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!

  • Recently, there has been an increased activity from spammers, which may result in you receiving unwanted private messages. We are working hard to limit this activity.

Monochrom and Leica lenses

Woody Campbell

Workshop Member
I thought that I would start a thread to collect experience with Leica and third party lenses on the Monochrom. My thesis is that at least some lenses that are fine with the M9 may be oversampled by the Monochrom (and presumably the M10).

It would be great if all of you could share your experience here.

I'm just sorting out how to approach my fairly extensive set of lenses without overwhelming myself or this forum. Suggestions would be welcome. I'll post results as I sort things out.

Subjective reactions are fine.
 

Hosermage

Active member
Sorry, but what are the effects of oversampling? I briefly googled it, but it seems to be a positive thing? Here is how someone describe the Nokia 41MP phone camera: "The answer is oversampling. At its default 5MP setting, every pixel in the finished image corresponds to about eight pixels on the sensor. This oversampling helps reduce noise, increase color accuracy, and increase sharpness."
 

docmoore

Subscriber and Workshop Member
May be correct...see here:

Oversampling - Wikipedia, the free encyclopedia


My assumption is that the premise of the thread is that some of the wonderful characteristics of Leica RF lenses may be nullified by the intense scrutiny that the MM resolution affords.

My personal experience in MF was that the Phase 20 made every lens look wonderful...as the pixel count went up the lenses appeared less attractive.
I hope this is not the case or that a mild Gaussian blur for background in a layer may correct for the intense acuity of the sensor.

As I seem to be fairly far down on my dealer's list with reference to MM delivery I will follow this thread with great interest.

Thank you Woody for posting it...look forward to the discussion.

Regards,

Bob
 

Woody Campbell

Workshop Member
May be correct...see here:

Oversampling - Wikipedia, the free encyclopedia


My assumption is that the premise of the thread is that some of the wonderful characteristics of Leica RF lenses may be nullified by the intense scrutiny that the MM resolution affords.

My personal experience in MF was that the Phase 20 made every lens look wonderful...as the pixel count went up the lenses appeared less attractive.
I hope this is not the case or that a mild Gaussian blur for background in a layer may correct for the intense acuity of the sensor.

As I seem to be fairly far down on my dealer's list with reference to MM delivery I will follow this thread with great interest.

Thank you Woody for posting it...look forward to the discussion.

Regards,

Bob
Bob - thanks for clarifying my shorthand.
 

CharlesK

New member
Hi Woody,

Excellent thread! Of course this will take some time, but just some initial observations, is the favourite 75mm on my M9-P is the 75 Lux. On the weekend I revisited the 75 Cron AA with the M-M and it feels a much different lens with character that I felt was lacking with the M9.

The 24 Lux, I feel also is sharper with the M-M, than with the M9-P. The 50 Nocti f/1.0, on the M-M seems to have wider tonality.

I don't really have the time to show comparisons, but these are just some initial observations:)

As a side note, I am finding I am just using LR 4 RC-2, to PP the raw DNG files. The default linear curve, is too passive, and needs the form, as in SEP II, curves in one of preset film types, that is stronger contrast, with the black and white points adjusted. At the moment I am not using SEP II, even though I love the conversion with it. I am learning the adjustments that are just needed in one step, to have these files PP'd.
 

BeeWee

New member
From a signal processing point of view, oversampling has no negative effect (other than creating bigger files with no new information). It just means that the sampling frequency (either in the spatial or temporal domain) is higher than the signal. The only effect is that you don't gain any fidelity (i.e. sharpness) by increasing sampling frequency (i.e. higher pixel density) since you can already capture all the signal that is present. In plain English, you're not going to gain anything from oversampling, but you're not going to lose anything either. On that note, by having a higher resolution sensor that doesn't require smoothing/interpolation (i.e. Bayer demosaicing), it allows you to take advantage of improvements in lens technology, meaning you can see tangible improvements in sharpness if you get a sharper lens.

On the flip side, sub-sampling leads to anti-aliasing. This is why many digital sensors usually have an anti-aliasing low-pass (aka. blurring) filter.

Bear in mind, with all of the above, oversampling due to defficient optics is different from oversampling due to low-pass filtering (anti-aliasing). Adding an anti-aliasing filter or low-pass filtering via Bayer demosaicing means that beyond a certain point as determined by the anti-aliasing or deomasicing filter used, making a lens sharper won't really improve image quality. On the other hand, oversampling due to defficient optics (i.e. sensor out-resolving the optics), means that you can use a sharper lens and will actually see tangible improvements in sharpness.
 

leicashot

New member
No offense but I think this thread is over-thinking and a classic example of why we all need to stop thinking so much. When does getting too technical get in the way of our photography? My answer is 'now'.

To sum up, the M Monochrom is a great camera that will work well with any lens, as long as the photographer does their job right. A bad picture with a top lens will still be a bad picture, and talking about such things will not change that fact.

I know I sound like I'm trolling here, but coming from a Monochrom user, trust me, this is over-thinking the technical way too much.
 

Paratom

Well-known member
My main thoughts when choosing lenses to use with the MM so far were:
a) in harsh light maybe rather use lenses which do not draw too contrasty
b) since the high ISO capability are so great fast lenses are mainly needed if you want super shallow DOF, but not so much to keep ISO down.

-> Besides some other lenses I have also used the 35 and 50 Summarit on the MM and results look fine to me. f2.5 is not that slow and I really like how those small lenses handle. (side note: I also like the metal lens cap...goes good side by side with the retro feeling when using the MM)

c) since it can make sense to use color filters it makes sense to combine lenses which do not have too many different filter diameters.
 

fotografz

Well-known member
I thought that I would start a thread to collect experience with Leica and third party lenses on the Monochrom. My thesis is that at least some lenses that are fine with the M9 may be oversampled by the Monochrom (and presumably the M10).

It would be great if all of you could share your experience here.

I'm just sorting out how to approach my fairly extensive set of lenses without overwhelming myself or this forum. Suggestions would be welcome. I'll post results as I sort things out.

Subjective reactions are fine.
By "oversampled" I presume you mean that the sensor out resolves the lenses, right Woody?

I wonder if that would be the case with any relatively recent M mount lens?

As far as character is concerned, you are right ... that will be subjective observation.

I do agree that the same principles that have been in operation to date concerning the use of lower contrast optics in harsh contrast situations will still be beneficial, but choices will have to be adjust based on the MM's tonal scale, and any highlight clipping that seems to be showing up in posted examples.

I will be curious to see how some of the older M lenses do on the MM since above 50mm I do not like how the ASPH and AA lenses render ... probably due to my preferred use (people) So, I hope the 90/2.8 keeps its character intact, and I see no reason why not.

Acuity may well magnify any slight optical/mechanical errors like focus shift and make them more visible, or maybe not. I have a well tuned V-1 35/1.4 ASPH that for my distance use demonstrates little focus shift if any, and hope that remains true on the MM or a higher meg M10 since I do not need the expense of changing a bunch of lenses as well as acquiring such an expensive camera as the MM/M10.

To add to the previous poster's observations, I'd say one of the key benefits of a faster optic combined with higher usable ISOs is gaining shutter speed in low light to eliminate or lessen subject movement when desired.

Good thread, and I'll be interested in following it.

-Marc
 

Woody Campbell

Workshop Member
Here's my lens inventory: 18mm super-elmar, 24mm lux, 28mm chron, 35mm lux (current version), 35mm chron IV, 50mm DR chron, 50mm lux, 50mm .95 nocti, 90mm elmarit and 135 apo-telyt. Most used are the 24mm, 35mm lux and the 50s and the 90.

There is a view out our dining room window which for about an hour each afternoon when it's sunny is a terrific brick wall torture test. I've included a sample below. I decided to run through the entire inventory at f4.0 - the fastest stop that works in sunny daylight - as a quick survey. If f4.0 is good then f5.8 and f8.0 are probably fine and f11 and beyond show effects of diffraction. I didn't do all lenses at all f-stops (as I have done with other cameras) because the amount of data simply overwhelms me. The idea was to do a survey then decide where to focus. I actually overlooked the 90 f2.8 - it was tucked in the corner of my carry around bag. I will be able to give it a close look toward the end of the week.

I will hold back any negative impressions until I've fully explored the lens.

The biggest surprise was the dual range summicron (I've had the close focus cam machined off so it mounts on a digital M). At f4.0 its shows no real softness in the corners, or for that matter any other bad characteristic. Overall the impression of acutance is slightly less than the Nocti at f4.0 but it's something that you would not notice except through side by side pixel peeping. The lens has a lovely signature - it or its cousin the rigid summicron were probably responsible for a majority of the iconic Leica images from the 50s and 60s (think JFK in the rocker with John John playing at his feet). From a "look" standpoint it mates beautifully with the MM.

(From prior experience the negatives are 1 meter close focus distance and some tendency to flare. There is a small amount of tame linear distortion that can be corrected in post if one desires to do so.)

So . . . I'll be exploring the dual range chron further for the next couple of days. I'll come back to the others in due course.

Here's the brick wall torture test and center and corner crops for the DR. I've included the crops because it's a little hard to believe that this lens does as well as it does.






 

Bob

Administrator
Staff member
I think that oversampling is a good thing from the signal processing point of view.
Unless there are at least 2 samples per period, a frequency cannot be accurately represented, and even then it is just a square-wave.
Unless the sensor over samples a lens, you cannot see how sharp it actually is. Of course, for the pixel-peepers (including yours truly) an oversampled lens may appear less that totally sharp everywhere, but that is life.
Undersampling on the other hand means you are leaving some of that potential resolution on the table.
-bob
 

Woody Campbell

Workshop Member
I've now had the chance to run my entire inventory through a brick wall torture test. In general terms the MM does not change the look and character of lenses. Most work exactly as they did with the M9. In this category I would include the Voigtlander 12 (but no color shift!), the WATE, the 24 Lux (but no purple fringing!), the 35mm Lux FLE, the f.95 Nocti, 90mm elmarit and the 135 APO Telyt. Perfect focus becomes harder for the longer lenses. I've done well enough with the Nocti but the 135 is a real issue.

The extra resolution may affect how you feel about some of the older designs. The 28 chron seems softer in the corners than I remember - at 2.0 the corners and the outer 20% or so are absolutely dreamy and there's still some softness at f4.0. The new 28 Lux that is being announced in a week or so should address this if you're dedicated to this focal length. The current 50mm Lux exhibited some softness in the outer 10% of the image through f2.8 or so. It looks to me like a curvature of field issue which I believe has been fairly widely discussed. Not a problem in most applications but it gets to the level of noticable to me with the MM (I had safely ignored it with the M9). This may be the rationale for the new super summicorn. The 35 Summicron V IV is positively dreamy. My Voigtlander 16 is not usable - the issue is asymetrical suggesting that it's a quality control issue with the lens. I've previously commented on my favorable impression of the 50mm Dual Range Summicron.

That's it.

I do hope that people will post any disagreements and experience with other lenses.
 

D&A

Well-known member
Woody, what I find interesting and at present am trying to contemplate your observations...is namely the change in performance (with certain lenses), in corner/SIDS/edge performance. In the case of the 36mp D800 DSLR, many otherwise excellent lenses on 12 mp bodies, revealed issues in similar areas of the image, due to increased resolution, pixel pitch and density. In the case of the MM, we're essentially dealing with a similar 18 mp sensor with the exact same pixel pitch and density as the M9, but increased resolution due to modifications of the sensor as compared to the one employed in the M9.

Therefore although similar deterioration of performance is observed with certain lenses as used on these two respective bodies (D800 & MM), I"m curious as to whether increase of resolution due to shear # of pixels ( and pitch & density) or increase of resolution as stated for the MM ( due to removal of the bayer aray),is most responsible for previous good edge performing lenses now deteriorating as such? Maybe the answer is "all of the above", but I wonder?

Dave (D&A)
 
Top