*SOLVED* Baffled by ND filter test results

[Chopstix Kid]

EDIT 10/3/2018: Problem figured out. Due to lighting conditions I had to run a very high ISO and that appears to be what caused it. This morning I was able to run the tests again at ISO-200 and everything aligned as expected. I'll post my test data and an overview later. Cheers.

==========
I was offered to demo ND filters from a few brands for the M2P. I accepted. Rather than the typical hey look wow these are great blah blah blah blah blah review, I wanted to run some tests to see if they indeed give the correct light reduction. After all, if you think you need to reduce light by 3 stops, you want to be sure that your ND is indeed 3 stops, right?

Anyhow, in AP mode I started at f/2.8 with a fixed ISO and ran a baseline on the camera to verify each full aperture stop indeed doubled the exposure time. Everything good...
today
2.8 - 1/320
4 - 1/160
5.6 - 1/80
8 - 1/40
11 - 1/20

Here is what baffles me. Holding at f/2.8, attaching an ND8 should simulate reducing the aperture to f/8. However The shutter speed only decreased to 1/80 not 1/40. The results were consistent with all three brands. I ran the complete test from ND4 to ND32 with all three brands and every filter was off by one stop. Thoughts?

EDIT: Trying a few different things I discovered that if I put it in manual and set it where is should be, it consistently reads EV +.3 confirming that these filters appear to be one stop lighter than they are labeled. Or quite simply I am analyzing the data wrong.

 

[SteelFlyer]

Interesting. That baffles me too. In your scenario ND8 should definitely lead to a 1/40 shutter speed. I’m guessing someone here may know.

 

[Hiflyer808]

Did you keep iso constant?

 

[Chopstix Kid]

Did you keep iso constant?

Yes. I believe I stated that above. I will edit that in if I did not.

 

[Lon Denard]

I was offered to demo ND filters from a few brands for the M2P. I accepted. Rather than the typical hey look wow these are great blah blah blah blah blah review, I wanted to run some tests to see if they indeed give the correct light reduction. After all, if you think you need to reduce light by 3 stops, you want to be sure that your ND is indeed 3 stops, right?

Anyhow, in AP mode I started at f/2.8 with a fixed ISO and ran a baseline on the camera to verify each full aperture stop indeed doubled the exposure time. Everything good...

2.8 - 1/320
4 - 1/160
5.6 - 1/80
8 - 1/40
11 - 1/20

Here is what baffles me. Holding at f/2.8, attaching an ND8 should simulate reducing the aperture to f/8. However The shutter speed only decreased to 1/80 not 1/40. The results were consistent with all three brands. I ran the complete test from ND4 to ND32 with all three brands and every filter was off by one stop. Thoughts?

EDIT: Trying a few different things I discovered that if I put it in manual and set it where is should be, it consistently reads EV +.3 confirming that these filters appear to be one stop darker than they are labeled. Or quite simply I am analyzing the data wrong.

Okay, I am NO EXPERT but this is the internet so here goes. I'm sure I'm wrong but isn't every double a stop and you count the first setting because F2.8 is the first stop from theoretical F0? As in ND8 is three, so 1. F2.8, 2. F4, and 3. F5.6 which equals 1/80. No? Oh well that was my theory.

 

[Chopstix Kid]

Okay, I am NO EXPERT but this is the internet so here goes. I'm sure I'm wrong but isn't every double a stop and you count the first setting because F2.8 is the first stop from theoretical F0? As in ND8 is three, so 1. F2.8, 2. F4, and 3. F5.6 which equals 1/80. No? Oh well that was my theory.

No. Doesn’t work like that, at least not to my understanding. ND filters gives you x-number of stops from any given setting.

 

[Goof]

F0 is an undefined aperture diameter, infinite in width that would theoretically let in infinite light, even despite the fact that the aperture somehow extends beyond the lens body...

 

[Lon Denard]

No. Doesn’t work like that, at least not to my understanding. ND filters gives you x-number of stops from any given setting.

Oh....right, duh! So if you were are F4 and added a ND8 you would go to F8 because you count the starting point as 1 stop? No? Oh well, I'll let the smart people get on with it!

 

[Lon Denard]

F0 is an undefined aperture diameter, infinite in width that would theoretically let in infinite light, even despite the fact that the aperture somehow extends beyond the lens body...

Then you could see the future!

 

[Chopstix Kid]

Oh....right, duh! So if you were are F4 and added a ND8 you would go to F8 because you count the starting point as 1 stop? No? Oh well, I'll let the smart people get on with it!

No. ND8 moves you 3 stops. If you are at f/4 an ND8 would reduce light equivalent to f/11.

 

[Lon Denard]

No. ND8 moves you 3 stops. If you are at f/4 an ND8 would reduce light equivalent to f/11.

I actually understand what it's supposed to be, I'm just trying to make the reality make sense. Hahaha!

 

[PetitFrereArmada]

Well, people on here who own or have owned and used good DSLR cameras know that one ND filter may cost way more than what we are here paying for 4 to 6 ND filters from a ND filter on this forum. So, I am not surprised if it is off. It cannot be of same quality and performance like what real professional cameras use.

Someone here can confirm what they probably pay for only 1 ND filter for their true powerful camera at home. But we are here on mavicpilots paying $100 for 4 or paying $130 for 6 to 8 ND filters. So, we should not expect top notch performance nor their claims to be exactly what they say their filters will do. Or maybe they can explain some facts that you may be overseeing. I personally do not know.

I am sure these filters for our drones work just fine, but are not of professional quality. But they are enough to improve from the out the box dji drone with only their uv filter on it.

 

[Goof]

Quick sanity check on the numbers, have you tried this on a bright sunny day? All those shutter speeds seem to be maybe 3.3 stops darker than "ideal" for direct sunlight...

Shouldn't matter too much what the ambient lighting levels are, the filters should still block the same percentage of light. Maybe it's a quirk of the camera, or perhaps the clouds are causing some weird polarisation issue?

@PolarPro @fstop.labs

 

[Chopstix Kid]

Quick sanity check on the numbers, have you tried this on a bright sunny day? All those shutter speeds seem to be maybe 3.3 stops darker than "ideal" for direct sunlight...

Shouldn't matter too much what the ambient lighting levels are, the filters should still block the same percentage of light. Maybe it's a quirk of the camera, or perhaps the clouds are causing some weird polarisation issue?

@PolarPro @fstop.labs

You are reading my mind . Overcast with a little bit of rain. I bumped up the ISO to compensate. What throws me is that without any filters installed, the base test runs true. So in theory, the amount of direct light shouldn’t matter since it is consistent with the base test. I’m hoping for a sunny day tomorrow and I’ll run it again and compare.

 

[noosaguy]

I was offered to demo ND filters from a few brands for the M2P. I accepted. Rather than the typical hey look wow these are great blah blah blah blah blah review, I wanted to run some tests to see if they indeed give the correct light reduction. After all, if you think you need to reduce light by 3 stops, you want to be sure that your ND is indeed 3 stops, right?

Anyhow, in AP mode I started at f/2.8 with a fixed ISO and ran a baseline on the camera to verify each full aperture stop indeed doubled the exposure time. Everything good...

2.8 - 1/320
4 - 1/160
5.6 - 1/80
8 - 1/40
11 - 1/20

Here is what baffles me. Holding at f/2.8, attaching an ND8 should simulate reducing the aperture to f/8. However The shutter speed only decreased to 1/80 not 1/40. The results were consistent with all three brands. I ran the complete test from ND4 to ND32 with all three brands and every filter was off by one stop. Thoughts?

EDIT: Trying a few different things I discovered that if I put it in manual and set it where is should be, it consistently reads EV +.3 confirming that these filters appear to be one stop lighter than they are labeled. Or quite simply I am analyzing the data wrong.

Do you mind me asking what were the brands?

 

[Chopstix Kid]

Do you mind me asking what were the brands?

I will once I get this sorted and post a review.

 

[Lon Denard]

Do you mind me asking what were the brands?

Well, I was also sent ND filters, although I hope his fit better than mine did. Let's see, FStop Labs, Skyreat and Freewell? I'm also wondering about something. Polar Pro sells a budget set of filters as well for $50 so there has to be a difference...

 

[noosaguy]

Well, I was also sent ND filters, although I hope his fit better than mine did. Let's see, FStop Labs, Skyreat and Pgytech? I'm also wondering about something. Polar Pro sells a budget set of filters as well for $50 so there has to be a difference...

I asked because I have ordered the Freewell all day kit of 8 filters. They f**ked up my order so I'm still waiting. They look very good though.

 

[Lon Denard]

I asked because I have ordered the Freewell all day kit of 8 filters. They f**ked up my order so I'm still waiting. They look very good though.

Thank you! Freewell was the third company I was thinking of! I edited my comment. I haven't gotten the Freewells or Polar Pros but I'm going to try the DJI vs Skyreats real quick.

 

[Lon Denard]

I was offered to demo ND filters from a few brands for the M2P. I accepted. Rather than the typical hey look wow these are great blah blah blah blah blah review, I wanted to run some tests to see if they indeed give the correct light reduction. After all, if you think you need to reduce light by 3 stops, you want to be sure that your ND is indeed 3 stops, right?

Anyhow, in AP mode I started at f/2.8 with a fixed ISO and ran a baseline on the camera to verify each full aperture stop indeed doubled the exposure time. Everything good...

2.8 - 1/320
4 - 1/160
5.6 - 1/80
8 - 1/40
11 - 1/20

Here is what baffles me. Holding at f/2.8, attaching an ND8 should simulate reducing the aperture to f/8. However The shutter speed only decreased to 1/80 not 1/40. The results were consistent with all three brands. I ran the complete test from ND4 to ND32 with all three brands and every filter was off by one stop. Thoughts?

EDIT: Trying a few different things I discovered that if I put it in manual and set it where is should be, it consistently reads EV +.3 confirming that these filters appear to be one stop lighter than they are labeled. Or quite simply I am analyzing the data wrong.

Does white balance stay the same with or without the filters? What happens if you adjust EV back to 0?