zeiss as “sehr gut”. Setting the camera manually to any Leica 35 other than the pre-ASPH f1.4 summilux was not recommended (and that setting only got a “gut” rating). Obviously this meant a choice between the best correction of vignetting/colour shift vs “correct” EXIF.... Hopefully as the sensors/software have over time been “tweaked” to reduce the issue this may no longer be the case - but you might want to try some “wrong” lens profiles to see their effects (if you don’t mind wayward EXIF data....) 1 Link to post Share on other sites More sharing options...
NigelG Posted June 21, 2020 Share #14 Posted June 21, 2020 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! Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! ' data-webShareUrl='https://www.l-camera-forum.com/topic/310751-coded-and-uncoded-m-lenses/?do=findComment&comment=3996987'>More sharing options...
farnz Posted June 21, 2020 Share #15 Posted June 21, 2020 I assume that the white boxes qualify by absence as 'keine sehr gut' then? Pete. Link to post Share on other sites More sharing options...
jaapv Posted June 21, 2020 Share #16 Posted June 21, 2020 59 minutes ago, Dennis said: Please help me to understand this. Sometimes RF system could be complicated. But once we master it, oh yeah ... I have two lenses at the moment: 50 Summicron V and the 2.8/35 Biogon. I'm trying these days different settings to better understand the concept of the lack of the 6-bit coding on some lenses. I created a user profile for each lens. The Cron is set to Auto, cause is 6-bit coded. The Zeiss is set to Manual (the 35 f/2 pre-Asph). And everything works fine when I manually change the profiles. But you're saying that there is no sense (or at least no significant changes) for my 35 Zeiss, because the problem is for 28 FL or wider, right? Is it always better to set to Auto and manual when is needed? Or is it better to save time and not do it? With only two lenses, let's say I create only one user profile called "ZZZ," and I set the lens detection to AUTO. When I put the 50 Cron, is correctly recognized by the camera and my user profile remain ZZZ. Awesome. When I switch to the 35 Zeiss, the Auto should be showing as 2/35, because I set it like this before, correct? But the ZZZ profile name disappears, and it shows the small gray user icon. Nothing changed in the camera settings except the lens, right? It's still ZZZ but modified? But once I put again the 50 Cron, the lens detection shows AUto, but the user profile icon remains gray. Is this the best practice? I would love to have the Exif and all corrections needed. What about I buy the CV 21 3.5 Asph soon... Can I still use ZZZ (only one user profile), or once I get a third lens, this modus operandi doesn't work anymore? I wouldn't worry about that Zeiss lens. The only corrections would be for cyan shift towards the corners and "Italian flag" As long as you don't see that on your images (and you probably won't) consider it irrelevant. Which lens choice in the menu works best for a non-Leica lens is a matter of experimenting anyway. If I were you I would forget about the whole business and use flat field correction in post-processing, if needed. 2 Link to post Share on other sites More sharing options...
NigelG Posted June 21, 2020 Share #17 Posted June 21, 2020 (edited) I think so Farnz. I thought there was another table with the “no good” bits marked but this is the one I had to hand. I remember when I used Voigtlander wides on the M9 there were some very “off” codes that worked best.... Edited June 21, 2020 by NigelG 1 Link to post Share on other sites More sharing options...
jaapv Posted June 21, 2020 Share #18 Posted June 21, 2020 39 minutes ago, NigelG said: Dennis To further “muddy the waters” for you - after questions from users re codings for their m-mount lenses - Zeiss issued recommendations for how best to set lens coding manually for M cameras....as users were finding that sometimes the best “match” between a Zeiss lens and the Leica software corrections was NOT the obvious closest match based on focal length+aperture. From the original table Zeiss produced the best profile for the 35 f2.8 Biogon was determined to be that for either the Leica 28 f2.8 model 11804 or the similar 11809! Both of these were rated by Zeiss as “sehr gut”. Setting the camera manually to any Leica 35 other than the pre-ASPH f1.4 summilux was not recommended (and that setting only got a “gut” rating). Obviously this meant a choice between the best correction of vignetting/colour shift vs “correct” EXIF.... Hopefully as the sensors/software have over time been “tweaked” to reduce the issue this may no longer be the case - but you might want to try some “wrong” lens profiles to see their effects (if you don’t mind wayward EXIF data....) As I recall that table was for lenses on the M8 when using IR cut filters. 1 Link to post Share on other sites More sharing options...
NigelG Posted June 21, 2020 Share #19 Posted June 21, 2020 1 minute ago, jaapv said: As I recall that table was for lenses on the M8 when using IR cut filters. I’ll try to find the original explanation from Zeiss... 2 Link to post Share on other sites More sharing options...
adan Posted June 21, 2020 Share #20 Posted June 21, 2020 12 minutes ago, Mikep996 said: Logically, (?) the lens correction must be applied by the processor before the data is converted to any sort of visible format. Correct. The Leica M in-camera lens corrections for color stains and vigetting are literal, numerical (1s and 0s) changes to the color/brightness data received from the sensor, according to a pattern based on the lens being used (if the lens being used is correctly indentified to the camera via coding, or as a menu selection). They are not profiles, or "how-tos" for later processing in computers, or algorithms - they are automatic gross changes to the RGB values for every pixel as it flows through the camera's CPU enroute to becoming a formatted image (.dng or .jpg). I.E. add a little more red (or remove blue/green) in these pixels and add even more red in these pixels, and add none in these pixels, and so on. Generally it will be a color/brightness gradient of some kind. It also varies (approximately) with the aperture used, which is why the digital Ms make an estimate of the aperture in use by comparing a direct reading of the scene brightness (little sensor above and to the left of the "red dot" on the front) compared to what the internal meter sees coming through the lens. Like this (amount of lens's color/brightness vignetting exaggerated, for clarity) 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! That is baked into the individual pixels' 1s and 0s before it is saved, in any format. It cannot be removed or applied after the fact. 1 2 Link to post Share on other sites Simply register for free here – We are always happy to welcome new members! That is baked into the individual pixels' 1s and 0s before it is saved, in any format. It cannot be removed or applied after the fact. ' data-webShareUrl='https://www.l-camera-forum.com/topic/310751-coded-and-uncoded-m-lenses/?do=findComment&comment=3997017'>More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now