Interrupt Driven Pitch Generator (Pitch VP)
Yendor
Posts: 288
·
Something that can be considered cool, obnoxious, and useful all in one!
·
Here is my ‘proof of concept’, yet simple·program for generating monophonic musical tones using a Piezo speaker through an interrupt.·
·
I’m sure there are some variants out there, but hopefully someone may find this helpful or put a smile on their face.
·
Pitch accuracy will be based on the interrupt time and the clock cycles.· I’ve enclosed a spreadsheet that will generate the SX/B·pitch table based on the desired·SX Clock and Interrupt rate, as well as·some idea on the accuracy of the pitches, so that some analysis can be made.·· Also, note that Piezo speakers also have a limited bandwidth, so not all pitches will be audible
·
The pitches are based on MIDI note numbers (0-127), as well as the MIDI·Baud Rate.· Feel free to use the Excel spreadsheet to come up with the pitch table for your clock & interrupt.·20MHz and 50MHz would probably be best.
Hook up the Piezo as described in FREQOUT or SOUND examples in the SX/B help file.
The Code:
········ RTCC Prescaler
········ And either the RATE or CYCLES of your interrupt, depending on your preference
o······ Right now the CYCLES (Cell D4) is copied over from Cell E3, but that’s just for calculation verification.
········ Then if you want, you can put your Bit Period Counter in to calculate the baud rate.
·
Hope this helps, and have fun!
·
Rodney
·
Post Edited (Yendor) : 8/6/2008 4:12:01 AM GMT
Something that can be considered cool, obnoxious, and useful all in one!
·
Here is my ‘proof of concept’, yet simple·program for generating monophonic musical tones using a Piezo speaker through an interrupt.·
·
I’m sure there are some variants out there, but hopefully someone may find this helpful or put a smile on their face.
·
Pitch accuracy will be based on the interrupt time and the clock cycles.· I’ve enclosed a spreadsheet that will generate the SX/B·pitch table based on the desired·SX Clock and Interrupt rate, as well as·some idea on the accuracy of the pitches, so that some analysis can be made.·· Also, note that Piezo speakers also have a limited bandwidth, so not all pitches will be audible
·
The pitches are based on MIDI note numbers (0-127), as well as the MIDI·Baud Rate.· Feel free to use the Excel spreadsheet to come up with the pitch table for your clock & interrupt.·20MHz and 50MHz would probably be best.
Hook up the Piezo as described in FREQOUT or SOUND examples in the SX/B help file.
The Code:
- A data table is required (256 bytes).· Can always scale back the table, as necessary.·
- Purposely simplistic. ·Improvements that can be made using arrays, assembly, bit flags, and using a lookup table for the tones/durations,
- Had the issue using the “Load” directive, so I put all the constants in the main file.
- Helps to visualize pitch variances with different levels of interrupts and clock speeds
- Generate the pitch table to put in your SX/B Code
- Can be used to give some idea of how accurate the pitches would be.· Most people start to hear pitch variations at +/- 12 cents, so you can use the spreadsheet as a guideline for your end goals.· Accuracy is highlighted in red using conditional formatting, if beyond +/- 12 cents.
- Spreadsheet Inputs (Green)
········ RTCC Prescaler
········ And either the RATE or CYCLES of your interrupt, depending on your preference
o······ Right now the CYCLES (Cell D4) is copied over from Cell E3, but that’s just for calculation verification.
- Spreadsheet Calculations (Yellow)
········ Then if you want, you can put your Bit Period Counter in to calculate the baud rate.
- The rest of the columns give the calculations to get to the counter for the frequency
- The ‘Rounded’ Column is the value that you use in your pitch table.
- Copy Column M and paste it into your program for your table, based on your values.
- The grayed rows (row 6-7) are not supported.· I just put them in to get an ‘A’ reference, for the Pitch Frequency.
·
Hope this helps, and have fun!
·
Rodney
·
Post Edited (Yendor) : 8/6/2008 4:12:01 AM GMT
Comments
Yes, was thinking about 2-3 voice polyphony, too, and was going to attempt that later this week - just seems a matter of tackling appropriate resistor values.
·MIDI compatibly is exactly what I had in mind, as I did this as a development 'helper' to send out note events to this subroutine instead of a MidiOut/TxByte routine (so I don't have to hook up my synth's each time I'm working on it).
Post Edited (Yendor) : 8/6/2008 4:52:04 AM GMT
My program is a clone of yours with the addition of two more tone generators and VP UART and handler code. It's running on a PDB using a simple resistor summing circuit for the tone output pins. The PDB amplifier has a filter circuit as shown for SOUND and FREQOUT in the SX/B help file; this helps smooth off the "rough edges" of the square wave output. The addition of an EEPROM would allow for a lot of song data, just in time for holiday decoration projects.
[noparse][[/noparse]Edit] Updated the PLAY_NOTE subroutine -- it's still compatible with Rodney's PlayPitch subroutine but now allows the selection of no timing (tone generator started but not stopped) and allows the user to select which tone generator (of 3) to use. By updating PLAY_NOTE the START_TGEN subroutine could be eliminated to save a bit of code space.
[noparse][[/noparse]Edit] Fixed "gotcha" in UART buffer check (thanks, Zoot)
Post Edited (JonnyMac) : 8/8/2008 4:32:04 AM GMT
I'd love to hear a polyphonic version of stairway to heaven.
Bean.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
"A government big enough to give you everything you want, is big enough to take away everything you·have."·· Thomas Jefferson
"It is our choices, Harry, that show what we truly are, far more than our abilities."·Dumbledore from Harry Potter
www.iElectronicDesigns.com
·
Note: I recalculated the pitch values based on a 4 uS interrupt cycle for better accuracy, and used a slightly different strategy than Rodney. My pitch calculations are based on a standard formula for converting MIDI note values to frequency. My Excel file is attached, too.
JonnyMac -- one thing I noticed in your Polyphonic program (and perhaps others where you've included this VP?) is this line for checking if the serial VP buffer is full:
JB rxBufCnt.4, RX_Done ' (2/4) skip if buffer is fullWhich seems to imply a 16 byte buffer; this program has an 8-byte buffer. Wouldn't this be correct?
JB rxBufCnt.3, RX_Done??
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
When the going gets weird, the weird turn pro. -- HST
1uffakind.com/robots/povBitMapBuilder.php
1uffakind.com/robots/resistorLadder.php
Looks like you've been busy, too!
That's great you got polyphony working (I'm jealous!)!·
I'm going to resist peaking to see what I can come up with and compare.· I agree that some buffer (maybe two) are needed for the notes.· One for Duration and the other for what note is waiting (see Tempo/BPM Post).
I'm using the Piezo Schematic from the SX/B file, so it's doesn't sound too bad.···
Did you use an R2R Ladder to combine the bits?· Was the sound consistant?
Not sure if you've thought about this, but here's a suggestion to save some song space.
Think of a song as a program·of repeatitive·information.· So simply break the song down into 'subroutines'!
For example:
And with each part you can have a sequence
- Verse 1:
- Sequence 1
- Ending 1
- Ending 2
- Verse 2:
- Sequence 1
- Sequence 2
- Chorus
Each sequence is a Pattern:- Sequence 1
- Pattern repeated·4 times
Then it's simply another table: number of elements, address, number of repeats.MySong: 3,Verse1,1,Verse2,1,Chorus,2
Verse1: 4,Sequence1,1,Ending1,1,Sequence1,1,Ending2
Verse2: 2,Sequence1,4,Sequence2,1
Chorus: 2, Sequence1,1,Sequnce2,1
Sequence1: 1,Pattern1,10
Sequence2: 1,Pattern2,4
Ending1: 1,Pattern3,1
Ending2: 1,Pattern4,1
etc..
Then just generate song files, and load them up with an 'Load' function.· Something like what Bean did on his MMVG driver. But, shhh, let's not tell the RIAA about this...
And Ooops!· You're right! Middle C (Note 60), is C4, not C5...So I'm an octave off.· I guess that's better than being off by one jack!· Thanks for catching this!
Good catch Zoot!· I knew this was an issue, but haven't dug into the details, so I just used the normal way for now.
Finally thanks for the supporting comments!·
Rodney
In looking at your song schema, may I suggest that you check out some of the excellent materials relating to storage and playback of music for the Stamp. There are a number of good examples in the WAM book by Parallax (chapter 8, http://parallax.com/Portals/0/Downloads/docs/books/edu/Wamv2_2.pdf). In particular there is a good article about parsing text ringtones. It would be easy enough to store music scores on an external EEPROM and play them back.
See also my (Stamp-based) code for reading scores from external memory and delivering on the fly for some I2C EEPROMS samples. http://forums.parallax.com/showthread.php?p=619826.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
When the going gets weird, the weird turn pro. -- HST
1uffakind.com/robots/povBitMapBuilder.php
1uffakind.com/robots/resistorLadder.php
Post Edited (Zoot) : 8/8/2008 11:14:44 PM GMT
In your column - would love to see if you can work in an EEPROM for the music, too.
I'm thinking about hacking an FC-4, myself! [noparse]:o[/noparse])
Thanks for the links Zoot! I'll check it out...·
As my end goal is MIDI, I'm just trying to code a·"Note-Duration" scheme on different 'tracks', if you will, and not worry about Note-Off events.
Thanks!
Rodney
Post Edited (Yendor) : 8/10/2008 2:15:54 PM GMT
Did you try increasing the PPQN in the spreadsheet, that I mentioned in the Tempo/BPM Thread?
Also, I still have some experimenting to do, but it seems that the data table for the clock can be eliminated to an algorithm if you use only even tempos.
Rodney