hardware/audio eng help needed 
Author Message
 hardware/audio eng help needed

Hardware hackers, I need some help. I apologize for crossposting to these
diverse groups but I know a lot of you hang out in these groups.

I agreed to prototype a circuit for $2000, that will make the sound of a
particular animal, for use in knickknacks, greeting cards, etc. It had to
be 3VDC, to run off 2 watch batteries. I built the prototype with a 555
for a 16KHz clock, 2 4040's to generate addresses, and a 27LV256 ROM on
which I programmed a sampled sound from my customer's audio tape.

2 problems. 1, he doesn't want to bother getting copyright permission from
the company that made the tape. So I have to analyze (I did) and
synthesize (no luck yet) the sound myself so it doesn't violate any
copyrights. 2, it's not loud enough. To make it loud enough, I'll probably
have to use PWM and somehow drive the 3V both ways to get enough drive
voltage for the transducer to sound really loud.

I underbid the job. I've already spent over 40 hours and $500 on it and
just want to get it over with. I'm willing to pay up to $1000 to someone
with the expertise to get this working to my customer's satisfaction.

beeper). Obhack for alt.hackers: the 4040's with the ROM. I probably stole
it from somebody but it's a neat hack nonetheless. - jc

--

Home page: http://www.*-*-*.com/ ~jcomeau/
Disclaimer: Don't risk anything of value based on free advice.
"Anybody can do the difficult stuff. Call me when it's impossible."



Fri, 21 Jul 2000 03:00:00 GMT  
 hardware/audio eng help needed

Quote:

>Hardware hackers, I need some help. I apologize for crossposting to these
>diverse groups but I know a lot of you hang out in these groups.

>I agreed to prototype a circuit for $2000, that will make the sound of a
>particular animal, for use in knickknacks, greeting cards, etc. It had to
>be 3VDC, to run off 2 watch batteries. I built the prototype with a 555
>for a 16KHz clock, 2 4040's to generate addresses, and a 27LV256 ROM on
>which I programmed a sampled sound from my customer's audio tape.

>2 problems. 1, he doesn't want to bother getting copyright permission from
>the company that made the tape. So I have to analyze (I did) and
>synthesize (no luck yet) the sound myself so it doesn't violate any
>copyrights. 2, it's not loud enough. To make it loud enough, I'll probably
>have to use PWM and somehow drive the 3V both ways to get enough drive
>voltage for the transducer to sound really loud.

>I underbid the job.

Been there, done that.

Here's a freebie for ya, so it's worth what you pay for it.

If you were to do a fourier analysis on the wave, taking it out to enough
harmonics to make it possible to re-create the original with at least SOME
accuracy, you could do just that and re-build a waveform that's CLOSE (but
not exactly the same).  It's now "synthesized" using a set of fourier
coefficients.  THEN, you can tweek each coefficient until you like the
result.

DFT (Discrete Fourier Transform) calculations are pretty simple to code, but
require lots of floating point math.  If you treat the entire sound clip as
"a cycle", you can divide the total # of data points by 2 to get the
theoretical maximum # of harmonics that you can actually USE.  Thus, a 64k
ROM would have a maximum of 32,768 harmonics in the DFT.  This would be a
rather painful calculation, sure, but if you want to let a P5 200Mhz crunch
it for a day or 2 you could come up with the fourier coefficients that
synthesize the result.  THEN, you plug it back in using sin/cos values and
voila, you re-create the original "more or less".  THEN, apply some "random
tweeking" to the parameters, and experiment until you come up with a
satisfactory result.

OBHACK:  I once had a hard drive that I needed to use, a 40Mb Seagate drive.
I really needed the storage space.  I didn't have enough cash on hand to get
a bigger drive (this was quite a while back), so I kept the one I had, and
used this 40Mb to supplement the storage.  The problem was, this 40Mb drive
had a tendency to freeze up if I powered the machine down for any length of
time.  SO, what did I do to fix the problem?  I ran cables out teh back of
the machine, with the drive on top and mounting hardware attached, and
screws into the mounting brackets that raised the drive up off of the case
just enough to prevent it from shorting out.  THEN, whenever I started the
machine up, I'd take a pair of pliers and spin the drive spindle from the
bottom by hand, to make sure it wasn't "frozen up".  I did this for about a
year until I was able to get a large enough hard drive that I could throw it
away.



Mon, 24 Jul 2000 03:00:00 GMT  
 
 [ 2 post ] 

 Relevant Pages 

1. Audio DSP hardware job opening in Sony in Silicon Valley

2. Help Please - Need a module for Audio programming

3. Video and Audio Compression (software and hardware)

4. Need help with Hardware setup?

5. We need to know if hardware acceleration can help our VRML performance

6. Hardware keyboard handling... help needed

7. Hardware keyboard handling... help needed

8. Help! -- Need access to hardware

9. Need help accessing hardware

10. Need a Jpn->Eng Translator?

11. ECAD SW start-up near Boston needs tcl/tk GUI Eng, UNIX

12. Kids Need Hands-On Tech on Real Audio

 

 
Powered by phpBB® Forum Software