Filter Accuracy


September 2015

ACCURACY OF NEUTRAL DENSITY FILTERS

Being used to hosting a regular program of workshops in the south of England and in particular Dorset, I find myself regularly tutoring photographers in the art and craft of using neutral density (ND) filters. For anyone who is unfamiliar with how or why neutral density filters can be useful for landscape photography, please see the brief explanation below. There are two types of neutral density filters:

ND Graduated Filters
Often called grads and used to balance an exposure when parts of the image are brighter than the range of light the camera can record. The top half of the filter has a dark opaque shade which blocks out the extreme light in the view. They are neutral in colour and graduated from top to half way down the filter. The bottom half is generally clear (see diagram below). There are two types, soft and hard graduated, both of which can be used in different circumstances. The soft grad extends softly from top to bottom and may be used when an image has something protruding through the horizon. The hard grad would be used when the horizon is more defined i.e. when a flat horizon exists on the coast.





Solid ND Filters
Often known just as "ND's" and used to slow the exposure down. The opaque shaded filter is solid all over with no graduation.

Aims of this test
To see just how big (if at all) the gulf is between cheaper and expensive filters. Also to see whether the gulf in quality control mirrors the gulf in the cost of the filter brands. I decided to take a closer look at two of the popular brands and see how accurately they perform.

Method
I used two cameras, a full frame Canon DSLR and mirrorless Fujifilm XT1 compact systems camera (CSC). This was to prove both cameras recorded similar results despite being different brands and camera types. Using both cameras in the semi automatic 'aperture priority' mode, I set a constant aperture on both cameras and allowed the camera to calculate the exposure time. I then attached a filter adaptor ring and a holder to the front of the lens.

Theory
I take a shot without the filter attached and record the speed of the shutter, eg. 1/100 sec. I then attach an ND filter to the lens and record the speed of the shutter. To be accurate the 1 stop filter should have halved to 1/50 sec. The 2 stop filter should have halved again to 1/25 sec and the 3 stop filter should have recorded a speed of 1/13 sec.

The Test
I first tested 12 x Cokin P (85mm) graduated and 12 x Cokin P (85mm)solid ND filters. A typical Cokin P filter costs between £10 & £15.

I then tested 10 x LEE (100mm) graduated and 3 x LEE (100mm) solid ND's, including the LEE big stopper. A typical LEE filter costs between £75 & £85.

The results

Cokin P Filters
Only two of the graduated filters recorded accurate shutter speeds, which means the strength of the neutral filters are completely wrong on 10 out of 12 of the filters! Some were as much as a whole stop of exposure out. The solid ND's were equally as bad, recording inaccurate times in 9 instances out of the 12! Again, in some cases recording inaccurate times of 1 to 1 1/2 stops. Shocking!

LEE Filters
All the graduated filters were 'spot on' and no inaccurate speeds were recorded. This was also the case with the LEE solid ND's with just one single exception - the big stopper. The big stopper slowed the shutter down by 11 stops instead of 10 (a workable and acceptable tolerance). I did the same test using a set of LEE Seven5 filters and the density was perfect right through the range of graduated and solid ND filters.

Conclusion
You pays your money and takes your choice!