Timeline for Design narrow bandpass filter for signal with high sampling rate
Current License: CC BY-SA 3.0
11 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 7, 2017 at 14:39 | vote | accept | lR8n6i | ||
| Aug 5, 2017 at 6:24 | comment | added | Marcus Müller | No, that indicates a problem with the software you wrote! | |
| Aug 4, 2017 at 22:38 | comment | added | lR8n6i | Ok. That was also my expectation. Thanks for your confirmation :) I still got memory issues when I used more than 5000 tabs. Does this indicate that my filter was unrealistically specified? | |
| Aug 4, 2017 at 22:34 | comment | added | Marcus Müller | well, length constraints might arise from the latency you can tolerate, or from processing effort. Ignore for now. 100*4000 samples is not a lot of samples. I thought I made that clear. | |
| Aug 4, 2017 at 22:24 | comment | added | lR8n6i | The other thing is that the capture duration is more than 100s. So we have quite a lot samples... | |
| Aug 4, 2017 at 22:16 | comment | added | lR8n6i | What is exactly meant by "length constraints"? What do I need to specify there? | |
| Aug 4, 2017 at 22:14 | comment | added | lR8n6i | Stop band attenuation: 40-60dB. Transition with: 1-3 Hz. Ripple: very low (like a Bessel filter) | |
| Aug 4, 2017 at 22:04 | comment | added | Marcus Müller | From a DSP perspective, generally, yes, it'd be more efficient. From a nexessity perspective: nah, totally superfluous, the rates you're dealing with allow you to use humongous filters, and your requirements certainly won't make things humongous. | |
| Aug 4, 2017 at 22:02 | history | edited | Marcus Müller | CC BY-SA 3.0 |
added 1479 characters in body
|
| Aug 4, 2017 at 22:00 | comment | added | lR8n6i | Hi Marcus. Thanks for this interesting answer. So you don't recommend an initial downsampling of the signal. | |
| Aug 4, 2017 at 21:49 | history | answered | Marcus Müller | CC BY-SA 3.0 |