ND2 is 1 stop and ND 400 is about 8 2/3 stop.
So the first marking is ND2 follow by ND4, ND8, ND16, ND32, ND64, ND100, ND256, and ND400. There should only be 9 marking on the ring.
But there are 13 marking on the ring; how do I use them correctly?
As with any variable neutral density filter the numbers are only approximate.
There are, in fact, 10 common ND ratings between ND2 and ND400: ND2 (1-stop), ND4 (2-stops), ND8 (3-stops), ND16 (4-stops), ND32 (5-stops), ND64 (6-stops), ND100 (6 2/3-stops), ND128 (7-stops), ND256 (8-stops), and ND400 (8 2/3-stops). But I wouldn't put any money on any of the 13 marks being exactly any of those 10 common densities.
A variable ND filter is basically two polarizing filters stacked on each other. The angle of polarization between the two filters determines how much light is allowed through.
So how do you use them correctly? You take good notes of which mark the ring was set when you took each photo along with what the scene metered before you put the filter on, what lens, focal length (if a zoom), and aperture you were using, etc. until you get an idea of how much the filter reduces the amount of light getting to the camera when set at each mark.
Years ago, the photo community set the f/stop increment as a doubling or halving of the exposing light energy. In other words, one f/stop = 2X and two f/stops = 4X etc.
We mount filters to gain an enhancement. The filter always blocks some percentage of the exposing light, and somehow we must set the camera to compensate for this light loss.
To aid in this regard, filter makers assign a filter factor (FF). The FF is a multiplier. To use we start by selecting a shutter speed without the filter applied. Next we multiply this value by the FF. As an example: Shutter speed without filter is 1/400 of a second. We mount a filter FF 4. The math is 1/400 X 4 = 4/100 = 1/100. We have discovered that we must slow the 1/400 of a second shutter to 1/100 of a second to compensate for the light loss induced by the filter. Now multiplying the fractional shutter speed by the filter factor is not easy; most of us have forgotten the math of fractions.
Another way to handle the FF is to count on your fingers by 2’s. Thus 2 – 4 – 8 – 16 -32 – 64 – 128 – 512. FF 2 = 1 finger – FF 4 = 2 fingers – FF 8= 3 fingers – FF 16 = 4 fingers etc. The number of fingers represents the number of stops of compensation needed. Mount a FF 8 filter, that’s 3 fingers = 3 f/stop compensation.
Another way to handle the FF, if using a hand-held light meter, is to divide the ISO by the FF and reset the meter to this value. If your ISO is 400 and you mount a filter FF 4, then we reset the handheld meter to 400 ÷ 4 = 100 ISO. This method allows the hand-held meter to do all the work.
In modern times, the camera features through-the-lens metering. In other words, we mount the filter, it reduces the exposing light, the camera senses this an automatically resets the exposure thus making the necessary compensation. This is your best bet!
Now the correct sequence for FF is: 2 – 4 – 8 – 16 – 32 – 64 – 128 – 512 – 1024 The shutter speeds which are traditionally based on the 2X increment use a modified set. 1 – 2 – 4 – 8 – 15 – 30 – 60 – 125 – 500 – 1000 The slight deviation from the true 2X increment is of no importance as it is impossible to make a shutter anywhere near the needed accuracy. Best we can do is 1/3 f/stop for both shutter and setting the diameter of the aperture. We must deal with gear train back lash and mechanical linkage.
What I am trying to say is, there should be only 8 FF markings or setting in 2X increments 2 thru 400. I think a variable ND made by rotating one of two polarizing filters is a crude approach that is difficult to calibrate. I think it won’t make much difference because you will be using this item with thru-the-lens metering and the camera automation will solve the exposure math for you