I've been flying my Mavic 3 Pro Cine, using all three lenses, with varying results.
I posted a while back about inconsistent image quality with the 166mm.
I now think some of my ND filters might be causing soft images, particularly with the 166mm.

The Freewell ND64/PL seemed to be the culprit today.
I was getting some shots off the coast and even on the RC screen I could tell the image was blurry.
I switched to the Hasselblad ND64 (no polarizer), and the image seemed notably sharper.

Has anyone else had a similar experience?

Could it be because the darker the ND and the longer the zoom, the steadier the shot needs to be.

ND64 is pretty strong and 166mm requires a very steady gimbal. If you put a dark ND on a drone then you will need to expose the shot accordingly.

Also, I always advise people that 3rd party ND filters are a bad idea. Stick to OEM for the best results.

Here's the test I did. Again, the lighting was bad. Nothing about these is pretty. I was just trying to compare quality. I think the shots beginning at :19 are clearer and sharper. The first two shots look muddy and soft.

Forgive me, but what is OEM?

Andrew Bland SF Posted at 3-8 11:31
Here's the test I did. Again, the lighting was bad. Nothing about these is pretty. I was just trying to compare quality. I think the shots beginning at :19 are clearer and sharper. The first two shots look muddy and soft.

Forgive me, but what is OEM?

OEM
Original Equipment Manufacturer

Here's a better test. I placed the drone in a spot where I could shoot at the side of a building, with hard, defined lines.
The Freewell ND64/PL is the clear loser. Really lousy quality. The Freewell ND32/PL looks better, but the Hasselblads are nice and sharp.
Thoughts on this?

And what about different parameters for cameras (frame rate, ex. time ...). I saw that some parameters available for Mavic Pro main camera can't be used/set for camera with zoom. And if you set it for the main camera, they may change when you switch to another camera.

Andrew Bland SF Posted at 3-8 13:47
Here's a better test. I placed the drone in a spot where I could shoot at the side of a building, with hard, defined lines.
The Freewell ND64/PL is the clear loser. Really lousy quality. The Freewell ND32/PL looks better, but the Hasselblads are nice and sharp.
Thoughts on this?

I can't see any difference, but I'm wondering what you are using the filters for anyway?

DronSap Posted at 3-8 13:54
And what about different parameters for cameras (frame rate, ex. time ...). I saw that some parameters available for Mavic Pro main camera can't be used/set for camera with zoom. And if you set it for the main camera, they may change when you switch to another camera.

The ND32 shots were at 60p with 1/120 shutter.
The ND64 shots were at 30p with 1/60 shutter.

Labroides Posted at 3-8 14:25
I can't see any difference, but I'm wondering what you are using the filters for anyway?

I find it difficult to believe that you don't see the quality difference in the Freewell ND64/PL shot.
Look at the left edge of the building. It's clearly soft. And watch it in 4K.
I agree the other 3 are pretty comparable.
But I think the Freewells diminish the quality for some reason.

As for why I'm using them? For controlling exposure. Have you never used NDs on a bright day?

Andrew Bland SF Posted at 3-8 14:39
I find it difficult to believe that you don't see the quality difference in the Freewell ND64/PL shot.
Look at the left edge of the building. It's clearly soft. And watch it in 4K.
I agree the other 3 are pretty comparable.

I find it difficult to believe that you don't see the quality difference in the Freewell ND64/PL shot.
Whatever it is that you see, it doesn't appear to be significant.

You are using a filter that stops 98% of the light with a camera that really needs plenty of light.
You are using the long tele close close to a 3-d subject where focus might be an issue

As for why I'm using them? For controlling exposure. Have you never used NDs on a bright day?
Do you think that you need to cut 6 stops to control exposure?
No .. I've never needed to use ND filters on a bright day.

Labroides Posted at 3-8 16:43
I find it difficult to believe that you don't see the quality difference in the Freewell ND64/PL shot.
Whatever it is that you see, it doesn't appear to be significant.

Can I ask how you control exposure then, if you're not using NDs? In my situation, if I hadn't used the 64, there would have been serious overexposure.

Andrew Bland SF Posted at 3-8 17:01
Can I ask how you control exposure then, if you're not using NDs? In my situation, if I hadn't used the 64, there would have been serious overexposure.

Do you need to ask?
The cameras all have shutter speeds that go up to 1/8000th and the main camera also has a controllable aperture.

Labroides Posted at 3-8 18:10
Do you need to ask?
The cameras all have shutter speeds that go up to 1/8000th and the main camera also has a controllable aperture.

First of all, this post is about the 166mm lens. I made that clear. It does NOT have a controllable aperture. And second of all, I'm shooting video.

If you crank the shutter up to 1/8000, you're going to get a super distracting Saving Private Ryan or 28 Days Later high-shutter-speed look. NDs are used to maintain natural motion blur and proper exposure at the same time.

I asked a simple question, looking for helpful advice. If you want to be an arrogant prick, do it somewhere else.

Andrew Bland SF Posted at 3-8 23:24
First of all, this post is about the 166mm lens. I made that clear. It does NOT have a controllable aperture. And second of all, I'm shooting video.

If you crank the shutter up to 1/8000, you're going to get a super distracting Saving Private Ryan or 28 Days Later high-shutter-speed look. NDs are used to maintain natural motion blur and proper exposure at the same time.

I can see that there's no need to engage further with someone who has such a commanding understanding of photographic principles.
Thanks for the (totally unneeded) education, arrogant prick.

Labroides Posted at 3-8 16:43
I find it difficult to believe that you don't see the quality difference in the Freewell ND64/PL shot.
Whatever it is that you see, it doesn't appear to be significant.

I know that you do like to push your opinion and I agree with you a lot of the time. But if you don't use ND filters and say "I've never needed to use ND filters on a bright day", then that means you don't know how to use them or don't care much for why they're needed. As pointed out in your next post which says that the camera goes to 1/8000th - I'd be laughed off set if I said that to a Director.

If you're happy to film at 30fps and 1/8000 shutter then that's your way of controlling exposure, but the correct way (if you want to be recognised/respected as a filmmaker), if all other means are exhausted, is to use an ND filter to achieve the 180 rule of doubling the shutter speed when compared to the frame rate.

The long lenses on the M3P don't have an aperture, so it's only the shutter left to control light, so that's where the ND's come in handy. I use them all the time.

Andrew Bland SF Posted at 3-8 13:47
Here's a better test. I placed the drone in a spot where I could shoot at the side of a building, with hard, defined lines.
The Freewell ND64/PL is the clear loser. Really lousy quality. The Freewell ND32/PL looks better, but the Hasselblads are nice and sharp.
Thoughts on this?

In your original post, I thought you meant that the photo quality was bad, not the video.

Personally I think it's down to the quality of the filter. OEM is the way to go and DJI has the benefit of being able to create something which is paired really well with their own equipment. You're adding a piece of glass in front of the lens and you're at the mercy of the quality that the manufacturer is willing to go with.

I've always thought that the 166mm lens was soft and it's probably easy to exacerbate the softness with a 3rd party filter. If the DJI ones are better, you're just going to have to stick to those and learn a $ lesson that DJI usually make the better quality accessories.

Andrew Bland SF Posted at 3-8 14:39
I find it difficult to believe that you don't see the quality difference in the Freewell ND64/PL shot.
Look at the left edge of the building. It's clearly soft. And watch it in 4K.
I agree the other 3 are pretty comparable.

I can see very clearly that softness on this particular ND.
Deffo something is not right with that Freewell

CloudVisual Posted at 3-9 01:14
I know that you do like to push your opinion and I agree with you a lot of the time. But if you don't use ND filters and say "I've never needed to use ND filters on a bright day", then that means you don't know how to use them or don't care much for why they're needed. As pointed out in your next post which says that the camera goes to 1/8000th - I'd be laughed off set if I said that to a Director.

If you're happy to film at 30fps and 1/8000 shutter then that's your way of controlling exposure, but the correct way (if you want to be recognised/respected as a filmmaker), if all other means are exhausted, is to use an ND filter to achieve the 180 rule of doubling the shutter speed when compared to the frame rate.

As the OP deduced, I'm a stills photographer.
I'm not wanting to be recognised/respected as a filmmaker.

But I'm well aware of how ND filters can be used in video,.
However the OP didn't give any clues to his level of expertise or what he was wanting to achieve, so I was trying to get him to give more details.

Andrew Bland SF Posted at 3-8 13:47
Here's a better test. I placed the drone in a spot where I could shoot at the side of a building, with hard, defined lines.
The Freewell ND64/PL is the clear loser. Really lousy quality. The Freewell ND32/PL looks better, but the Hasselblads are nice and sharp.
Thoughts on this?

YouTube is a bad way to show video issues, due to the compression, but, there's something in these comparisons which has caught my eye.

There is easily double the noise and artefacts in the shadow of the window pane to the right of the video in the Freewell ND comparisons. DJI's filters show a much cleaner (less noisy) shot.

Putting my tinfoil hat on here, but I do know that DJI put magnets in their filters to tell the drone when certain filters are attached. Perhaps there's some kind of change to the settings when it knows it's got DJI filters on...

I have found from experience, both as a professional photographer shooting landscapes using traditional 35mm land cameras, and now aerial photography and video with drones, that it never pays to go for the less expensive option when it comes to ND Filters... or any lens filter for that matter.

NiSi filters, both ND and their non-circular Polarizer filter are made from glass and are a cut above DJI filters made from plastic, however they are not available for all of the different DJI Camera Form Factors.  The NiSi 'Filmmaker' filter set made for the standard Mavic 3, and Mavic 3 Cine have produced excellent results for me.  The NiSi 'Filmmaker' for Mavic 3 also has the added benefit of working on my M3E camera.

I have found that DJI ND filters to be both adequate, and necessary.  I use them on Mavic 3 Pro, Classic and Phantom 4 Pro v2 and they do what they were designed for.

There appear to be allot of cheaper filter options out there, however rarely do they produce good results.  I would expect that the Freewell filters fall into that category.

Fly Safe.

Labroides Posted at 3-9 04:22
As the OP deduced, I'm a stills photographer.
I'm not wanting to be recognised/respected as a filmmaker.

When I posted video clips trying to demonstrate the problem, that should have been your first clue that I was shooting video, and that 1/8000 shutter was not an option.

I'm well aware of using shutter speed to adjust exposure for photography. I prefer not to put an extra layer of glass over the lens if I can avoid it, but to maintain the 180º shutter, it's often necessary.

So yes, to clarify, this is about video.

For those of you recommending OEM filters, are the Hasselblads considered to be OEM? If so, I'll just stick with those.  

Andrew Bland SF Posted at 3-9 07:23
When I posted video clips trying to demonstrate the problem, that should have been your first clue that I was shooting video, and that 1/8000 shutter was not an option.

I'm well aware of using shutter speed to adjust exposure for photography. I prefer not to put an extra layer of glass over the lens if I can avoid it, but to maintain the 180º shutter, it's often necessary.

DJI purchased Hasselblad a while back. They put the logo on the front of some of their drones, but generally means nothing in terms of hardware. Don’t be fooled by the H logo on the front, it was a great marketing idea and lots of people fall for it!

The filters you’re using with Hasselblad on the front are DJI OEM.

Andrew Bland SF Posted at 3-8 14:36
The ND32 shots were at 60p with 1/120 shutter.
The ND64 shots were at 30p with 1/60 shutter.

1/60 shutter speed probably need more stabilization and all cameras share the same gimbal. ND filter needs slower shutter, slower shutter means more blurry (reason why ND is used). And plus more required stabilization with zoom camera. What do you think? I read more comments regarding hi zoom camera on Mavic Pro. One group of comments says it is awesome how it is stable (in flight), The second says good but it needs more stable platform.  What about to do test without fly, without engine running? Then we will try to check if it is problem with stabilization or camera itself. Just put dron on stable solid desk and try it with high ND filter.
If you see difference - between static test and test in flight, then we will know that vibrations are reason.
Then we could try something more. E.g. change propellers, because may be one of them is very unbalanced. But as the first. Try static test.