FOTGA variable IR filter: A review

During an IR photography workshop I recently attended, a fellow participant brought to the group’s attention an affordable variable IR filter available on Amazon. The advantage of a variable is that it allows us to experiment with IR photography at various parts of the spectrum with a single chunk of glass that could replace a wallet full of single-wavelength IR filters. For less than $25 (US), I thought this was a pretty low-risk thing to try, so bought one myself.

Before I discuss my photography results from using this gadget, some thoughts on mundane aspects of it:

So with that all out of the way, how well does it work?

I don’t have any lab equipment handy that can test the filter’s accuracy or transmission vs. wavelength, so I started with some hands-on work by taking it out for a spin (pardon the pun) at a local car show.

First, a shot at the “short” end of the filter’s scale (labeled as 530nm):

Did I mention the car show was held in a cemetery? At any rate, I like how this shot turned out — all I had to do to it was the usual channel swap and a bit of tweaking of highlights and shadows. A very useable setting, in my judgment — and the result looks about right for a filter in the shorter IR wavelengths, plenty of color to work with.

The “long” end of the filter’s scale is labeled as being 750nm — and as you’d expect, given that it’s hard to pull colors out of a 720nm photo, it’s even harder at (what’s nominally) 750:

This is very nearly a monochromatic image, but that’s not a surprise in this wavelength range.

So those are my first-use thoughts on the variable filter. But of course, one doesn’t need to stop there. How accurate are the labeled end-points (nominally 530 and 750 nm) of the filter’s adjustment scale? Can I make some sense of the 10 hash marks between the end points (i.e., is there anything like an even change in wavelength with each mark?).

It would be nice to be able to produce transmission curves along the lines of what Kolari publishes for their filters (image courtesy of Kolari Vision):

Unfortunately, producing data like this requires a spectrometer (and a pretty good one at that) — that kind of gear starts at about $2000, awfully pricey to primarily satisfy my curiosity. For what it’s worth, rental units are not cheap either ($200+ for a few days’ rental).

But after mulling this over, I realized that I was overthinking the problem. I didn’t really need transmission curves for the filter, I just needed a way to compare settings on the variable filter with some reliable (known) IR filters. And of course, I already have access to the hardware required to do that — a camera converted to full-spectrum use, and a handful of IR filters. So here’s what I did…

With each of the good, fixed IR filters that I own (all are Kolari standard single-wavelength filters), I did the following:

The end result (with a little help from a spreadsheet) is a plot of measured color temperature vs. cut-off wavelength for the single-wavelength IR filters:

Next, I followed the same procedure with my FOTGA variable IR filter, taking data at each mark on its dial (for good measure, I added four additional marks “below” the low end of its scale — more on that later):

Fair warning: I don’t know how well-controlled FOTGA’s filter production is, so this plot is reasonably accurate for the filter I purchased, but I can’t guarantee that every copy they sell behaves identically.

But back to the story.

I added marks to the low end of the scale after preliminary testing showed me that the filter’s low end (marked as being equivalent to a 530 nm filter) didn’t get anywhere close to the advertised wavelength. Maybe it continued to lower wavelengths beyond the end of the marked scale? But as you can see from the above plot, it doesn’t get much lower — the low end of the filter’s scale isn’t near 530 nm, but it’s about as close as you can get with the unit.

I think the simplest way to show the “space” that the FOTGA filter occupies is to overlay measurements from the Kolari fixed filters on top of the FOTGA filter’s curve:

As I mentioned at the start of this article, the FOTGA filter doesn’t produce any obvious reductions in image quality. But what if we do a little pixel peeping?

I visited a local park and took photos of a scene containing good details — with no filter (full spectrum), with the Kolari 665 nm filter, and with the FOTGA variable filter set to approximately 665 nm. All images below are unprocessed (no lens distortion correction applied, or RAW import processing for color balance, etc.) — I did a straight RAW-to-JPEG conversion followed by cropping.

Here’s the scene at 42mm / f11 / no filter:

And now here’s a 400 pixel square from near the center of the image:

From what I can see, the FOTGA filter only slightly reduces image center detail compared to the Kolari filter. But what about vignetting? Let’s look at the top of some relatively wide (14mm, the lens’ widest setting) shots.

I’m not seeing vignetting either — the sky is already a bit different in IR (vs. full-spectrum), and any vignetting isn’t strong enough to be obvious beyond that.

The FOTGA variable filter can be a fun little gadget to play with. Working within its advertised /marked range, I saw no image artifacts, and experienced no major impacts on photo sharpness, vignetting, etc. And of course, it has the advantage of being affordable.

But the range of equivalent IR wavelengths it can produce is much smaller than is advertised, and is skewed to longer wavelengths. While the filter is advertised as covering 530–750 nm, my experiments indicate that its real range is closer to 640–800 nm. And the upper half of its variable range covers sufficiently long wavelengths (730+ nm) that you’ll produce very similar monochromatic images regardless of the filter position in that part of the scale. While the FOTGA filter promises to replace a pocket full of IR filters, at best it can cover two or three.

You’ll have to decide for yourself if this filter is worth even its modest price (<$30 US) to you. If you hadn’t planned on doing much shooting at shorter IR wavelengths, you can happily play with the lower half of the filter’s dial and get a range of effects. As for me, I’ll likely stick to single-wavelength filters.