Why Variable NDs Are Problematic
A variable ND filter works by rotating two polarizers against each other. At low stops (ND2-ND8), this works well. At high stops (ND32+), two problems emerge: cross-polarization (the overlapping polarizers create rainbow patterns, especially on reflective surfaces and water) and color temperature shift (the filter introduces a blue, green, or magenta cast that varies across the frame).
Manufacturers rarely publish this data. So we measured it.
Our Test Method
White balance target (X-Rite ColorChecker) photographed at each stop from ND4 to ND1000, at 5600K constant light source. We measured color temperature deviation from the target in degrees Kelvin, and cross-polarization on a reflective aluminum panel at each stop.
Filters tested: 8 popular variable NDs from $35 to $280, tested on Sony A7IV and Canon R6II.
Results
Below ND16: All filters performed acceptably. Color temperature deviation under 200K, no visible cross-polarization in most conditions.
ND32-ND64: The separation. At ND32, filters below $80 began showing visible cross-polarization patterns on reflective surfaces. Color temperature deviation exceeded 400K on several budget filters.
ND64 and above: Only 3 of 8 filters maintained neutral color. The rest introduced shifts of 600-1200K (read: your footage will look greenish or magenta).
Best performers: Breakthrough Photography X4 (ND32, $230), Haida M15 II (ND32, $180), and Freewell VND (ND32, $85). The Freewell is the value winner — its color neutrality at ND32 matches filters 3x its price.
Recommendation
If you shoot above ND32 regularly: buy two fixed NDs (ND8 and ND64) instead. Variable NDs above ND32 are unreliable for professional work. If you need a variable ND, limit its use to ND32 maximum and test it on a color checker before every shoot.