Both lights are compared in 2xAA round-up thread:
https://www.candlepowerforums.com/threads/214890
Realistically, there's isn't a huge difference in the performance or beam - although the LD20 is a bit more focussed with a slightly narrow spillbeam width.
As for the reduced runtime on Turbo, that may have more to do with variability among samples (especially for emitter Vf, etc.). I notice on mine that there is little output/runtime difference on Hi mode, for ex.
P.S.: You can't take % differences of raw lux, since it is a non-linear scale (i.e. light decays by an inverse square law). So for example, a light that has 4000 lux at 1m would only have 1000 lux at 2m - not 2000 lux as you expect if it were linear. So, you can't use % changes, or you will incorrectly amplify the difference (in the example above, the light is NOT 4x brighter at half the distance, obviously). To get accurate results, you need to compare the square-roots of raw lux data for % difference calculations.
https://www.candlepowerforums.com/threads/214890
Realistically, there's isn't a huge difference in the performance or beam - although the LD20 is a bit more focussed with a slightly narrow spillbeam width.
As for the reduced runtime on Turbo, that may have more to do with variability among samples (especially for emitter Vf, etc.). I notice on mine that there is little output/runtime difference on Hi mode, for ex.
P.S.: You can't take % differences of raw lux, since it is a non-linear scale (i.e. light decays by an inverse square law). So for example, a light that has 4000 lux at 1m would only have 1000 lux at 2m - not 2000 lux as you expect if it were linear. So, you can't use % changes, or you will incorrectly amplify the difference (in the example above, the light is NOT 4x brighter at half the distance, obviously). To get accurate results, you need to compare the square-roots of raw lux data for % difference calculations.
Last edited: