24” x 36” x 1/2” Clear Acrylic
2 x AlexP 850nm 10mW Lasers @ ~+5.02V mounted on tripods
FireFlyMV with 850nm +/- 40nm bandpass filter with #13VM308ASIR Tamron 1/3” Aspherical IR Vari Focal lens
Camera is facing the acrylic directly
Software:
AlexP’s modified TouchLib
Here’s the problem:
In the video I setup a test with a strip of Tracing Paper, Rosco Grey, and the bare acrylic. The test materials are taped perfectly flat up against the underside of the acrylic.
The first part shows the raw camera output, with the bandpass filter in place. The blobs on the bare acrylic are visible, the tracing/rosco are extremely feint.
The second part shows the TouchLib config screen, all brightnesses are on max. Anytime you see the bright white blobs, thats from touching the bare acrylic. The blobs on the bare acrylic are perfect! The blobs on the Rosco are too feint to really be seen. The tracing paper just barely makes it through, but not bright enough to make a bright enough blob for touchlib to see.
My question:
It’s clear that I’m having brightness issues, I cannot get the IR to penetrate the surface of the projection materials.
Does anyone see anything that’s out of place or strange? Does the raw output look similar to other people using the FireFly with lasers?
I’m not sure yet how to go about applying different of the TouchLib filter graphs. I did a search and used a slightly modified FilterGraph and it seemed to help, but only on the bare acrylic, I posted the XML config file I’m using below.
When you apply different filters is this done by directly modifying the Config.xml file?
Any help would be much appreciated! Thanks to all in advance.
First of all is the Lense on the FireflyMV non ir coated and there is no IR filter on the MV?
Secondly , Rosco is known to feighten(sp?) blobs , try just the Tracing paper for projection , your not using any mirror this time right?
You shouldn’t be neading to use any type of filter apart from contrast and rectify in touchlib for your setup , the blobs should be coming in really bright.
Try my DI config , i’ve atached it to this post , talk to AlexP to see what he uses as a projection screen.
First of all is the Lense on the FireflyMV non ir coated and there is no IR filter on the MV?
Secondly , Rosco is known to feighten(sp?) blobs , try just the Tracing paper for projection , your not using any mirror this time right?
You shouldn’t be neading to use any type of filter apart from contrast and rectify in touchlib for your setup , the blobs should be coming in really bright.
Try my DI config , i’ve atached it to this post , talk to AlexP to see what he uses as a projection screen.
Hey Glenn it seems to me that you are using some unnecessary filters in your touchlib.
Take a look again at my version of touchlib here. Do not modify the config.xml.
You’ll also benefit from faster blob processing speed and lower CPU usage.
What is your capture frame size 320x240?
Rosco would not be the perfect projection surface in this case since it blocks some of the IR.
Hey Glenn it seems to me that you are using some unnecessary filters in your touchlib.
Take a look at my version of touchlib here.
You’ll also benefit from faster blob processing speed.
What is your capture frame size 320x240?
~Alex
Yeah I was using your version of touchlib originally, I believe i was getting very weak blobs, that’s why I saught out some new options. Ill revert back to your version and post the results tonight.
When I startup configapp.exe the PGR Streaming Config Property Page comes up.
I click on Enable Format 7, and for the video I used the stock settings of Top: 0 Left: 0 Width: 752 Height: 480.
But I did try the 320x240 and it worked the same, much faster.
The reason why I left the default option is that it’s a much bigger window and its easier to see while testing blobs.
Hey Glenn it seems to me that you are using some unnecessary filters in your touchlib.
Take a look at my version of touchlib here.
You’ll also benefit from faster blob processing speed.
What is your capture frame size 320x240?
~Alex
Yeah I was using your version of touchlib originally, I believe i was getting very weak blobs, that’s why I saught out some new options. Ill revert back to your version and post the results tonight.
When I startup configapp.exe the PGR Streaming Config Property Page comes up.
I click on Enable Format 7, and for the video I used the stock settings of Top: 0 Left: 0 Width: 752 Height: 480.
But I did try the 320x240 and it worked the same, much faster.
The reason why I left the default option is that it’s a much bigger window and its easier to see while testing blobs.
Did you adjust the camera brightness and shutter speed?
I would look into just running graphedit and tweak the camera setting by lookint at the camera output.
Hey Glenn it seems to me that you are using some unnecessary filters in your touchlib.
Take a look at my version of touchlib here.
You’ll also benefit from faster blob processing speed.
What is your capture frame size 320x240?
~Alex
Yeah I was using your version of touchlib originally, I believe i was getting very weak blobs, that’s why I saught out some new options. Ill revert back to your version and post the results tonight.
When I startup configapp.exe the PGR Streaming Config Property Page comes up.
I click on Enable Format 7, and for the video I used the stock settings of Top: 0 Left: 0 Width: 752 Height: 480.
But I did try the 320x240 and it worked the same, much faster.
The reason why I left the default option is that it’s a much bigger window and its easier to see while testing blobs.
Did you adjust the camera brightness and shutter speed?
I would look into just running graphedit and tweak the camera setting by lookint at the camera output.
~Alex
Yikes, I did not do that, I just hooked it up and went!
Is Graphedit bundled with the Point Grey Software? I googled it but all I see is this program that creates filtergraphs for Directshow.
Hey Glenn it seems to me that you are using some unnecessary filters in your touchlib.
Take a look at my version of touchlib here.
You’ll also benefit from faster blob processing speed.
What is your capture frame size 320x240?
~Alex
Yeah I was using your version of touchlib originally, I believe i was getting very weak blobs, that’s why I saught out some new options. Ill revert back to your version and post the results tonight.
When I startup configapp.exe the PGR Streaming Config Property Page comes up.
I click on Enable Format 7, and for the video I used the stock settings of Top: 0 Left: 0 Width: 752 Height: 480.
But I did try the 320x240 and it worked the same, much faster.
The reason why I left the default option is that it’s a much bigger window and its easier to see while testing blobs.
Did you adjust the camera brightness and shutter speed?
I would look into just running graphedit and tweak the camera setting by lookint at the camera output.
~Alex
Yikes, I did not do that, I just hooked it up and went!
Is Graphedit bundled with the Point Grey Software? I googled it but all I see is this program that creates filtergraphs for Directshow.
Thanks Alex
Using graphedit you can quickly setup video capture/rendering and at the same time have access to all the camera settings.
I went into graph edit and watched the output, but I just ended up cranking everything to max. Not really any change. I think it may have already been maxed out.
Alex-- is your setting “Absolute Mode” checked off?
I also reverted back to the original binaries, and same ol dull blobs, and extremely feint blobs behind the tracing paper.
This is driving me nuts!
The only other thing I can think of testing is swapping out my lens unit for the stock microlens, just to see for some crazy reason I have an IR coated lens or something. I’m going to order that JIC #00 screwdriver so I can unscrew the CS mount and try it.
Plan B is to try hopping on the 780nm bandwagon and hope for brighter blobs.
Anyone else have any other suggestions before I lose my mind!?
Gw
“Exact science, Mr Angier, is not an exact science.”
I went into graph edit and watched the output, but I just ended up cranking everything to max. Not really any change. I think it may have already been maxed out.
Alex-- is your setting “Absolute Mode” checked off?
I also reverted back to the original binaries, and same ol dull blobs, and extremely feint blobs behind the tracing paper.
This is driving me nuts!
The only other thing I can think of testing is swapping out my lens unit for the stock microlens, just to see for some crazy reason I have an IR coated lens or something. I’m going to order that JIC #00 screwdriver so I can unscrew the CS mount and try it.
Plan B is to try hopping on the 780nm bandwagon and hope for brighter blobs.
Anyone else have any other suggestions before I lose my mind!?
Gw
“Exact science, Mr Angier, is not an exact science.”
Can you answer the following:
1. Do you have B/W or color version of camera?
2. What is your shutter speed seting?
3. Maybe there is an IR blocking filter on your lens?
4. Did you focus the laser light?
780nm is a bit brighter than 850nm, but not that much that it will make a difference in your case.
I went into graph edit and watched the output, but I just ended up cranking everything to max. Not really any change. I think it may have already been maxed out.
Alex-- is your setting “Absolute Mode” checked off?
I also reverted back to the original binaries, and same ol dull blobs, and extremely feint blobs behind the tracing paper.
This is driving me nuts!
The only other thing I can think of testing is swapping out my lens unit for the stock microlens, just to see for some crazy reason I have an IR coated lens or something. I’m going to order that JIC #00 screwdriver so I can unscrew the CS mount and try it.
Plan B is to try hopping on the 780nm bandwagon and hope for brighter blobs.
Anyone else have any other suggestions before I lose my mind!?
Gw
“Exact science, Mr Angier, is not an exact science.”
Can you answer the following:
1. Do you have B/W or color version of camera?
2. What is your shutter speed seting?
3. Maybe there is an IR blocking filter on your lens?
4. Did you focus the laser light?
780nm is a bit brighter than 850nm, but not that much that it will make a difference in your case.
~Alex
Yes it’s the B/W Version.
The shutter looked the brightest at the max setting, 66.23ms.
There are other people on the boards using the #13VM308ASIR, but just to be on the safe side I’m going to just swap in the microlens and holder when I get the special screwdriver to remove those rear mounting screws.
Yes I put a white poster board in the center of the screen at 90 degrees and rotated the line lens until I had the skinniest, brightest line.
Have you played around with the placement of your Rosco. How close is it to the lasers? How thick is the acrylic? Do your blobs look the same if your set up is laser, rosco then acrylic as apposed to laser, arcylic then rosco?
Have you played around with the placement of your Rosco. How close is it to the lasers? How thick is the acrylic? Do your blobs look the same if your set up is laser, rosco then acrylic as apposed to laser, arcylic then rosco?
As of right now I’m not using any projection material, just the clear acrylic, because I’m trying to figure out why the blobs are so weak. I used a a strip of Rosco and a strip of tracing paper just for testing. Tracing paper works the best, but theres barely enough IR light shining through to see a blob. It’s not enough to trigger a blob in the config app, too weak.
The placement of the lasers is shown above, I’m using the stock AlexP 120 degree line lenses.
The acrylic is 24” x 36” x 1/2” Clear Acrylic. I also have a sheet of thin glass the same size which I also tested the lasers on with the same results.
With the Rosco in place, in either position the blob is way too feint, barely visible.
I would test your camera with a known good sorce of IR light.
Wire up a single IR LED and point it at the camera and see how it responds.
It should flood the CCD and go white. You’ll know right away if there are any IR filters in the optics.
Once you verify that the camera is seeing IR, you can move forward and check that component off your list.
Troubleshooting can drive you nuts! Here’s my philosophy:
Test components and use a process of elimination. Check them off the list as you go.
Only change one thing at a time and check results.
Look for the simple causes.
I would test your camera with a known good sorce of IR light.
Wire up a single IR LED and point it at the camera and see how it responds.
It should flood the CCD and go white. You’ll know right away if there are any IR filters in the optics.
Once you verify that the camera is seeing IR, you can move forward and check that component off your list.
Troubleshooting can drive you nuts! Here’s my philosophy:
Test components and use a process of elimination. Check them off the list as you go.
Only change one thing at a time and check results.
Look for the simple causes.
Thanks for the response Tinkerman,
I have a IR LED rigged up in a pen from the Wiimote Whiteboard testing I was doing.
I’ll shine it at the lens with the bandpass filter and take a video of the result.
I have the exact same camera and lens, plus a similar bandpass filter (880, +-40nm). I can assure my B/W FireflyMV and Tamron lens have no IR coating, and the bandpass filter allows a lot of IR through (75% at the center wavelength according to Thorlabs).
It seems to me that your problem is either with the diffuser or camera settings. Rosco is surely a bad choice for LLP setups, so perhaps you could try another kind of tracing paper.
Regarding the camera, if you have another machine around, try installing the PGR FlyCapture drivers, which allow better control over the camera settings. I would avoid doing it at the computer you’re currently using, because the FlyCap drivers could mess up FlyStream. It’s possible to fix that with the driverControlGUI utility, but it could make things worse.
But, if you manage to configure your camera through FlyCap, you can save the settings to the camera’s memory and bring it back to your machine.
I’m sorry I can’t help you further, cause I’m currently running a FTIR setup and I’m lovin’ it.
EDIT: just one more thing I noticed. My bandpass filter is placed at over the top of my lens cap, not inside the cylinder as your photo shows. Might make a difference.
