Mine also came today. Its a really cool design and well put together so I’m excited to get it dialed in. I also have a JoskK sphere for comparison, so I ran some quick numbers as well
Light/host | Spec | Sphere | TA Tube | Diff |
HDS 170 | 170 | 168 | 234 | 71% |
HDS 200 | 200 | 203 | 327 | 62% |
ReyLight Lan | N/A | 234 | 327 | 62% |
Swm V10r 219c | N/A | 315 | 474 | 66% |
Haiku 219c | N/A | 390 | 609 | 64% |
Zebra SC5c mk2 | 475 | 427 | 648 | 66% |
Hanko 219b | N/A | 615 | 890 | 69% |
Zebra SC62w | 930 | 891 | 1362 | 65% |
Zebra SC63w | 1150 | 1086 | 1748 | 62% |
Zebra SC600w mk2 HI | 1126 | 1120 | 1738 | 64% |
So a few important points:
- Like Jason I didn’t bother to test cell voltages because we’re comparing measuring devices here, looking for relative numbers, so it really shouldn’t matter just to try and find a trend. But maybe it will when trying to compare to other users’ numbers so I may need to re-do this when I have more time.
- The HDS lights are the most interesting to me, because according to Henry at HDS each and every single light is custom calibrated on his sphere to give the promised output. He tweaks the levels in the firmware so that the lumens are exactly as promised, so they should make pretty good “reference” lights for these sorts of tests. Unfortunately (for this application) I’ve modded most of mine which invalidates that work, but these two are still stock.
- I’ve included “spec” numbers for factory, unmodified lights. They’re just there for reference and not included in the math. Lights marked with N/A are modded in some way so factory numbers, if they ever existed, would be useless.
- I’m concerned that my differences don’t show a more precise trend. A range of over 9% difference, between the lowest and highest variations, seems like a lot to me. I suspect my “technique” is to blame there which makes all my numbers questionable. I’m still trying to figure out the best way to use this thing.
- I didn’t include any “high-powered” lights for this test because the JoshK sphere can’t handle them, so I’d have nothing to compare to. If the tube’s response isn’t linear then it may not be off as much at the higher end. I have no way to test that. I did order those numbers above in lowest -> highest order in the hopes of seeing some sort of trend in the differences, but I don’t.
- More data is definitely needed but I think its safe to conclude that the tubes are reading high, its just a matter of determining how much high. Mine seems to be up by ~35%. I’ve not yet tried to bring it down in any way and look forward to what some of the rest of you come up with.