PDA

View Full Version : Devices to enhance Kinect capability for mocap



Ryan Roye
01-24-2014, 07:59 AM
Something I'm going to be working on over Febuary in regards to Nevron is basically to answer this question:

"What devices could one use to enhance motion capture even further? Are there devices or parts of devices (preferably low cost) that could be feasible and practical for the animator's workflow?"

Currently, I'm looking at the feasibility of using a Wiimote motion plus (because you can get it insanely cheap) to allow motion capture that tracks the rotation of the user's wrists and head, as I feel it would greatly cut down on hand-animated "post work" required after recording full body mocap. If it works in a usable fashion, it'll mean for about $40 users of Nevron can add more rotation tracking capability to their kinect setup. I will report my findings here after I receive my hardware in the mail. I know a bit of working around will be required to get this functional and the possibility it may not work at all so we'll see.

I do know the PS move devices work well from what others have said... but from what I understand that route is incredibly expensive and only really benefits people who already own a PS3 and the corresponding camera. EDIT: Ok, so from the post below by Greenlaw, it appears only the PS move controller is needed... still more pricey than the wii motion controller, but knowing this makes it much, much more feasible.


OTHER DEVICES:

-GAMEPAD:

This can be used to control the character's hands if pre-animated constraints like cyclist are used. Thus far, my tests with some junker playstation 2 controllers (with a USB converter) have yielded some pretty nice results as they are decent quality devices... the most useful of course being the thumb axis rotational sticks. My current idea is that, just like with facial capture, hand animations driven by controllers could be made much easier to create if the mocap clip is playing at 1/3rd of its original speed... this gives the user more time to react and manipulate the character appropriately.

That's the only device i've really tried out thus far... the only other devices I have only return boolean values so i'm not sure how I could make use of that. If I could find a way to refresh layout and make the calculations apply repeatedly over time, theoretically booleans could be used, for instance, to make it so that when a button is held down, a value will increase until the button is let go, then it will decrease until it reaches the "floor". I don't know how to make this happen yet or if its even possible though...

Greenlaw
01-24-2014, 10:19 AM
I think this could be a very informative thread. Thanks for starting it.


Currently, I'm looking at the feasibility of using a Wiimote motion plus...to allow motion capture that tracks the rotation of the user's wrists and head...I do know the PS move devices work well from what others have said... but from what I understand that route is incredibly expensive and only really benefits people who already own a PS3 and the corresponding camera.

Technically speaking, you don't need the game console or camera to use PS Move if you're only recording rotation--all you need is a bluetooth adapter. That's how iPi Soft did it for their mocap system to record head and wrist rotations, using only the PC and a cheap USB bluetooth adapter. For mocap you don't need the position data because you already have it in the bones position--all you need is to overwrite the rotation data. I don't know how you would do this for LightWave but I imagine it's possible. BTW, iPi Mocap studio can use the WiiMote in the same way. The PS Move's magnetometer is more accurate though. To record wrist motions, some users have made gloves that hold the device to the back of the hand--I prefer to hold mine because I like to use the buttons to remote control the software. For the head rotations, I modified a hard hat--you can see the rig here:

My DIY Mocap Helmet (http://littlegreendog.blogspot.com/2013/04/mocap-helmet-update.html)

The PS Move data is recorded simultaneously with the body capture. After the body motion has been tracked and edited, you simply apply the rotation data from each device to the appropriate joints (I use three, one for each hand and one for the head.) It works surprisingly well--I used it recently for background character mocap for a feature film, as well as in our Brudders music video.

Here's another option: there's a video on YouTube that shows a user recording rotation in LightWave using an iPod Touch:

Using the iPod Touch in LightWave (http://www.youtube.com/watch?v=rfZe6UQ7yWg)

The Myo looks really cool. It's an arm band that uses EMG sensors to read what your muscles are doing and it's supposed to accurately interpret forearm, wrist and finger movements. iPi Soft has been looking into that but the device is not officially out yet. Soon though and it's surprisingly inexpensive. Here's more info:

Myo (https://www.thalmic.com/en/myo/)

The existing Kinect system for LightWave Nevron Motion shows interesting possibilities for alternative data input. I've been curious to use it puppet other channels in LightWave besides the usual ones but haven't had time to explore this yet.

G.

Ryan Roye
01-24-2014, 10:54 AM
Thanks for the wealth of info! And the DIY helmet setup is really slick and I'll see if I can dig up my old hardhat from the days when I worked as a road traffic operator (basically, stand there and hold a stop sign for 6+ hours). If not, I know someone who has a bunch in their closet :)

I'm mainly seeing the feasibility of wiimotes first due to their low cost. The biggest question there would be whether or not it is good enough to simply track the bank rotation (or calculated equivalent via formula) of the user's hand and nothing else... I can settle for less accuracy if the results are still usable. Of course, I'll also test it for head rotation as well. I got my test Wii motion plus as low as $14. If I get usable results, I'll post a video showing my findings.

Greenlaw
01-24-2014, 01:33 PM
Sure. I'm hoping others will add to this thread--it could become a very interesting and useful resource.

Re: the hard hat--it's really cheap. I bought this one at a hardware store for about $6 I think. I had to track down the chin strap on the web though--with shipping it probably cost more than the hard hat itself.

Another iPi Mocap Studio user found a removable suspension that you can wear without the outer shell, and he just tie-wraps the motion device to that. It's probably less bulky but the hard hat version I have works well for me.

Tip: if you use this hat with Kinect, paint it a flat gray so it doesn't bounce the IR rays. I unfortunately used a paint that doesn't bond to the plastic very well, and the paint is chipping and flaking now. I think Krylon Fusion spray paints will bond to plastic--I'm going to try that soon for the hard hat and for painting 3D printed parts.

My bicycle helmet actually worked better, and it looked cooler too, but I needed to use it as, well, a bicycle helmet. :p

Here's what that version looked like:

DIY Mocap Helmet - Bicycle Helmet Version (http://littlegreendog.blogspot.com/search?q=bike+helmet)

G.

Greenlaw
01-24-2014, 02:01 PM
I think any wrist motion should look better than no wrist motion (i.e. 'paddle hands',) so long as the wrist is moving in a natural direction and with reasonable limits. With iPi Mocap Studio, it's remarkable how much realism you gain by including the wrist and head rotations.

Here's a little more info about how it works in iPi Mocap Studio. When recording with a Move or Wii device, you need to point it directly at one of your Kinect devices for a few frames. This gives the software a reference for device heading, relative to the scene and performer's arm/head. Then, after the body has been tracked, you parent a 'virtual' version of the device that has the recorded motion to the appropriate skeleton bone (hands or head) and orient the device to how it was positioned in the real world relative to that body part. You can do this on any frame where the correct orientation is very clear (having the RGB video data from the Kinect is helpful here.) When you're ready, simply click the Apply button and the data is transferred to the bone the device is parented to. Play your animation to check the results. Usually, it should work fine the first time you apply it but if the rotation axis looks off, you can make an adjustment and apply it again--the software will overwrite the previous motion for you.

I'm not sure how this method can be translated for Lightwave/Nevron Motion but it works very well in iPi Mocap Studio.

G.

spherical
01-24-2014, 04:29 PM
We have yet to get into NM. Right now, my head is swimming with all of the terms and concepts being thrown about, but I'm sure things will fall into place once some experience has been gained. A couple of initial questions, though. We have a PS3 and a Move Controller. Up above you mention a Bluetooth adapter. Adapting from what to what? Do you have any shots of the wrist set up and can more than one Move Controller be used at the same time?

Greenlaw
01-24-2014, 05:57 PM
Oh, sorry...I was speaking theoretical as far as LightWave/Nevron Motion goes. The whole bluetooth/PS Move/WiiMote thing works with iPi Mocap Studio, not Nevron Motion.

I haven't spent much time with Nevron Motion yet but I believe the only direct input currently available is what's available from the Microsoft API. I don't know enough about LightWave's Virtual Studio to tell you what's possible there either.

G.

Greenlaw
01-24-2014, 06:19 PM
But as far as how it works in iPi Mocap Studio, it's as described above. I don't have a video of a workflow posted online but the B2 Music Video Excerpt (http://vimeo.com/channels/littlegreendog/68543424) we released last fall features wrist and head motions captured with 3 PS Moves for each character. I have to admit the wrist movements aren't very interesting because the characters are performing mostly repetitive motions, but you can see what it did for head motions fairly well.

A much better example would be the stuff I did recently for background characters in a film but I can't show that. When I have time, I'll make a quick workflow video with a more generic example.

G.

Edit: wait, I take that back. I have a partial workflow video online. The Sister Mocap Test (http://www.youtube.com/watch?v=jASC8IOsIqY) from last spring shows a little bit of the PS Move in action. If you look at the live action in the first part of the video, you can see that I'm holding two PS Move controllers. The bluetooth adapter is on the end of a repeater cable on the floor near the capture space. The wrist motion was applied to the character in the second and third parts. It's not perfect because this was literally the first time I used the feature and I didn't know what I was doing. If I was doing it again, I would correct the orientation and reapply the motion, but at the time I was just excited that it worked at all.

I don't have the head rig in the video because that feature was added a couple of weeks after I shot this.

LW_Will
01-25-2014, 11:46 PM
Hmm... seems like some Duck Tape would do okay to me. Just a bit sticky after the chin... ;-)

My one question is, how can I adapt the output of LW to work on my iDevice or Nexus7? If I could make a floating camera rig, THAT would be amazing!

Greenlaw
01-26-2014, 01:47 AM
If you manage to make your iDevide or Android device work with LightWave, I think you will only get rotation data, no translation. Apart from the current LightWave Virtual Studio system using a PS Move and game console, I'm not sure. Without the console, you can get device rotation but the tricky part is getting translation, which is what the game console and cameras are doing. Without that, you will need an additional controller to navigate position.

A few years ago in the Box, our programmer experimented with using a Wii Motion Plus in Maya to behave like a camera. The Wii uses its game console and an infrared sensor to capture translation. This system has potential as an alternative to PS Move console and cameras and it's a lot cheaper. Anyway, he somehow got it working in Maya and then the motion was ported to LightWave. I don't know if you can use Wii Motion Plus with LightWave but I imagine some smart person could probably make it work with Virtual Studio. The downside is that the Wii is old tech and I don't think anybody will want to put much effort into this. Sorry, I have no further information about this.

The Space Navigator gives you movement in all six directions (XYZHPB) and it does work with LightWave right out of the box. It's not quite the same thing though because it doesn't have the 'handheld' interaction you get with a PS Move. Maybe somebody could combine the two devices? That is, use the PS Move with bluetooth to capture rotation and use the Space Navigator (or even just the keyboard) for XYZ motions? Just a thought.

In the foreseeable future, I think the best bet for a 'low cost' handheld virtual camera, beside the PS Move with console, might be Leap Motion, or maybe even the Kinect 2--but it's too early to say.

I have to confess, I actually enjoy setting up my own camera choreography, so virtual camera isn't for me. But I know a lot of other users have been asking for this for some time.

G.

Greenlaw
01-26-2014, 02:01 AM
I just remembered this:

https://www.lightwave3d.com/assets/plugins/entry/w_gamedevice/

Don't know if it still works but maybe somebody will want to try it out.

G.

LW_Will
01-27-2014, 04:12 PM
Looks very cool... except that it doesn't use the Virtual Studio. ;-(

Also, the thing I want is the monitor on the iOS/Android (N7) to show a (presumably VPR) view!

I've seen a video from 3rd Powers, awhile ago, where he was manipulating LW on a tablet AND a screen. I want to do that! ;-)

dwburman
01-29-2014, 08:53 PM
We have a PS3 and a Move Controller. Up above you mention a Bluetooth adapter. Adapting from what to what? Do you have any shots of the wrist set up and can more than one Move Controller be used at the same time?

The bluetooth adapter would be to use the Move Controller without the PS3 by connecting it directly (wirelessly) to the computer like you would a mouse or keyboard or any other bluetooth device.

- - - Updated - - -


I just remembered this:

https://www.lightwave3d.com/assets/plugins/entry/w_gamedevice/

Don't know if it still works but maybe somebody will want to try it out.

G.

I used that a bit a few years ago, but I don't think you need it anymore with the virtual studio tools. It might be slightly easier to set up, though.

spherical
01-29-2014, 11:07 PM
The bluetooth adapter would be to use the Move Controller without the PS3 by connecting it directly (wirelessly) to the computer like you would a mouse or keyboard or any other bluetooth device.

Ok, so the PS3 isn't needed at all or, if you have one, does it help any?

Greenlaw
01-30-2014, 10:41 AM
As explained above, the PS3 is needed for capturing positional data of the PS Move controllers. This is done using the PS3 Eye cameras to track the illuminated balls I guess. (Sorry, I don't have a PS3 so I have no experience with this.) In theory, because all of these devices can connect to a PC, I think somebody could write software to track the balls on a PC while simultaneously recording the rotation data by bluetooth but I'm not aware of anything like that. (In the case of the Wii, it does this using the IR bar that sits in front of you.)

What iPi Mocap Studio does is it records only the rotation data from one to three PS3 Move controllers via bluetooth directly to the PC, while at the same time it records a point cloud of the body with one to three Kinects or up to eight PS Eye cameras. It doesn't need the PS3 or Wii consoles for this, just the PC and a bluetooth adapter because the PS Move and Wii Motion Plus are bluetooth devices. Then, after a skeleton has been tracked, you apply the rotation data to the wrist and head bones. There is no need to record positional data in this case because you already have the position from the bones.

The position data is needed for 'free floating' objects, like a virtual camera.

I guess theoretically speaking, you could mocap a person to for the positional data and then apply the 'camera' rotation but IMO, the data isn't nearly accurate enough to make this work for accurate camera motions. (Not that there's a setup for that anyway.)

To summarize, it appears the best low cost solution for LightWave for a virtual camera that behaves like a real hand held camera (rotation and position) is still the PS Move with a PS3 console. You can use cheaper controllers like a gamepad or Space Navigator but obviously that's not the same as using a 'hand held' camera--it's more like running a video game.

Then there's the rotation only method (a device with accelerometer or magnetometer, like iPod Touch, Wii Motion Plus, or PS Move) which can get some natural motions but with obvious limitations (but you might combine this with another controller or maybe a 'nunchuk' attachment for position.) I don't know anything about how to do any of this though.

G.

spherical
01-30-2014, 06:01 PM
Hmmmm, may be too early to tell but I would seem that iPi would have been a better choice than NM. Didn't do enough research.

Greenlaw
01-30-2014, 06:34 PM
It depends on what you need. For human motion capture with multiple Kinects, I'd say yeah, of course, but also keep in mind that iPi Mocap Studio costs a lot more than Nevron Motion and it is not a mocap editing tool (like Nevron Motion/Genoma.) iPi Mocap Studio also does not have face capture, nor does it allow you to input Kinect data directly into LightWave's channel inputs the way Nevron Motion can.

Ideally, you might use both--but I would talk to somebody who is actually doing that for accurate information. (In my case, I'm using iPi with Motion Builder to LightWave, but a friend of mine says iPi plays well with Nevron Motion on his projects.)

G.

erikals
01-31-2014, 02:56 AM
i'm just waiting for Myo...

http://forums.newtek.com/showthread.php?139587-Myo-5-finger-motion-capture

Greenlaw
02-05-2014, 05:31 PM
Just got an email from Thalmic labs announcing that they are shipping Myo alpha units to developers now. Yay! It's not vaporware! :)


All kidding aside, this is exciting news.


G.

Ryan Roye
02-07-2014, 08:28 AM
Just an update for you all. I tested out the Wii remote and wasn't able to get it to even connect via bluetooth despite looking up the compatibility lists beforehand and testing out all of the various device drivers (called "stacks" apparently). So, either the USB bluetooth is the issue, or it is the fact I purchased a 3rd party Wii remote. I will have a second USB adapter coming in the mail soon which might work (longer range one basically), but I'm skeptical at this point. If it works great, if not, losing $15 isn't going to kill me :)

Since i've gathered a bunch of data on the PS3 move controller, I also have one coming in the mail; it seems you can get them fairly cheap on Ebay as well. If it works as expected, I may be able to put out some interesting test videos with it... hopefully!

Ryan Roye
02-11-2014, 10:53 AM
UPDATE:

I got my ps3 move controller in the mail yesterday (~$24 on ebay) and got a chance to try it out. I have mixed results:

- I cannot get the gyroscope functionality/values to work in a usable fashion... the values seem to operate relative to the orientation direction of the device, but it jumps/jitters between values ranging between 150-14000 degrees. I am using the motioninJoy drivers at the moment, and I cannot get the calibration functions to operate at all.

- I can, however, utilize all of the buttons on the device, including the pressure-sensitive trigger which can provide a means of allowing the character's wrists to rotate gradually as it is pressed.

With a bit more research, I may be able to get the gyroscopic functions to operate, but for now I'll need to settle with the back trigger functionality.

Greenlaw
02-11-2014, 11:04 AM
Hmm. I can only speak of my experience with iPi Mocap Studio but when using the PS Move devices, the user is required to aim each device directly into a Kinect for a few frames to give the software a reference orientation for each. My understanding it that this is for locking in the heading (I guess pitch and bank can be assumed once you have the heading.) I believe they're using their own drivers for Windows 7 but they I think they suggest using MotionInJoy for Win 8. I know that's not specifically useful info but hopefully it will give you an idea or two for your project.

G.

Ryan Roye
02-11-2014, 02:06 PM
UPDATE:

Ok, I got the ps3 move device to work without the PS3 or the eye camera using a program called "JoyEmu.exe". This generates a generic driver that outputs values in a usable fashion including the gyroscope/magnetometer values (minus the jittering and value flipping). It works quite nicely! I'll try and get a demo video up soon.

Greenlaw
02-11-2014, 02:40 PM
That sounds fantastic! Looking forward to seeing your results. :)

Ryan Roye
02-11-2014, 11:01 PM
Results of ps3 move with Nevron without using a ps3 or extra move camera. Works quite nicely. (I cross-posted this on my videos thread as well to keep all the video content in one place).


http://www.youtube.com/watch?v=hLZ3wJCTYyk

Ryan Roye
02-12-2014, 12:27 PM
The next thing I need to verify, which will be sometime in the next week or two, is whether or not this driver will let me use two PS3 move controllers at the same time (it's really hard to avoid having to do excessive cleanup/editing with only 1 controller doing both hands). I'll update again when I get it in the mail.

Greenlaw
02-12-2014, 01:06 PM
That's pretty cool! Thanks for reports and demonstration...and please keep 'em coming. :)

G.

LW_Will
02-13-2014, 10:53 PM
Great work, Ryan. You are OUR E.I.C. ... Explain-er In Chief! ;-)

Also, if you want to get your hands on a Motion Move Controller, you can run down to the local GameStop and pick one up for just $20. I'm going to do it.

Ryan Roye
02-23-2014, 12:00 PM
UPDATE:

Regarding dual PS3 move controllers... after about 10 or so hours of fighting with drivers, lots of programs, and experimentation I cannot get the gyroscopic functions of both controllers working at the same time.

DETAILS

So, to get one controller working properly, I use the following programs:

- MotioninJoy is required to pair up the controllers with a bluetooth device. I have read that you can natively pair the controllers with the windows bluetooth stack (win7), but i've had no luck with it and the general consensus is that it is a real pain in the butt to get working. To be clear, windows bluetooth stack can see my controllers and connect to them, but the PS3 controller can't "pair", making the windows bluetooth stack useless. MotioninJoy's drivers do not produce usable gyroscopic values in Virtual Studio (specifically, the value appears to be read as an 8-bit integer, or 0 thru 255).

- PPJoy is needed to get usable values out of the PS3 move (used in conjunction with motioninJoy). PPJoy + MotioninJoy works great for 1 controller, but the second controller essentially gets ignored... sources i've been able to read up on say that MotioninJoy currently only supports 1 controller's gyroscopic functions at a time. I can get both button sets of both PS move controllers working, but again it is the rotation functionality I'm after.

- JoyEmu.exe works with PPJoy and feeds data into its HID entries. Again, a single controller yields 100% functionality, but the second controller is ignored and the values in both MotioninJoy and virtual studio look frozen in place.


OTHER EXPERIMENTS

- I have also tried other methods in attempt to get dual PS move controllers working. The MoveAPI program works when both PS moves are plugged in, but it works off of the windows bluetooth stack so despite the fact both controllers work with it... I can't unplug them. Persuing this option further would require me to figure out a way to force the PS move to pair with the bluetooth windows stack... which I don't see happening. I have multiple bluetooth adapters and have tried them all (again, the controllers get seen, but will not pair).

For reference, I'm on Windows 7 home premium (64-bit)

EDIT: I want to see if I can inject IPI's drivers into HID; don't expect it to work but its worth a try.

spherical
02-23-2014, 04:15 PM
Looking at this from a different perspective, the symptoms may indicate that it is only seeing one and locking onto it. I mean actually seeing. That the buttons work on both would indicate that it's driver can deal with two but visually locks onto only one and/or thinks that it is large in one dimension, because there are two sources separated by some distance like a two-point poly. So, this leads me to the possibility that changing the color of one of them far enough so that the sensor can differentiate between them may work. I wouldn't think that the color range would be severely narrow, although it probably has limits so that it sees the correct object. If that is the case, changing the color of both, one off in one direction from the original, say yellower, and the other off in the other direction, say bluer, both still within the range, may get some positive results.

Ryan Roye
02-24-2014, 11:49 AM
I was able to utilize the windows bluetooth stack yesterday with the help of ipi recorder (which basically spams your connection until it pairs)... it still was really tedious to do and apparently restarting results in the user having to go through the process all over again. I can confirm from this however that the problem isn't the stack I'm using; it's the software/drivers/setup so that at least narrows things down.

To be clear, I'm not using any cameras; just the PS3 Move controllers by themselves. I only care about the rotational data from the gyroscopes/magnetometer.

Greenlaw
02-24-2014, 12:23 PM
Hi,

I'm not sure what 'spams your connection until it pairs' means but in iPi Recorder you normally only need to pair each device once and then it remembers all the devices for following sessions, unless you change your devices by pairing them with something else. I normally have three PS Moves paired for iPi Recorder and they all work simultaneously.

I'm not sure how they do that though. iPi Soft might be using custom drivers for Win 7, although for Win 8 users I believe they're recommending MotionInJoy at this time. If they are using their own drivers, they probably just haven't gotten around to updating them for Win 8 yet.

Naturally, I don't know anything about this for certain...it's just a guess.

Please keep us posted on your development. This is very interesting. :)

G.

Ryan Roye
02-24-2014, 05:35 PM
Hi,

I'm not sure what 'spams your connection until it pairs' means but in iPi Recorder you normally only need to pair each device once and then it remembers all the devices for following sessions, unless you change your devices by pairing them with something else.

I guess it can depend on other devices that are plugged in or the bluetooth dongle used. Ipi recorder can make the windows bluetooth stack easier to connect to because it basically re-tries rapidly until it succeeds, but at least I know I can stick with the MotioninJoy bluetooth stack since it seems to work more reliably on my end, and I know it isn't the stack itself that is causing the issue; its the drivers and software. This is further evidenced by the fact that test apps i've downloaded easily let me use two controllers, but don't inject any values accessible in virtual studio.

Anyhow, it looks like ipi uses drivers external to HID, so that makes that route a no-go. At this point, it appears I'd need to learn how to work with the PSmove API (https://thp.io/2010/psmove/) in order to get multiple PS move controllers working in Lightwave without a PS3; something that is way beyond my experience level in terms of programming.

For the time being, I'll manage with a single move controller... I'll need to put the goal of multiple moves on the backburner for a while as I need to refocus on some other projects.

Ryan Roye
02-25-2014, 07:16 PM
With the second PS3 move controller put on the backburner for now... I want to experiment with/research something called "air mice", many of which have gyroscopic sensors and are HID-compliant.

A quick ebay search (http://www.ebay.com/sch/i.html?_trksid=p2050601.m570.l1313.TR11.TRC1.A0.H0 .Xair+mouse+6+axis&_nkw=air+mouse+6+axis&_sacat=0&_from=R40) shows they aren't horribly expensive, but of course there is the possibility that they produce unusable results... I wouldn't suppose anyone has, or has experience with devices like this in reference to motion capture usage? I did some searching around and I can't find a single motion capture example usage demo... hmmm...

Ryan Roye
02-26-2014, 08:20 AM
I want to experiment with/research something called "air mice", many of which have gyroscopic sensors and are HID-compliant.

Right now, it is looking like none of these mice have magnetometers. In the PS3 move, the magnetometer it has acts like a compass and prevents "drift" by providing a constant reference of orientation. The biggest questions on the table in reference to gyroscopic air mice now are:

1) How significant of a problem is this "drift" issue without the benefit of a magnetometer? Does the average air mouse provide an easy way to reset orientation? How often will this have to be done? Greater error with faster movements? Is slowing down actions a way to reduce drift errors?

2) How easy will it be to pair it up with Virtual Studio? (this is more dependent on the device than Lightwave).

3) Quality of rotation tracking: Is it good enough for use with virtual studio? Or will it be too jittery and sloppy to work with?

erikals
02-26-2014, 11:12 AM
the PS3 prize isn't what it used to be though, maybe just get a PS3, or a used PS3 on eBay...?

Greenlaw
02-26-2014, 11:25 AM
You know, I had no intention of getting a PS3 unit but now you have me thinking about it. Will have to look into that. :)

Ryan Roye
02-26-2014, 11:45 AM
the PS3 prize isn't what it used to be though, maybe just get a PS3, or a used PS3 on eBay...?

It's still about $150 (used, on ebay)... that's asking a lot for simply getting wrist and head rotations in place. Obviously... it'd save a hell of a lot of headaches. Next month I'll try buying into one of these air mice and we'll see how usable it is then. Worst case scenario; I end up with a remote-controller that I can use with whatever computer... we do have an entertainment center hooked up to a PC downstairs.

I will say though that getting a single PS3 move controller working without a PS3 in virtual studio is not for the faint of heart... it isn't user friendly and even if you do get the controller working you need solid knowledge of node editing in order to know how to plug in the values properly. Some node editing knowledge is required for all solutions, but not having to go through the drivers/software hassles would go a long way to eliminate the time required to get it going in the first place.

Greenlaw
02-26-2014, 11:58 AM
Oh, I was thinking more for the virtual camera system, to get all six axis of movement. I already have several PS3 Move and PS Eye cameras to support this feature--just missing the console because I couldn't justify the cost. (I stopped having time to play games when we started Little Green Dog. Sometimes I miss it but not too much.)

For wrist, head rotations, etc., yes, a PS3 is probably overkill and there are probably better, more economical ways.

G.

erikals
05-20-2014, 09:59 AM
guess you guys saw the leap hand mocap thread... ?

http://forums.newtek.com/showthread.php?141628-Hand-motion-capture-with-leap-motion

Greenlaw
05-20-2014, 03:22 PM
Yes, that's pretty cool! :)

G.