Roman Black Sound (btc)

Here are some number sound files and Darrel Taylor's example program that allow your PIC chip to talk out numbers to you. The RB sound files are pretty scratchy, so if you want a professional sounding product, this is not the way to do it. But it is understandable, and sure beats counting the flashes of an LED. It also uses a lot of storage space for the sound files. But the fact that you can store them on a chip, is pretty cool. Darrel's RB_Sound include file was used, as well as his testRBsound program . Check out post # 5 for his post in this thread.

Here is the suggested filter (Roman Black) that may make it sound a little better.

; R = 25968 ohms
;
; Digital -----------R-----*----- Analogue
; | out
; |
; |
; C = 0.22 uF
; |
; |
; |
; Gnd
;

Code:

---

'*********************************************************************
'*  Name    : TestRBsound.pbp  from Darrel Taylor                    *
'*  Author  : Darrel Taylor, modified by Walter Dunckel for numbers  *
'*  Date    : 5/24/2010                                              *
'*  Version : 0.1                                                    *
'*  Notes  : This version is limited to 18F's                      *
'*          : with only 1 sound at a time                            *
'*********************************************************************
;-- Target = 18F4550 --

asm
  __CONFIG    _CONFIG1L, _PLLDIV_5_1L & _CPUDIV_OSC1_PLL2_1L & _USBDIV_2_1L
  __CONFIG    _CONFIG1H, _FOSC_HSPLL_HS_1H
  __CONFIG    _CONFIG2L, _PWRT_ON_2L & _BOR_OFF_2L & _VREGEN_ON_2L
  __CONFIG    _CONFIG2H, _WDT_ON_2H & _WDTPS_512_2H
  __CONFIG    _CONFIG3H, _PBADEN_OFF_3H
  __CONFIG    _CONFIG4L, _LVP_OFF_4L & _XINST_OFF_4L
endasm

DEFINE OSC 48
btc_Pin1  VAR PORTC.1  'sound output pin
btc_Pin2 var portC.3      'dummy pin to keep DT_RBSound.pbp happy
LATE_LED var PORTD.0

DEFINE  btc_DATA  db
INCLUDE "DT_RBSound.pbp"

@  ImportSound n0,    "zero.asm"
@  ImportSound n1,    "one.asm"
@  ImportSound n2,    "two.asm"
@  ImportSound n3,    "three.asm"
@  ImportSound n4,    "four.asm"
@  ImportSound n5,    "five.asm"
@  ImportSound n6,    "six.asm"
@  ImportSound n7,    "seven.asm"
@  ImportSound n8,    "eight.asm"
@  ImportSound n9,    "nine.asm"
@  ImportSound n10,    "ten.asm"
@  ImportSound n11,    "eleven.asm"
@  ImportSound n12,    "twelve.asm"
@  ImportSound n13,    "thirteen.asm"
@  ImportSound n14,    "fourteen.asm"
@  ImportSound n15,    "fifteen.asm"
@  ImportSound n16,    "sixteen.asm"
@  ImportSound n17,    "seventeen.asm"
@  ImportSound n18,    "eighteen.asm"
@  ImportSound n19,    "nineteen.asm"
@  ImportSound n20,    "twenty.asm"
@  ImportSound n30,    "thirty.asm"
@  ImportSound n40,    "forty.asm"
@  ImportSound n50,    "fifty.asm"
@  ImportSound n60,    "sixty.asm"
@  ImportSound n70,    "seventy.asm"
@  ImportSound n80,    "eighty.asm"
@  ImportSound n90,    "ninety.asm"
@  ImportSound n100,    "hundred.asm"
@  ImportSound n1000,    "thousand.asm"

Main: 'says 140,259
    @ PlaySound n1
    pause 100
    @ PlaySound n100
    pause 100
    @ PlaySound n40
    pause 100
    @ PlaySound n1000
    pause 100
    @ PlaySound n2
    pause 100
    @ PlaySound n100
    pause 100
    @ PlaySound n50
    pause 100
    @ PlaySound n9
    pause 3000
goto main

---

The attached audio sample audio-sound-mp3.zip using a 18f4550 in PIC-USB-STK hardware from Olimex. But you can also get similar sound out of a non amplified piezo speaker using a 5uf capacitor.

I had not seen this before. I have not played the files from a PIC yet, but it sounds interesting as well. This C code :( uses a "simplified Adaptive Differential Pulse Code Modulation (ADPCM) algorithm". The files look slightly larger than the Roman Black files, but not by much, comparing to the numbers files sizes. The source code also has an executable file called winspeech.exe which encodes wave files.

Attachment 4694

Source code: http://ww1.microchip.com/downloads/e...otes/an643.zip

Quote:

---

Originally Posted by scalerobotics

I had not seen this before. I have not played the files from a PIC yet, but it sounds interesting as well. This C code :( uses a "simplified Adaptive Differential Pulse Code Modulation (ADPCM) algorithm". The files look slightly larger than the Roman Black files, but not by much, comparing to the numbers files sizes.

Source code: http://ww1.microchip.com/downloads/e...otes/an643.zip

---

See M25P32 PIC Audio.
Options include 16 bit ADPCM compression to 4 bit in Basic.

Sound quality far better than Roman Black's.
The M25P32 may not be as easily found by now as its transfered ownership.

Norm

Is it just me, or is anyone else having problems finding Norms code? Might be a firewall issue, I can click on the attachement, but I can't download it: http://www.protonbasic.co.uk/forum/a...1&d=1169530365 When I click on it IE says "Done" but I have a blank screen. And when I look on protons NEW site, i don't see any attachements, which does seem kind of strange. I became a member of their new forum, just in case only members could view attachements. Maybe I have to wait a little while?

In any case, thanks Norm! Very interesting work. I look forward to seing your proton code ... but not sure when I will be allowed to. :( Maybe when I get home to a less restrictive firewall I will have better luck.

Quote:

---

Originally Posted by scalerobotics

Is it just me, or is anyone else having problems finding Norms code? Might be a firewall issue, I can click on the attachement, but I can't download it: http://www.protonbasic.co.uk/forum/a...1&d=1169530365 When I click on it IE says "Done" but I have a blank screen. And when I look on protons NEW site, i don't see any attachements, which does seem kind of strange. I became a member of their new forum, just in case only members could view attachements. Maybe I have to wait a little while?

In any case, thanks Norm! Very interesting work. I look forward to seing your proton code ... but not sure when I will be allowed to. :( Maybe when I get home to a less restrictive firewall I will have better luck.

---

I'll update my website to include all the files.

The ADPCM basic code was ported from the full C ADPCM code.

Norm

Quote:

---

Originally Posted by scalerobotics

Is it just me, or is anyone else having problems finding Norms code? Might be a firewall issue, .

---

Hi, Walter

Lol ...

But ... would this help ???

Alain

Quote:

---

Originally Posted by Normnet

The ADPCM basic code was ported from the full C ADPCM code.

---

Ooooh! That sounds fantastic. I have some reading to do.

Uh..., yes, that does help a lot, thank you Alain! Thankfully, I am permitted to view them and download them as well! Grrrr to Corporate, Proton, or me. Not sure who to blame, but I will start with myself.

Thanks again Norm and Alain.

Walter

Quote:

---

Originally Posted by scalerobotics

Ooooh! That sounds fantastic. I have some reading to do.

Walter

---

No luck , the WavPlayer explanations are too big ... 1.6 Mb.

I can post it to you @ your private mailbox ...

Alain

Walter

Do you have the files?

Norm

Yes, thank you. I have all the files that Alain posted here. I do not have the 1.6 meg file he mentions, but I am sure that he will send it. ADPCM has me intrigued, so thanks again for your code.

Walter

Quote:

---

Originally Posted by scalerobotics

Yes, thank you. I have all the files that Alain posted here. I do not have the 1.6 meg file he mentions, but I am sure that he will send it.
Walter

---

Hi, Walter

Looks a little problem with your Mail :o

Quote:

---

This is the mail system at host mwinf5c07.orange.fr.
I'm sorry to have to inform you that your message could not
be delivered to one or more recipients. It's attached below.
For further assistance, please send mail to postmaster.
If you do so, please include this problem report. You can
delete your own text from the attached returned message.
The mail system
<[email protected]>: host gmail.com[74.125.43.27] said: 552
5.7.0 review our attachment guidelines. f21si12297915bkf.11

---

...
sorry ...

Alain

It sounded like you don't want to use a 16f because they either don't read their own memory and even if they do the memory is so small that it won't hold long sounds. My question is would a 16f88 for example, be able to pull the audio from a 1meg flash thorugh I2C or is that too slow? If it would work what size oscillator would you need? I hate asking these questions because I am normally for doing my own homework but in this case I don't even know where to start. I have owned PBP for about 10 years and I only use it periodically for simple toggling of IO to test out my hardware. This is probably my most agressive project so far.

Quote:

---

Originally Posted by Acetronics

Hi, Walter

Looks a little problem with your Mail :o


...
sorry ...

Alain

---

It could probably be made to work in pic16 chips .... but why not pic18? The include files in his post#5 of this thread were written for PIC18 chips.

Darrel says I2C is too slow:

Quote:

---

Originally Posted by Darrel Taylor

I did some experimenting with storing the sound in external EEPROM.
And for sure, you have to be using the MSSP module to read the data.
SHIFTIN or I2CREAD can't keep up with the higher bitrates.

---

Roman Blacks board answers the question for me. He is selling a board that takes in the output from his program. It runs on a 16f micro at 20mhz and with a 1 megabit I2C eeprom can store 24 seconds of sound.

http://www.blackrobotics.com/Brains_...ontrollers.htm

Quote:

---

Originally Posted by scalerobotics

It could probably be made to work in pic16 chips .... but why not pic18? The include files in his post#5 of this thread were written for PIC18 chips.

Darrel says I2C is too slow:

---

Here is an update to my project. I brought the roman black board and it does work. I tried many band pass filters and the quality is terrible. The only way to even get decent quality was to have a speaker in an enclosure. That did more to filter out the noise, but it still wasn't great. I settled on a windbond device. I wouldn't waste time or money with this method.

Look also at the 4D SOMO Module here: http://www.4dsystems.com.au/prod.php?id=73 .

Ioannis

schu4647-

Quote:

---

Here is an update to my project. I brought the roman black board and it does work. I tried many band pass filters and the quality is terrible. The only way to even get decent quality was to have a speaker in an enclosure. That did more to filter out the noise, but it still wasn't great. I settled on a windbond device. I wouldn't waste time or money with this method.

---

Hi, the BTc system was never designed for *high quality* sound! It's strangths are in other ways, mainly that the PIC microcontroller can also be used to do other tasks, all in one self contained module, and that is quite memory efficient.

Which speaker were you using that sounded terrible? For better sound quality I suggest a speaker with a larger diameter, ie 4" or 6". The tiny speakers that are supplied (free of charge!) with the product are not meant for good sound (they are "scratchy") and are only for small self contained applications like little robots, talking remote controls etc.

The BTc module sold is optimised for general purpose control applications and has source code, so you can use that one module to operate servos and play speech clips, or control a small device with some talking feedback. Or be used as a complete talking remote control all with the one PIC in the one module with its own 5v regulator and tiny low-sound quality amp already built in.

If all you want is something to play a sound, or specifically need high quality sound then there are plenty of options including the Winbond chips and $10 MP3 players from the supermarket.

Darrel,
I was reading your article on Wiki and decided to test it. I have a proto hardwares done (18F452 + 4MHZ cristal) and I would like very much to use this tool with my circuits. Firstlly I want to understand better some things.
I tested using RB_Sound Folder files (TestRBSound.pbp, DT_RBSound.pbp and the asm files like HitNew.asm, FascinatingNew.asm...)
in the TestRBSound.pbp is assigned:
btc_Pin1 - PORTC.2
btc_Pin2 - PORTC.3
LATE_LED - PORTC.5
well, I wired a led to the PortC.5. It is done
What I need to do with btc_Pin1 and btc_Pin2. How to wire and where?
...others files (Number folder) into the .asm files has a draw:
Digital label + resistor R = 25968 ohms and a 0.22uF capacitor to the ground and Analogue label on junction. Need I to put a speeker here or a buzzer? Or is necessary to use a pre-amplifier? Do you used the BTc.exe (BTC Sound Encoder from Roman Black)? I teste it using a Ding.wav from windows and exported to a ding.asm, but into the files is diferent comparing with your files...Into the one.asm for example you coment "File converted to .BSM format with the freeware btc_convertor by Darrel Taylos...Is it other converter diferent of the BTC Soundd Encoder?

Well...
Then I decided to wire a 26k resistor to portc.2 and a 0.22uF capacitor like the draw, and puted a buzzer on junction only to test. The portC.3 NC.
I compiled with sucess but it don't worked!
The led blink just one time and nothing occur.
I don't know what I need to use, if a buzzer, a speeker, a amplifier, how to wire it with certainly but the led should to blink basead on program!
Then I alterate the program to blink 5 times, and then the led blinked 5 times but after it stoped again...I percever the after @ PlaySound BigGun line, nothing occur (led always low, buzzer low)
If you can help me to use it I would be very happy.
I really apreciate it and woul like to use in my projects