amandaghassaei

Yeah there's not a lot you can do to get the file size down - the detail in the grooves that the code currently spits out was the bare minimum I found that I could get away with without impacting quality. I think the best bet is to process a shorter audio file, which will generate a shorter groove with less geometry.

38 804 kb? yeah that's gonna be tough - not sure you will find a slicer to handle that unfortunately. Maybe if you write your own :)

No problem! Either turn up the laser power to cut the grooves a bit deeper, or decrease the amplitude of the waveform so that the loud parts don't cause the needle to rock back and forth in the groove so much. In the code:float amplitude = 10;//in pixels

yes, the code generates pdf's which can be opened in corel draw.

yeah for a 78rpm record you could set the cutoff filter to a higher frequency, and for a 33rpm you should set the frequency lower. None of this is super critical, but it may improve your results a bit. here's the formula:samples/sec = samples/rev * rev/min * min/secsamples/sec = 6000 * 45 * 1/60 = 4.5kHzthen you want to divide the samples/sec by 2 to get the max frequency that you can reproduce with the record - this is your cutoff frequency. 2.5kHz ~= 4.5kHz / 2

Hey, can you describe the problem a little more? I'm not sure what you mean. When you download the code, open the file called SugarCubeLibraryFiles.ino and upload that to the Arduino.

no, that's a big bottleneck in this project. Ideally, you could just send a bitmap of the layers directly to the 3D printer. But I don't know of a printer that supports that.

oh man, it's bees so long I'm not sure I can track down something to help. Where are the differences? I think you should go by the schematic if in doubt. The schematic should be correct.

Yeah that should work, Either change the processing code to write svg:https://processing.org/reference/libraries/svg/ind... or (easier) use a vector editing program (Illustrator, Inkscape) to convert to SVG.

beautiful!

cool! It should still sound pretty much like the song after audacity - just maybe a little bass heavy, not as much detail in the highs. Not noisy.

file looks good to me, is it in the same folder as the processing sketch?

hey sorry for the late reply - when I opened this up the document size was very small (0.5in width and 0.7in height) so there was data on it, it was just much bigger than the document and not visible. I selected all and did Ibject>Release Clipping Mask in illustrator and the vector lines showed up. Then I made to document bigger. Attached.

you probably know better than me at this point - I did this project in 2010, that was the last time I looked into IR grid related stuff.

yeah absolutely! always meant to work up to this, but never found the time.

yes that would work with this ir receiver setup.

that should work fine - I would just take the working code and copy the LSR lines so that you are measuring signal and sending midi from all 5 sensors. Yes, flex sensors would be similar code, though you might have to scale things differently depending on the numbers coming in from analogRead.

yes definitely. The only limitation is the number of pins on the arduino - there are only six analog pins so you may be limited there, but there are plenty of digital pins to work with.

sorry for the late reply! hairless expects 115200 baud rate, so it is not surprising that it doesn't work at 9600. are you sure you're not sending a note on message later in the code that is never followed by a note off message?

hey have you resolved this? what's the error?

I ran into this problem too - I was using the slicer in a program called Objet Studio (it's specific to the printer I was using, it doesn't actually make g code). I found that if I put over 1 min of audio on a record the slicer would crash. How big are the files you're trying to slice?

awesome! Hope you were able to finish your project!

I was just sending midi out of my computer. I haven't tried this with a midi keyboard, but I'd love to hear if you found a way to solve it.

if you have an instrument like that in your midi software then yes. MIDI is just a language that electronic instruments use to communicate, the part where MIDI commands get interpreted into sound happens outside the arduino (at least for the process I've described in this tutorial)

you need to read in the pot as an analog input, scale the analog input val from 0-1023 range to 0-127 range, when you create a MIDI note on command, you output this number as the "note", the second parameter in the note on message.in note on and note off commands, the "note" parameter is the pitch. Here's a table of how note corresponds to pitch:https://newt.phys.unsw.edu.au/jw/notes.html

Awesome, I'm glad the cnc worked better! I'd love to see/hear the final product!ps I think you're confusing air assist with the fume extractor (ventilation fan). You want to keep the ventilation on. Air assist is an extra thing that blows a high pressure air on the part right where it's cutting. Shouldn't make a huge difference either way though.

what kind of cnc? I think it's going to be nearly impossible to get this to work without doing it in vector format.

nice!!

air assist is the compressed air that some lasers blast at the cut to help when you're trying to cut through the material. Glad it's at least getting a little better. I've had luck cncing with a little engraving bit on hdpe - about the same sound quality as the laser. Other people have also had success with machinable wax. Good luck! Let me know how it goes!

I think you need to back off on power - it looks like you're burning the material too much. Here's what I recommend - focus the laser on the surface of the material (I like to use the manual focus to make sure it's really dialed in), make sure the material is really flat so the beam doesn't go out of focus. You're putting too much heat in the material, so bump the speed up to 100 and lower the power. You want to take the power down so that you're making a deep enough groove so the needle doesn't fall out, but no deeper. Any more heat just widens the groove and increases noise. If you're doing this on plywood, the surface shouldn't look scorched https://cdn.instructables.com/FUC/9WSF/HFSHURAQ/FU...the groove should be a thin line and and there should be clear separation between the ...

I think you need to back off on power - it looks like you're burning the material too much. Here's what I recommend - focus the laser on the surface of the material (I like to use the manual focus to make sure it's really dialed in), make sure the material is really flat so the beam doesn't go out of focus. You're putting too much heat in the material, so bump the speed up to 100 and lower the power. You want to take the power down so that you're making a deep enough groove so the needle doesn't fall out, but no deeper. Any more heat just widens the groove and increases noise. If you're doing this on plywood, the surface shouldn't look scorched https://cdn.instructables.com/FUC/9WSF/HFSHURAQ/FU...the groove should be a thin line and and there should be clear separation between the groovesdifferent lasers will require different settings, so don't worry if your settings are different from mine. But keep the speed up and lower power. I assume you that laser is actually moving pretty slow at 100 speed like it was for me?are you using air assist? I believe I turned that off for these cuts.also did you do the low pass filtering in audacity that I recommended?hope it works out! The records you've made so far look sweet if that counts for anything!

can you post some pics and a recording?

I would do some tests with the power and speed settings and cut it to whatever depth that keeps the needle in the groove. This is something you have to tune with your machine and materials.

there is the green button that says "clone or download"

yeah change the diameter and also change outerRad.outerRad is the radius of the outermost groove, you'll need to bring this in to fit on a ten inch record with a little room to space it from the edge, outerRad = 4.8" should be good.

awesome! unfortunately, it's not possible to buy it anywhere. Maybe one day I'll design a PCB for it at least, but just haven't had time.

this code uses python 2, you may need to download a different version pf python.

I'd go with the one on the left, the smaller one.

you should be able to load it into illustrator as vectors and change it there. You can also try changing lines in the code that say:strokeWeight(0.001);to strokeWeight(0.25);

yeah this code is old... maybe try changing line 24 to:frameOneChannel = list([0]*numframes)or change it toimport npframeOneChannel = np.zeros(numframes)

ah sorry that'simport numpy as npframeOneChannel = np.zeros(numframes)

yes, I actually already wrote a little code to do that. If you download the github repo you'll see a file in there for linear processing:https://github.com/amandaghassaei/LaserCutRecord/b...hope that helps!amanda

glad it helped!

just saw this, that's awesome! It's been so long since I made this project, that I can't think of why high bpm would stop playback, that's an annoying issue. Do you have any plans to mess around with the code?

not sure, did you sort this out? maybe try on a different comp?

cool! One thing you should consider is that the surface speed of the needle past the paper will be slower for lower diameter records. So for best results, play at 78rpm and choose bass-heavy audio.hope it works!amanda

maybe try using an older version of processing.

yes, I think a newer build of python 2 should work.

whoa, I never saw this before, so cool phil!

you mean one resistor for all the LEDs? No - that will give you variable current through your leds depending on how many leds are on at once. Best case scenario, you will not be able to control the brightness of your LEDs, worst case scenario, you could blow an LED.

hey sorry for the late reply - it seems like you have your wav in the wrong place. It should be in the same folder as the python script. Hopefully you figured that out on your own by now!

oh wow that's cool! Do you have them set up for pwm-ing? would love to see more videos as you progress with this!

yeah I think it's possible. I haven't implemented any midi receive stuff yet, but you can do it with arduino's serial library:https://www.arduino.cc/en/reference/serialhttps://www.instructables.com/id/Send-and-Receive-...

arduino isn't really set up to store large amounts of audio data and output analog, but you can get external modules that you can connect to arduino to make that work (though the code gets a little tricky). My only concern is that this project already uses a lot of io pins and you might find that you run out of other resources on the board - timer/counters etc.

nice, glad to see this worked! what are you going to do with it next?

you mean you want to be able to control the leds to make them do other things? They won't be able to change color or brightness bc the circuit isn't set up for that, but you could write your own app to control them in different ways. this is the basic keyboard app:https://github.com/amandaghassaei/Sugarcube-Arduin...see how you can add code to different events? I would start from there if you want to program something yourself.

you could definitely write an app that caused the sugarcube to send a program change message:https://en.wikipedia.org/wiki/General_MIDI#Program...the sugarcube class in this code has a method to send a generic midi command:void sendMIDI(byte command, byte param1, byte param2);this could be used to send the change program message.

For those asking about the mph calc:mph = (56.8*float(circumference))/float(timer);"circumference" is measured in inchestime per revolution of the wheel ("timer") is measured in milliseconds (0.001s)miles/hr = 1/(inches per mile) * (miliseconds per hr) * (circumference / timer)there are 63360 inches in a mile and 1000*60*60 = 3,600,000 ms in an hr:mph = 1/63360*3600000*circumference/timermph = 56.8*circumference/timerto do this in km/hr:km/hr = 1/(inches per km) * (miliseconds per hr) * (circumference / timer)km/hr = 1/39370*3600000*circumference/timerkm/hr = 91.4*circumference/timer

use a text to speech chip, it will be much easier:https://www.instructables.com/id/Twitter-Enabled-Text-to-Speech/

awesome! yeah go for it!

https://github.com/amandaghassaei/3DPrintedRecord

you'll have to look through the reference pages:https://docs.cycling74.com/max5/refpages/max-ref/serial.html

you could hack something together to make that work over midi, but it would probably be easier to use serial. You might also look into osc, i think it could support something like that - not sure though.

yeah at some point I think there was formatting to make that list more clear. I fixed it. glad to hear you solved it!

leds could be wired backwards? do they turn off when you would expect them to be on?

in Max if you send a "bang" to the serial object and then look at what's printed it will tell you which port - a, b, c, d, e, etc You can set the port by clicking on the one of the buttons attached to the serial object.

ok, use a multimeter to check the voltage across the leds to see if they're even getting power. Are you sure the leds are installed in the correct orientation?

hmm that is really strange. What is the error message? usually I just run the code with the wav in the same folder.

the connections look fine. what are the symptoms? no leds are turning on? In my experience, the ics used to drive the leds are pretty robust, I haven't seen one get fried in normal circumstances - I'm not generally too careful about esd. Do you have an oscillascope to debug the output from the ics?