| Main > Lightguns |
| Lichtknarre: Unmodified Wii remote as a sight accurate Lightgun using 2/4 LEDs |
| << < (10/68) > >> |
| lilshawn:
just my 2 cents regarding the peculiarities of aming/aligning/seeing the appropriate LED's... your "viewing angle" is really bad... the led shape has a drastic effect on how much of an angle you are going to be able to see them at. typical "normal" shaped LED's have about a 10 or 20 degree viewing angle. a shape commonly named "straw hat" LED's have about 90 ish degrees of emission... where a flat topped and even inverse cone shaped LED lenses have a viewing angle gaining on 160 to even 180 degrees or more. the original wii sensor bars used 5 "regular" LED's on each side ... but arranged in an arc to improve the angle in which the IR "dots" can be seen. when going with a single source IR emission like this, you should probably look into "straw hat" or flat top LED's to improve the angles in which the sources can be seen by a camera. |
| RandyT:
--- Quote from: Fusselkroete on November 03, 2022, 11:04:28 am ---tbh i think normal 980nm IR lights which are in the wii sensor bar could also work fine enough for normal size monitors/tvs. The video i made was only my way i did it. :D Some ppl told me that 980nm works fine. I used high power LEDs because i use this tool in beamer/projector from far distance. But i didn't validated this. Your choice xD --- End quote --- You are on a site where people take games which can be played on a cell phone and put them into huge, furniture-sized boxes. If something can be done to excessive levels, we'll usually take that route. Seriously though, the wii lightbar is more powerful than a lot of people probably think it is. There are a total of 10 LEDs in that thing (4 of which aid in angles of incidence) just to give it two blobs to work with. There's also no such thing as a perfect filter, so if there's a 15% loss from the one in front of the LEDs and an additional %15 from the one on the controller, there's a 30% reduction of the available IR right off the top. Then, if you run everything at the longer wavelengths, the camera becomes less sensitive by about half (based on the IR sensitivities of most silicon-based sensors). So, my current belief is that, if people are willing to deal with the dim red dots, performance and range can be enhanced by a decent margin with your kind of setup. But yeah, unless one has issues while playing, it's probably not necessary to go too crazy for a simple setup where the LEDs are exposed and the screen isn't huge. :) --- Quote from: lilshawn on November 03, 2022, 12:45:21 pm ---just my 2 cents regarding the peculiarities of aming/aligning/seeing the appropriate LED's... your "viewing angle" is really bad... ... when going with a single source IR emission like this, you should probably look into "straw hat" or flat top LED's to improve the angles in which the sources can be seen by a camera. --- End quote --- I mostly agree. However, the particular LEDs we've been tinkering with have a stated viewing angle of 140degrees, so at any usable distance from them, they should be good-to-go for this type of application. |
| lilshawn:
--- Quote from: RandyT on November 03, 2022, 01:00:14 pm ---I mostly agree. However, the particular LEDs we've been tinkering with have a stated viewing angle of 140degrees, so at any usable distance from them, they should be good-to-go for this type of application. --- End quote --- yeah 140 degrees... but you can see in the first pic you posted that the top one's emission has become barely "visible". so they probably have in the datasheet that it's 80% in the first 20 degrees of angle... but could be down to 1% by the time you hit 140 degrees. manufacturers are weird like that. this should be easily doable with four 5mm LED's running at 30ma.... with 3 watt LEDs... your camera should be literally BLINDED by them. just spitballing here... i wonder if you placed a small square of frosted plexi (about the size of a square of shredded wheat) in front of the LED... if it's detection would be better. |
| RandyT:
--- Quote from: lilshawn on November 03, 2022, 01:27:02 pm ---yeah 140 degrees... but you can see in the first pic you posted that the top one's emission has become barely "visible". so they probably have in the datasheet that it's 80% in the first 20 degrees of angle... but could be down to 1% by the time you hit 140 degrees. manufacturers are weird like that. this should be easily doable with four 5mm LED's running at 30ma.... with 3 watt LEDs... your camera should be literally BLINDED by them. just spitballing here... i wonder if you placed a small square of frosted plexi (about the size of a square of shredded wheat) in front of the LED... if it's detection would be better. --- End quote --- If the manufacturer's graph is to be believed, these should be giving 35% at 140 degrees, which isn't bad. A couple other data points are 50% at 120, 75% at 90 and it goes up from there. I also don't know where the 3watts comes from, unless that is the maximum they hit before they catch fire. I've always thought P=VI, which in this case comes out to 1.26w = 1.8v x 700ma, and the goal is not to run them that hot, both to maximize longevity and minimize the need for cooling. I think the sweet spot is where you start to see diminishing returns. I just tried some good diffusing material (not saying it's best for this) and wasn't able to achieve anything at any angle which didn't just make a softer, but dimmer spot. Even at very acute angles, my cell camera with a good IR filter on it picks up the naked LED very well. |
| RandyT:
Finally got this set up on a 50" plasma for testing. Being wireless and sight accurate is great. After setting up the LEDs with my dumb little program, and with no IR offsets (all left at 0), the cursor sits right in the sights of the gun, pretty much anywhere on the screen. I'd say it was at least as good as a traditional lightgun for accuracy. But I am seeing an issue. Everything which follows was done watching the mouse cursor at the desktop, so no game calibration or influence is at play. I'm seeing an oddity, possibly related to the angle compensation code. If I pan the screen quickly and repeatedly to the left and right , I see the blob pattern in the preview window tilting clockwise when moving left and counter-clockwise when moving to the right, even though I am being very careful not to tilt the gun. This incorrectly perceived tilt by the software seems to trigger the tilt compensation, which then gets confused and seems to place the cursor randomly on the screen as the motion continues and the software seems to become more confused. The interesting thing is that this does not occur when doing the same thing slowly. As long as I don't exceed a certain, somewhat slow speed, the pattern stays dead straight and the cursor stays where it should. I do not expect that this is linked to Bluetooth or system speed issues, as I can crank up the inputs per second cutoff to 120, which makes the cursor move extremely smoothly, but with no change in this odd behavior. I should also note that distance does not seem to make a difference. Fusselkroete, is it possible to add an option to disable the tilt compensation code entirely? While this may not be the root cause of the issue, it seems like the lower speed at which it operates causes error to compound, leading to the cursor ending up momentarily in places it shouldn't. I'm also seeing an occasional, but also rare, minor offsetting of the cursor. Usually, just going off the edge of the screen and coming back will correct it, or it will sometimes just manage to fix itself. No idea as to what might be causing this one. It almost feels as if the cursor ends up "stuck" in a small offset and then eventually frees itself. And as long as I'm asking for stuff, I'd also like to see a troubleshooting feature added. Namely, a basic, full-screen preview. It would be handy for larger screens to be able to double-click the preview window to bring it to full-screen so it's easier to see what is happening at longer distances from the screen. The code seems already to be present for this, so I wouldn't expect that it would be too difficult to add and it would be a nice feature. One final oddity I found. In the 3. Default Gun calibration mode, I was only able to shoot one of the blue squares. It would progress to the next screen, but it would no longer register a button press to proceed. If I used the mouse to exit the mode, the trigger would again function normally. This seems like a bug, unless I am doing something wrong here. I did as the screen stated, having only two of the proper color dots on the screen, but could not get past the screen no matter what I did. The 2. default LED-calibration was no problem at all. Edit: Whoops, forgot one. I also noticed that through some currently unknown combination of events, the blue dots in the preview window disappear, and only re-appear when less than 4 are present. Whenever there should be 4 dots displayed, it shows only the numbers and no dots. A re-boot of the software will fix this temporarily, so it appears to be a bug of some sort. |
| Navigation |
| Message Index |
| Next page |
| Previous page |