Yes ... you're right - it needs a preload block. We have one, but due to different latencies on different systems, it obviously needs to be longer. Once everything is loaded, all is well, but this crops up during the initial load from time to time.
Thanks for the feedback on the instruments that seem to work best. Another approach might be to have kids sample sounds and vary the playback rate to let them create their own instruments.
I'm going to try to work on the color dots on the rotating carousel next to see if I can get anything that is even close to being in the ballpark with respect to timing. (fingers crossed)
You're welcome, but the real point of that paragraph was to say that the piano and the organ don't sound to me like they're playing the same notes. Am I crazy?
I am not your best informant with respect to the pitch of notes. However, I asked a couple of other people to compare; they said that it sounded like the two instruments are playing the same notes to them.
I then wasted an hour trying to troubleshoot the audio settings on a Windows 11 computer because the system didn't seem to recognize my microphone. It turned out that the USB connection to the microphone had come unplugged. Once that was resolved, I recorded the two instruments.
It turns out that the two instruments are playing the same notes in different octaves. Both are supposed to be in the fourth octave, but as you can see, the piano (bottom track) is twice the frequency of the organ (top track). So clearly both are not in the fourth octave. Good catch!
But I'm going to wait until Monday to delve into this issue.
We have now taught the art & music course for six years for a total of twelve consecutive semesters. A colleague uses TuneScope in a psychoacoustics course each spring. And Monty Jones teaches a spin-off course at VCU focused solely on music that has now expanded to three sections.
The reason that none of us noticed that the organ is mis-indexed may be due to a phenomenon known as "Octave Equivalence." The auditory system seems to treat notes an octave apart as similar due to the 2:1 harmonic relationship. The brain appears to process notes separated by an octave as the same pitch class, making the error less perceptible.
Clearly, from now on, we will have to refer to you as the "man with the golden ears."
Haha. What I heard was that the piano sounds kinda tinny, especially compared with the organ's boomboxiness, and I had to think about it to understand it as an octave shift. Honestly it's the piano that sounds wrong to me! Like 1950s transistor radios.
About "golden ears," when I was younger I was a hifi snob (Crown amp and preamp, Magnaplanar speakers, etc.), but of course like everyone else I'm losing the high end of my hearing as I age.
I'm not a musician, but...
The organ diagram looks like a square wave combined with the 8..10 times higher frequency.
The piano seems to be the triangle + higher frequency.
It's hard to see it clearly, but the higher sound may be the perceived note.
Yeah, organs are entirely odd harmonics, like square waves.
I don't think the piano can really be based on anything as simple as a triangle; what you're seeing is a bunch of sine waves (i.e., the Fourier analysis of the sound) all of whose peaks coincide at certain points.
...but I'm afraid the resolution isn't good enough for any common factor to leap out at me. (Ignore the "x:259," which is an absolute screen coordinate; the others are distances from that one, and ought to be multiples of some constant, which would be the fundamental frequency (measured in pixels per Hz or something).)
P.S. There are a few cases of nearby twin peaks:
... leading me to wonder if the distance between siblings represents the fundamental, in which case there are way too many harmonics for me to make sense of. But I'm hoping those are just an artifact of the way each piano note has two or three strings that are tuned to slightly different frequencies.
I created an initial proof of concept in which the turtle plays a note as it moves across a series of colored dots. Each color is mapped to a note that the turtle plays as it moves across the dot:
The white space is treated as a rest between notes. Even though I added an additional one-second wait after TuneScope initializes, there's still an issue with the first note sometimes failing to play during the first run-through. After the initial load, it seems to be OK in subsequent run-throughs.
The next step will be to replace the colored circles on the stage with colored dots viewed with a web cam. Once that's working satisfactorily, I'll place the dots around the circumference of a carousel so that the notes will play as the carousel turns.
This brief video is a test of the ability to translate a color viewed through a USB camera into a note. The background color white is translated as a rest. Squares of construction paper were then placed beneath the camera to trigger notes. It quickly became apparent that the color temperature of room lighting affects the color picked up by the camera. It also seems likely that even lighting will be important. However, it appears that in concept, this could potentially work.
Here's the first proof-of-concept of the completed project. As the carousel turns, the USB camera displays the colored dots on the Snap! stage. The dots trigger notes that play as the carousel turns. There's quite a bit of cleanup and work to be done at this point, but the proof-of-concept indicates that this will be feasible.
Note: we're using a MicroBit microcontroller connected to a DC motor to turn the carousel. In the final version, Snap! will be connected to the MicroBit via the MicroBlocks library.
An Adafruit Circuit Express, a color sensor (TCS34725 )and just 5 blocks of code. The Neopixels from the Circuit Express change to the color that the color sensor detects.
If I would use a board that has BLE I could send the value to Snap! and do something on the screen. You could make a kind of virtual keyboard. I would need a piece of cardboard with colored piano keys. Problem is the distance to the cardboard while sensing is important. If you go to close everything becomes white for the sensor because of the LED light.
