The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Focusing the M9

Paratom

Well-known member
As a long time M-user and someone who recently posted how good the M9 works for me I am starting to get annoyed with slight focus problems.

This seems to turn into a neverending story and I am questiong somewhat if the rangefindermetering is really suited well for high resolution digital photography.

For example one ofmy most used lenses - the 50asph/1.4.
Backfocused on my M8 when I boght the lens new, sent it in and then it worked fine on the M8. When I got my M9 and had to send it in for a repair (together with the 50) I got it back and for some reason the 50 seems backfocus slightly specially at infinity. This is enough to make an image "not sharp".
So I send the M9 together with 5 lenses in another time , it comes back, looks fine at short to medium distance but stil not correct at long distance.
Then I buy a new 35/1.4asph, and it backfocuses slightly from everything longer than 1m distance.
My 75 Summarit however seems to work fine (even if I take into account the larger DOF I can see where the focus point is).
My 75Lux, which also had been calibrated by Leica also seemed to be very slightly off and I finally gave up and sold it. I am sure the lens was fine but I just couldnt focus it allsways 100% and I believe it might suffer very slightly from focus shift.(but focus shift is a different problem eliminated step by step by Leica with newer designs)

I have to mention here that I also struggeled with AFaccurancy from some cameras in the past. The Canon 7D, and also lately the K5 (I dont know yet how much is caused by lens problem and how much camera AF-system).

Slight focus errors do destroy each and avery advantage of a fast good lens IMO. A 200 Euro zoom focused propperly will lead to an sharper image than a slightly OOF 50/1.4asph.

Nikon works very fine here (in my experience), and the S2 also seems to work very accurate except maybe in certain special situations.

I think I will give it another try and send it all in another time. But its the first time I really having thoughts to eventually sell my Leica M stuff and stay just with S2, a DSLR and a X1 (or X100).


I am question the rangefinder system for high resolution digital photography with fast lenses. Also talked to Leica about this and they said they allready to a lot to calibrate lenses before they leave the company but that it is quite tricky to get everything perfect.

If I wouldnt like the sensor and the user interface and the lenses of the M-system so much I would have given up long time ago.

I am quite happy that at least the S2 seems to work for me quite well-specially for a MF camera.
 
From what I have personally experienced and learned from other users experience I'm reluctant to blame the rangefinder only for all this out of focus issue.
In my view the problem is with the combination of very fast lenses and high resolution sensors for which not even the AF built in the most advanced DSLR seems to be any more sufficient to fullfill our expectations.
I have seen for instance exemples of poor auto focus performances of the most recent Nikon and Canon fast Lenses (24/1.4 and 35/1.4) when mounted on the top of the class respective DSRL's.
There is also some interesting reading on the digilloyd blog about this subject.
Ario
 

Paratom

Well-known member
From what I have personally experienced and learned from other users experience I'm reluctant to blame the rangefinder only for all this out of focus issue.
In my view the problem is with the combination of very fast lenses and high resolution sensors for which not even the AF built in the most advanced DSLR seems to be any more sufficient to fullfill our expectations.
I have seen for instance exemples of poor auto focus performances of the most recent Nikon and Canon fast Lenses (24/1.4 and 35/1.4) when mounted on the top of the class respective DSRL's.
There is also some interesting reading on the digilloyd blog about this subject.
Ario
Quite an advantage of contrast based focusing a la m4/3.
If speed of contrast AF becomes even better (the gh2 seems allready pretty good here) thats maybe really the way to go except for action and sports photography.

On the other side my Nikons could reliably focus my 50,85 f1.4 lenses, and also the 200 at f2.0, and my S2 also seems to focus pretty good.
 

Paratom

Well-known member
Just curious, have you tried the same lenses on a film body?
Most of them not, except the 90/2.8
Film seemed for sure more forgettable.
Also most lenses worked fine on my M8.
I assume it is now the infinity calibration of my M9 which migh be wrong and probably the 35asph/1.4II which might need some calibration.
The frequency lenses and/or body need calibration seems just to often and nearly each time one adds a piece new body or new lens the confusion begins again.
 

Bob

Administrator
Staff member
Remember that it is an open-loop system, so there is nothing to compensate for slight errors. Each and every component in the chain must be calibrated just-so since there is no sensor that can tell you if the ACTUAL image is on-focus.
They are pretty sensitive and many folks really do not consider it an interchangeable-lens system. Perhaps a changeable lens system, but not necessarily between all bodies and lenses.
When it is right though it works pretty well.
-bob
 
Remember that it is an open-loop system, so there is nothing to compensate for slight errors. Each and every component in the chain must be calibrated just-so since there is no sensor that can tell you if the ACTUAL image is on-focus.
They are pretty sensitive and many folks really do not consider it an interchangeable-lens system. Perhaps a changeable lens system, but not necessarily between all bodies and lenses.
When it is right though it works pretty well.
-bob
The same limitation is applicable also to DSLR conventional (non-LV) AF where the sensor on which the focus is being evaluated is different from the capture sensor.
 

Paratom

Well-known member
Remember that it is an open-loop system, so there is nothing to compensate for slight errors. Each and every component in the chain must be calibrated just-so since there is no sensor that can tell you if the ACTUAL image is on-focus.
They are pretty sensitive and many folks really do not consider it an interchangeable-lens system. Perhaps a changeable lens system, but not necessarily between all bodies and lenses.
When it is right though it works pretty well.
-bob
Thats how I feel-if its right it works well.
Specially as someone who has not only 1 or 2 lenses, and many of them fast /(24,35,50 Lux) it is even more difficult to get everything set.
 

Bob

Administrator
Staff member
The same limitation is applicable also to DSLR conventional (non-LV) AF where the sensor on which the focus is being evaluated is different from the capture sensor.
That is not true.
At least in the case of dslr-like focusing systems, the location of the focusing sensor and its relative position to the the imaging sensor can be adjusted. Once that is done, variations in lenses are in most systems accounted for as the servo moves them enough to achieve detected focus at the focusing sensor. In the M9 lens variations are completely not compensated. Also, in the M9 style of focusing arrangement, the focus sensor is the eye and the servo mechanism includes the brain and hand. I know that my eye is not as good as a focusing sensor in a dslr and my hand trembles and my attention wanders, so my bit as part of this servo is not well implemented either.
-bob
 
Also In the M9 system, once the rangefinder of the camera and the elicoids of the lenses are all calibrated within the correct tolerances everything is supposed to work as expected. In both cases you need adjustments based on the observed performances.
I do not argue in general about the supposed higher reliability of the electronic system vs the brain/eye/hand driven system, but, based on my personal experience, I would say the if the DSRL type AF is in may instances faster and may be easier to use, it is not more accurate and consistent then the RF.
 

Bob

Administrator
Staff member
Also In the M9 system, once the rangefinder of the camera and the elicoids of the lenses are all calibrated within the correct tolerances everything is supposed to work as expected. In both cases you need adjustments based on the observed performances.
I do not argue in general about the supposed higher reliability of the electronic system vs the brain/eye/hand driven system, but, based on my personal experience, I would say the if the DSRL type AF is in may instances faster and may be easier to use, it is not more accurate and consistent then the RF.
I wouldn't be so sure.
It depends on the specific cameras in question, but if accuracy includes repeatability I can measure the hysteresis in the rangefinder system wich is well taken up with a well designed servo system.
Both systems need to be used properly I will grant, but open loop systems have insurmountable shortcomings.
It is just engineering. If the result is good enough for you, fine, but it cannot be more accurate or more consistent because it fails in theory.
-bob

p.s.
don't argue with the mods, it makes us cranky :ROTFL:
 

jonoslack

Active member
Hi Tom
I've been thinking about this one, after a week away with just my M9s. I still think that given time I get a better hit rate with the M9 than with AF, even though my eyesight is less than perfect.

Of course, Lieca's problem here is that there is no such thing as 'right', because of the adjustment of both the rangefinder and the lenses. . . . with the added wobbly of focus shift.

I can certainly make myself very unhappy testing the focus on my different lenses (9) on the two bodies, but I don't really find that this apparent variation is borne out in real life, where I do seem to get it right most of the time.

I always find that the misleading size of the focus points in dSLR's (with relation to the actual focus sensor) create a number of oof images, even though all has looked good.

Perhaps the search for perfection in focus only leads to misery. Of course, if your M9 makes you miserable, then the solution is easy!
 

Paratom

Well-known member
Hi Jono,
yes, perfection never happens.

However to be frienk I believe there are many photographers who blame their eyes but do just have a focusing system not working right. And thats sad.

However I need the infinity thing fixed and I am frustrated that Leica didnt fix it even though I sent everything in.
I can overcome this by turning the focus wheel just sligtly over the point I focus on, but thats not the solution for me.
The S2 and the Nikons show me that a good AF is up to the task, and I start to lust for a M9 with a good AF with AF fineadjust.
Buying a 35mm prime for 3500 Euro and then having to send it in is just no fun and not right IMO. Having to send a 50/1.4asph in 3 times until it works ok isnt funny either.
My M9 should have been adjusted after the repair (I sent the 50 with it), and then it should have been fine after I sent it in another time and exactly explaining what was wrong. Now I have to send it in another time.
I will include a long letter and also cc it to some people I know at Leica.




Hi Tom
I've been thinking about this one, after a week away with just my M9s. I still think that given time I get a better hit rate with the M9 than with AF, even though my eyesight is less than perfect.

Of course, Lieca's problem here is that there is no such thing as 'right', because of the adjustment of both the rangefinder and the lenses. . . . with the added wobbly of focus shift.

I can certainly make myself very unhappy testing the focus on my different lenses (9) on the two bodies, but I don't really find that this apparent variation is borne out in real life, where I do seem to get it right most of the time.

I always find that the misleading size of the focus points in dSLR's (with relation to the actual focus sensor) create a number of oof images, even though all has looked good.

Perhaps the search for perfection in focus only leads to misery. Of course, if your M9 makes you miserable, then the solution is easy!
 
X

xywriter

Guest
Hi,
This is my first post here, but I've been intrigued by this thread, as I've been spending a good deal of time worrying about focus myself. I'm especially concerned with three lenses. 1. the 135 APO Telyt. 2. the 50 Summilux ASPH. 3. the 90 Elmarit M.

Of course, Leica doesn't "support" the APO Telyt on the M9, but I so love the particular quality it has that I've wanted to make it work. I've tried everything, including buying both magnifiers -- which has led me to conclude that I do better with the naked eye (however wretched mine are.) In the end, what it's taken to get an acceptable number of "hits" is practice. In my back yard, I have a buried phone cable marked by a post with a warning label; I've taken more shots of that label than you can believe. It's taught me that the 135 slightly front focuses, at least on my M9, so that, focussing toward infinity, I have to accept the very first "in focus" position. The 90 is almost exactly the opposite -- so that, in focussing this lens, I always do so turning away from infinity. The 50 is dead on.

From all this, I conclude that the OP is right, in a way. I expect it would be very difficult, if not impossible, to calibrate the rangefinder in a particular camera to give reliably accurate focus with a number of lenses, given the unforgiving nature of the "digital film plane." On the other hand, you can work it out on an individual basis, however annoying this can be. I've found this to worth it, if only to let me use the APO Telyt, which is a wonderful lens.

anthony hyde
 

jonoslack

Active member
Hi,
This is my first post here, but I've been intrigued by this thread, as I've been spending a good deal of time worrying about focus myself. I'm especially concerned with three lenses. 1. the 135 APO Telyt. 2. the 50 Summilux ASPH. 3. the 90 Elmarit M.

Of course, Leica doesn't "support" the APO Telyt on the M9, but I so love the particular quality it has that I've wanted to make it work. I've tried everything, including buying both magnifiers -- which has led me to conclude that I do better with the naked eye (however wretched mine are.) In the end, what it's taken to get an acceptable number of "hits" is practice. In my back yard, I have a buried phone cable marked by a post with a warning label; I've taken more shots of that label than you can believe. It's taught me that the 135 slightly front focuses, at least on my M9, so that, focussing toward infinity, I have to accept the very first "in focus" position. The 90 is almost exactly the opposite -- so that, in focussing this lens, I always do so turning away from infinity. The 50 is dead on.

From all this, I conclude that the OP is right, in a way. I expect it would be very difficult, if not impossible, to calibrate the rangefinder in a particular camera to give reliably accurate focus with a number of lenses, given the unforgiving nature of the "digital film plane." On the other hand, you can work it out on an individual basis, however annoying this can be. I've found this to worth it, if only to let me use the APO Telyt, which is a wonderful lens.

anthony hyde
HI Anthony
and welcome - what a pragmatic first post!
I think that Tom (understandably and correctly) feels that if he's sent all the lenses in to Leica together with the camera body, they ought to be able to get the calibration right, and it's hard to argue with.
On the other hand, I'm pretty much in your camp - I understand my lenses, and what I'm going to get with them - and like you I focus some lenses from infinity, and some lenses from the other end.
I don't actually think that the quality of one's eyesight is nearly as important as practice - and I think there is a real truth that if the M9 isn't your primary camera system, it's hard to keep in practice.
all the best
 

Paratom

Well-known member
HI Anthony
and welcome - what a pragmatic first post!
I think that Tom (understandably and correctly) feels that if he's sent all the lenses in to Leica together with the camera body, they ought to be able to get the calibration right, and it's hard to argue with.
On the other hand, I'm pretty much in your camp - I understand my lenses, and what I'm going to get with them - and like you I focus some lenses from infinity, and some lenses from the other end.
I don't actually think that the quality of one's eyesight is nearly as important as practice - and I think there is a real truth that if the M9 isn't your primary camera system, it's hard to keep in practice.
all the best
I agre Jono,
the other thing is when having used an AF camera with accurate focus (like my S2) you realize how easy life can be ;)
If just the S2 and its lenses were in the size (and price) of the M9 and lenses.
By the way the A900 and 24-70 also seems to focus pretty reliable.
 
Top