I see ~20% drop in output when sliced. This is a rough measurement (celling bounce) and doesn’t seem to matter whether you slice at 1.2 mm or completely dedome.
I also see ~30% reduction in hotspot width when sliced. Again, the difference from any slice to dedome is very small. Once you make the surface flat (slice or dedome) I see this change.
These are estimates and it will depend on reflector or TIR design. The CCT and duv changes are much more noticeable but I do like the tighter beam.
FB
20% drop is still kind of a lot. I don’t dedome or slice unless it’s absolutely necessary. In fact I glued most of dedomed ones back together just to restore their original output lol.
Your ceiling bounce measure could be off. Since the ceiling absorbs some light and a tighter, more energetic hotspot should results in a larger loss due to absorption.
I’m not sure if a tighter hotspot causes more light to be lost via absorption–the beam profile does nothing to the albedo of your ceiling. In fact, the error is the other way around: when performing a ceiling bounce measurement, a throwy beam will get a higher reading than a floody beam at the same output as more of the light will bounce straight back rather than to an oblique angle that does not reach the sensor.
In TIR-based lights, another consideration is that smaller LES results in higher transmission efficiency, which partially counteracts the loss from dedoming.
I want a cool-neutral quad, would 519A 5000K domed or 5700K half-dedomed look better? They would be similar CCT but the half-dedomed would be more rosy (-0.004 duv?) As I know the lower duv would look better for warm and neutral white but not sure if that’s true for the 4500-5000K range.
I have a single 5700K sliced. CCT is lower than 5000K and duv can be as low as -0.0060 on turbo, but it looks nice. The main point you should consider is that slicing may result in some artifacts under clear optics. 5000K and green AR lens maybe can fit your needs.
Maybe mixed leds for rosier 5000K outcome? Like 2x5700+2x4500.