Sensors seem to be a factor when considering variable ND, IRND, or stacking. For the purpose here I'm using a GH4 (GH5 will come soon) for video. Maybe that doesn't matter, maybe it does?

There's a fair amount of acknowledgment that variable NDs cause a loss of sharpness in telephoto lenses. I often use the 35-100 2.8 Panny at the long end and have observed as much using the Fotga variable ND. I'm looking for something better.

There appears to be a recent set of arrivals from China touting Schott glass that some find to be solid performers while others dismiss. This would be appealing if they are any good.

Pricing. I don't have an unlimited budget but am willing to spend money to solve problems (like sharpness and color cast). Spending 2x the money for 10% improvement is beyond what I'm willing to do, however.

Thanks to all who contribute here. I read a ton, comment little. I look forward to your thoughts and experiences on this topic.

KIPON released auto-focus adapter with interchangeable plug-in graduated ND filters(four steps).

We will start to deliver this new adapter before March 15, the market price would be 428USD.

• Designed for using All Canon EF lenses on Sony E Mount Cameras
• Including all full frame and APS-C Cameras
• Interchangeable plug-in variable ND filters/four steps *High-speed auto-focus function and Exposure data
• Phase detection AF comaptible
• Maintains all electronic connection between lens and camera
• ND range from 1.5X to 7X
• Securely Mates Lens to Camera
• Precision Machined Parts
• Maintains ‘Infinity’ Focus with ND adapter
• Firmware upgradable
• $428 price

Available at https://www.amazon.com/Kipon-E-Mount-Integrated-Filters-Adapter/dp/B01F6ICSA0/

Is there any good way of adjusting for variable light conditions while keeping aperture, ISO and shutter angle locked?

Vari ND's with gears? Is it ever done in cinema, and if so, how?

I found this: http://www.fujifilm.co.jp/corporate/news/articleffnr_0727.html very cool and usable looking vari nd from fujinon, but it looks like its for smaller sensors than m43

So, any suggestions? Or do I have to just face the facts and start playing around with auto shutter/aperture?

I'm guessing this has something to do with the polarising effects of Variable ND filters and the effect that has on highlights, reflective surfaces etc.

Any thoughts on this? Are there any Variable ND filters out there which minimise the 'polarising' effect as they're certainly more convenient, but right now I just can't trust the look i'm getting from them and straight NDs give me a much more 'realistic' image (which I can apply looks to in post to take the edge off).

I'm not sure if I'm going to use the ND filter. But if I do, how do I set the white balance?

Variable ND filter tends to change the colors and I've read that automatic WB is not recommended as it loses that golden glow from golden hour.

This filter has also polarizer. Any sense to use that feature?

So, for this test, which is probably not well done, I set up my GH1 with the reliability hack and set it to smooth with everything -2. I manually white balanced on the white door with each test after I set exposure. I pulled focus using a loupe.

To my eye, the one with the filter on it and the extra LED light does look, perhaps, a little less sharp. When I sharpened, it became a little more clear that there may be a slight loss of sharpness. But you guys look at it and tell me if my eyes need to be reset.

I ran three cameras on my doc (two GH1s and an HMC150) and was able to find an angle every time that worked. But I did not properly pay attention to color on some shots and tried to dial it in. After this test, I'm going to manually do it moving forward. Actually, I did manually balance with my video camera (the HMC150) on each interview, which did make a difference.

My main question is: now that Light Craft has a 37mm variable ND available at Adorama, based on this test, how much better would it be?

Finally, I realize that the casting and focus issues would become more pronounced as I cranked the ND down (darker), however, mine was set in the middle of the scale on the side of the filter, which I thought would be good for this test.

Via: http://cinescopophilia.com/fujinon-nd-p01-optical-variable-nd-filter/

Text is quite bumpy, but main idea is that bu turning two gradient wheels in opposite direction you can get even ND from 90% transparancy up to 1/64 original light only.

And no cross and color shift issues associated with reverse poliriser filters. :-)