[sdiy] Reconstruction Filters

[Olivier Gillet]

It's a different matter. Here the problem is not reconstruction
filtering (killing everything above sr / 2), but more importantly
killing the nasty PWM carrier due to the fact that PWM is used instead
of proper DAC.

If I had used proper DACs, I wouldn't have bothered with a
reconstruction filter on the oscillator signals, and would have used a
simple 1 pole for the control signals too.

It's a deliberate choice for being cheap and somehow lo-fi :)

On Mon, Feb 14, 2011 at 12:10 AM, James Hughes <james at virtualjames.com> wrote:
> This excerpt from the Shruthi-1 site talks about their filtering strategy:
> mutable-instruments.net/static/documentation/smr4_analysis.pdf
>
> -------
>
> All input signals, be it the raw oscillators signal or the control
> voltages, are 0 / 5V PWM signals
> with a carrier frequency of 20M Hz = 39062Hz/512.
>
> This implies that the control signals have to be smoothed. You
> shouldn’t feel bad about it: the
> MCU generates the control signals at a “slow" rate of 976Hz anyway, so
> a simple 1-pole low-pass
> with a cutoff frequency of 1.25kHz kills the PWM carrier by 30dB while
> still tracking fast enough
> the fastest transitions the MCU can create.
>
> The oscillators signal also has this 39kHz carrier. It is attenuated
> by the main 4-pole LPF itself,
> and by an extra 1-pole LPF with a fixed cutoff frequency nearing 20kHz
> at the final output buffer.
> In the worst case (cutoff set to its maximum value), the carrier is
> thus attenuated by 30dB. In
> most cases, however, the cutoff frequency is set to a lower value, and
> the carrier is attenuated more
> strongly. Anti-aliasing filters at the input of your soundcard,
> speakers with a limited bandwidth, or
> your ears will be doing the rest of the filtering.
>
> -------
>
> Does that help? Synth noob here, so disregard if it looks like I have
> no idea what I'm talking about (I don't).
>
> Best,
> James
>
> On Sun, Feb 13, 2011 at 1:52 PM, Matthew Smith <matt at smiffytech.com> wrote:
>>
>> Quoth Tristan at 13/02/11 22:40...
>>>
>>> If you don't want to dynamically change the filter frequency after the note has started, only require a small number of frequencies and/or a steep filter with a high number of poles, then then switching resistors may be acceptable. But for something like a 4 pole LPF the VCF and DAC might actually work out to be less complex and much more flexible.
>>
>> I'm actually treating reconstruction filtering and any 'creative' filtering separately - the reconstruction filter being part of the DCO, module. Just to tidy-up the waveform before it goes to the VCF module for fancy stuff.
>>
>> --
>> Matthew Smith
>>
>> Blog: www.smiffysplace.com  Business: www.smiffytech.com
>> www.linkedin.com/in/smiffy  www.flickr.com/photos/msmiffy
>> twitter.com/smiffy
>> _______________________________________________
>> Synth-diy mailing list
>> Synth-diy at dropmix.xs4all.nl
>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>