3 months of M240 - am I keeping it?

[01af]

Advertisement (gone after registration)

However, I've found that focusing the 135 with the M is giving me the same kind of hit rate I expected with a well calibrated M9 and a 90.

You do? That's interesting. Maybe I should let Customer Care take another look at my 135 mm M lens ... for me, focusing the 135 mm via EVF often (not always) gives better results. Maybe my lens' focus calibration is slightly off.

 

 

FWIW I think that with the tighter tolerances of the new rangefinder the ability to focus with longer lenses has also improved (and lets' face it, that's hardly surprising).

It definitely has. Still, the question remains, when is EVF focusing better? Maybe it's not from between 90 and 135 mm as I stated in my previous post but, umm, say, between 135 and 180 mm. If so then it'll be hard to prove in practice, as rangefinder-coupled lenses beyond 135 don't exist. But then, the exact answer to this question isn't important anyway—in principle, we agree, don't we? For short lenses, rangefinder focusing is better; for long lenses, SLR/EVF focusing is better.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[jaapv]

I yet have to try the comparison because I have no 240 yet, but I have managed to focus the M9 with a reasonable hit rate (better than 75 % acceptable, at least 50% good) with the Komura 2x extender up to 180/5.6 and even 270/6.8. The last was hard and fiddly but I came up to 50 % acceptable.

[UliWer]

Yes, on this video at 10:46. It's funny that he says although the structure was taken from the M9, it was "curiously" improved. Interesting choice of words.

 

...

 

The sound of the video isn't too good and Mr. Daniel comes from a region in germany where they are used to swallow some syllabes. I think he says "the accuracy was improved" which might sound like "curiously improved".

[tashley]

Hi, Tim. I appreciate the write up, as usual. I've read your thoughts on the RF vs. live view thing, and I'm still struggling to come to grips with what you're saying.

 

I certainly prefer a rangefinder over live view in use, but I can't grasp what you're talking about in terms of accuracy. If you can't focus using live view more, or at least as, accurately than the rangefinder when using the M, it seems that it's the live view implementation that is poor. Of course, I'm talking about in a static situation. Rangefinders are faster and maybe more generally accurate in "regular" use, IMO.

 

Rangefinders still have to find a happy medium when dealing with focus shift and focus distance (not to mention calibration of the body and lenses,) and that varies so much from lens to lens. There is no way to calibrate a rangefinder so that it focuses dead on at every aperture and every focus distance on every lens. I've spent a lot of time tooling around inside my M9, and I was amazed at the variance between my lenses. In fact, I even sent a 35mm lens to DAG, and he asked me what aperture I wanted the lens calibrated to, which surprised me. Steve Choi told me that some models of 35 Cron focus past infinity digital, which was less of an issue with film. There is a ton of variance in these lenses.

 

Take focus shift, throw in lens variance, and you've got a lot of possibility for error with a rangefinder. Now, I'm not talking about major errors, but I am talking about slight errors, which should be avoidable with a good live view setup.

 

I'm sure Leica improved the tolerances of the M's rangefinder, and my bet is that they're also coming from the factory more accurately set up than the M9 in the first place, but, even with a newly setup rangefinder and lens, a little bit of error is the name of the game, and I'm ok with that.

 

Interestingly, it seems Leica is moving a bit away from field curvature with the 50AA. That lens is pretty flat field.

 

Douglas, let me try to explain what I mean but I caution that the best way, if you have the time, is to read my review of the 35FLE (especially the part added on April 13th) and my article on How to Focus a Tricky Lens.

 

In short:

Many (most?) cameras that use contrast detect AF (which is in effect an automated version of what you do when you use the 'shimmer' on the M240 EVF) do their focussing full open and then stop down for the exposure. This is because, at stopped down apertures, the DOF can be too great to focus accurately due to the ambiguous number of depths into the scene at which elements can appear to have good edge contrast.

 

Clearly M240 users don't have Auto AF or Auto Aperture so both these things must be done manually but the same logic follows, and you will experience less ambiguity as to the correct placement of focus for a centrally located subject if you focus wide open using Live View.

 

They key phrase here is 'for a centrally located subject' because, for many shots, the subject is either not located centrally or is larger than the central region.

 

Let us assume this subject is planar, and fills the entire frame. Say, you are on a cherry picker ten meters in the air and ten meters from a very very large wall that fills the frame. Let us assume initially that there is a sign in the middle of the wall and that you want to accurately capture the lettering on it. In this case, the EVF will always give you a good result though you might still get slightly better sharpness if focussed wide open, as long as the lens has no focus shift at all. If it has some focus shift and you are shooting apertures smaller than maximum but less than about F5.6 then you will generally be better focussing at shooting aperture. Of course this depends on subject distance, focal length etc etc. and will vary per lens and shooting situation.

 

But all this refers to the case where you want the sign lettering in focus and the rest of the wall doesn't matter. But what if the sign is off centre and you want to avoid the possibility of cosine error in focus and recompose? Or what if you want the entire wall in focus? Maybe you've been paid to do an architectural shoot of a brave new building for a glossy double spread page in a book. Something similar happened to me recently.

 

In this case, it doesn't just matter where the point of central focus is placed: it matters where the entire zone of good focus is placed. If that zone is curved or wavy, then there is the possibility that you can place focus at, for example, two different places, both of which satisfy the Circle of Confusion requirements for your output on centre but only one of which positions the entire focus zone so as to capture the entire wall sharply.

 

Here is an idealised example showing a lens with a wavy field of focus at, say, F5.6. Both points of central focus will get your sign in great resolution but only one will get the whole wall in focus.

 

 

 

This field shape can be curved, wavy, cone shaped, and it can face forwards or backwards and it might vary its shape with aperture, subject distance and even at different cross sectional heights into the frame. Depending an all shooting parameters there may or may not be a focus setting on the lens that can get the entire wall sharp enough for your needs: but if there is such a setting, then you are less likely to find it by focussing stopped down because of the number of possible 'depths' into the field that can look good in magnified view on centre. Overall, you might find that you are tending to 'home in' on best shimmer, which will tend to place the POF bang in the middle of the zone of good focus. That might or might not suit the field shape of the lens.

 

Live View knows only one thing: is the contrast of edges high enough to activate the shimmer? It has no idea about the lens and its other characteristics, not does it know the aperture or shooting distance. Whereas the properly calibrated rangefinder has more information. It knows (via the shape of the lens cam) for example, the exact shooting distance. It also knows in general where the filed of best focus is placed at different shooting distances though it would ideally like to know more, such as the shooting aperture.

 

But it does seem from my investigations that it knows 'enough'. Time and time again, with a variety of lenses, I have found that careful RF focus will not only get the central subject sharp but will do a better job of placing the entire zone of best focus. Shooting the same planar scenes with stopped down EVF focus is much, much more hit and miss. Shooting again with wide open EVF focus is damned nearly as good as focussing with the RF but is more fiddly because of the need to constantly be changing aperture.

 

As I say this is an idealised argument to show the principles - many steps, caveats and exceptions are left out for brevity. But in general, it explains the situation.

 

Of course, focus shift throws all this out the window. I long ago winnowed from my collection all lenses that had serious focus shift (other than the F1 Noctilux) because a RF simply cannot focus them reliably because it doesn't know the aperture. But it has become very clear to me that the RF on the M240 is capable in a way that the M8 and 9 were not. Not only does it not 'drift' (so far!) over time but it also gives me consistent performance across all my lenses. It is as if the cam shape in each lens is brilliantly designed, but the previous version of the RF could not understand what each individual cam was saying all the time. For my current collection of lenses, the RF in the M240 pretty much can!

 

This stuff afflicts SLR users quite badly. Many of the lenses that D800 users use can appear actively faulty in that they cannot resolve the edges of a simple landscape well even at F5.6 and this has led to a lot of people returning perfectly good copies of, for example, the 28mm F1.8g which is in fact a very nice lens if focussed correctly. This is particularly daft because actually the D800 knows everything: shooting distance, wide open AF best placement, aperture... but it doesn't have the smarts to compare these parameters to an internal lookup table containing information about the field characteristics of the lens and jiggle its focus accordingly.

 

Of course, this can only ever help if the subject is all at a uniform distance. For most subject shapes, it is down to the photographer to know the field shape of his/her lens and to shoot according to how it maps onto the shape of the subject field. I, for example, would always use the 28mm F1.8g Nikkor if I wanted to shot an alley of trees because it has a cone-shaped zone of best focus.

 

HTH!

Tim

[Photoskeptic]

However, I've found that focusing the 135 with the M is giving me the same kind of hit rate I expected with a well calibrated M9 and a 90.

Today I went out for a couple of hours and ONLY shot with the 135 Apo Telyt - I don't think any of the pictures were out of focus - and they were almost all shot at f3.4.

FWIW I think that with the tighter tolerances of the new rangefinder the ability to focus with longer lenses has also improved (and lets' face it, that's hardly surprising).

 

Back in my film days I never had a problem focusing a 135, but then I used an M6 TTL w/.85 viewfinder. I never could figure out why Leica used such a small viewfinder magnification with their digital products.

[3D-Kraft.com]

Tim, thank you for another informative and well written review. I agree on nearly all aspects that you cover (including the long list of drawbacks of the M 240...). I also see the field curvature is a complex issue with many lenses especially in "full frame" format and in extreme designs optimized for wide open apertures.

 

But I can not follow your argumentation that a rangefinder will be more helpful than live view when shooting lenses with field curvature. You write in your article: "But with Live View, you have little chance of finding that distance unless you always focus wide open and then stop down to shoot - because Live View will encourage you to place the central subject in the middle of the field of focus, rather than in the 'best' place overall when the rest of the frame is considered."

 

The main problem with the M here is that you are not only "encouraged" to focus on the central subject with live view - you are FORCED to. But you are forced to live with the same restriction also when using the range finder. As you wrote in an earlier review, this is a waste of potential of the EVF by the M. At least, working with live view allows you to control the depth of field (and its potential curvature) better already in preview.

 

From my experience, the best way to take the field curvature into account as well is to measure/adjust the focus (using a corresponding focus point or panned magnification position) on the object that should take peak sharpness or at least on a point that will be a good compromise for best overall sharpness when the camera is already fixed in it's final position for the desired composition. The M does not allow to work this way - neither with rangefinder nor with live view/EVF.

 

I took a lot of comparison shots with the M and Summilux 50/1.4 ASPH vs. NEX-7 with a HyperPrime CINE 35/T0.95 (which gives you comparable field of view and depth of field at the same 24 MP resolution). As mentioned in my article about these issues (you find the link here and the discussion here), I took about 47 shots of the same scene (where the focus target was composed outside the center on purpose) with the M and Summilux using different focusing approaches (including "focus bracketing" by moving the upper part of my body). Only a few shots gained best possible sharpness with the M (which of course was still disappointing when you compose the object to be focused significantly outside the center with the 50 Summilux as you confirmed in your article as well). On the other hand, all shots taken with the Sony NEX-7 using EVF and magnification moved to the desired point were as sharp as the HyperPrime can deliver in that area although the HyperPrime has a signficantly stronger field curvature than the Summilux.

[3D-Kraft.com]

Advertisement (gone after registration)

Live View knows only one thing: is the contrast of edges high enough to activate the shimmer? It has no idea about the lens and its other characteristics, not does it know the aperture or shooting distance. Whereas the properly calibrated rangefinder has more information. It knows (via the shape of the lens cam) for example, the exact shooting distance. It also knows in general where the filed of best focus is placed at different shooting distances though it would ideally like to know more, such as the shooting aperture.

 

Of course I understand what you want to say but to be more precise: The rangefinder of the camera does not have any more information then just a mechanical feedback from the lens about the focused distance. The additional information may be contained in the way, how the lens is calibrated. You seem to assume that the manufacturer of the lens takes the field curvature into account to provide a better overall compromise. That may be the case for some manufacturers and some lenses and in that case may help you in the situations that you mentioned (shooting architecture at F5.6 or F8). It will not help with lenses that have a strong curvature because there you will propably never find that overall compromise. Here you have to decide by yourself where you want to have the optimal sharpness and live view with movable magnification will help you more than any range finder. Also the "shimmer" will appear in a certain range around the peak sharpness (as it is only detected at live view resolution) and may help indicating the curved DOF in order to choose the appropriate overall focus point.

 

An extreme example for this situation may be when you want to shoot tilt/shift lenses (which may help you to get better architecture shots). There is no chance to shoot these lenses on an M - neither with RF nor with its center-limited EVF.

 

Shooting again with wide open EVF focus is damned nearly as good as focussing with the RF but is more fiddly because of the need to constantly be changing aperture.

 

Shooting wide open typically does not require to change aperture and so is not more fiddly. And also the first part of the sentence should be reversed: Shooting wide open with EVF will provide you best focus results by concept whereas RF can get close (at least in the center) when you have good eyesight and a perfectly calibrated camera+lens combination without any focus shift... You propably know about the reality when shooting lenses wider f/1.4 in a world of adding mechanical tolerances.

[tashley]

 

[/i]The main problem with the M here is that you are not only "encouraged" to focus on the central subject with live view - you are FORCED to. But you are forced to live with the same restriction also when using the range finder. As you wrote in an earlier review, this is a waste of potential of the EVF by the M. At least, working with live view allows you to control the depth of field (and its potential curvature) better already in preview.

 

 

Again for the sake of brevity and because I have covered all this in much more depth at the links I gave, I did not explain to the nth degree but let me try: I am NOT talking about which area of the viewfinder one can focus on: clearly it is only the central area whether in LV or RF. I am talking about how 'deep' into the scene you focus. It is that which determines where the zone of best focus gets placed and it is that which can differ markedly between focussing with the RF and focussing with LV when stopped down, for the reasons I carefully gave. And the RF does indeed 'know' more that the LV system because it at least knows the focus distance and it may, depending on how carefully the cam has been designed and machined, know something about where the field of focus moves to at different focus distances and where best to place focus within it.

 

Whatever the exact reasons, and only Leica can tell us, this method works. I have tried it again and again at a lot of distances with a lot of lenses and I have written up and presented the results in some ten articles at my blog. Not only does it work - it makes sense!

[tashley]

Of course I understand what you want to say but to be more precise: The rangefinder of the camera does not have any more information then just a mechanical feedback from the lens about the focused distance. The additional information may be contained in the way, how the lens is calibrated. You seem to assume that the manufacturer of the lens takes the field curvature into account to provide a better overall compromise. That may be the case for some manufacturers and some lenses and in that case may help you in the situations that you mentioned (shooting architecture at F5.6 or F8). It will not help with lenses that have a strong curvature because there you will propably never find that overall compromise. Here you have to decide by yourself where you want to have the optimal sharpness and live view with movable magnification will help you more than any range finder. Also the "shimmer" will appear in a certain range around the peak sharpness (as it is only detected at live view resolution) and may help indicating the curved DOF in order to choose the appropriate overall focus point.

 

An extreme example for this situation may be when you want to shoot tilt/shift lenses (which may help you to get better architecture shots). There is no chance to shoot these lenses on an M - neither with RF nor with its center-limited EVF.

 

 

 

Shooting wide open typically does not require to change aperture and so is not more fiddly. And also the first part of the sentence should be reversed: Shooting wide open with EVF will provide you best focus results by concept whereas RF can get close (at least in the center) when you have good eyesight and a perfectly calibrated camera+lens combination without any focus shift... You propably know about the reality when shooting lenses wider f/1.4 in a world of adding mechanical tolerances.

 

Honestly, if I want perfect architectural shots I use an IQ180 on an Alpa with Rodenstock lenses on T/S adaptors*. My reference to 'fiddly' is intended to refer to the process of focussing using LV wide open and then stopping down to whatever shooting aperture you require. That is fiddly! As for the rest, we seem to agree.

 

*EDIT but for certain kinds of subjects the M is perfectly good enough: I am after all talking in my examples about this sort of planar architectural shot and I have indeed found that the M240 with the right glass and the right approach to focus can get a good planar result across the field, whereas the exact same equipment focussed with stopped down EVF will much more often give a poor result.

[3D-Kraft.com]

Tim, as our posts intersect here and you propably did not read my second response, let me repeat that I understand your point and position in theory (where it can make sense) but do not share it from my practical experience. It may depend on lenses, cameras, situations, experience etc. so I appreciate that you found your way with your M and your lenses but please respect that I do not share that as a general rule and came to a different result and decision (selling my M).

[Paul J]

Nice article. You have a writing style that makes large blocks of text easy to read.

 

OK...the increased RF accuracy has me interested in this camera now. Hmmm....maybe i'll be eating hats too soon.

 

I agree with the wavy field issues of some lenses. It can be disappointing at times but I've not really used a system that doesn't already have this problem at some point.

 

I have to please some exceptionally discerning editors and art directors though and they have never once mentioned it.

 

Thanks for the article. I also found it interesting in the sense that I rarely use my Phase One P65 and H Blad now.

[tashley]

Tim, as our posts intersect here and you propably did not read my second response, let me repeat that I understand your point and position in theory (where it can make sense) but do not share it from my practical experience. It may depend on lenses, cameras, situations, experience etc. so I appreciate that you found your way with your M and your lenses but please respect that I do not share that as a general rule and came to a different result and decision (selling my M).

 

My friend I had read both both responding to either. Of course I respect your experience and, having only tested a certain range of Leica M glass I cannot speak to any other lenses (especially third-party ones which may have different focus cam design aims) though I found the same results with all of those I tested. But my response was not intended to indicate that I think you are wrong, but that you had mis-read my material: indeed you were in some places voicing a disagreement with things I had not said! Now this might be partly because I am trying to be brief and partly because it is a very complex subject and not all the terms are defined in a way that is universally unambiguous so I do apologise for that! But the main thing I was clarifying is that was referring the the 'depth' at which focus is placed rather than the lateral position across the frame at which it was placed since, unfortunately, that is not an option with M240 LV as you quite rightly say.

[tashley]

Nice article. You have a writing style that makes large blocks of text easy to read.

 

OK...the increased RF accuracy has me interested in this camera now. Hmmm....maybe i'll be eating hats too soon.

 

I agree with the wavy field issues of some lenses. It can be disappointing at times but I've not really used a system that doesn't already have this problem at some point.

 

I have to please some exceptionally discerning editors and art directors though and they have never once mentioned it.

 

Thanks for the article. I also found it interesting in the sense that I rarely use my Phase One P65 and H Blad now.

 

Thank you Paul!

 

The two lenses that I like most for 'flat field' types of work at the wider end are the one on the Sony RX-1 (there is a gentle wave that can make some midfield zones a touch weaker sometimes but the OOF area of the wave rarely seems to intersect the subject plane) and the Rodenstock 40 HR-W which seems to have extremely nice manners! But the Rodie is an F4 design whereas the little Sony/Zeiss lens is phenomenal in that it actually gives across the frame sharpness from F2 onwards. I have been using it or some quite serious work rather than using the D800E and it passes with flying colours. Great lens/sensor combo.

[Paul J]

Thank you Paul!

 

The two lenses that I like most for 'flat field' types of work at the wider end are the one on the Sony RX-1 (there is a gentle wave that can make some midfield zones a touch weaker sometimes but the OOF area of the wave rarely seems to intersect the subject plane) and the Rodenstock 40 HR-W which seems to have extremely nice manners! But the Rodie is an F4 design whereas the little Sony/Zeiss lens is phenomenal in that it actually gives across the frame sharpness from F2 onwards. I have been using it or some quite serious work rather than using the D800E and it passes with flying colours. Great lens/sensor combo.

 

Yes the Rodenstock is legendary. Quite a different set up for what I shoot though! I prefer working small and unencumbered so I can move quickly and freely which is the reason I don't shoot with the medium format so much now days although there is still an overlap as lot of my work is printed large enough to be limited by 24MP. The M9 has really done away with quite a lot of MF use which I'm happy with.

 

I'm yet to try out the RX1 but agree it really seems a marvel of a thing. Great to hear you've been using it in place of the Nikon! Interesting times.

[gpwhite]

Tim, thank you for posting the informative graphics regarding field curvature. Lloyd Chambers posted an impressive review on this topic with actual images shot on a tile mosaic, but I think your diagram is probably more on point for the calibration error introduced by the M RF when "(brick) wall shooting." I think your simulated wave of DOF probably understates the behavior of 35 FLE and 50 SX ASPH, at least at f/2 or so.... which leads me to my question.

 

Like many, I try to shoot at maximum aperture or sometimes 1 stop down. This is a stylistic preference. But the snap and nuance of colors from many Leica M lenses, IMHO, is noticeably greater at maximum than any other aperture-- even 1 stop down. In particular, I find my 18 SEM, 24 EL, 50 AA and 135 APO show brilliance wide open that diminishes just as soon as you crank down the diaphragm... why is this?

 

The upshot is that one must sacrifice the apparent lovely gamut in order to crispy up the focus point a little. This seems more complicated on the M240 because of what appears to me to be improved peripheral detail captured by the sensor (perhaps because the M240 microlenses are different?). In other words, detail in the image periphery seems superior in M240 images shot at large lens apertures compared to M9. For example, with my cherished 18 SEM (a lens you use beautifully), I am now leaning toward f/5.6 even though f/3.8 has more snap... those settings presume composition involving perspective setting and not trying to capture the detail of a brick at the upper right corner of a wall shoot;).

 

Peter

[tashley]

Peter, thanks for the kind words. I am not sure I can answer this but I'll have a go with some theories though we share few lenses in common.

 

Firstly, these lenses are really designed to give great results when wide open. Their maximum apertures are not just a marketing department box-tick, they are for real, even though many of them reach their peak sharpness on centre stopped down a stop, and often need F5.6 or 8 for best performance across the width of the frame. Corners may require more!

 

But when wide open, the 'snap' may please you so much because of the separation of subject from background being at its greatest due to the narrowest possible DOF. This might make the subject look really sharp in a relative sense, even though it would in fact generally be sharper when stopped down a touch. As for the colour. Hmmm. I'm a bit lost on that one. Could it relate to the way that colours look when OOF? In other words, it might be a sort of 'beauty of colour bokeh' thing. Or are you referring to the colour rendition of the in focus areas? Possibly you just prefer your colours rendered a little more gently and the wider apertures, being a touch softer and touch more liable to veiling flare type effects, provide this?

 

Of course many of these lenses reach their best micro contrast in the F2.8 to F4 range but that would tend not to explain your observations. Diffraction is probably starting at or around F5.6, very slightly, and taking the edge of the best micro contrast there before moving on to subtly reduce larger detail at F8.

 

Or it could have something to do with astigmatism and the way it tends to be more pronounced wide open. Depending on which MTF lines it affects most, this might play a part. Maybe it's just a Leica Look thing and, like the recipe for Coke, we'll never know exactly...

 

I would be fascinated to hear what other people think.

[JHellow]

TIm:

 

Just a quick thank you for all the work you have done on the M240 and related optics. My M is back at Leica for the lug issue and I miss it dearly. I do find the files better than my R8/DMR. So far I only have the 35mm FLE and 24mm Elmarit and I am extremely pleased with their characteristics. Images with the 35mm most folks find stunning in their clarity and dimensionality. I look forward to your continued work to help us all decide what is worth adding to our lens stables. I have a full boat of R glass I cannot wait to use once the R adapter is available. I purchased an Adapter on Amazon so I could use my 280mm f/4 for a graduation shoot, but there was too much play in the adapter for consistently sharp results. So hopefully Leica will get theirs out to us soon.

 

Please keep up the great work.

 

JHellow

[tashley]

Thank you - I rely on the good advice I get here for lens choices! I am, for example, currently shooting a lot with a walkaround kit of 18/35/90 but sometimes I think I'd like to slot in a 24. The 28 Cron isn't 'doing it' for me on the M240 and in any event its not so nicely spaced in the focal length range so I might sell mine (I know, I know, never sell a Leica lens: I once sold my MATE and regret it bitterly!)

 

If I do I'll have an interesting choice, so it is good to hear from owners of the lenses I don't know - especially from people who have used them on the M240. I am a bit torn between the 24 Elmarit and the Lux... my feeling is that when walking around, you only really need one fast wide and the rest can be slower if that makes them lighter and smaller, as long as they are still good!

[Advertisement]

Hallo,

Look here: Leica M

[jaapv]

The maximum size for forum images is 1024 pixels, thank you. Please use the upload facility.

[Jager]

Tim, thanks much for your terrific insights. You're a wonderful, natural writer and it is a deep pleasure to read your reviews.

 

I don't suppose you have the 50 APO on order, do you? I'd love to hear your thoughts on it...