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!

P45+ back, body & lens kit: learning curve thread.

carstenw

Active member
I've got a chrome 35 Lux ASPH with almost no focus shift, but somehow the entire front section (from the focusing ring and forwards) worked its way loose during an Italy trip, so I had to send it back to Leica for repairs, which ended up costing about 280 Euro. I have not had a chance to check if for focus shift yet, but will do so soon. Hopefully it stayed a good copy...
 

Bob

Administrator
Staff member
80mm f/2.8 D P45+ samples at 2.8.
I went shooting around Napa with Jack today and found this nifty old wall which reminded me of this thread.
These have been processed with default C1 sharpening, and then sliced to make the 100% crops. Results of the omitted corners were similar so in sloth I didn't bother posting.
First the full image at f/2.8
View attachment 9093
100% Center crop
View attachment 9094
100% Upper right corner
View attachment 9095
100% Lower left corner
View attachment 9096
-bob
 

Dale Allyn

New member
Bob, big difference between f/2.8 and f/5.6 here IMO. Great subject to shoot for this discussion. I shot the shots at f/5.6 because I had it in my mind that Tim had specified that aperture as part of his process. I may have mis-read that. Perhaps I should have included some f/2.8 files as well.

In any case, your samples look great, and hopefully they will help Tim in getting immediate resolution to his issues.

(It's cool that you and Jack hooked up in Napa this week. Hope you're having a great time.)

Edit to add: It looks like your 80mm exhibits less barrel distortion than mine, but we should probably compare at similar distance to subject, etc. to be certain.
 

Bob

Administrator
Staff member
Dale,
After his posting, I just had to try my sample. Yes there is a big difference between 2.8 and 5.6, but 2.8 is still pretty usable. I normally use this lens at f/8-f/11 and occasionally at f/16.

Jack and I had a great time running around Napa valley looking for targets of opportunity. Of course we had to concentrate on lunch, and tried out Buchon in Yountville. I introduced Jack to the simple pleasure of a bone dry Hendricks martini. After that, steak frits, cheese, and a bottle of wine, we were well prepared for the rest of the afternoon.
-bob
 

Bob

Administrator
Staff member
Dale,
If I had only known.
Where is California/Thailand? Near the Dumbarton Bridge? :D
-bob
 

carstenw

Active member
In a conversation with Thierry, he mentioned the possibility of sensor misalignment, rather than lens problems. He thought that with very bad results all over the frame like this, the sensor might be more likely to be at fault than the lens. Something to look into, perhaps by trying other lenses.
 

tashley

Subscriber Member
80mm D f/2.8 demonstration, continued
Now for tthe same lens and subject at f/5.6
Whole image

-bob
Bob, thank you in the extreme for the time you took to do this: to me it definitively shows that either your lens or your technique is significantly better than mine! Seriously, your corners at both those apertures are notably better than mine and this helps me abandon any last vestige of fear of hypochondria!

Best

T
 

tashley

Subscriber Member
Dale,
After his posting, I just had to try my sample. Yes there is a big difference between 2.8 and 5.6, but 2.8 is still pretty usable. I normally use this lens at f/8-f/11 and occasionally at f/16.

Jack and I had a great time running around Napa valley looking for targets of opportunity. Of course we had to concentrate on lunch, and tried out Buchon in Yountville. I introduced Jack to the simple pleasure of a bone dry Hendricks martini. After that, steak frits, cheese, and a bottle of wine, we were well prepared for the rest of the afternoon.
-bob
My that sounds good. Yum yum...

Great opportunity for photos in California today: I understand there's earth quake practice on a huge scale?

T
 

tashley

Subscriber Member
In a conversation with Thierry, he mentioned the possibility of sensor misalignment, rather than lens problems. He thought that with very bad results all over the frame like this, the sensor might be more likely to be at fault than the lens. Something to look into, perhaps by trying other lenses.
I'd wondered about that but it seemed to me that if it were the case, there'd likely be one or two corners sharper than the others and in all my tests so far they look similarly soft. I guess theoretically I could have a persistent bias towards orientating the camera at a slight skew to the image plane and that might be consistently correcting for any sensor skew...
:angry::cry::sleep006:
 

Guy Mancuso

Administrator, Instructor
I tend to doubt the sensor is off but of course there is a remote chance in that. Only real way to test that is with more lenses and see if there is a pattern somewhere in the files. I lean much more on the lens as the issue.
 

carstenw

Active member
Yeah, hard to know, but definitely something worth checking out. Testing a known good lens will show what is up with the back, if anything. Tim, it could be simply at the wrong distance, lending a softness to the entire image. It doesn't have to be tilted.
 

tashley

Subscriber Member
Yeah, hard to know, but definitely something worth checking out. Testing a known good lens will show what is up with the back, if anything. Tim, it could be simply at the wrong distance, lending a softness to the entire image. It doesn't have to be tilted.
Hi Carsten,

Not sure I understand you there... if the sensor is parallel to the image plane (or close to) and the focus is correct, what 'wrong distance' do you refer to?

Best

T
 

Bob

Administrator
Staff member
Bob, thank you in the extreme for the time you took to do this: to me it definitively shows that either your lens or your technique is significantly better than mine! Seriously, your corners at both those apertures are notably better than mine and this helps me abandon any last vestige of fear of hypochondria!

Best

T
Tim,
No trouble,
I didn't use much technique at all, tripod, mup, 3 sec self timer, auto focus.
No matter the cause, you CAN to get sharper results.
-bob
 

Jack

Sr. Administrator
Staff member
A few more with the 80, along with crops. Not tests per se, just images FWIW...

First was this private driveway (yeah, life sucks for some eh?). FWIW, there is so much fine detail in this file that downrezzing it is almost impossible to do without leaving it looking like crap as it does here. I didn't spend a huge amount of time trying though, because the main point is the crop:



Focus point is about 1/3 way in or about at the red grape leaves on the right edge of this crop. This is the 80 at f16, note the telephone lines in the background. Here we are seeing the beginning effect of diffraction, but IMO not horrible for the DoF gained over f11, so I use f16 - f22 when added DoF is needed:



~~~

Next is a funky building we came across. Looks like a mini grain-elevator/livestock-barn so maybe it was a small dairy? Anyway, this one is the 80 at f11:



The jaggies on the insulators are in the original tiff too, but worse here in the jpeg, accentuated and more of them:



~~~

Bottom line is a few observations:

1) I don't think the 80 should suck -- if it does, return it for a replacement.

2) Processing these high-rez MF files for web is difficult! First off, 8-bit sRGB color is doing them no favors -- a *bunch* of the best color gets lost (especially high yellows and reds) with the web conversion. Next is the amount of jpeg compression needed to keep them under 400K (so they load fast) is significantly more than what I ever had to do with the typical DSLR file -- Normally, 80% jpeg was fine for up to a 900x1200, but some of these are down to jpeg 50% to get them under 400K for a 750x900... Lastly, I definitely need a new/better downrezzing routine to try and keep the file looking "right" --- off to work on that now ;).
 
Top