Thanks for the link, Dennis. I've read Ed's reviews.
Notice that Ed doesn't actually measure field curvature in those reviews. Anybody, including you, can measure the field curvature of the 8x32 SE quite easily, because the right eyepiece diopter adjustment hash marks are at approximately 1 diopter increments. To measure the field curvature simply place an object at the bottom edge of the field of the right side and carefully focus using only the right eye. Then, still using only the right eye, move the object to the center and carefully refocus using the diopter adjustment ring, not the center focus knob. Now count how many hash marks difference there is between focus at the edge and focus at the center.
Henry, actually I do measure field curvature. I simply don't report it like you do. You report it in diopters. I actually measure it in the amount of aberration that can be removed from the view, in arcseconds. Arcseconds aberration seems a much more useful indicator to me, and that's how I've been reporting this aberration (and others) for many years now. I'll explain further below.
Frankly Henry, the Nikon SE 8x32 has one of the lowest recorded readings of total aberration I've ever seen in any binocular. And the curvature is only a small percentage of the total aberration. Certainly in the top ten of over 70 binoculars, and IIRC, as for curvature, in the top 5.
10/36 = 10 arcseconds curvature at 50% out out of a total 36 arcsec total aberrations
15/50 arcseconds at 70% out
20/60 arcseconds at 90% out.
As I said above, I measure in arcseconds, since everything we see in the view is either measured in arcsec or arcmin. It's the same parameter I've been using to report aberrations in all binoculars and in fact in all telescope eyepieces (I've also tested and written reviews on perhaps 30 to 40 eyepieces) for many years. This allows a user (mostly astronomers that I report too) to judge whether or not there is enough error correction availbale (by refocus) to clearly see that object such as a double star in their target window. Arcseconds correction is a value they can relate too.
An example: Nu Draco, a perfectly even magnitude double star separated by almost exactly 1 arcminute, (60 arcseconds), would not be seen as two stars in the Nikon SE 8x32 at 90% out, as the total binocular aberration is 60 arcseconds and the pair appears blended into one blob. However, with the ability to focus out 20 arcseconds of curvature, you can indeed see Nu Draco at 90% out in the 8x32, since Nu Draco is a 60 arcsec pair of stars, and after focusing out the curvature, there is only 40 arcseconds of residual aberration. That extra 20 arcseconds that can be focused out allows you to see the pair. Not sure anyone would be able to interpret the data with a reference given in diopters.
the only binoculars I seen that even come close or better than the correction of the SE 8x32 are
Nikon SE 12x50
Fujinon FMT-SX 10x50
Celestron Regal 8x42 Roof
Nikon ProStar 7x50
a few others that come pretty close are
Zen Ray Summit 10x42
Leupold WRMesa 10x50
Fujinon FMT-SX 10x70
Nikon Action Extreme 12x50
One of the benefits I have of using this method is, not only can I tell you the position and extent of curvature, but also, I can tell you the extent of the residual aberrations in the view. For instance, the Celestron Regal, a binocular with a field flattener lens has 25 arcsec curvature at 90% out out of 50 arcseconds total aberration. (Note this is significant. Even with a field flattener lens it has more curvature than the SE 8x32). At 90% out, exactly half the error is due to curvature. The other half is due to any combination of coma, astigmatism and spherical aberration, although at only 25 arcsec residual aberration that is very very small error.
FWIW, binoculars with very little curvature (just at the point of maximum 90% out from center) have 20 to 50 arcseconds of curvature. 20 is the lowest I've ever measured (at 90% out). Binoculars with moderate curvature have 50 to 100 arcseconds of curvature. Binoculars with lots of curvature have 100 to 150 arcseconds curvature. 150 is the most curvature I've ever measured.
just another note, although I have no idea if the tic marks on any particular binocular actually do correspond to 1 diopter, I do know that there is often inconsistency between brands. For instance without my eyeglasses on some binoculars my right eye requires an adjustment of 2 tic marks, while 5 tic marks on others. It's not my eye that is causing the difference. Obviously they are not both giving diopters per each tic mark. Furthermore, some binoculars don't have any tic marks at all, just a + or - sign. So I don't rely on the tic marks as a reliable indicator. That's another advantage of actually measuring the degree of arcseconds error in the image, rather than in tic marks. There is no assurance that the tic marks represent diopters, so there is no assurance that any two binoculars can ever be compared. Whereas, measuring arcseconds of error in the image provides the same real measure for any binoculars, and then they can all be compared.
edz