Camera Labs

After only ONE week (you knew this was inevitable ):
Directly from his dark and secluded labs somewhere in the bavarian forest in an earthquake-proof bunker...

"Hollido! Here is your's truly again with more data than you can shake a stick at! Now it's time to put the Nikon 18-200 VR to the test at how good it is to stop shake'n'blur from unsteady hands, old age, earthquakes, and bursting laughter.
The spreadsheet was almost empty this morning but is by now already filled with 118 data-points for 1/125 sec at 200mm. Testing commenced at 18:00 sharp under sunny conditions and light westerly winds. It included 3 batches with VR=off, VR=normal and VR=active on a Siemens-star target after some warm-up shots into the open air (no ducks were harmed in the course of the action, Patti).
There are still the test-runs at 1/250, 1/60 and 1/30 sec to be concluded. So there are only tidbits from the lab-reports that I can share, but at least I can say as much:
The distribution of sharp shots with the Nikon 18-200VR at 200mm and 1/125sec is similar to the distribution of shots with the Sigmas at 400mm and 1/250sec. Which was to be expected. So average sharpness is also almost identical. Interestingly the increase in sharpness when VR is activated is exactly the same as with the Sigma 120-400mm OS, namely around 0.3 "stars" - which is not much by any measure. And there was no statistically significant deviation in the results found between VR=normal and VR=active.

So this first batch of tests seems to support the conclusion that the effectiveness of the OS of the Sigma 120-400mm at 400mm and the VR of the Nikon 18-200 at 200mm is approximately equal."

And with these famous last words he donned his lab-coat and descended the steep stairs down into the inscrutable bowels of the secret research facility...

----To be continued----
Oh, NOOO

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses

Keep up the good work Thomas!

Great stuff Thomas, thanks for sharing!

....all the numbers are a little overwhelming though

_________________
Graham
Flickr

Just had to eliminate 160 (!) test-shots from the batch as they came out with f/32 and were so blurry from diffraction that they were unusable for my shake test

At least those with f/22 are ok
So there is something to be learned here:
- loss of resolution from stopping down from f/22 to f/32 is equivalent to a loss from shake between a 4 star and a 0 star shot
- if I interpret the results from photozone correctly the stopping down from f/22 to f/32 should reduce resolution on almost all lenses by about 30%. This can also be confirmed by measuring the grey disk of my Siemens-stars. Btw.: stopping down from f/16 to f/32 can halve your resolution

So beware the shake- and stop-down-shock

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses

Just another observation as my tests of the 18-200 VR at 200mm continue (currently 426 testshots). It seems that 1/30 is the most effective shutter speed for the effectiveness of VR. At least preliminary sorting establishes a clear lead of shots with VR=on in the sharpest categories. at 1/60 sec and clearly at 1/125 sec the quality-distribution of shots with and w/o VR is much more similar.
That resonates with some other results I've produced (e.g. with the micro-Nikkor 105mm f/2.8 VR). Surely the speed of VR/OS/IS-correction is limited by the mass that has to be moved and the energy the manfucturer is allowed to draw from the battery. That could explain why VR seems more effective at slower shutter speeds.

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses

After another week in a dark and secluded place:

As the reporter descended the dimly illuminated staircase a cold shiver ran down his spine: What if Dr. Frankenthom was not amused to see him intrude on private property? Deep in thought he did not notice the slippery slime on the last flight of the steep stairs to the secret lab..
BAMM
He lost his grip of the rusty rail as he tumbled the last meters down the stair and went CLONG against the heavy stainless steel door, the entry to the testing facility deeply hidden in the bavarian forest.
As the last echoes of his contact with the massive door died ringing in his ears, with a SWISH and a WOOOSH some unseen mechanism awakened and finally made the door move. Bright light pierced his eyes as he shielded them against the looming figure stooping strangely over him.

"Welcome, I might say. Welcome!" The strange croaking voice made his skin crawl as he struggled to get to his feet. There he was: Dr. Frankenthom, the crazy scientist, who was at the edge of discovering the truth about what steadies the shaky waves of the ever-powerful photons.
"Er, em. You might as well come in! I have a surprise for you!" With these words Dr. F turned on his heels and limped away.
Like if a horrible spell has been cast upon him, he could not hesitate to follow Dr. F into his lab, shivering in anticipation of the news he might be the first to bring to the world. "What is it?" he gasped.
"Well I've turned my interest to the other VR-lens in my arsenal, the aptly named micro-Nikkor 105/2,8 VR. After only 160 test-shots the results seem to converge. So I'm able to give a peek preview of my findings. But please sign here on the dotted line of this non-disclosure agreement - if you want to live to learn...
I have finally discovered that some VRs seem to work better than other OSs and I'd like to share my preliminary findings. So, now here's the meat, lumping together all 4- and 5-star results, as they are only distinguishable by the trained connoisseur:
- @1/125sec 55% of shots w/o VR were 4+5 star quality whereas 95% were of that quality with VR=on
- @1/60sec 45% of shots w/o VR were 4+5 star quality whereas 80% were of that quality with VR=on
- @1/30sec 20% of shots w/o VR were 4+5 star quality whereas 90% were of that quality with VR=on
Talking in average quality, the shots with VR=on were all at 4.2-4.3 stars (remember 5 stars are only awarded to the highest possible quality) regardless of shutter-speed, whereas the shots w/o VR went from 3.6 stars (at 1/125 sec) via 3.0 stars at 1/60 sec down to 1.8 stars at 1/30 sec.
From this you can clearly see that the improvement from the VR-system is larger than 2 stops as the average quality of shots with VR=on at 1/30sec is better than the average quality of shots with VR=off at 1/125sec."

Suddenly the sound of a large explosion reverberated through the lab and the reporter made a quick escape before he could post that burning question that he had come here with: "What confidence level are we talking about here?". We will never know...

----To be continued----

_________________
Thomas (beware: Nikon-fanboy and moderator!) My Lens Reviews, My Pictures, My Photography Blog
D800+assorted lenses