My A7s.KMod is finally here! 5-week wait seems like forever (the normal turn-around is 1-1.5-week, I chose to wait for the arrival of this Mark 2 glass). It is equipped with the Mark 2 glass with thinner profile (same thickness as the M9) and humidity-resistant feature. The glass, however, has the same transmission property as the previous one. I just paraphrased what Ilija gave me.
Let's get the test method out of the way first. I have both the normal A7s and modded A7s. I used the Jobu Algonquin tripod (Gitzo and RRS Series 4,5 equivalent) together with the Sunwayfoto XB-44DL head (low profile tripod head to improve stability). There was a very light breeze but I am confident that this set-up would be able to handle it so I just shot away. I used IR remote to eliminate any possible user-induced vibration (typical with the wired remote).
To maintain the same framing during the test, I used the battery grip. So I just simply and carefully detach/attach the camera body from/to the grip. I believe the framing was maintained as consistently as possible. Both cameras' WB was set at AWB mode, so there was a different in WB (more cyan for the normal and magenta for the modded). I left it there on purpose for easy ID and to see the effect of changing the filter. There was about 0.5 stop difference in exposure between the two cameras (the normal one was a bit under-exposed). I did not normalize the exposure since I don't believe the lighting shifted at all. It is an overcast day today. Any discrepancy was just the different intrinsic properties between the two cams. There is no additional sharpening. Vignette is corrected (100 amount, 0 Midpoint).
This is with the
Zeiss Distagon ZM 15 f/2.8. I shot two aperture-series tests for each camera. I include full-res JPEGs at the end of this post.
The first set is to simply study the improvement as-is without any DOF and field curvature optimization.
Center: I don't see much difference in the center except for a slightly sharper and more moire/false-color with the modded A7s. This is understandable since the modded A7s doesn't have AA filter. Therefore, I just show the f/4 comparison here.
Center f/4 Comparison (Left - Modded, Right - Normal)
Corners: This is where things get exciting.
The order is the same as above, Left - Modded, Right - Normal
f/2.8
f/4.0
f/5.6
f/8.0
This is the second set. The purpose is to find the optimized performance. In Lloyd's test, he mentioned about the extended DOF (i.e. thicker DOF) with the modded camera. Interesting point that I have not seen mentioned anywhere else. With this premise, I took another series with the focus at the corners. I hope the extended DOF would give me a sharp corner-to-corner picture at f/5.6 instead of f/8 or f/11 (to avoid diffraction).
Anyway, the normal A7s with its extremely curved field toward the corners would be useless for this method (only works around f/11). Nonetheless, I do include the results in the rar file. What I am showing right here is the difference between placing the focus at the
center vs.
corners.
Center: Left - Corners Focus, Right - Center Focus
f/2.8
f/4.0
f/5.6
f/8.0
Corner: Left - Corners Focus, Right - Center Focus
f/2.8
f/4.0
f/5.6
f/8.0
My own conclusions: The KolariVision mod helps tremendously. It greatly reduces not only the field curvature, but also the astigmatism. By doing so, it increases the effective DOF that the lens has. As seen from the results above, I can happily shoot at f/5.6 by simply focusing at the corners/edges. The center, by then, is identical. I doubt M9 and M240 users could do better than this, at least on the ZM 15. Below is the link to the full-res version. Again, you can tell these two cameras apart by their WB (cyan-normal, magenta-modded). There are ten pictures for each cam; the first 5 are center-focus, and the last 5 are corner-focus. I also include an aperture series I took with the Samyang 14 (a copy with the least de-centering out of 6 copies I have tried). In the center, the Samyang needs to stop down to f/5.6 or f/8 to match what the Zeiss can do at WO. In the corners, they match at WO, but the Zeiss pulls ahead after that with the optimization. I didn't optimize for the Samyang shot.
ZM15 Aperture Series
Samyang Aperture Series
DSC02879 by unoh7, on Flickr
So for awhile I was wondering about the CV 35/1.4 (pre-asph 35 lux knock off, as sebboh at FM notes
