Cv1 feels about 10% better than DK2. The hype train on CV1 is ridiculous...
After the crap Palmer Lucky & crew have pulled the past months... closing Oculus to Rift only users is terrible marketing plan & above all ... the system is still relying on optical tracking. I've had the Rift a few hours and feel I'll be selling it on FleaBay and order myself the HTC Vive. Valve and HTC are open to all dev. platforms & are simply embracing VR not trying to corner of a section of the market. I've tried all systems from DK1,Dk2 CV1 & now the Vive & it's clear that the Vive shines above the CV1 with or without the controllers... Vive can easily be used in a seated manner as well. Why restrict myself.
You can't easily do finger tracking with the Lighthouse, because you'd need to wear gloves packed with sensors, at least one for each joint. That's the tradeoff between the two systems.
That said, while my Leap Motion is quite good at tracking my fingers, it isn't perfect, and suffers from inevitable occlusion problems due to only having one camera. It may work better when I actually have a headset to stick it to, as I've currently only been able to test it by holding it in front of my face.
The Leap Motion actually has two cameras in it. The Diagnostic Visualizer found in the Leap Motion Control Panel shows both camera views, which can be verified by hovering a finger over each half of the sensor.
I've heard that the only system that can possibly do finger/hand tracking is the CV1's constellation system. I've heard that the Vive's light house isn't accurate enough at all for that sort of thing and that is why the Leap Motion uses the same sort of tracking the CV1 uses.
Can you confirm that?
The Leap Motion uses a stereo camera rig inside its sensor. CV1 wouldn't be able to adequately support finger tracking unless the tracking camera was upgraded to a stereo camera. The reason the DK2 and CV1 Rifts can get away with one camera is because the headset and IR emitters on it are of known dimensions and placement. Measuring the distance between two emitters is enough to calculate distance from the camera.
On the other hand, everyone has different sized hands which change shape as joints are flexed. Because of that, the Leap Motion needs two side-by-side cameras to help get the needed depth information, and even then the Leap Motion software has quite a bit of trouble reliably tracking fingers. Occlusion is a huge problem with the Leap Motion. To counteract that sort of thing would require additional cameras, which would increase bandwidth and processing requirements considerably.
For now, gloves like the Manus VR are probably the best bet for reliable hand tracking. Overall hand tracking can be done using either Constellation or Lighthouse, with the finger positions being tracked by sensor strips that run along the fingers. No extra cameras would be needed, and since the sensors directly represent fingers, there's no heavy processing needed to figure out what the sensor data means. The values reported by the sensors can be used more or less as-is.
What makes me laugh is we have three reviews saying three different things about CV1. Just look over the forums and you'll see what I mean. We have people saying: "CV1 FOV is smaller. CV1 FOV is Bigger. CV1 FOV is the same as DK2." I think everyone should just judge for themselves, because going by these reviews the FOVs all three and that's impossible LoL.
