I wanted to see whether the expensive L-eXtreme would provide a significant improvement over the cheaper Duo Band, which I already own. Thanks to an awesome poster (who is free to name himself if he wants) I got a loaner L-eXtreme to test. Since I had just shot the Wizard with the Duo Band, I decided to use that as my test subject.
This is completely unscientific. The data were taken on different nights with different Moon phases and potentially different transparencies. The ZWO data was taken at ~46% Moon, while the L-eXtreme data was taken last night with 68% Moon that set at a couple hours later. Both stacks are about 6 hours. I am in Bortle 5 with an artificial to natural brightness ratio of 4.5:1. Subs for the ZWO were 2.5 minutes, 3 minutes for the Optolong.
Even though the two datasets are not apples to apples, if the Optolong filter performed radically better than the ZWO it might be worth the expense.
On the left is the Duo Band, on the right is the L-eXtreme. I've done nothing but stack and autostretch in SiriL. Top is the wide field, bottom is a 100% crop. The images are subject to jpeg compression, of course.
The first thing to note is that the L-Extreme has tighter stars. I don't believe this is a significant difference in the seeing, although i can't prove it. The L-eXtreme should produce tighter stars and appears to do so, by 15-20%. The second thing is that the L-eXtreme definitely pulls fainter nebulosity.
I would not say there is a significant improvement in the noise. This is making me wonder if my subs were too short; they were set at the start of the evening when the target was low and may not have been swamping the read noise adequately. Any thoughts on this are welcome.
To within that last caveat, though, I would not say that the L-eXtreme provides a radical improvement over the Duo Band, at least not for Bortle 5. The ZWO should take the brightness ratio from ~4.5:1 down to about 1:1. Improvement beyond that is diminishing.