Anaglyphic 3D with Lasers

[marksmartus]

I bought a MOTU Ultralight audio interface to use for this project specifically because it has DC coupled outputs.

https://www.sweetwater.com/sweetcare...re-dc-coupled/

[marksmartus]

I already have the beam blanking working. You can do almost ANYTHING in Reaktor. LOL In the video, sometimes the blanking is on, like when the cube is showing. But for getting the continuous curly lines I turn the blanking off. With it on, you get dotted curly lines, which are also cool.

At the moment I have 4 outputs going out of the sound card: X, Y, blanking, and eye selection. The blanking and eye selection is done by TTL chips on a protoboard in response to these signals, using an inverter for selecting the other eye, and 2 AND gates to control what color happens in each eye state. But I realized I can accomplish the blanking and eye selection in software by just turning the R, G, and B voltages on and off. I didn't realize that initially. So I will be able to eliminate the TTL chips.

[marksmartus]

Here, you can see it drawing curvy lines with the blanking on, then I hit the button on the screen to turn blanking off and they become continuous.

https://www.youtube.com/watch?v=4AtjZAmS8Gk&t=145s

I just realized the sound doesn't change because the audio is taken from a point in the chain before the blanking happens. I should try listening to the signal with blanking and see what that sounds like. Hopefully not as nasty as the 10K pwm.

[marksmartus]

I think if you download the free version Reaktor Player, then you can download this thing I posted in the user library a while back for doing 2D vector images:

https://www.native-instruments.com/e...try/show/6857/

I did this first, then in 2020 figured out how to compute left and right images for 3D. In December, a friend suggested I could do this with lasers, so in the last week I got the MOTU audio interface and the laser projector.

My blanking signal IS lined up with the selector that is picking which 3D coordinates are in use. The "on" (not blanking) signal goes low, then a new line gets selected, then there is a short delay waiting for the mirrors to move, then the laser is turned on again. Without that, this cube had extra slew marks on it:

https://www.youtube.com/watch?v=4AtjZAmS8Gk&t=119s

[marksmartus]

I've used Reaktor to send and receive MIDI from a Teensy, but don't know if you can do audio.

[Cassy0110]

This is a very interesting experiment, thank you for sharing, no I haven't tried it.

[marksmartus]

Hi all.
I got some dimmer circuitry working for the Laserworld projector (which normally just reacts to R, G, and B inputs on the ILDA connector as either on or off). It uses a single LM339 comparator chip. Pretty cool. You can make an oscillator out of one comparator stage:

https://theorycircuit.com/lm339-oscillator-circuit/

Then I use the other 3 stages on the chip to do the pulse width modulation for the red, green, and blue signals. I tweaked the above circuit to run at a much higher speed and put out a triangle-like waveform that extends from about 0-5V. 0-5V control voltages coming from Reaktor via the MOTU audio interface decide the threshold for the PWM on the other 3 stages of the chip. The laser diodes have a very small region of pulse widths within which the transition from off to full brightness happens, so with the Reaktor software I am setting the maximum and minimum values, which aren't very far apart. You get an effect kind of like the transporter on the original Star Trek. LOL. The lines start out as cloud of flickering dots and then materialize.

Somehow in the process of doing that, I broke the 3D effect, but I'll figure that part out. I'm hoping to use this at a concert I have in April with Joy Yang's group The Interdisciples:

https://www.visitchampaigncounty.org...ts?event=17281

At the above gig in December, I used Reaktor to project 3D shapes through a video processor, but it was kind of monotonous. With that setup it was 3D, but I could only do straight lines. I should be able to do much more nuanced things with the laser projector.

I'll keep you posted.

Mark Smart

[marksmartus]

The big concert was today! The Laserworld projector worked right up until the beginning of the gig, then died so I couldn't do anything! Better luck next time. I'm pretty sure it's the projector itself that went bad, I'll have to debug it. I might need to get a better projector.

[kecked]

I only read the start of this thread so…

something newer to try for 3d is called a structured light modulator. Basically it’s an LCoS panel. SLM

you pass video to it to control the phase of te wavefront and it produces a digital hologram and that can then be projected and it can move. There is now software to form the phase images so it’s literally plug and play. Look it up and give a try. I have not seen anyone diy this yet but the principle is really simple. Off the shelf is 10k.

so go get a panel for 30.00 and have a play at it. Ps you can also use it to control your lasers and remove aberrations as well as correct your laser on the fly. Do adaptive optics, make optical tweezers, make gratings etc…have fun.

I’ll help a little. Pas you laser through a pbs to select just one polariization. Bounce off the lcos back through the pbs straight through. Pass this through a pinpoint spatial filter and collect just the fist order diffraction. Your image is on the other side of this filter. The pinpoint is the Fourier plan where the phase image forms.

im using it for a special type of microscopy.

stupid money for this thing
https://www.thorlabs.com/newgrouppag...group_id=10378

[CountFunkula]

I only read the start of this thread so…

something newer to try for 3d is called a structured light modulator. Basically it’s an LCoS panel. SLM

you pass video to it to control the phase of te wavefront and it produces a digital hologram and that can then be projected and it can move. There is now software to form the phase images so it’s literally plug and play. Look it up and give a try. I have not seen anyone diy this yet but the principle is really simple. Off the shelf is 10k.

so go get a panel for 30.00 and have a play at it. Ps you can also use it to control your lasers and remove aberrations as well as correct your laser on the fly. Do adaptive optics, make optical tweezers, make gratings etc…have fun.

I’ll help a little. Pas you laser through a pbs to select just one polariization. Bounce off the lcos back through the pbs straight through. Pass this through a pinpoint spatial filter and collect just the fist order diffraction. Your image is on the other side of this filter. The pinpoint is the Fourier plan where the phase image forms.

im using it for a special type of microscopy.

stupid money for this thing
https://www.thorlabs.com/newgrouppag...group_id=10378

Did you see the little 1024x768 (720p-ish) SLM I built at SELEM? I had it on my optical breadboard . I've worked with some stupid-expensive DMD-based SLMs (they're the only game in town for UV wavelengths), and that inspired an active investigation into the lowest quality (read: cheapest) LCoS panels from which I can successfully eek out a simple HOE like a zone plate. I WISH you could pull this off for $30 ...it's realistically more like low $$$ hundreds.

SLMs aren't new, but there's been a recent explosion in new applications for them. They used to be popular for automatic optical inspection and some niche high-speed computation systems (<cough> terrain-following guidance systems <cough>), but now they're driving everything from resin-based 3D printers, HUDs and VR devices, optical computing, and even digital inter-cavity beam shape control. At a super basic level, overhead transparencies and film negatives or slides can be considered static spatial light modulators. LCoS panels are convenient because of their polarization properties, but there are other types of electronic SLMs such as transmission-mode TN LCD and digital mirror array devices. I vaguely recall some researchers also doing exotic things like using surface acoustic waves to create SLMs with engineered metamaterials. Optically-addressable SLMs also exist.

The basic input "image" for your wavefront is the 2D Fourier transform of your desired output image. Computing the SLM input for
arbitrary phase and/or amplitude-based images is trivial, but the materials and build quality have a huge effect on ability to accurately form a wavefront and usually require blemish-mapping for defect correction. That necessitates access to a setup for characterizing the SLM. The special sauce in a lot of commercial packages is the extremely low-defect panel and some really good correction algorithms. One disadvantage of an SLM-based holographic image is the need to focus it onto a plane for viewing. You need some serious resolution to produce anything approaching the quality of a film-based hologram. However, low-res is still very cool for making things like optical tweezers and simple structured light.