Hello,
I'd like to calculate the amplitude loss of a signal when it's run
through a bandpass filter, currently using butterbp.
The rudimentary observations are:
The narrower the bandwidth, the less amplitude the output signal will have.
Also, the gain loss is not equal across the frequency range.
Basically, lower frequencies have more gain loss at the same bandwidth
than higher frequencies.

I tried to adjust empirically, but I don't seem to find anything that
gives a close to unity gain for all (any) center frequency and
bandwidth combinations.
I could use balance to do it, but since the amplitude variation can be
quite large, balance does not work really well either. As balance
needs a short amount of time (ihp) to find rms and do the balancing,
excessive transients are likely to bleed through, and adjusting ihp to
a very low value can cause distortion due to rapid gain changes.

Maybe there's another bandpass filter I could use that has a close to
unity gain output regardless of cutoff and bandwidth setting ?

best
Öeyvind


Send bugs reports to this list.
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

... I'm aware that the output level of course would depend on the
spectrum of the input signal, e.g. if there's any energy in the
passband at all, so assume white noise as input.
best
Oeyvind

2009/12/4 Oeyvind Brandtsegg :
> Hello,
> I'd like to calculate the amplitude loss of a signal when it's run
> through a bandpass filter, currently using butterbp.
> The rudimentary observations are:
> The narrower the bandwidth, the less amplitude the output signal will have.
> Also, the gain loss is not equal across the frequency range.
> Basically, lower frequencies have more gain loss at the same bandwidth
> than higher frequencies.
>
> I tried to adjust empirically, but I don't seem to find anything that
> gives a close to unity gain for all (any) center frequency and
> bandwidth combinations.
> I could use balance to do it, but since the amplitude variation can be
> quite large, balance does not work really well either. As balance
> needs a short amount of time (ihp) to find rms and do the balancing,
> excessive transients are likely to bleed through, and adjusting ihp to
> a very low value can cause distortion due to rapid gain changes.
>
> Maybe there's another bandpass filter I could use that has a close to
> unity gain output regardless of cutoff and bandwidth setting ?
>
> best
> Öeyvind
>


Send bugs reports to this list.
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

You can try eqfil.

Victor

On 3 Dec 2009, at 23:03, Oeyvind Brandtsegg wrote:

> Hello,
> I'd like to calculate the amplitude loss of a signal when it's run
> through a bandpass filter, currently using butterbp.
> The rudimentary observations are:
> The narrower the bandwidth, the less amplitude the output signal  
> will have.
> Also, the gain loss is not equal across the frequency range.
> Basically, lower frequencies have more gain loss at the same bandwidth
> than higher frequencies.
>
> I tried to adjust empirically, but I don't seem to find anything that
> gives a close to unity gain for all (any) center frequency and
> bandwidth combinations.
> I could use balance to do it, but since the amplitude variation can be
> quite large, balance does not work really well either. As balance
> needs a short amount of time (ihp) to find rms and do the balancing,
> excessive transients are likely to bleed through, and adjusting ihp to
> a very low value can cause distortion due to rapid gain changes.
>
> Maybe there's another bandpass filter I could use that has a close to
> unity gain output regardless of cutoff and bandwidth setting ?
>
> best
> Öeyvind
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"



Send bugs reports to this list.
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Reading  again, perhaps eqfil will not do what you want. It boots or  
attenuates a given
band, but lets other pass unchanged. It looks like you need to remove  
all but the band.
On 4 Dec 2009, at 08:16, Victor Lazzarini wrote:

> You can try eqfil.
>
> Victor
>
> On 3 Dec 2009, at 23:03, Oeyvind Brandtsegg wrote:
>
>> Hello,
>> I'd like to calculate the amplitude loss of a signal when it's run
>> through a bandpass filter, currently using butterbp.
>> The rudimentary observations are:
>> The narrower the bandwidth, the less amplitude the output signal  
>> will have.
>> Also, the gain loss is not equal across the frequency range.
>> Basically, lower frequencies have more gain loss at the same  
>> bandwidth
>> than higher frequencies.
>>
>> I tried to adjust empirically, but I don't seem to find anything that
>> gives a close to unity gain for all (any) center frequency and
>> bandwidth combinations.
>> I could use balance to do it, but since the amplitude variation can  
>> be
>> quite large, balance does not work really well either. As balance
>> needs a short amount of time (ihp) to find rms and do the balancing,
>> excessive transients are likely to bleed through, and adjusting ihp  
>> to
>> a very low value can cause distortion due to rapid gain changes.
>>
>> Maybe there's another bandpass filter I could use that has a close to
>> unity gain output regardless of cutoff and bandwidth setting ?
>>
>> best
>> Öeyvind
>>
>>
>> Send bugs reports to this list.
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe csound"
>
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"



Send bugs reports to this list.
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"

Another thing to try is a higher-order butterworth (I assume you are  
using 2nd order).
On 4 Dec 2009, at 08:27, Victor Lazzarini wrote:

> Reading  again, perhaps eqfil will not do what you want. It boots or  
> attenuates a given
> band, but lets other pass unchanged. It looks like you need to  
> remove all but the band.
> On 4 Dec 2009, at 08:16, Victor Lazzarini wrote:
>
>> You can try eqfil.
>>
>> Victor
>>
>> On 3 Dec 2009, at 23:03, Oeyvind Brandtsegg wrote:
>>
>>> Hello,
>>> I'd like to calculate the amplitude loss of a signal when it's run
>>> through a bandpass filter, currently using butterbp.
>>> The rudimentary observations are:
>>> The narrower the bandwidth, the less amplitude the output signal  
>>> will have.
>>> Also, the gain loss is not equal across the frequency range.
>>> Basically, lower frequencies have more gain loss at the same  
>>> bandwidth
>>> than higher frequencies.
>>>
>>> I tried to adjust empirically, but I don't seem to find anything  
>>> that
>>> gives a close to unity gain for all (any) center frequency and
>>> bandwidth combinations.
>>> I could use balance to do it, but since the amplitude variation  
>>> can be
>>> quite large, balance does not work really well either. As balance
>>> needs a short amount of time (ihp) to find rms and do the balancing,
>>> excessive transients are likely to bleed through, and adjusting  
>>> ihp to
>>> a very low value can cause distortion due to rapid gain changes.
>>>
>>> Maybe there's another bandpass filter I could use that has a close  
>>> to
>>> unity gain output regardless of cutoff and bandwidth setting ?
>>>
>>> best
>>> Öeyvind
>>>
>>>
>>> Send bugs reports to this list.
>>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>>> "unsubscribe csound"
>>
>>
>>
>> Send bugs reports to this list.
>> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
>> "unsubscribe csound"
>
>
>
> Send bugs reports to this list.
> To unsubscribe, send email sympa@lists.bath.ac.uk with body  
> "unsubscribe csound"



Send bugs reports to this list.
To unsubscribe, send email sympa@lists.bath.ac.uk with body "unsubscribe csound"