One of the most common questions I get when teaching workshops or talking about long exposures is “Who makes the best neutral density filters?” In contrast to just five or six years ago, most ND filters now are really high performing. But I must confess I too have been curious about which ones are the best. So I took 10- and 6-stop* ND filters from four popular companies (Lee, Progrey, Nisi, and Breakthrough) and put them to a head-to-head-to-head-to-head test. In the test I compared color tint, actual density, vignetting, and sharpness to a baseline, non-filtered photo.
Disclaimer: I am not sponsored by any of these filter companies. However, with the exception of the Lee filter which I purchased, each company sent me their filters for free to test. Nevertheless, this comparison is as unbiased and objective as I can make it. On that note, this is not a scientific, lab-based test. I simply looked at the characteristics that are important to me in a filter and evaluated them based on how I would use a filter. You want science, go follow Neil deGrasse Tyson. You want a real-world usability test, read on!
* Progrey doesn’t make a 6-stop, only a 7-stop, so that’s what I used.
10-Stop ND Filter Comparison
The baseline photo was taken at f/8, 1/6 second, ISO200, WB 3030K, and 18mm. I chose ISO 200 to keep my shutter speeds reasonable as I was shooting indoors. In theory a 10-stop filter requires you to increase your shutter speed by a factor of 1024, meaning I shot my filtered photos at 1024*1/6 sec = 170.66666 sec = 2 min, 51 sec. The 3030K (which LR interpreted as 3100K) was chosen specifically to stay constant throughout the test, though in the second part of the test I chose Auto WB to see if the camera could compensate for any color shifts of the filters.
Here are the first five photos, straight out of camera:
You can instantly see that the Progrey and NiSi filters are excellent, Breakthrough is great but a tiny bit dark, and the Lee is the worst performer (though to be fair my Lee Big Stopper is 5 years old, so perhaps they have a new recipe by now). Let’s take a look at each category in detail.
Arguably the most important feature of a strong ND is how color-neutral it is. And if it is not color neutral, can the color cast be corrected in post? I compared each photo and made adjustments to the white balance until I felt the colors were accurate enough that the average person wouldn’t notice a difference. As you might expect, the Lee required the most adjustment, but the Breakthrough, NiSi, and Progrey almost none at all. Here are the tints, the corrections required, and the color-corrected photos.
|Filter||Apparent Color Tint||WB Correction Required|
|NiSi 10-Stop||A tiny bit warm||-100K from 3100K to 3000K. No tint correction.|
|Lee 10-Stop||Noticeably cool and green||WB: increased 1300K from 3100K to 4400K|
Tint: increased 42 from -9 to +33
|Progrey 10-Stop||A tiny bit warm||-100K from 3100K to 3000K. No tint correction.|
|Breakthrough 10-Stop||A tiny bit green||No White Balance correction.|
Increased tint 4 from 4 to 8
Are these 10-stop filters actually 10-stops? That’s the question here. To analyze this I looked at the center of each photo and adjusted the exposure of each till it matched the baseline exposure as closely as possible. The reason I only looked at the center was so that I could separate out vignetting from density as much as possible. Here the Breakthrough filter was the farthest out, at 10.25 stops. The rest performed excellently, each being withing +/- 0.10 stops of 10 stops in the center of the image.
|Filter||How Close to 10-Stops?|
|NiSi 10-Stop||Just slightly darker than 10-stops. Exposure increase of 0.10 stops required.|
|Lee 10-Stop||Very slightly darker than 10-stops. Exposure increase of 0.05 stops required.|
|Progrey 10-Stop||Just slightly lighter than 10-stops. Exposure decrease of 0.10 stops required.|
|Breakthrough 10-Stop||Slightly darker than 10-stops. Exposure increase of 0.25 stops required.|
At this point most of the images are getting very close to the baseline photo. The last main adjustment needed is vignetting. Although the centers of each image have been match in terms of exposure you can see that there’s still some variance in the corners. This, along with color cast, is where the Lee shows its worst performance. The other filters do a much better job of controlling vignetting. I attempted to equalize vignetting in each case by using the lens correction tab in Lightroom, and though I couldn’t get them perfect, they are close.
|Filter||Degree of Vignetting / Correction Required|
|NiSi 10-Stop||Slight vignette. +24 correction.|
|Lee 10-Stop||Strong vignette. +90 correction with a shift in the midpoint of -49.|
|Progrey 10-Stop||Minimal vignette. +11 correction.|
|Breakthrough 10-Stop||Slight vignette. +20 correction.|
Now we’re within a gnat’s ass hair of the baseline photo. Let’s take a look at the last important feature of a neutral density photo: sharpness. In other words, does the filter have some optical deficiencies that blur out the details of your photo? For this test I took the camera outside where it could be on a firm surface (asphalt) to minimize any external vibrations. Here I did quick color and exposure corrections to each photo then compared the corners and center. And to be honest with you I’d be stretching to find significant differences between any of the filters. Heck, I can’t even find any insignificant differences. Maybe if you were in a lab with one of those resolving charts you would see something but in this real-world example I found all of the filters to have no noticeable degradation of sharpness.
Now let’s take a subjectively objective look at the performance of each filter in each category to determine the best performer. In each category I thought about what would make a filter unusable for me and created a measurement scale based on that, from 0 to 10. For example, I decided that if a filter induced a color shift of 10,000K or a tint shift of 100 it should score 0 on the color scale. That made it easy to determine a linear relationship between color shift and performance. For exposure I decided that every tenth of a stop out of spec would deduct 1 point from the performance score. For vignetting it’s judged out of the maximum possible LR correction of 100. For sharpness there was no need to define a scale since they all performed excellently. Then I summed all the points and divided by the number of categories to get a normalized final score out of 10. These are not standard measurements of filter performance; they are derived by me based on what would make me not want to use a filter for my own work.
* Scored as 10 – ((WB Change/1000) + (Tint Change/10))
** (Scored as 10 – (Exposure Change * 10))
*** (Scored as 10 – (Vignette Correction/10))
**** Total points / 4
From this we can see that the Progrey filter is the best overall performer, followed closely by NiSi and the Breakthrough not far behind. As expected the older Lee Big Stopper is not on the same level, though as you’ve seen in this comparison its faults are still correctable, making it a completely usable filter.
As a final point let’s take a look at cost, since you might expect an expensive filter to out-perform a cheap one. These prices are the most current retail prices, as pulled from each filter’s website on September 12th, 2017. Prices are for 100mm square filters for NiSi, Lee, and Progrey, and the 77mm screw on for Breakthrough.
|Filter||Retail Cost (USD. Not including shipping or tax.)||Value* (out of 10)|
* Performance Normalized by Price (Final Score * Lowest Price / Filter Price)
The takeaway here is that not only is the Progrey filter the best performer but it’s also the best value, with an outstanding performance for the 2nd to lowest price. The NiSi is the next best value with excellent performance to justify its high price. The Lee is 3rd, since even though its performance is the weakest it also has the cheapest price point. Finally, Breakthrough is last in terms of value as even though it has great performance it’s also the most expensive in this list. The caveat there is that the Breakthrough is a screw-on filter, meaning you don’t need to buy an adapter and holder like the other filter systems, which obviously raises their overall cost. This metric will be factored in in the 3rd part of this analysis that looks at filter holder systems.
Which Filter Should I Buy?
If you already own a filter holder system and simply want the best filter, get the Progrey. Even if you are on a bare bones budget it’s worth spending the extra $20 on the Progrey over the Lee since it will save you lots of time and headaches in post. If you don’t own a filter system already and don’t intent to buy one, or you want to keep your kit small and simple, go for the Breakthrough.
Stay tuned for the next part of this review where I look at 6 stop ND filters and filter holder systems.