[david94903]

Wow! Great photos Twitch! Inspires me to break out my ND and rethink the world around me.

I use a Marumi DHG Light Contol 8 ND filter. I discovered the Marumi brand here in the US recently, it's now well known here but I took a chance a while back and have discovered that I really like their filters, and it's an affordable alternative to the other major brands.

[Docrwm]

Wow, those are inspiring photos. In looking at ND filters, because I do not own one, I ran across a new one from Tiffen that is a 2 to 8 stop variable filter. Any experience or thoughts on something like it that covers a wide range?

[wasser]

Wow, those are inspiring photos. In looking at ND filters, because I do not own one, I ran across a new one from Tiffen that is a 2 to 8 stop variable filter. Any experience or thoughts on something like it that covers a wide range?

I'm curious about these variable ones too. panoguy linked another one earlier:

Light Craft Workshop - Fader ND mark II

and I came across this one too:

RainbowImaging 58mm Adjustable Fader ND Filter Neutral Density (ND2 - ND400)

ND2 - ND400

[TOUGEFC]

I have a simple 3 stop ND, never really use it, and right now im away on a long weekend getaway and I was plannng to pack and use it, and I completely forgot to bring it

[Docrwm]

I'm curious about these variable ones too. panoguy linked another one earlier:

Light Craft Workshop - Fader ND mark II

and I came across this one too:

RainbowImaging 58mm Adjustable Fader ND Filter Neutral Density (ND2 - ND400)

ND2 - ND400

Thanks for the links. Its something that this thread started me thinking about (really awe inspiring pics BTW) and had not researched it, so those links are very helpful!

The Fader ND MkII looks good too but nearly all the sizes are out of stock at that link but Adorama has then in stock. The cost is bout the same. The Tiffen that I found is only listed as 77mm and no other sizes. There is also a Heliopan ND filter which is adjustable from 1 to 6.6 stops at Adorama. The Rainbow images model is ND2 - ND400 and costs about 1/3-1/2 at Amazon.

[newarts]

I put a circular polarizer on the camera then stack a linear polarizer on top of it. When one polarizer is turned with respect to the other the light intensity varies by about 8 stops.

It is easy to do and not very expensive as you probably have a circular polarizer anyway.

It is the same thing as the variable nd filters referred to earlier Docrwm http://www.lightcraftworkshop.com/site/page1000.html

You probably will notice some tinting as the crossed polarizers go to complete extinction; it doesn't matter - you'll never go to complete extinction in practice - even if you do, white balance will take care of it.

The second polarizer can be made by buying a second circular polarizer, removing its element, flipping the element over, & replacing it. This has the slight advantage of not interacting with light that's already polarized (if one counts this as an advantage.) Using linear on circular combination allows one to do reflection reduction, sky darkening etc. at the same time as reducing overall brightness.

[Rainy Day]

I just got my first circ. pol and have just started to look for NDs and ND grads... thanks guys! Quite helpful.

[northcoastgreg]

Any experience or thoughts on something like it that covers a wide range?

The so-called "Fader NDs" are a bit controversial, as they don't seem to be all from the same manufacturer, and Light Craft claims to make the highest quality one. The main claim against them is that they affect resolution when used with telephoto lenses. Light Craft maintains that their "genuine" versions don't do this, but I can't verify that from my own experience. I have one of the non- Light Craft versions and have noticed no loss of resolution with it; but I've never used it with a telephoto lens.

Whether some of them affect resolution or not, all them suffer limitations when used on ultra wide angle lenses. Since they use polarizing to achieve variable ND affects, on wide angle lenses you can get banding issues if you dial up the effect. So if you shoot with them on a 15mm lens, you might only get two or three stops before banding problems start to become noticeable. All the variable ND filters suffer from this, even the very expensive Singh-Ray version.

[newarts]

The so-called "Fader NDs" are a bit controversial, as they don't seem to be all from the same manufacturer, and Light Craft claims to make the highest quality one. The main claim against them is that they affect resolution when used with telephoto lenses. Light Craft maintains that their "genuine" versions don't do this, but I can't verify that from my own experience. I have one of the non- Light Craft versions and have noticed no loss of resolution with it; but I've never used it with a telephoto lens.

Whether some of them affect resolution or not, all them suffer limitations when used on ultra wide angle lenses. Since they use polarizing to achieve variable ND affects, on wide angle lenses you can get banding issues if you dial up the effect. So if you shoot with them on a 15mm lens, you might only get two or three stops before banding problems start to become noticeable. All the variable ND filters suffer from this, even the very expensive Singh-Ray version.

If "Banding" refers to a band of the sky being darker due to the fact it is polarized, that can be prevented by using DIY stacked Circular Polarizers - the top one in the stack having the element flipped in its mount (it works because polarized light from the sky going backwards through a Circular Polarizer is effectively depolarized before getting to the actual polarizing component of the filter.)

Any effect stacked polarizers might have on resolution should not depend on focal length so far as theory is concerned. Any effect I'm aware of depends on the optical quality of the filter and is not inherent; ie. it is possible to have stacked polarizers that approach optical perfection. The main optical problem is the light passes through at least 4 surfaces so good coating is important.

Stacked circular polarizers used on wide angle lenses may cause vignetting problems if the diameter is too small - that can be avoided by using large enough diameter filters.

Here's a pair of small scissors being used as a spanner wrench to remove the ring holding a polarizing filter element in place:


Remove the ring, remove the element, turn the element over, replace it and the ring.
Stack this reversed CPL atop another CPL; The combination will be a variable ND filter not subject to Banding.

[tm01]

The so-called "Fader NDs" are a bit controversial, as they don't seem to be all from the same manufacturer, and Light Craft claims to make the highest quality one. The main claim against them is that they affect resolution when used with telephoto lenses. Light Craft maintains that their "genuine" versions don't do this, but I can't verify that from my own experience. I have one of the non- Light Craft versions and have noticed no loss of resolution with it; but I've never used it with a telephoto lens.

Whether some of them affect resolution or not, all them suffer limitations when used on ultra wide angle lenses. Since they use polarizing to achieve variable ND affects, on wide angle lenses you can get banding issues if you dial up the effect. So if you shoot with them on a 15mm lens, you might only get two or three stops before banding problems start to become noticeable. All the variable ND filters suffer from this, even the very expensive Singh-Ray version.

I think even LCW admits that on telephotos 200mm or greater their Fader may decrease image quality. Does the DIY method with say B+W filters work better? I've read yes elsewhere, but if that's the case why doesn't LCW build theirs this way?

[newarts]

I think even LCW admits that on telephotos 200mm or greater their Fader may decrease image quality. Does the DIY method with say B+W filters work better? I've read yes elsewhere, but if that's the case why doesn't LCW build theirs this way?

I cannot think of a reason that such filters would affect image quality for long lenses any different than they would affect short lenses. The filter has no knowledge of the lens.

Perhaps image degradation is made more obvious with long lenses because more enlargement might be used with longer lenses - In my experience my long lens is often not long enough so I've got to enlarge/crop to get the final image size I want. This extra enlargement makes image degradation obvious.

I'd love to see some examples or learn more about this if anyone has a reference! I've no doubt it occurs - I just can't imagine why: here's a quote from http://lensporn.blogspot.com/2010/01/sakar-circular-polarizer-filter-cpl-650.html

UPDATE: I have recently noticed strange behaviour with these [sakar aka vivitar] filters when using lenses with longer focal lengths, for some reason where a filter will give perfect results on a wider lens the image will turn mushy on telephoto lenses, I am investigating the matter further and for now I can only recommend these filters for use on lenses 50mm or lower in focal length where they have performed and still are still performing perfectly well.

Also - does "Banding" refer to skylight polarization? - that's implied in the discussions I've read http://philipbloom.net/2009/07/22/loving-the-fader-nd-and-singh-ray-vari-nd-life-savers/

Using CPLs should solve that problem - I'll try to demonstrate it today.

[panoguy]

I cannot think of a reason that such filters would affect image quality for long lenses any different than they would affect short lenses. The filter has no knowledge of the lens.

Perhaps image degradation is made more obvious with long lenses because more enlargement might be used with longer lenses - In my experience my long lens is often not long enough so I've got to enlarge/crop to get the final image size I want. This extra enlargement makes image degradation obvious.

"Affect image quality" is the key term here, not enlargement. The (logical) truth is that *any* filter (CPL, UV, ND) will degrade image quality by placing another not-entirely-transparent layer in front of the lens. Some brands are worse than others.

These variable NDs are made from two stacked filters (which are polarizers made from sandwiched sheets to begin with), and so they already double the potential for distortion due to irregularities in the thickness of the material, which is typically plastic. The cheaper the filter (in mfr cost) the more likely that distortion will be compounded.

So why is it more obvious in a 200mm lens than a 12mm lens? My understanding is that the image from the longer lens is made from light "almost parallel to the sensor" (narrow FOV, even if the front elements and filter glass is the same size). This highlights any small irregularities in the optical quality of the variable ND filter with distortion (which makes things blurry when the distorting element is so close to the lens). So a wider FOV means that these minute distortions are not as apparent in the final image because the relevant light is coming from many different angles.

This is purely for filters, of course, as the lens elements themselves are purpose-designed for that FOV from the beginning. (Unless they are in a zoom lens, but I think Pentaxians know the value of a prime lens...)

[newarts]

....
So why is [image degradation due to cpl] more obvious in a 200mm lens than a 12mm lens? My understanding is that the image from the longer lens is made from light "almost parallel to the sensor" (narrow FOV, even if the front elements and filter glass is the same size). This highlights any small irregularities in the optical quality of the variable ND filter with distortion (which makes things blurry when the distorting element is so close to the lens). So a wider FOV means that these minute distortions are not as apparent in the final image because the relevant light is coming from many different angles....

Thanks for the response. It makes sense that long lenses use light closer to axial - but it isn't clear to me why that matters regarding deviations....(not that it doesn't matter - it is just that I'm slow )

Aha! R Clark at http://www.clarkvision.com/articles/evaluating_filter_quality/index.html says part of the problem is that the long lens has a larger absolute aperture than a shorter lens at the same f-stop therefore sees more of the filter. And the long lens magnifies more (but I haven't seen the math yet to be completely convinced...convinced of the reason for, but not the observation that, filter flatness and lack of optical defects are more important for longer lenses.)

EDIT Here is a simple optical reason why image degradation from filter defects depends on focal length.

1) All light rays going through a lens at a particular angle to the optic axis strike the image plane at the same point. That point's location in the image plane is equal to the ray's angle with respect to the optic axis times the distance from lens to image.
2) A typical filter defect is a small region that acts like a prism and slightly changes the direction of light entering the lens; therefore the affected light rays will be deviated from their proper focus location in proportion to the focal length (for a scene far from the camera.)

This implies a faint fingerprint on a long lens has a greater effect on image quality than the same fingerprint on a short lens. Easy to test with a slight grease smear on a UV filter!

[northcoastgreg]

Also - does "Banding" refer to skylight polarization?

Yes. It's the result of using polarizers to achieve the variable ND effect. As I noted before, this problem affects all the variable ND filters, even the very expensive Singh-Ray.

I'd love to see some examples or learn more about this if anyone has a reference!

Light Craft Workshop, which markets the "genuine" Fader ND, claims that there new "Fader ND Mark II" "has increased the image resolution at tele focal length (70mm or above on 135 format camera)," making their version "more outstanding when compared with those imitation."

[Ben_Edict]

Thanks for the response. It makes sense that long lenses use light closer to axial - but it isn't clear to me why that matters regarding deviations....(not that it doesn't matter - it is just that I'm slow )

Aha! R Clark at Clarkvision.com: Evaluating Filter Quality says part of the problem is that the long lens has a larger absolute aperture than a shorter lens at the same f-stop therefore sees more of the filter. And the long lens magnifies more (but I haven't seen the math yet to be completely convinced...convinced of the reason for, but not the observation that, filter flatness and lack of optical defects are more important for longer lenses.)

EDIT Here is a simple optical reason why image degradation from filter defects depends on focal length.

1) All light rays going through a lens at a particular angle to the optic axis strike the image plane at the same point. That point's location in the image plane is equal to the ray's angle with respect to the optic axis times the distance from lens to image.
2) A typical filter defect is a small region that acts like a prism and slightly changes the direction of light entering the lens; therefore the affected light rays will be deviated from their proper focus location in proportion to the focal length (for a scene far from the camera.)

This implies a faint fingerprint on a long lens has a greater effect on image quality than the same fingerprint on a short lens. Easy to test with a slight grease smear on a UV filter!

I think the reason for the problem, that poor filters affect long fl lenses more, is quite simple:

A wide angle lens has a poor angular resolution - and it doesn't need a high angular resolution, as it "cramms" a very wide angle onto the given sensor. A telephoto obviously has a much higher angular resolution (if it is a decent lens), thus will make the image degradation of a poor filter very visible.

Another point is, that filters quite usually do not have a single bad surface area (unless damaged by a hit). Cheap filters have an undulation glass surface, not a level one. This is due to a simple production fact: cheap glass is flame polished, which leaves a smooth but rippled surface. Better filters are optically ground and polished, which leaves a smooth and plane glass surface.

Ben