[Csnd] Issues with dcblock, dcblock2
| Date | 2022-08-25 14:39 |
| From | Steven Yi |
| Subject | [Csnd] Issues with dcblock, dcblock2 |
| Attachments | image.png image.png image.png image.png vco2_D0.csd |
Hi All,
Csound mailing list
Csound@listserv.heanet.ie
https://listserv.heanet.ie/cgi-bin/wa?A0=CSOUND
Send bugs reports to
https://github.com/csound/csound/issues
Discussions of bugs and features can be posted here
I'm trying to figure out some oddities with dcblock and dcblock2. I'm using a vco2 sawtooth tuned to D0 and I see: * With dcblock, it seems to introduce DC * With dcblock2, it rolls off the low end I've pasted screenshots below of time domain and spectral plots from Audacity. I've also attached my test CSD to this email. Is there something I'm missing? Thanks! Steven Time Domain Plain VCO2: dcblock dcblock2 |
| Date | 2022-08-25 14:53 |
| From | Steven Yi |
| Subject | Re: [Csnd] Issues with dcblock, dcblock2 |
| Attachments | image.png image.png image.png image.png |
Following up, I should have looked around for some info about frequency response of DC blockers: https://www.dsprelated.com/freebooks/filters/DC_Blocker.html That said, dcblock2's low end response seems very steep compared to dcblock. I'm not really sure what happened with dcblock for it to cause DC either. On Thu, Aug 25, 2022 at 9:39 AM Steven Yi <stevenyi@gmail.com> wrote:
|
| Date | 2022-08-25 19:08 |
| From | ST Music |
| Subject | Re: [Csnd] Issues with dcblock, dcblock2 |
| Attachments | image.png image.png image.png image.png 20220825_135053.png Screenshot_20220825-131138.jpg |
A few notes, not that this resolves anything. As can be seen in the attachments, the DC blocks are literally And dramatically altering waveshapes. Whether this would be the case with far more complex waveforms as opposed to simple ones I don't know, it perhaps warrants investigating. Note I attached 2 images. The middle saw wave example does not in fact appear to have a DC offset, to me it looks as it should for a saw wave that has a convex tail so to speak. I've seen this before from waveform generators. I cropped the middle rendered section and there is no reported DC offset (I checked in two editors and got the same). By appearances, the section of the saw wave marked A would probably have the same area as B - they have different offsets from 0 but also substantially different widths. I can't mathematically prove this so just taking a guess. The image of the square waves which I rendered would possibly to support the idea that DC is not being introduced. In the end, as it stands, it seems that using the DC block becomes more of an issue of practical necessity and not so much as a safety net so to speak as it seems to have it's own side effects. I know my reply isn't particularly helpful but wanted to point out a few things I noticed, apologies if they seem redundant. As an aside, I had to laugh when I saw the third rendered square wave as it closely resembles a well known shape that wave folders produce referred to as The Batman shape. I actually did a little video recently that shows this: On Thu, Aug 25, 2022, 9:40 AM Steven Yi, <stevenyi@gmail.com> wrote:
|