Zeiss objectives were not tested using 250mm tube lens as recommended by manufacturer; Nikon 200mm f4 AiS was used instead. This brought two advantages: measurements were slightly better and objectives could be directly compared to my old faithful infinite Nikons, CFI LU Plan 5x N.A. 0.15 and CF Plan 10x N.A. 0.30 EPI.
For a start working distances were checked. Those were 16.5mm for Zeiss 6.3x (at actual 5x), 9mm for Zeiss 12.5x (at actual 10x) and 14.5mm for Amscope mounted on 150mm tube.
Later could be extended to 16.5mm if external shell is removed, and all the way to 29mm if you dare to machine out tube-like front part.
Zeiss GF Panachromat 6.3x N.A. 0.12 was a bit dissapointing when it comes to resolution, expecially if you consider it was measured at 5x instead of 6.3x: only 1300 LW/PH at center and 1240 at the APS-C edge. CA area was modest at 0.44 pixels; visible but easy to remove. Comparable Nikon CFI LU Plan 5x N.A. 0.15 provides no less than 60% higher resolution both in center and at the edge. Yes, it costs twice as much but at $185 as advertized on Ebay these days it is still a bargain.
In contrast to it's little brother, Zeiss GF Panachromat 12.5x N.A. 0.25 produced much better results even in absolute numbers. Considering it's two times higher magnification, it's resoution figures of 1460 LW/PH at center and 1310 at the edge looks very respectable. Only a bit excessive CA at 0.82 pixels spoil the general picture. If converting from raw in Capture One, tool for killing CA works just fine and can cope nicelly with such high values. Again infinite Nikon gives some higher values but only 20% more instead of 60% seen above. Considering much lower price for Zeiss GF, it turns out to be a solid bargain agains Nikon.
Third contender here is well known „knurled ring“ Amscope, one of dozens of 4x Chinese lenses sold on Ebay and other places. Urban legend says it performs much better than it's bargain price suggests. So lets see: Central resolution is 1570 LW/PH which falls to not impressive 1170 at the edge. In absolute numbers all of this might not sound shabby untill competition comes into picture. Which is similary priced Russian Lomo 3.7x N.A. 0.11. Although it's production ceased some time ago, it is still widely available on Ebay for about $50. I have tested few samples and all of them produce excellent resolution on 150mm tubes: Over 2100 LW/PH in center and 1900 at the edges. Even CA is very low at 0.17, about 4 times lower than Amscope. When the plain tube is swapped for 150mm tube lens, Lomo performs even better.
Although Lomo was my favorite 4x objective for some time, simple tube lens scaling down provided much better solution: previously mentioned Nikon CFI LU Plan 5x N.A. 0.15 when relayed via vintage Rodenstock Rodagon 150mm f5.6 gives 3.75x magnification and exceptional resolution.
At center you get 2317 LW/PH and 2276 LW/PH at APS-C edges. Mitty 5x might be even better but untill I finaly buy one or some good soul decides to borrow me one, I wouldn't know! Similar to Amscope objective, LU Plan doesn't shine when it comes to CA, but as said previously, this drawback is solved easily.
So here are all the numbers aligned into one simple spreadsheet making it easier to compare. If you put it side by side to my previously published measurements, you'll notice how the latest figures dropped significantly below 1x - 3x magnification range lenses. This is because microscope lense work deeply into difraction zone. If you leave convolution presharpening at the same standard level used for ordinary lenses, you won't get same sharpness. This is why I have established different sharpening figures for each and every high magnification lens/objective I use for stacking. But that is a topic for another ocassion...
Next time: Filip's scanner lenses.
