• Open Menu Close Menu
  • Apple
  • Shopping Bag
  • Apple
  • Mac
  • iPad
  • iPhone
  • Watch
  • TV
  • Music
  • Support
  • Search apple.com
  • Shopping Bag

Lists

Open Menu Close Menu
  • Terms and Conditions
  • Lists hosted on this site
  • Email the Postmaster
  • Tips for posting to public mailing lists
Re: Problems with a Low Pass Filter process
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Problems with a Low Pass Filter process


  • Subject: Re: Problems with a Low Pass Filter process
  • From: Brian Willoughby <email@hidden>
  • Date: Sat, 04 Jun 2011 12:19:20 -0700

Herve,

I suggest that you grab some basic audio DSP texts and study the background information. But I can quickly point out a few mistakes:

The BitCrunch noise is due to clipping. Your gain weighting factors are huge. The multiple 2* factors represent significant gain that you never correct for, and this distorts the audio. The common solution is to pre-multiply your weighting factors such that they all add up to < 1.0, and then you won't have distortion. The algorithm that you're using to calculate alpha, beta, and gamma is convoluted enough to make this harder that it should normally be, and a simple low-pass is usually easier than this.

The cutOff problems are probably related to your filter algorithm. Check your references on this. Some filters are only defined for fs/ 4 and below, some are only defined for fs/4 and above. Since digital audio only carries frequencies between 0 and fs/2, this fs/4 frequency is right in the "middle" as you say. Other filters have different usable ranges for their cutOff, so if you want to use the full range from 0 to fs/2 then you'll need a different filter algorithm.

Brian Willoughby
Sound Consulting


On Jun 4, 2011, at 08:34, herve.noury wrote:
float beta = 0.5 * ((1.0 - theSin) / (1.0 + theSin));
float gamma = (0.5 + beta) * cos(theta);
float alpha = (0.5 + beta - gamma)/4.0;

left[frame] += 2*(alpha*(leftB[frame] + 2*leftB[k] + leftB[j]) + gamma*leftB[k] - beta*leftB[j]);

The result is just noisy : when the cutOf is in the middle, no effect, but when I want to filter the additive synth, it sounds like a BitCrunch effect. When Qval is too high, it crashes...


_______________________________________________
Do not post admin requests to the list. They will be ignored.
Coreaudio-api mailing list      (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden


References: 
 >Problems with a Low Pass Filter process (From: "herve.noury" <email@hidden>)

  • Prev by Date: Re: Can MIDIPackets be modified after they are sent?
  • Next by Date: Re: Create MIDI device programmaticly without a MIDIDriver bundle
  • Previous by thread: Problems with a Low Pass Filter process
  • Next by thread: Re: Problems with a Low Pass Filter process
  • Index(es):
    • Date
    • Thread