Have a probem -sharpness - DMR or what reason?

[Nitnaros]

OK, after approx. 400 shots on my DMR, I think I have a sharpness problem.

I had over-sharpened my initial batch of shots, I think (now);
After revisiting all files, I believe there is a sharpness issue, looking at the file without any sharpening, they look "crappy".

I am posting four images incl. a 100% crop of 900x900 for each one:

- first image, f2.8, 1/1000, ISO 100, 60mm macro
- second image, f8, 1/1500, ISO 400, 60mm macro
- third image, f1.4, 1/180s, ISO 100 ... 60mm macro (see next post)
- fourth image, f5.6, 1/125, ISO 400 ... 35mm-lux (see next post)
All files are processed in C1 4.1 with sharpening turned off.

Do you agree?

Possible reasons:
- DMR ... distance issue; needs shimming or similar
- Focus screen; but images 1-3 have been shot with the stock-DMR focus screen; image 4 with the Maxwell screen, and shows same problem. I believe the screen is lying correctly in its frame (but who knows...).

Is Leica-NJ capable of investigating - and if necessary - fixing this?
Or would the R9-DMR need to go to Solms? In that case, it is advisable to ship directly to speed up the process?

How about the lenses ... they would not need to be checked out, I guess; three lenses all show the same problem (60mm, 35mm-lux, 180mmf2.8).

Thank You for any advise; not happy right now....

Peter

 

[Nitnaros]

Two more images with crops;

The building in Chelsea Manhatten looks good in the overall view, but it was noon, harsh sunlight, high contrast - that fooled me first; the unsharpened crop does not look that great.

The red machine looks really crappy in the crop.

Focusing itself seems easy, is not a real "operator" challenge.

Peter

 

[EH21]

I've not experienced the trouble you have but I'd think a well sorted DMR could do better than that. Definitely NJ can handle this and will do it more quickly if you walk it in. Hopefully some others will chime in but I know that they can adjust the camera to fit the DMR in terms of focus, etc.

 

[glenerrolrd]

This is not up to the DMR s capabilities. I know my R9/DMR went thru several rounds of adjustments to the focusing screen position before the original owner was satisfied. NJ can definitely do this as they have a senior technician from germany that was on the original R8 development team.

 

[robmac]

If the diopter is set correctly, the back attached securely to the R8/9 frame and is clean and all three lenses have the issue with the stock focus screen, I'd think a trip to NJ is in order. The odds of all three lenses being wonky would be slim to say the least.

I would the basic 45 deg angle shots of a focus chart/ruler via tripod first to see if it is a back/front focus issue (hard to tell from the jpegs) or an overall softness problem. Might narrow things down.

What you're seeing is no where near what a DMR should deliver.

 

[Guy Mancuso]

It sounds certainly like the focusing screen is out of placement. The DMR should deliver a lot better than that. This is something that you need to take out the screen and put back in again and make sure your clicked in place and just do a simple focus test. Find a intersection on a 45 degree angle of a side walk crack. Focus on the middle and see if you are back or front focusing. make sure your diopter is okay also. This may just take a few tries. Been there and have the T shirt. Or buy a micro prism screen which I found better on the DMR. Brightscreen sells them. I would try a few things before sending to NJ

 

[Nitnaros]

OK, thanks, folks;

I will reset the focus screen, check the diopter, put the baby on the tripod for more testing and post back here.

For education: What effect has an incorrectly located focus screen?
A front/back focus shift?

I am wondering why the shot with the building in Chelsea is not sharp; it is at f8 and far away; wouldn't DOF compensate there even for a misaligned focus screen?

Peter

 

[woodyspedden]

Make sure you test on a tripod as well. I am 71 with shaky hands and it is sometimes hard for me to avoid camera shake. Tension can create the same issue. The only way to know for sure is to put the camera on a tripod to see if the sharpness improves in a major way.

Woody

 

[robmac]

To get accurate manual focus, the distance from the lens mount to the sensor must be IDENTICAL to the distance from the mount to the focus screen.

I don't recall if the R8/9 has shims (don't seem to remember any), but in a camera with FS shims (40D, 5D, etc), you need thinner shims (move FS back from sensor) to cure front focus and thicker (move FS closer to sensor) to cure back focus.

 

[doug]

I don't recall if the R8/9 has shims (don't seem to remember any), but in a camera with FS shims (40D, 5D, etc), you need thinner shims (move FS back from sensor) to cure front focus and thicker (move FS closer to sensor) to cure back focus.

The R8/R9 do not use shims. I haven't opened the R8 or R9 but if it's anything like the Leicaflexes the screen's position and planar alignment is adjusted with adjustment screws. On the Leicaflexes the screen can be adjusted so that it's in precise focus at all point of the screen, not just in the center. Shims are a shortcut used by some camera makers to cheapen assembly costs and from my perspective as a mechanical engineer this design/assembly technique is one of the things that separate consumer equipment from the good stuff.

Regarding the sharpness problem, I agree the DMR should do much better than this. Are these test photos from RAW files? How processed? The ruler at 45 degrees should help narrow the problem down.

 

[Nitnaros]

OK, I run focus tests:

- Tripod
- Focus chart taped on table (A4/letter size)
- Remote capture to Flexcolor
- I checked that I am not closer than the minimum focus distance
- All processing with sharpening to zero
- Still using Maxwell screen (like yesterday)
- Checked diopter, standard setting is best for me (I wear glasses).

I tested three lenses:
- 35mm-lux at f1.4
- 60mm macro at f2.8
- 180mm APO at f2.8.

I post images from 35 first (two different runs), then 60, then 180.

Take a look; interesting result - what is your conclusion?

Peter

 

[Nitnaros]

Now the 60mm macro results;

again, two independent runs at slightly different angles.

 

[Nitnaros]

Now the 180mm APO result:

 

[Nitnaros]

Observations:

a) The 35-lux is a heavy front-focusing sample; about 4cm when shooting this test (test distance was a little bit further away than the closest focus distance of 50cm).

b) The 60 and 180 seem to do OK.

Nevertheless they are not really sharp either!
... I post below another tripod comparison from today - R9/DMR/60mm-macro at f2.8 and M8 50mm-lux at f2.8 (shot so that they covered approx. same amount FOV). First file is full image with DMR, then DMR crop, then M8 crop. The M8 looks better (although not great either; it has not been at DAG yet like my other M8 lenses).

What are the conclusions beyond that the 35 needs adjustment?

Does this allow to say anything about the focus screen?

Peter

 

[woodyspedden]

Your last test showed close to symmetrical performance. But then you are at nearly parallel to the test chart so it is harder to tell if there is a front or back focus problem. It still could be the focusing screen IMHO. As Guy said earlier, make sure the screen is seated then "CLICKS" into place. I sometimes found there was a mushy closure and this usually meant the screen was not properly in place and results showed it.

Woody

 

[doug]

OK, I run focus tests:

- Tripod
- Focus chart taped on table (A4/letter size)
- Remote capture to Flexcolor
- I checked that I am not closer than the minimum focus distance
- All processing with sharpening to zero
- Still using Maxwell screen (like yesterday)
- Checked diopter, standard setting is best for me (I wear glasses).

I tested three lenses:
- 35mm-lux at f1.4
- 60mm macro at f2.8
- 180mm APO at f2.8.

If I'm reading these test photos correctly, the viewscreen and the sensor are not in the same virtual plane, whether that's because the screen isn't in the right spot or the sensor, I can't tell. Or it could be the mirror isn't exactly at 45 degrees to the screen and sensor. The one thing you can do is be sure the screen is properly seated in its frame and clicks in properly.

What I see is that the 35 is farthest out of focus, the 60 less so, and the 180 the least. This is consistent with a longer lens-to-sensor distance than lens-to-screen distance. This would suggest that the screen is too low, i.e., it's not properly seated in the frame.

 

[robmac]

Something is out of wack - I could focus a 180/2 APO at MFD with a $600 Canon 30D and Katzeye screen dead on - with any of those lenses, especially the 180 and on a DMR, you should be CLEARLY seeing the texture and fibers in the paper at the focus point.

Start with Woody's suggestion (using stock screen) and work your way thru the 'image chain' to try and isolate. Make sure your at roughly a 45deg angle to the test chart.

The tests would seem to indicate a tendency to front focus, but the 180 shot is just mush.

 

[Nitnaros]

I changed back to the DMT stock screen - same result.

The screen snapped into the frame with a clear sound; so that must be OK.

Peter

 

[doug]

Also check to be sure that the sensor is in the correct plane. The sensor's carrier has a bit of 'float' w.r.t the back and it's supposed to press against the film rails so that it is positioned in the correct plane. Be sure there are no obstructions that would come between the film rails and the sensor's carrier. If this is OK and the problem persists it's time to call Leica in NJ.

 

[Nitnaros]

I checked the back, took it off and on again, no difference in images.

I'll try to drop it off on Tue at Leica NJ - I am only 30 car minutes away.
--

General question:
If the focus screen is not in the right position, does it create uniform unsharpness at all f-stops and focal lengths, or what is the consequence?

Just want to learn at least something from this episode....
(spent today 5h testing, arghhh:wtf.

Peter