DoglegDave
Active Member
- Messages
- 320
- Reaction score
- 0
- Location
- Houston, TX
Spent the better part of three weekends chasing "accuracy issues" on my camera-based unit before I finally figured out it was my garage lighting the whole time.
Backstory: 12x20 garage, two 4ft LED shop lights I grabbed at Costco, plus whatever light bleeds in from the door. Every 6-8 shots the thing would throw a no-read. And when it did read, driver spin was bouncing around by 800-1000 rpm between identical swings. I'm sitting there reading forum posts about firmware updates and radar interference like an idiot.
Then a buddy who does machine vision for a living came over, took one look at my setup, and laughed. Camera-based monitors are basically taking high speed photos of a ball that's moving 150 mph. Not enough light = motion blur = garbage reads or no read at all. Too much directional light from one side = hard shadow that the ball detection can't tell apart from the ball. Same reason a white hitting mat will wreck a lot of these units, the contrast is garbage.
What fixed it for me:
- Ditched the two mismatched shop lights for four 5000K panels spread evenly overhead, no single hotspot
- Pulled the mat back so the ball sits in even light instead of half shadow
- Put a dark gray mat under the hitting strip so the white ball actually pops
No-reads went from like 15% down to maybe 2%. Spin numbers stopped wandering. Same unit, same firmware, zero dollars on new hardware.
I get why nobody talks about this. It's boring and it doesn't sell launch monitors. But if you're getting weird numbers indoors, check your lighting before you blame the unit or ship it back. What are you guys running over your hitting bays? Wondering if I'm overkill now or if there's a better bulb temp for these cameras.