Originally posted by Adam If you have any suggestions for evaluating this lens, let me know and I will forward them to Ole, who's doing the review. I'm planning on extending our loan of the 560mm by a month to ensure that the review evaluates it fairly.
My suggestion is as follows:
1. Do shots with MLU and the K-5IIs locked/clamped to a heavy stone table top or concrete block (no tripod). The idea is to increase the bodies' effective mass by 10x, not stability.
2. Do focus bracketing and select the sharpest. Maybe, use a rail slider (under the stone table top) to move the body, not the focus ring. Use increments of 10µm divided by magnification^2. Contrast-detect AF or magnified zoom focus is NOT good enough. Ask the makers of photozone ...
The distance should be 50x focal length (say, 20m; 10m at least). At 10m, 10µm / M^2 would be 3 mm shifts. 10 steps on either side (you may be able to get away with 5 steps if you shift around the contrast detect AF position). You may want to have a look at section 3.1 at
http://www.falklumo.com/lumolabs/articles/D800AA/D800AAFilter.html. But 10m is probably too short, knowing the 560 is advertized as a far focus lens. OTOH, there are atmospheric turbulences kicking in at medium distances already (and in the optical tube as well if too hot). So outdoor, don't shoot near the desert ground ...
3. Shoot a target with multiple black squares (or multiple targets) on white ground tilted at 5°, exposed for a range 10...90% gray. The target is tilted at the wall, not the elements on the chart. Inkjet printer. Illuminate avoiding any reflections, best would be a strobe.
4. Expose with no sharpening from LR (best would be LR3 or LR4 PV2010 with all sliders to zero (you may leave contrast at default if done consistently), except for white balance and exposure). No lens corrections. Export TIFFs from crops around edges, at the center at various distances from it.
5. Compute MTF charts using the software from quickmtf.com. Use regions with at least 100 pixels on the edge. It has a batch mode and the crops prevent memory overflows ...
6. Compare the lp/mm resolution for 50% and 20 or 10% contrast (MTF50 and MTF20). IMHO, MTF20 figures are more meaningful for tele lenses than MTF50 (if not sharpened) because they tend to have reduced contrast w/o loosing the details. MTF50 would probably not provide a good comparison of results when sharpened.
7. Do this for the DA560/5.6, FA600/4, Sigma 500/4.5 at F4, F4.5, F5.6, F6.3, F8.
That's my proposed procedure, YMMV
Originally posted by goubejp Hi Falk,
if I remember well you have published an empiric formula to predict the price of a lens vs its length and front element diameter ? Can you remind us the predicted price of the DA560 ?
That was in
Falk Lumo: A hypothetical Pentax DFA* 500mm F5.6 ED(IF) SDMii based on the then current list prices for a lens w/o VR and fluorite front elements (2011 April). It would give you 2800$. BTW, the same formula gives you 5800$ for a 500/4 and 10k$ for a 600/4.
The more recent prices (e.g., 500 with VR or 600 and 800 with fluorite) are more expensive. But there is no reason why the DA560/5.6 exceeds 3k$, maybe 3.5k$ if allowing for 2 years of inflation.
In fact, given its simpler construction, it shoud be below 2.5k$. Really.
Originally posted by eurostar Comparing lenses which are double the price of the DA560, or discontinued by a long time, seems... meaningless. I believe a person which is able to buy a Nikkor 800mm can buy a D4 to complement it.
That wasn't my point.
Buying a super tele lens more than anything else is an investment into a system. For the price of a DA560, I can buy a D600/D7100 plus a used Nikkor 500/4D.
So, the question to me really is if the lens' development at Pentax has still kept their teeth. In the past, Pentax super tele lenses have always been technically excellent. Now, we'll see.