Correcting Neutral Density Filter "X" Factor

Started Dec 26, 2012 | Discussions
Chrisd999
Chrisd999 Senior Member • Posts: 1,380
Correcting Neutral Density Filter "X" Factor

I recently acquired a variable ND filter for my 5D3, which I am planning to use mostly for landscape still pictures. The version I have is the Rodenstock Digital Vario ND MC. Below are a set of pictures I took with varying shutter speeds to test the filter out. The first shot is without the filter, and the rest are with the filter on at increasing levels of darkness. After pixel peeping at 200%, I was amazed at how this filter maintained sharpness throughout the range, although there was a small loss of contrast noticeable after about 5 stops.

However, to my dismay, there appears to be some severe "X" shaped vignetting over the entire image which becomes noticeable at 4-6 stops. Beyond that it is horrible. It is more of an "hourglass" shape in the samples below, and I noticed about a 2 stop different in exposure between the center and sides of the image in the 6-7 ND stop range. I had read this was an issue with all variable ND filters, particularly at wide focal lengths, but I did not realize the effect would be this bad.

I know this can be manually corrected in Photoshop, but that seems like tedious work, and I was wondering if there are any easy solutions to correct this in PP?

Canon 17-40 at 17mm f/11 1/6s - No filter

Canon 17-40 at 17mm f/11 0.4 s - with filter - about 1.5 stops

Canon 17-40 at 17mm f/11 1s - with filter - about 3 stops

Canon 17-40 at 17mm f/11 4s - with filter - about 4.5 stops

Canon 17-40 at 17mm f/11 15s - with filter - about 6.5 stops

Canon 17-40 at 17mm f/11 30s - with filter - about 7.5 stops

 Chrisd999's gear list:Chrisd999's gear list
Sony a7R V Sony 1.4x Teleconverter (2016) Venus Laowa 12mm F2.8 Zero-D Zeiss Loxia 25mm F2.4 Sigma 35mm F1.2 DG DN +8 more
Canon EOS 5D Mark III
If you believe there are incorrect tags, please send us this post using our feedback form.
mikes
mikes Veteran Member • Posts: 3,451
Re: Correcting Neutral Density Filter "X" Factor

I barely see the vignetting versus the flash exposure. To be fair, re-do the test with no flash on an evenly toned background, or wall inside the house. Vignetting will then be apparent.

-- hide signature --

Only when you can criticize yourself, should you criticize others. Mikes.

Chrisd999
OP Chrisd999 Senior Member • Posts: 1,380
Re: Correcting Neutral Density Filter "X" Factor

mikes wrote:

I barely see the vignetting versus the flash exposure. To be fair, re-do the test with no flash on an evenly toned background, or wall inside the house. Vignetting will then be apparent.

I know that looks like flash, but flash was not used with any of these pictures. What you're seeing in the high stop ND pictures is the "X" effect I am describing.

 Chrisd999's gear list:Chrisd999's gear list
Sony a7R V Sony 1.4x Teleconverter (2016) Venus Laowa 12mm F2.8 Zero-D Zeiss Loxia 25mm F2.4 Sigma 35mm F1.2 DG DN +8 more
mikes
mikes Veteran Member • Posts: 3,451
Re: Correcting Neutral Density Filter "X" Factor

It is still best to shoot a big wall for these tests. If no flash was used then yes, I would be concerned as well. I have never seen vignetting like that before, and I have many polarizers and neutral density filters including a 10x. But none that exhibit that!

-- hide signature --

Only when you can criticize yourself, should you criticize others. Mikes.

tony brown Veteran Member • Posts: 4,388
Re: Correcting Neutral Density Filter "X" Factor

Hundminen wrote:

I recently acquired a variable ND filter for my 5D3, which I am planning to use mostly for landscape still pictures. The version I have is the Rodenstock Digital Vario ND MC. Below are a set of pictures I took with varying shutter speeds to test the filter out. The first shot is without the filter, and the rest are with the filter on at increasing levels of darkness. After pixel peeping at 200%, I was amazed at how this filter maintained sharpness throughout the range, although there was a small loss of contrast noticeable after about 5 stops.

However, to my dismay, there appears to be some severe "X" shaped vignetting over the entire image which becomes noticeable at 4-6 stops. Beyond that it is horrible. It is more of an "hourglass" shape in the samples below, and I noticed about a 2 stop different in exposure between the center and sides of the image in the 6-7 ND stop range. I had read this was an issue with all variable ND filters, particularly at wide focal lengths, but I did not realize the effect would be this bad.

I know this can be manually corrected in Photoshop, but that seems like tedious work, and I was wondering if there are any easy solutions to correct this in PP?

There is an inherent problem with a double polarising filter due to selective colour alteration, internal reflection between the two filters and the fact that polarisers have directional characteristics. A circular polariser no less so. A circular polariser consists of a conventional linear layer at the front followed by a bi-refringent layer nearest the camera which outputs the light in two preferential waves at right angles to each other. (Rough description only - plenty on-line.)

In Rodenstock's catalogue, you can read the inevitability of a cruciform output exposure distortion and colour shift as follows:-

I use a homemade version using two polarisers - a linear followed by a circular (nearest the camera) but only for motion blurr effects as in waterfalls, etc. and only weak enough to get the effect and only VERY occasionally. I don't even know what stops reduction I get, as I just rotate to taste. Certainly, the colour bias does shift towards blue if I turn it up too much.

If the outputs are crucial to you, you would be better taking a small set of plain (non-polarising) Neutral Density filters and just adjusting the apertures to get any required intermediate values.

Cheers, Tony.

christiangrunercom Regular Member • Posts: 468
Re: Correcting Neutral Density Filter "X" Factor
1

You can correct this issue with a LCC shot through a white plate.

This can be fed into Capture One, and it will remove the x-shaped vignette, colorcasts, and dust specs.

More info here: http://www.luminous-landscape.com/techniques/iq180_lcc_correction.shtml and here: http://help.phaseone.com/en/CO7/Editing-photos/Lens-Correction.aspx#item6

-- hide signature --

/Christian Grüner http://christiangruner.com/

 christiangrunercom's gear list:christiangrunercom's gear list
Nikon D3X
Chrisd999
OP Chrisd999 Senior Member • Posts: 1,380
Re: Correcting Neutral Density Filter "X" Factor

tony brown wrote:

Hundminen wrote:

I recently acquired a variable ND filter for my 5D3, which I am planning to use mostly for landscape still pictures. The version I have is the Rodenstock Digital Vario ND MC. Below are a set of pictures I took with varying shutter speeds to test the filter out. The first shot is without the filter, and the rest are with the filter on at increasing levels of darkness. After pixel peeping at 200%, I was amazed at how this filter maintained sharpness throughout the range, although there was a small loss of contrast noticeable after about 5 stops.

However, to my dismay, there appears to be some severe "X" shaped vignetting over the entire image which becomes noticeable at 4-6 stops. Beyond that it is horrible. It is more of an "hourglass" shape in the samples below, and I noticed about a 2 stop different in exposure between the center and sides of the image in the 6-7 ND stop range. I had read this was an issue with all variable ND filters, particularly at wide focal lengths, but I did not realize the effect would be this bad.

I know this can be manually corrected in Photoshop, but that seems like tedious work, and I was wondering if there are any easy solutions to correct this in PP?

There is an inherent problem with a double polarising filter due to selective colour alteration, internal reflection between the two filters and the fact that polarisers have directional characteristics. A circular polariser no less so. A circular polariser consists of a conventional linear layer at the front followed by a bi-refringent layer nearest the camera which outputs the light in two preferential waves at right angles to each other. (Rough description only - plenty on-line.)

In Rodenstock's catalogue, you can read the inevitability of a cruciform output exposure distortion and colour shift as follows:-

I use a homemade version using two polarisers - a linear followed by a circular (nearest the camera) but only for motion blurr effects as in waterfalls, etc. and only weak enough to get the effect and only VERY occasionally. I don't even know what stops reduction I get, as I just rotate to taste. Certainly, the colour bias does shift towards blue if I turn it up too much.

If the outputs are crucial to you, you would be better taking a small set of plain (non-polarising) Neutral Density filters and just adjusting the apertures to get any required intermediate values.

Cheers, Tony.

Thanks for the info Tony. Nice to see the manufacturer forthcoming with the issue, but I wish the camera shop salesman shared that with me when he advised me to buy the variable ND. In hindsight, I think a fixed stop ND  would have been a better choice.

 Chrisd999's gear list:Chrisd999's gear list
Sony a7R V Sony 1.4x Teleconverter (2016) Venus Laowa 12mm F2.8 Zero-D Zeiss Loxia 25mm F2.4 Sigma 35mm F1.2 DG DN +8 more
Chrisd999
OP Chrisd999 Senior Member • Posts: 1,380
Re: Correcting Neutral Density Filter "X" Factor

christiangrunercom wrote:

You can correct this issue with a LCC shot through a white plate.

This can be fed into Capture One, and it will remove the x-shaped vignette, colorcasts, and dust specs.

More info here: http://www.luminous-landscape.com/techniques/iq180_lcc_correction.shtml and here: http://help.phaseone.com/en/CO7/Editing-photos/Lens-Correction.aspx#item6

Thanks. Making my own lens filter correction profiles at a few specific FLs and filter stops to counteract the exposure variance is what I am looking for, except I would want to do that in LR4 or CS6, because that is where I am invested. Not sure if that is possible as I am inexperienced with this kind of thing.

 Chrisd999's gear list:Chrisd999's gear list
Sony a7R V Sony 1.4x Teleconverter (2016) Venus Laowa 12mm F2.8 Zero-D Zeiss Loxia 25mm F2.4 Sigma 35mm F1.2 DG DN +8 more
MagicAngel Veteran Member • Posts: 4,504
Re: Correcting Neutral Density Filter "X" Factor
1

Not only that, but I think you'll be really disappointed in variable ND overall.  ND is very useful, when it is needed, but one or two ND filters is really all you will ever need, IMO.  I think you will find that you either don't need it, or you need it on its full setting more often than not.

Personally, I always find myself thinking I need a bit more ND than I actually have.  I don't think I've ever said, "Wow, if I only had a little less ND...".

As for the X, I'm thinking that it looks far worse on that scene than it will in general use.  Just really bad luck that the X feel on that tree the way it did.

MagicAngel Veteran Member • Posts: 4,504
Re: Correcting Neutral Density Filter "X" Factor

Correcting it in LR4 is fine, but the big danger in that is that you are losing dynamic range in those areas.  Some scenes are going to fall outside of your ability to fully correct them as you will lose the ends of your curve.

The Rising Tide New Member • Posts: 1
Re: Correcting Neutral Density Filter "X" Factor

I have a Promaster Variable ND filter for my Canon Rebel T5 and once I get over 5 stops the X starts to appear. I've scoured the net for advice and I think the best thing to do is just stick with a ND filter that is not variable. I would just buy a 8, 9, or 10 stop ND filter and only use that, just like the person above me said, you are always wishing for more stops, so just get the highest and stick with that.

Keyboard shortcuts:
FForum MMy threads