I just discovered something else recently related to this that is pretty important. When you use adapters they are almost all made a little short to guarantee infinity focus. However, lenses are designed for exact measurements of the flange distance, so when you move this even a little it can cause a lot of problems with some lenses. Also different sensor stacks make a difference, the m240 was the first M to ditch the offset micro-lens design I believe, and the sensor stack has improved even more with new models. Pretty much every M cam will show slightly different lens performance.
Thanks Alik, yes that was my thought re. new Ms improving and others agreed. I focused via LV as yes infinity focus on lenses is not always accurate too.
Great video, Matt! One thing to note is that the sensor stack size, while it does affect corner sharpness a bit, doesn’t affect it nearly as much as the sensor micro lenses. The M sensors have micro lenses on them that correct the very angled light from non-retrofocus lenses.
Hi Matt - I always click on your videos as soon as I see them, they are always worthwhile. You used a term with which I'm not familiar and didn't explain what it means. What is this "focus stack" to which you refer several times? After I understand the term, I'll re-watch the video and get far more out of it. Thank you as always and Cheers!
Hi Chris, thanks and sorry I should have explained. From my understanding..and excuse the simple/ bad explaination.. it's the thickness of the glass on the sensor. If you can visualise the sensor on a camera.. some cameras have thicker glass over them as part of the design so when the rear element of the lens protrudes into the camera if it sits very close to this "stack" of sensor + glass layers then light can't travel as easily to the corners of the sensor giving softer focused corners. SLR types lenses don't mount as close to the sensor so it's only an issue with some RF lenses (often wides) so Leica tried to develop there cameras to make sure they could still get great results from their vintage and existing lenses on digital cameras. M's have thinnest "stack", then SL series, then other brands.. as I understand. Hence nice Leica glass can perform badly on some Sonys. (If that confused you more Google probably tells it better and more accurately!)
Possibly worth mentioning that with anything wider than 35mm you need a external viewfinder or evf on the M, whearas easy accurate TTL focusing with the SL may more than compensate for any deficiencies in the sensor stack or even the lens.
Matt: Thank you for the thought provoking tests. Some observations: 1. In Poland you were judging sharpness from the left side of the image. In the tests at home you’re using the right side of the image. Could the Apo Lanthar or the adapter be tilted or decentered slightly - making the left side of image less sharp (and the right side more sharp)? 2. It seems like you chose the portion of your scene most lacking in detail, in the shade, to test your lenses. When you showed the whole photo - there’s a more distant grayish pink house just to the right of center. It has a complex roof - a roof slanted toward you with a double paned dormer window popped out of it and a telephone pole just in front of it; and it was in full sunlight. Wouldn’t that have had more detail to test lens sharpness? Here's a test just for the 50 Apo on the Leica SL to see if it's decentered. Place the (smallish) test area in the center of the viewfinder and focus. Take a shot. Then move the camera so the test area is in upper left corner(shoot), then the upper right corner(shoot), then lower right corner(shoot), then lower left corner(shoot). Now shoot the 4 midpoints between the center and the corners. Then shoot the test area in the center again (it helps to confirm that the focusing ring hasn’t moved at all). Don't refocus for any of the shots - use the initial focus you established for the first centered shot. Also use the same exposure for every shot. Develop the 10 raws. Then open them in photoshop. Take one of the corner shots and draw a small marquee over it to cover the test area. Save the selection so you can select the same sized area in the other shots. Copy the selected corner and paste into the appropriate corner of the shot with the centered test image. Go through the other corner and mid-zone frames and select those test areas and paste them into their proper positions on the centered test image. You’ll end up with test images in an X surrounding the centered test image. Move them closer to the center image so you can more easily look from one to the other to judge differences in sharpness - but keep them in the X formation so you know what part of the sensor they came from.
Thanks CF! You are clearly smarter than me :) I am partly limited by time and try to give a "basic test" to help the masses. The APO was poor in all areas of the photo in Poland.. I did look for sharp windows in focus to compare.. and then yes normally compare edges for worse case sharpness as most lenses are sharp in the centre. Normally this will be a darker area from the vignetting. If I ever have time to do a more complex test as you share I will make a video to give the results. (My only explanation for Poland thinking now is I used the 7A M-L adapter not Leica M-L adapter as I use both. The 7A adapter is amazing for portraits but wont give absolute infinity.
Another reason to use the SL2 for wide angle lenses, no need for accessory VF , SL2 has among the best EVF, so you see what will be the end result framed right, in focus and you get IBIS too. Thanks Matt my next lens with be the Sigma 65mm F2.
These results are really interesting. I noticed that strong magenta cast with a color-skopar 35mm f2.5 (I started to dislike this lens) At first I thought the lens was de-centered but sharpness was even across the frame, then I thought maybe something misaligned because the previous owner of the Leica bumped it into something. Seeing that we both have the same M (Typ 240) and the same color cast only at the right portion of the frame makes me feel relieved that that’s the way some lenses interact with it… sucks but at least it’s not my copy of the camera/lens.
very interesting comparison! What I had in my head was, how would they compare to your Lumix S5. Being not a Leica product. Would it be as close as the SL and the M240? I would guess so, but I am not sure at all. And very interesting Comments here about vintage Lenses with floating element. I have just to adapters both from novoflex (FD to Lm and M39 to Lm) did not have any problems like that, but I never tested it that exact.
Having owned SL/CL/M cameras, I agree that the SL offers vastly superior focusing for fast aperture lenses. I've got my joystick set to zoom to 100% in MF mode (when not set to BBF). Of course, we never have to worry about rangefinder drift with the SL either.
When I was comparing the SL vs. the SL2 I read that leica redesigned the sensor to work better with wide angle M lense. They made the stack thinner and optimized for the 47mp SL2 sensor. So even though the SL2 was far more expensive I bought it knowing the IQ was better.
Awesome video and very informative. I use the 35mm Summilux-M FLE and the 50mm Summilux-M ASPH on my SL2 and SL(typ601) using the Leica M to L adaptor and have seen some of the similar softness on the edges but it doesn't bother me as I normally shoot from wide open (f/1.4) to no smaller than f/2.8. I've also noticed that the bokeh is different when comparing my 50mm Summilux-M ASPH vs my 50mm Summilux-SL ASPH when shooting on both of my Leica SLs. Thanks again for doing this comparison.
Excellent video but are you sure you are using the latest version of the Leica Elmarit m 28 ASPH? 2 types look almost identical. The Leica product name of the old one is 11606 and the new one is 11677. The newer one has a screw-on metal hood instead of the plastic lens hood. Some people say the new one performs better on digital M cameras. Same story as the Summicron M 28.
Caro Matt, in due o tre video sei riuscito a farmi venire il mal di testa 😄 perchè, alla fine, tutto sembra uguale al contrario di tutto. Cerchiamo di fare un po’ di ordine. Ho fatto delle lunghe prove (centinaia di foto) con obiettivi “M” sulla Leica SL2-S. Per quello che riguarda i miei obiettivi M, acquistati nella seconda metà degli anni ‘90, e in particolare Elmarit 24/2.8 Asph, Summicron 35/2 asph e Elmarit 90/2.8, non si tratta di sola evidente differenza centro-bordi, ma di inferiore nitidezza all’infinito. Il problema è proprio la resa all’infinito, non eccezionale come visto su apparecchi “M”. Se si mette a fuoco a distanze comprese all’incirca tra 3/5 metri e 10/12 metri la nitidezza, invece, c’è tutta, come per magia. Perchè? Non lo so, ma è evidente, come hai detto tu, che si tratta di distanze diverse tra lenti posteriori e sensori, complicate da quella “prolunga” costituita dall’adattatore per ottiche M su corpi SL, con incostanza di rendimento da obiettivo a obiettivo. E’ chiaro che i tecnici Leica hanno voluto rendere compatibili decine e decine di obiettivi vintage con i nuovi corpi SL, ma con un accettabile compromesso. Ancora inspiegabilmente sulla SL2-S funzionano globalmente meglio a mio giudizio i miei obiettivi “R”, anche e soprattutto all’infinito. Sarà merito della diversa distanza tra la lente posteriore e sensore? Anche qui non lo so. Quello di cui sono certo è, invece, che, attualmente, i migliori obiettivi Leica tra tutte le corrispondenti focali fisse 28,35,50, 75 e 90 sono (e questa è la famosa scoperta di Colombo 😂) sono di gran lunga gli “L”. Ma questo lo sapevi già.👋
Hello thanks! Sorry I had to wait to find time to translate before I could read your message. If you are using a non-Leica M-L adapter there is a chance you will not have sharpness at infinity. I see same real problem with 7Artisan M-L adapter. With Leica adapter it is no issue. Corners, yes this is the common issue know when using mirrorless verses RF camera. Yes L lenses should be the highest performing (both designed by Leica to be better than their M lenses and possible by other brands that use software to improve their not perfect image quality)(Distortion on Sigma lenses for example corrected by the software). Real metal and glass older lenses don't use software so they have this disadvantage (if you want to call it this!) :) Sorry for the headache! :) ..and yes I enjoy R lenses too but there are not as sharp as M lenses (all models like for like) from my testing (even compared to Voigtander M glass). I like how they render though + the easy close focus option.
From what I see, the differences are negligible to my eyes... Maybe there will be a greater difference with a more recent M camera. But no big deal, at least for me. For the color cast, I don't know for the SL, but my Sigma fp has a color cast correction option in the camera. Very handy.
Hi Matt. I don't see any mention of which lens codes you used in the M camera. I don't have an SL so don't know if you can also select them in that body. Selecting specific lenses in the M menu allows the camera to auto-correct vignetting and colour cast to a decent degree so you generally don't have to worry about it in post processing. The whole idea of Leica coding their lenses and using slightly curved sensor stacks in the M is precisely to deal with the issues you are looking at here resulting from the proximity of the back of the lens to the sensor which is a consequence of the rangefinder camera and lenses small design. This was not such an issue in the film days because film is transparent to light shining on it from shallower angles, whereas a sensor really prefers to have oblique angles of exposure as close to perpendicular as possible. I'm sure you know that lenses can be selected digitally in the M camera menu (and saved as separate User Profiles for quick access in the Menu) even if you don't want to physically code the lenses manually on the lens mount. e.g. for the ZM21 you could select code 11134 (code for the Leica 21/2.8 pre-ASPH). This is a really important part of your test, so please add the lens codes in the comments. If you missed them, please make another video showing any difference you see with the correct lens codes selected. Cheers, and thanks again. Stu
Hi Stuart, great question. I used no profiles for any lens in this test on either camera so this is the "worst case" scenerio. (Yes agreed when I first go into Leica cameras I selected the closest profile. In the last 4-5yrs I don't use any profile as I shoot mostly portraits).
@@MattOsborne-MrLeicaCom Okay, that makes sense. I use the first version Voigtlander 12mm ltm/M39 with the 11134 code and get less colour fringing than you saw with the ZM21, so I know that code corrects pretty aggressively. I shoot mostly landscape, travel, family and some macro (yup, classic happy amateur) so I do use the profiles all the time and have manually coded my screw mount lenses on the ltm-M adapters to make it easier. Cheers again, and thanks for the inspiring work. Stu
Hi Matt, I would like to see what about Canon R5 compare with SL. I have some voigtlander lenses and use these on EOS R. With 21/3.5, 28/2.0II, 50/1.2 I have color cast. On 35/1.4II i haven't. I test a little on EOS R5 21/3.5 and 50/1.2 and I think was ok.
Hey Pawel, I can't help you directly as I don't own the R5 ..but you might like the next video which is somewhat related. Maybe it is something I can cover in the future if I get access to cameras.
You make a very good point which is often overlooked. M Lenses do not work well on modern Sony FE cameras. Centre sharpness is excellent, performance at the edges falls off really quite a lot for the quality of the lens. Something to bear in mind if you want to adapt beautiful M glass to Sony.
@@sebastiang7183 there’s no doubt the sigma is great but I think the APO is also better than I showed. The Sigma is chipped do pulls the lens profile automatically so a combination of optics and software seem to make for great images from the 65.
I tested the voigtländer 50 apo, summicron-c 40 and summicron 50 on M10, M240 and Sony 7R5 with techart autofokus adapter and an manual adapter. On the A75R all lenses perform really worse. Sharpness and resolution are much better on both 24MP M compared to the 62MP R. the center does not differ as much, in the edges the 3 M Lenses are completely unusable on the sony. Beside the sensor on the M11 is the same as on the Sony A7R 4/5, the sensor glas is much thinner. older Leicas have shifted microlenses (M9, not sure about M240 and M10). If you want a small lens you habe to buy a M.
@@MattOsborne-MrLeicaCom So you pick auto in the menu on the SL. I chose a Leica lens from the M menu most closely to the lens I’m using. Example: when I shoot the Voigtländer 21mm f1.4, I select Leica 21mm f1.4 from the list. Seems to work better.
@@MattOsborne-MrLeicaCom I have programmed that front button next to lens release on the SL to lens profiles. Your fingers are there to release lenses for any lens change, so it’s only natural to make profile change while your fingers are there.
Hi Matt - your results are exactly as I have found. My M lenses definitely are not as sharp on the SL as my M240. The most disappointing lens of them all has been my Leica 135 APO F3.4 lens - superb on the M240 but is pretty mediocre on the SL. My old 135 F4 Elmar is superb on the SL and much better than the 135 APO. Conversely my Leica R lenses are stellar on the SL and seem to perform better than on a Leica R film camera! Regards.
A very timely topic for me since I own both SL and M cameras. I found focusing is typically easier on my SL2 than on my other M cameras. With lenses wider than 28 mm, I always use those on SL2 because of the frame line issue. Nevertheless, your suggestions are much appreciated. One question - focusing aside, is the 50 mm Summilux ASPH better on the M than the SL? Thanks.
Hi Wilfred, I am happy to use wides on the SL as easier to compose but image quality at the edges is better with the Ms. I would use SL for Lux for accurate focusing at 1.4
My experience with the Zeiss ZM Biogons (21, 25, 28 & 35mm) on Leica digi-M cameras is that the newer the model the less edge-of-frame magenta coloring you'll see. On an M8 the color casts can be quite strong even with the smaller sensor. Better on the M9, better still on the M10 and all but gone on the M10-R. (I have no experience with the Typ 2xx cameras.)
Thanks David! I'm surprised you saw issues with the M8 being a crop sensor but as I shot B&W with mine I can't remember either way. I did wonder if the newer M had corrected some of these issues with the older Ms. Thanks for confirming.
@@MattOsborne-MrLeicaCom Matt, I find with the M8 the color casts are a problem with some wider lenses but not all. The original Voigtländer 15mm f/4.5, for example, has no color issues. The old Voigt 28mm f/3.5, though, causes fairly heavy magenta-ing. 🤷🏼♂️
When you have the same lens on the SL vs the M, and you focus on the same object at infinity does the focus scale stand exactly equal or does it differ? What I try to establish is if your adapter has the exact correct length. Lenses with floating elements are extremely sensitive if the mount to sensor distance is a little wrong and that shows especially in the corners. My Minolta MD 24/2.8 had bad corners on my Sony even at f16, it is a floating element construction. On a film body the lens had the focus scale standing at infinity when I focused on infinity but on the Sony the focus scale went beyond infinity, meaning the adapter was to short. When I shimmed it 0,2 mm to the correct length the lens got much better corners. I hope that mess of a text is understandable.
Doesn't help that a lot of adaptors out there are of indifferent quality, and many are prone to warping and shifting due to both cheap materials and poor construction quality.. The good ones are properly CNC machined, and tend to be more consistently reliable, but like you I've had to fiddle with several to get the adapted lenses to focus consistently and correctly, or correct for interference issues with the rear elements (this is especially a problem with M42 adaptors for some reason). Urth is perhaps the only brand I've found to not need corrections.
Thanks James, I ignore the infinity scale on most lenses as many are not accurate / hard stop. (and that accounts for adapters too).. I focus via LV where possible or check the lens works if using RF camera.
@@MattOsborne-MrLeicaCom Live view focusing in itself is correct, but before you start your testing you need to do a long distance (infinity distance) check and make sure that the focus scale on the lens for both cameras ends up at the lens infinity mark. If it only does that on the M body but not on the SL body, lets pretend that when it looks sharp in live view the lens focus scale stands at 10 meters, then the adapter is to short on the SL by a few tenth of millimeters (this is a rather common problem, adapter manufacturers often goes on the minus side of their tolerances since a too long adapter will not hit infinity at all). On lenses with a floating element in a too short adapter will affect corner sharpness since the lens floating element internally will be positioned for 10 meters distance and not for infinity as your scene is at and this will give worse corner sharpness. Not super much, but it will show when pixel peeping, which you are when testing. (For lenses without a floating element in them this doesn't matter though, so it is not all lenses that will be affected by a too short adapter, it depends on the construction of the lens).
J'utilise les deux ( M 240 et SL2S ( 24 MPX) avec les mêmes objectifs surtout le 35 mm Summicron. En terme de résultats vraiment peu de différence avec cette touche Leica irremplaçable. Avantage du SL2S pour le bonus capteur avec une montée en Iso très supérieure et un focus picking beaucoup plus performant que celui du M permettant de figer la vitesse au 1/500 à F 9 et iso flottant. MAIS avantage du M dont je n'ai pas pu me séparer : la légèreté et l'irremplaçable télémétrique. En fait j'adore les deux.
I get slightly different takeaways from these results: (1) In terms of sharpness, a lens generally performs better on a "native" camera (one for which it was designed) than on a camera to which it's adapted. (2) In general, a newer lens (e.g. Ultron II, TTArtisans) loses less performance when adapted to a "non-native" camera than an older one does, possibly because lens designers now give more weight to cross-platform usage. (3) On a given camera, a modern-design native lens will always perform better than an older-design adapted lens, and usually will perform better than a modern-design adapted lens. It's too bad there isn't an M-mount-native version of that 65mm Sigma, so you could make an apples-to-apples comparison with the L-mount version. Unfortunately, I don't think any manufacturer has that. Voigtlander offers several of its lenses in either M-mount or E-mount for Sony, but that would mean the only way to compare would be to use a Sony camera (ewww...)
Thanks JL, yes that's a nice summary and agree third party brands are making lenses to perform on multiple sensors (or should if they are smart) which explains why the new Voigtlanders/ TTA etc work well on the SL and M.
A nice test but there is a major flaw and that is that a lens is 'tuned' for its wides aperture and the sweet spot which is around 5.6 to 8 So a F1.4 lens is designed to make a great shot at F1.4 and around 8. In between it can perform better cause when you stop down towards 8 but then other variables are coming in place. Ideal would be a solid performance over the whole band from 1.4 to 8 (think of a flat MTF curve) So a F2 lens could perform less at F2.8 than a F1.4 lens but will/could be be better at F2 then a F1.4 cause that's where it is designed for. So testing lenses only at 2.8 is okay to get an idea about how they perform at that aperture but making a decision wether a lens is better based on an aperture for which it is not designed is totally wrong ! If want to do a shortcut test always test at wides aperture !
Interesting idea RS, thanks! Some lenses are designed to be optimised at set apertures (like the famous Sonnar 50 1.5 where they started to optimise it at 2.8 so it was crap wide open.. I tried to re-buy 2 times new and both useless as soft wide open. My older original copy was sharp at 1.5).. BUT this is for RF cameras. I focused all images via LiveView to avoid the usually RF focus flaws. If you look at MFT charts for any lens from any company you don't see the performance start high then drop as stop downm the go high by f8. They steadily improve to F5.6-F8 then drop. Expensive lenses start "higher"/better on the chart, often with faster max apertures. (Unless i'm totally missing your point, that's how i've always seen it).
Matt, your review ends up being far less useful than it could have been. Two main issues. First, the M11 is far superior to the M10 which is far superior to the M240 that you tested these lenses on. The sensor glass cover plate on the M11 is 0.85mm, much, much thinner than the plate in the M240, so you should really be doing these tests with an M11. As for the lenses, it makes no sense to test an f/1.4 or f/2 lens at f/2.8, IMHO. If you get a chance, please redo this, using an M11 and also shooting all the lenses at their max aperture. TIA...
Thanks Roy, I appreciate your view. Actually some photographers prefer the M10-R to the M11 when owning both so don’t assume newer is better for everything. This was the M240 vs SL test. At some point I’ll do a M11 test too but that’s a different video. See the individual lens tests as that it’s supposed to be covered here.
@@MattOsborne-MrLeicaCom - In that case, perhaps you should consider changing the title of your video to "Leica M240 vs. Leica SL" or something similar. The current title is misleading: "Lenses DO Perform Better on Leica M Cameras!". From this, I would conclude that a high performance lens like the Leica 50/2 APO Summicron-M or Voigtlander 50/2 APO Lanthar performs better on a 60 MP M11 than it does on a 47 MP SL2. That could very well be true, but it would be nice to get a validation on that. BTW, I appreciate and enjoy all your reviews on UA-cam, so thank you for making these and sharing your insights.
💫 MRLEICA LR PRESETS - mrleica.com/presets/
⭐ PATREON: 100+ Videos & Zoom - mrleica.com/patreon/
🚀 EBOOK: mrleica.com/ebooks/
▶ WATCH NEXT: See All Leica Videos - bit.ly/33H7pPq
📞 ZOOM: Not sure what camera or lens to buy? Arrange a call today - bit.ly/3lBkdgq
✅ CHECK PRICES: (US) ebay.to/2F0HoxY (UK) ebay.to/3ijzle2 (DE) ebay.to/3iqDNYs (FR) ebay.to/2ZJ3E75 (HK) ebay.to/32A8xBu
🛎 SUBSCRIBED?: Monthly Giveaways! bit.ly/3qET0ZO
📷 SEE MY KIT: mrleica.com/kitlist/
💫 MRLEICA LIGHTROOM PRESETS - mrleica.com/presets/
🚀 LEICA WORKSHOPS: mrleica.com/workshops/
▶ VIDEO PLAYLISTS: mrleica.com/youtube/
🎞 NEED FILM?: mrleica.com/do-you-need-film/
☕ COFFEE: Thank Matt with a coffee - www.paypal.com/paypalme/MrLeica
I just discovered something else recently related to this that is pretty important.
When you use adapters they are almost all made a little short to guarantee infinity focus. However, lenses are designed for exact measurements of the flange distance, so when you move this even a little it can cause a lot of problems with some lenses.
Also different sensor stacks make a difference, the m240 was the first M to ditch the offset micro-lens design I believe, and the sensor stack has improved even more with new models. Pretty much every M cam will show slightly different lens performance.
Thanks Alik, yes that was my thought re. new Ms improving and others agreed. I focused via LV as yes infinity focus on lenses is not always accurate too.
Great video, Matt! One thing to note is that the sensor stack size, while it does affect corner sharpness a bit, doesn’t affect it nearly as much as the sensor micro lenses. The M sensors have micro lenses on them that correct the very angled light from non-retrofocus lenses.
Hi Ruel, thank you. Sorry I missed this. Thanks for explaining it better than me!
Hi Matt - I always click on your videos as soon as I see them, they are always worthwhile. You used a term with which I'm not familiar and didn't explain what it means. What is this "focus stack" to which you refer several times? After I understand the term, I'll re-watch the video and get far more out of it. Thank you as always and Cheers!
Hi Chris, thanks and sorry I should have explained. From my understanding..and excuse the simple/ bad explaination.. it's the thickness of the glass on the sensor. If you can visualise the sensor on a camera.. some cameras have thicker glass over them as part of the design so when the rear element of the lens protrudes into the camera if it sits very close to this "stack" of sensor + glass layers then light can't travel as easily to the corners of the sensor giving softer focused corners. SLR types lenses don't mount as close to the sensor so it's only an issue with some RF lenses (often wides) so Leica tried to develop there cameras to make sure they could still get great results from their vintage and existing lenses on digital cameras. M's have thinnest "stack", then SL series, then other brands.. as I understand. Hence nice Leica glass can perform badly on some Sonys. (If that confused you more Google probably tells it better and more accurately!)
Possibly worth mentioning that with anything wider than 35mm you need a external viewfinder or evf on the M, whearas easy accurate TTL focusing with the SL may more than compensate for any deficiencies in the sensor stack or even the lens.
Wider than 28mm* but yes good point. I use LiveView for wider.
Matt: Thank you for the thought provoking tests. Some observations:
1. In Poland you were judging sharpness from the left side of the image. In the tests at home you’re using the right side of the image. Could the Apo Lanthar or the adapter be tilted or decentered slightly - making the left side of image less sharp (and the right side more sharp)?
2. It seems like you chose the portion of your scene most lacking in detail, in the shade, to test your lenses. When you showed the whole photo - there’s a more distant grayish pink house just to the right of center. It has a complex roof - a roof slanted toward you with a double paned dormer window popped out of it and a telephone pole just in front of it; and it was in full sunlight. Wouldn’t that have had more detail to test lens sharpness?
Here's a test just for the 50 Apo on the Leica SL to see if it's decentered.
Place the (smallish) test area in the center of the viewfinder and focus. Take a shot. Then move the camera so the test area is in upper left corner(shoot), then the upper right corner(shoot), then lower right corner(shoot), then lower left corner(shoot). Now shoot the 4 midpoints between the center and the corners. Then shoot the test area in the center again (it helps to confirm that the focusing ring hasn’t moved at all). Don't refocus for any of the shots - use the initial focus you established for the first centered shot. Also use the same exposure for every shot. Develop the 10 raws. Then open them in photoshop. Take one of the corner shots and draw a small marquee over it to cover the test area. Save the selection so you can select the same sized area in the other shots. Copy the selected corner and paste into the appropriate corner of the shot with the centered test image. Go through the other corner and mid-zone frames and select those test areas and paste them into their proper positions on the centered test image. You’ll end up with test images in an X surrounding the centered test image. Move them closer to the center image so you can more easily look from one to the other to judge differences in sharpness - but keep them in the X formation so you know what part of the sensor they came from.
Thanks CF! You are clearly smarter than me :) I am partly limited by time and try to give a "basic test" to help the masses. The APO was poor in all areas of the photo in Poland.. I did look for sharp windows in focus to compare.. and then yes normally compare edges for worse case sharpness as most lenses are sharp in the centre. Normally this will be a darker area from the vignetting. If I ever have time to do a more complex test as you share I will make a video to give the results. (My only explanation for Poland thinking now is I used the 7A M-L adapter not Leica M-L adapter as I use both. The 7A adapter is amazing for portraits but wont give absolute infinity.
I will be able to do a comparison on M11 coming home early next week. I also have the Leica SL2S
Thanks Frank, sorry I missed this. M11 colours should be richer, better I expect.
@@MattOsborne-MrLeicaCom I decided to hold off on the M11. I need stabilization and my s2s have that. I wait to see what comes out form Leica next./
Thank you. Your video's are very informative and quite inspiring.
Thanks! See the MrLeica.com blog for all the content
Another reason to use the SL2 for wide angle lenses, no need for accessory VF , SL2 has among the best EVF, so you see what
will be the end result framed right, in focus and you get IBIS too. Thanks Matt my next lens with be the Sigma 65mm F2.
Thanks, yes I like the SL series for wider lenses for the ease of composition
These results are really interesting.
I noticed that strong magenta cast with a color-skopar 35mm f2.5 (I started to dislike this lens)
At first I thought the lens was de-centered but sharpness was even across the frame, then I thought maybe something misaligned because the previous owner of the Leica bumped it into something.
Seeing that we both have the same M (Typ 240) and the same color cast only at the right portion of the frame makes me feel relieved that that’s the way some lenses interact with it… sucks but at least it’s not my copy of the camera/lens.
Sorry to hear Ricardo. I've not seen it on my 35 Skopar but i've not looked closely (and shoot mostly B&W)(and now use the Ultron more).
very interesting comparison! What I had in my head was, how would they compare to your Lumix S5. Being not a Leica product. Would it be as close as the SL and the M240? I would guess so, but I am not sure at all. And very interesting Comments here about vintage Lenses with floating element. I have just to adapters both from novoflex (FD to Lm and M39 to Lm) did not have any problems like that, but I never tested it that exact.
Hi Nils, thanks! Yes I must try to do SL vs S5 test like this one day.
Having owned SL/CL/M cameras, I agree that the SL offers vastly superior focusing for fast aperture lenses. I've got my joystick set to zoom to 100% in MF mode (when not set to BBF). Of course, we never have to worry about rangefinder drift with the SL either.
Thanks JK, YES! The SL kicks ass :) ..I rarely need optimum edge sharpness at f2.8 with my work so i'm more than happy with the SL.
When I was comparing the SL vs. the SL2 I read that leica redesigned the sensor to work better with wide angle M lense.
They made the stack thinner and optimized for the 47mp SL2 sensor. So even though the SL2 was far more expensive
I bought it knowing the IQ was better.
Great info thanks, that would make sense if they knew the SL was not as "good" as the M series.
I've used the Elmarit M 28 ASPH on the M9, the M240 and the M10-P and have never experienced a magenta shift in DNG files.
I'll be honest I wasn't expecting it either. I think I got mine after the M9 but I shoot B&W so never saw it.
Awesome video and very informative. I use the 35mm Summilux-M FLE and the 50mm Summilux-M ASPH on my SL2 and SL(typ601) using the Leica M to L adaptor and have seen some of the similar softness on the edges but it doesn't bother me as I normally shoot from wide open (f/1.4) to no smaller than f/2.8. I've also noticed that the bokeh is different when comparing my 50mm Summilux-M ASPH vs my 50mm Summilux-SL ASPH when shooting on both of my Leica SLs. Thanks again for doing this comparison.
Thanks Michael! Yes i'm the same, I shoot portraits so edge sharpness rarely matters to me.
Excellent video but are you sure you are using the latest version of the Leica Elmarit m 28 ASPH? 2 types look almost identical. The Leica product name of the old one is 11606 and the new one is 11677. The newer one has a screw-on metal hood instead of the plastic lens hood. Some people say the new one performs better on digital M cameras. Same story as the Summicron M 28.
Hi Alexander, sorry my mistake. I check, yes I have 11606. That makes more sense. Thanks for spotting my oversight.
Caro Matt, in due o tre video sei riuscito a farmi venire il mal di testa 😄 perchè, alla fine, tutto sembra uguale al contrario di tutto. Cerchiamo di fare un po’ di ordine.
Ho fatto delle lunghe prove (centinaia di foto) con obiettivi “M” sulla Leica SL2-S. Per quello che riguarda i miei obiettivi M, acquistati nella seconda metà degli anni ‘90, e in particolare Elmarit 24/2.8 Asph, Summicron 35/2 asph e Elmarit 90/2.8, non si tratta di sola evidente differenza centro-bordi, ma di inferiore nitidezza all’infinito. Il problema è proprio la resa all’infinito, non eccezionale come visto su apparecchi “M”. Se si mette a fuoco a distanze comprese all’incirca tra 3/5 metri e 10/12 metri la nitidezza, invece, c’è tutta, come per magia. Perchè? Non lo so, ma è evidente, come hai detto tu, che si tratta di distanze diverse tra lenti posteriori e sensori, complicate da quella “prolunga” costituita dall’adattatore per ottiche M su corpi SL, con incostanza di rendimento da obiettivo a obiettivo. E’ chiaro che i tecnici Leica hanno voluto rendere compatibili decine e decine di obiettivi vintage con i nuovi corpi SL, ma con un accettabile compromesso. Ancora inspiegabilmente sulla SL2-S funzionano globalmente meglio a mio giudizio i miei obiettivi “R”, anche e soprattutto all’infinito. Sarà merito della diversa distanza tra la lente posteriore e sensore? Anche qui non lo so. Quello di cui sono certo è, invece, che, attualmente, i migliori obiettivi Leica tra tutte le corrispondenti focali fisse 28,35,50, 75 e 90 sono (e questa è la famosa scoperta di Colombo 😂) sono di gran lunga gli “L”. Ma questo lo sapevi già.👋
Hello thanks! Sorry I had to wait to find time to translate before I could read your message. If you are using a non-Leica M-L adapter there is a chance you will not have sharpness at infinity. I see same real problem with 7Artisan M-L adapter. With Leica adapter it is no issue. Corners, yes this is the common issue know when using mirrorless verses RF camera. Yes L lenses should be the highest performing (both designed by Leica to be better than their M lenses and possible by other brands that use software to improve their not perfect image quality)(Distortion on Sigma lenses for example corrected by the software). Real metal and glass older lenses don't use software so they have this disadvantage (if you want to call it this!) :) Sorry for the headache! :) ..and yes I enjoy R lenses too but there are not as sharp as M lenses (all models like for like) from my testing (even compared to Voigtander M glass). I like how they render though + the easy close focus option.
From what I see, the differences are negligible to my eyes... Maybe there will be a greater difference with a more recent M camera. But no big deal, at least for me.
For the color cast, I don't know for the SL, but my Sigma fp has a color cast correction option in the camera. Very handy.
Thanks Yoann, I've not seen a colour correction option on Leicas
Would be interesting to see comparison also with the Leica Q vs these wide angle beauties on M240! Think the Q will smash it?
Hey Thomas! Yes nice thought.. perhaps a 28mm test to be done :)
I'm thinking of trading in my M10-P for a SL3 just for ease of use with the built in EVF. I have the M lens adaptor from my old SL. Thoughts?
Yes some of my students find SL series much easier to use, especially with fast lenses.
Color cast with vignette can be hard to correct but edge sharpness might be fixable with Topaz. Thanks for showing some of the differences.
CornerFix does a good job
Thanks Ryan!
On my M240 I get edge colour cast with the Skopar 21mm but nothing with the 21mm Elmarit M. Both fine on my M2.
Thanks Nick! Yes film will be fine. Yes Skopar is worse than some others for this on digital.
Hi Matt. I don't see any mention of which lens codes you used in the M camera. I don't have an SL so don't know if you can also select them in that body. Selecting specific lenses in the M menu allows the camera to auto-correct vignetting and colour cast to a decent degree so you generally don't have to worry about it in post processing. The whole idea of Leica coding their lenses and using slightly curved sensor stacks in the M is precisely to deal with the issues you are looking at here resulting from the proximity of the back of the lens to the sensor which is a consequence of the rangefinder camera and lenses small design. This was not such an issue in the film days because film is transparent to light shining on it from shallower angles, whereas a sensor really prefers to have oblique angles of exposure as close to perpendicular as possible. I'm sure you know that lenses can be selected digitally in the M camera menu (and saved as separate User Profiles for quick access in the Menu) even if you don't want to physically code the lenses manually on the lens mount. e.g. for the ZM21 you could select code 11134 (code for the Leica 21/2.8 pre-ASPH). This is a really important part of your test, so please add the lens codes in the comments. If you missed them, please make another video showing any difference you see with the correct lens codes selected. Cheers, and thanks again. Stu
Hi Stuart, great question. I used no profiles for any lens in this test on either camera so this is the "worst case" scenerio. (Yes agreed when I first go into Leica cameras I selected the closest profile. In the last 4-5yrs I don't use any profile as I shoot mostly portraits).
@@MattOsborne-MrLeicaCom Okay, that makes sense. I use the first version Voigtlander 12mm ltm/M39 with the 11134 code and get less colour fringing than you saw with the ZM21, so I know that code corrects pretty aggressively. I shoot mostly landscape, travel, family and some macro (yup, classic happy amateur) so I do use the profiles all the time and have manually coded my screw mount lenses on the ltm-M adapters to make it easier. Cheers again, and thanks for the inspiring work. Stu
Also SL vs Panny S5 would be interesting.
Yes I see i've set myself up haven't I haha :)
Hi Matt, I would like to see what about Canon R5 compare with SL. I have some voigtlander lenses and use these on EOS R. With 21/3.5, 28/2.0II, 50/1.2 I have color cast. On 35/1.4II i haven't. I test a little on EOS R5 21/3.5 and 50/1.2 and I think was ok.
Hey Pawel, I can't help you directly as I don't own the R5 ..but you might like the next video which is somewhat related. Maybe it is something I can cover in the future if I get access to cameras.
You make a very good point which is often overlooked.
M Lenses do not work well on modern Sony FE cameras. Centre sharpness is excellent, performance at the edges falls off really quite a lot for the quality of the lens. Something to bear in mind if you want to adapt beautiful M glass to Sony.
Thats especially with wide angle lenses since M wide angle lenses work best on M cameras due to the M cameras micro lenses on their sensors.
Thanks for confirming Andrew (that was my thought but I've never shot Sony)
Thanks Bruno
Well done. Thank you for doing this suite of tests. I am surprised the 50 Apo-Lanthar did as well on the SL as the M. However, the results were clear.
Thanks Sebastian, yes trying to serve the Leica community! :)
@@MattOsborne-MrLeicaCom So does that mean the Sigma does indeed beat the Apo-Lanthar?
@@sebastiang7183 there’s no doubt the sigma is great but I think the APO is also better than I showed. The Sigma is chipped do pulls the lens profile automatically so a combination of optics and software seem to make for great images from the 65.
I tested the voigtländer 50 apo, summicron-c 40 and summicron 50 on M10, M240 and Sony 7R5 with techart autofokus adapter and an manual adapter. On the A75R all lenses perform really worse. Sharpness and resolution are much better on both 24MP M compared to the 62MP R. the center does not differ as much, in the edges the 3 M Lenses are completely unusable on the sony. Beside the sensor on the M11 is the same as on the Sony A7R 4/5, the sensor glas is much thinner. older Leicas have shifted microlenses (M9, not sure about M240 and M10). If you want a small lens you habe to buy a M.
Thanks for feeding back on your Sony experience. Noted if anyone asks me. Thanks
How did you handle the lens profiles on your SL.
I tend not to use profiles on any camera as I use a lot of vintage lenses (and swap lenses a lot)
@@MattOsborne-MrLeicaCom So you pick auto in the menu on the SL. I chose a Leica lens from the M menu most closely to the lens I’m using. Example: when I shoot the Voigtländer 21mm f1.4, I select Leica 21mm f1.4 from the list. Seems to work better.
@@williaminbody205 thanks, yes I used to when I first started with Leica.
@@MattOsborne-MrLeicaCom I have programmed that front button next to lens release on the SL to lens profiles. Your fingers are there to release lenses for any lens change, so it’s only natural to make profile change while your fingers are there.
Hi Matt - your results are exactly as I have found. My M lenses definitely are not as sharp on the SL as my M240. The most disappointing lens of them all has been my Leica 135 APO F3.4 lens - superb on the M240 but is pretty mediocre on the SL. My old 135 F4 Elmar is superb on the SL and much better than the 135 APO. Conversely my Leica R lenses are stellar on the SL and seem to perform better than on a Leica R film camera! Regards.
Hi Francis, sorry I missed this. Interesting about your 135mm, thanks for sharing. Yes R lenses are amazing on the SL!
What about using the L wide angle lens on the SL camera ?
Yes I use L39 and M wide lenses on the SL - 15mm Heilar, 20mm, 21mm etc.
Hi can you do the same test with the Leica M10 and M11, you may see a big different in the sensor, just to let you know.
Thanks, if I ever have both cameras to test at once then yes for sure.
A very timely topic for me since I own both SL and M cameras. I found focusing is typically easier on my SL2 than on my other M cameras. With lenses wider than 28 mm, I always use those on SL2 because of the frame line issue. Nevertheless, your suggestions are much appreciated. One question - focusing aside, is the 50 mm Summilux ASPH better on the M than the SL? Thanks.
Hi Wilfred, I am happy to use wides on the SL as easier to compose but image quality at the edges is better with the Ms. I would use SL for Lux for accurate focusing at 1.4
I wonder if you'd see the same color casts on the newer m10 and m11?
Yes my thoughts too.. in theory I think it's the same camera design but perhaps with software..
Hope one day you get to compare the Leica 28f5.6 Summaron vs that TTartisan copy.
Thanks Carlos, yes! Hopefully one day.
@@MattOsborne-MrLeicaCom there’s one for sale like new here for $2,000. I’m tempted. 😀
My experience with the Zeiss ZM Biogons (21, 25, 28 & 35mm) on Leica digi-M cameras is that the newer the model the less edge-of-frame magenta coloring you'll see. On an M8 the color casts can be quite strong even with the smaller sensor. Better on the M9, better still on the M10 and all but gone on the M10-R. (I have no experience with the Typ 2xx cameras.)
Thanks David! I'm surprised you saw issues with the M8 being a crop sensor but as I shot B&W with mine I can't remember either way. I did wonder if the newer M had corrected some of these issues with the older Ms. Thanks for confirming.
@@MattOsborne-MrLeicaCom Matt, I find with the M8 the color casts are a problem with some wider lenses but not all. The original Voigtländer 15mm f/4.5, for example, has no color issues. The old Voigt 28mm f/3.5, though, causes fairly heavy magenta-ing. 🤷🏼♂️
@@davidkieltyka9 ah thanks. I used the 15mm so that’s probably what I’m thinking of.
When you have the same lens on the SL vs the M, and you focus on the same object at infinity does the focus scale stand exactly equal or does it differ? What I try to establish is if your adapter has the exact correct length. Lenses with floating elements are extremely sensitive if the mount to sensor distance is a little wrong and that shows especially in the corners. My Minolta MD 24/2.8 had bad corners on my Sony even at f16, it is a floating element construction. On a film body the lens had the focus scale standing at infinity when I focused on infinity but on the Sony the focus scale went beyond infinity, meaning the adapter was to short. When I shimmed it 0,2 mm to the correct length the lens got much better corners. I hope that mess of a text is understandable.
Doesn't help that a lot of adaptors out there are of indifferent quality, and many are prone to warping and shifting due to both cheap materials and poor construction quality.. The good ones are properly CNC machined, and tend to be more consistently reliable, but like you I've had to fiddle with several to get the adapted lenses to focus consistently and correctly, or correct for interference issues with the rear elements (this is especially a problem with M42 adaptors for some reason). Urth is perhaps the only brand I've found to not need corrections.
Great question. I focused both cameras by LV (magnified view with focus peaking) as some lenses are not accurate at infinity setting.
Thanks James, I ignore the infinity scale on most lenses as many are not accurate / hard stop. (and that accounts for adapters too).. I focus via LV where possible or check the lens works if using RF camera.
@@MattOsborne-MrLeicaCom Live view focusing in itself is correct, but before you start your testing you need to do a long distance (infinity distance) check and make sure that the focus scale on the lens for both cameras ends up at the lens infinity mark. If it only does that on the M body but not on the SL body, lets pretend that when it looks sharp in live view the lens focus scale stands at 10 meters, then the adapter is to short on the SL by a few tenth of millimeters (this is a rather common problem, adapter manufacturers often goes on the minus side of their tolerances since a too long adapter will not hit infinity at all). On lenses with a floating element in a too short adapter will affect corner sharpness since the lens floating element internally will be positioned for 10 meters distance and not for infinity as your scene is at and this will give worse corner sharpness. Not super much, but it will show when pixel peeping, which you are when testing.
(For lenses without a floating element in them this doesn't matter though, so it is not all lenses that will be affected by a too short adapter, it depends on the construction of the lens).
@@lattesweden thanks, interesting stuff!
Would love the S5 thrown in for fun (and for those that can’t decide before they stop making batteries for them) 🤣
I will have to just buy you one (and send the bill to your wife*) if you don't decide soon! :P ... but yes that could be another video.
J'utilise les deux ( M 240 et SL2S ( 24 MPX) avec les mêmes objectifs surtout le 35 mm Summicron. En terme de résultats vraiment peu de différence avec cette touche Leica irremplaçable.
Avantage du SL2S pour le bonus capteur avec une montée en Iso très supérieure et un focus picking beaucoup plus performant que celui du M permettant de figer la vitesse au 1/500 à F 9 et iso flottant.
MAIS avantage du M dont je n'ai pas pu me séparer : la légèreté et l'irremplaçable télémétrique.
En fait j'adore les deux.
Merci, yes I still keep M + SL etc as each camera has benefits.
I get slightly different takeaways from these results: (1) In terms of sharpness, a lens generally performs better on a "native" camera (one for which it was designed) than on a camera to which it's adapted. (2) In general, a newer lens (e.g. Ultron II, TTArtisans) loses less performance when adapted to a "non-native" camera than an older one does, possibly because lens designers now give more weight to cross-platform usage. (3) On a given camera, a modern-design native lens will always perform better than an older-design adapted lens, and usually will perform better than a modern-design adapted lens.
It's too bad there isn't an M-mount-native version of that 65mm Sigma, so you could make an apples-to-apples comparison with the L-mount version. Unfortunately, I don't think any manufacturer has that. Voigtlander offers several of its lenses in either M-mount or E-mount for Sony, but that would mean the only way to compare would be to use a Sony camera (ewww...)
Thanks JL, yes that's a nice summary and agree third party brands are making lenses to perform on multiple sensors (or should if they are smart) which explains why the new Voigtlanders/ TTA etc work well on the SL and M.
Voigtlander started making ZF Nikon mount version s and Fuji X mount versions of their latest lenses
@@ofeykalakar1 yes I saw that 👍🏻
A nice test but there is a major flaw and that is that a lens is 'tuned' for its wides aperture and the sweet spot which is around 5.6 to 8
So a F1.4 lens is designed to make a great shot at F1.4 and around 8. In between it can perform better cause when you stop down towards 8 but then other variables are coming in place.
Ideal would be a solid performance over the whole band from 1.4 to 8 (think of a flat MTF curve)
So a F2 lens could perform less at F2.8 than a F1.4 lens but will/could be be better at F2 then a F1.4 cause that's where it is designed for.
So testing lenses only at 2.8 is okay to get an idea about how they perform at that aperture but making a decision wether a lens is better based on an aperture for which it is not designed is totally wrong !
If want to do a shortcut test always test at wides aperture !
Interesting idea RS, thanks! Some lenses are designed to be optimised at set apertures (like the famous Sonnar 50 1.5 where they started to optimise it at 2.8 so it was crap wide open.. I tried to re-buy 2 times new and both useless as soft wide open. My older original copy was sharp at 1.5).. BUT this is for RF cameras. I focused all images via LiveView to avoid the usually RF focus flaws. If you look at MFT charts for any lens from any company you don't see the performance start high then drop as stop downm the go high by f8. They steadily improve to F5.6-F8 then drop. Expensive lenses start "higher"/better on the chart, often with faster max apertures. (Unless i'm totally missing your point, that's how i've always seen it).
Not owning a Leica lens for my SL is driving me crazy >.<
But I think I need to be content with what I have for now
Hey Vic, I was going to message you yesterday.. I will today!
Thanks Matt😃
No problem!
Most interesting……😎
Thanks Michael! Agreed
Love it 👍🏼✨
Thanks Maurice!
it's called "filter stack", not "focus stack".
Thanks Lorenzo, yes sorry a slip of the tongue/ me being dumb!
So the 65mm f2 is a junker? Haha JK. I bought it.
Haha not at all! That lens proved itself.. i'm just trying to work out why the APO didn't do as well as it should have. You're safe! :)
@@MattOsborne-MrLeicaCom Now if only Sigma would give us the 50mm f2 we all want to buy already...
Matt, your review ends up being far less useful than it could have been. Two main issues. First, the M11 is far superior to the M10 which is far superior to the M240 that you tested these lenses on. The sensor glass cover plate on the M11 is 0.85mm, much, much thinner than the plate in the M240, so you should really be doing these tests with an M11. As for the lenses, it makes no sense to test an f/1.4 or f/2 lens at f/2.8, IMHO. If you get a chance, please redo this, using an M11 and also shooting all the lenses at their max aperture. TIA...
Thanks Roy, I appreciate your view. Actually some photographers prefer the M10-R to the M11 when owning both so don’t assume newer is better for everything. This was the M240 vs SL test. At some point I’ll do a M11 test too but that’s a different video. See the individual lens tests as that it’s supposed to be covered here.
@@MattOsborne-MrLeicaCom - In that case, perhaps you should consider changing the title of your video to "Leica M240 vs. Leica SL" or something similar. The current title is misleading: "Lenses DO Perform Better on Leica M Cameras!". From this, I would conclude that a high performance lens like the Leica 50/2 APO Summicron-M or Voigtlander 50/2 APO Lanthar performs better on a 60 MP M11 than it does on a 47 MP SL2. That could very well be true, but it would be nice to get a validation on that.
BTW, I appreciate and enjoy all your reviews on UA-cam, so thank you for making these and sharing your insights.