Frequency multiplication required

Hi,

I need to multiply a low frequency (human heart beat measured by a piezo
crystal)
of aprox one hertz by say 200, so that it would be audible. Realtime, with
minimum delay.
I wish to use it for biofeedback.
All its harmonics should be multiplied too.

How do I do it ? On PC or using a processor like Atmel AVR@20Mhz ?

Regards and thanks,

Devidas

devidasbhonde wrote:
> Hi,
> 
> I need to multiply a low frequency (human heart beat measured by a piezo
> crystal)
> of aprox one hertz by say 200, so that it would be audible. Realtime, with
> minimum delay.
> I wish to use it for biofeedback.
> All its harmonics should be multiplied too.
> 
> How do I do it ? On PC or using a processor like Atmel AVR@20Mhz ?
> 
> Regards and thanks,
> 
> Devidas

Let's see if I get this straight. You have a signal that is periodic at 
approximately one Hz. You want to speed it by a factor of 200, so that 
it becomes a pitch approximately G or G# below middle C. Right so far?

This is easy to do by collecting samples at a suitable rate and playing 
them back 200 times faster. What does "real time" mean here?

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

devidasbhonde wrote:
> Hi,
> 
> I need to multiply a low frequency (human heart beat measured by a piezo
> crystal)
> of aprox one hertz by say 200, so that it would be audible. Realtime, with
> minimum delay.
> I wish to use it for biofeedback.
> All its harmonics should be multiplied too.
> 
> How do I do it ? On PC or using a processor like Atmel AVR@20Mhz ?
> 
> Regards and thanks,
> 
> Devidas
> 
> 
> 
Heart sounds are audiable without any processing.  I think you want to 
extract the beat to beat pediod and convert that to a tone but when you 
mentioned harmonics, I don't really understand what your looking for.

Also healthy hearts have a fair amount of beat to beat variability.

devidasbhonde wrote:

> Hi,
> 
> I need to multiply a low frequency (human heart beat measured by a piezo
> crystal)
> of aprox one hertz by say 200, so that it would be audible. Realtime, with
> minimum delay.
> I wish to use it for biofeedback.
> All its harmonics should be multiplied too.
> 
> How do I do it ? On PC or using a processor like Atmel AVR@20Mhz ?
> 
> Regards and thanks,
> 
> Devidas
> 
> 
> 


As he mentioned "biofeedback" and you can not "control something 200 
times faster than it occurs", I wonder if he suffered confusion of terms.

Interpretation #1
He was really trying to describe AM modulation of a tone.

Interpretation #2
He was really trying to describe FM modulation of a tone by the average 
period between beats.

I'd lean towards #2 as it would seem to match what I vaguely recall of 
the goals of biofeedback.

OP needs to specify goal not method.

Richard Owlett wrote:
> devidasbhonde wrote:
> 
>> Hi,
>>
>> I need to multiply a low frequency (human heart beat measured by a piezo
>> crystal)
>> of aprox one hertz by say 200, so that it would be audible. Realtime, 
>> with
>> minimum delay.
>> I wish to use it for biofeedback.
>> All its harmonics should be multiplied too.
>>
>> How do I do it ? On PC or using a processor like Atmel AVR@20Mhz ?
>>
>> Regards and thanks,
>>
>> Devidas
>>
>>
>>
> 
> 
> As he mentioned "biofeedback" and you can not "control something 200 
> times faster than it occurs", I wonder if he suffered confusion of terms.

Maybe more than just terms. There may be some unwarranted assumptions 
that need examining.

> Interpretation #1
> He was really trying to describe AM modulation of a tone.
> 
> Interpretation #2
> He was really trying to describe FM modulation of a tone by the average 
> period between beats.
> 
> I'd lean towards #2 as it would seem to match what I vaguely recall of 
> the goals of biofeedback.

I wouldn't want to superpose my assumptions on his. I'll wait for 
clarification.

> OP needs to specify goal not method.

That's always good. In this case, one needs to distinguish heartbeats as 
events from the frequencies generated by those events. Raising the 
frequency without creating virtual tachycardia may prove difficult. 
Creating a pitch inversely proportional to the period between beats is 
simple, but not directly applicable to harmonics.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

devidasbhonde wrote:
> Hi,
> 
> I need to multiply a low frequency (human heart beat measured by a piezo
> crystal)
> of aprox one hertz by say 200, so that it would be audible. Realtime, with
> minimum delay.
> I wish to use it for biofeedback.
> All its harmonics should be multiplied too.
> 
> How do I do it ? On PC or using a processor like Atmel AVR@20Mhz ?

If I understand you correctly, you are looking at the same thing that
kreuters is asking about in the thread "frequency shifting
(compressing?)" in this newsgroup except that you want to scale up in
frequency rather than down.  As I responded to him, if you find a
solution at that extreme a scale value please come back here and share
the results.


Bob
-- 

"Things should be described as simply as possible, but no simpler."

                                             A. Einstein

Jerry Avins wrote:
> devidasbhonde wrote:
> > Hi,
> >
> > I need to multiply a low frequency (human heart beat measured by a piezo
> > crystal)
> > of aprox one hertz by say 200, so that it would be audible. Realtime, with
> > minimum delay.
> > I wish to use it for biofeedback.
> > All its harmonics should be multiplied too.
> >
> > How do I do it ? On PC or using a processor like Atmel AVR@20Mhz ?
> >
> > Regards and thanks,
> >
> > Devidas
>
> Let's see if I get this straight. You have a signal that is periodic at
> approximately one Hz. You want to speed it by a factor of 200, so that
> it becomes a pitch approximately G or G# below middle C. Right so far?
>
> This is easy to do by collecting samples at a suitable rate and playing
> them back 200 times faster. What does "real time" mean here?

I humbly think that your solution wouldn't work here since, as I
understand it, it's supposed to play a different frequency depending on
the heart beat rate, and that real time here means that it should be
"live", for example, you would hear "in real time" the signal changing
as the heart would change it's rhythm

Stan Pawlukiewicz wrote:
> devidasbhonde wrote:
> > Hi,
> >
> > I need to multiply a low frequency (human heart beat measured by a piezo
> > crystal)
> > of aprox one hertz by say 200, so that it would be audible. Realtime, with
> > minimum delay.
> > I wish to use it for biofeedback.
> > All its harmonics should be multiplied too.
> >
> > How do I do it ? On PC or using a processor like Atmel AVR@20Mhz ?
> >
> > Regards and thanks,
> >
> > Devidas
> >
> >
> >
> Heart sounds are audiable without any processing.  I think you want to
> extract the beat to beat pediod and convert that to a tone but when you
> mentioned harmonics, I don't really understand what your looking for.
>
> Also healthy hearts have a fair amount of beat to beat variability.

I think that by harmonics he means he doesn't want it to be ~1 Hz + 200
Hz but rather ~1 Hz * 200. Thus a change of 0.1 Hz in the heart beat
rhythm would be heard as a change of 20 Hz

Michel Rouzic wrote:
> Jerry Avins wrote:
> 
>>devidasbhonde wrote:
>>
>>>Hi,
>>>
>>>I need to multiply a low frequency (human heart beat measured by a piezo
>>>crystal)
>>>of aprox one hertz by say 200, so that it would be audible. Realtime, with
>>>minimum delay.
>>>I wish to use it for biofeedback.
>>>All its harmonics should be multiplied too.
>>>
>>>How do I do it ? On PC or using a processor like Atmel AVR@20Mhz ?
>>>
>>>Regards and thanks,
>>>
>>>Devidas
>>
>>Let's see if I get this straight. You have a signal that is periodic at
>>approximately one Hz. You want to speed it by a factor of 200, so that
>>it becomes a pitch approximately G or G# below middle C. Right so far?
>>
>>This is easy to do by collecting samples at a suitable rate and playing
>>them back 200 times faster. What does "real time" mean here?
> 
> 
> I humbly think that your solution wouldn't work here since, as I
> understand it, it's supposed to play a different frequency depending on
> the heart beat rate, and that real time here means that it should be
> "live", for example, you would hear "in real time" the signal changing
> as the heart would change it's rhythm

Indeed. It appears that your understanding of his problem exceeds 
Davidas's. I hope to get him to think more about what he really wants to do.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

Michel Rouzic wrote:
> Stan Pawlukiewicz wrote:
> 
>>devidasbhonde wrote:
>>
>>>Hi,
>>>
>>>I need to multiply a low frequency (human heart beat measured by a piezo
>>>crystal)
>>>of aprox one hertz by say 200, so that it would be audible. Realtime, with
>>>minimum delay.
>>>I wish to use it for biofeedback.
>>>All its harmonics should be multiplied too.
>>>
>>>How do I do it ? On PC or using a processor like Atmel AVR@20Mhz ?
>>>
>>>Regards and thanks,
>>>
>>>Devidas
>>>
>>>
>>>
>>
>>Heart sounds are audiable without any processing.  I think you want to
>>extract the beat to beat pediod and convert that to a tone but when you
>>mentioned harmonics, I don't really understand what your looking for.
>>
>>Also healthy hearts have a fair amount of beat to beat variability.
> 
> 
> I think that by harmonics he means he doesn't want it to be ~1 Hz + 200
> Hz but rather ~1 Hz * 200. Thus a change of 0.1 Hz in the heart beat
> rhythm would be heard as a change of 20 Hz

He wants to multiply all frequencies by 200 without dividing all periods 
by the same amount. Can you think of a way to do that? I can't.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������