nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpq4ns08s2g76tu7v2pnxtupq363rtw2p34jxturt74u7apqfdh7ctqu2mc3s (nprofile…mc3s) Here's a relevant video. People call this effect "filter ringing", but it's just a natural consequence of applying a phase change.
https://youtu.be/H3ZyaXgTqjM?si=Gvz1HEuomidFuMon
The DFFT relationship is interesting to think about. The length of the DFFT spectrum is equal to the length of the equivalent FIR. That means that the more DFFT bins you have (the more "graphic EQ bands" you have), the longer the resonance can be. For example, with 1000 "EQ bands" you can create a reverb that lasts up to 1000 samples.
In my example above I use a single EQ filter, but it's a complicated IIR filter so it doesn't fit neatly into few DFFT bins. You'd need a lot more EQ bands in a "graphic EQ" to replicate it. That's why the "reverb" (or "filter ringing" if you like) is audible. If I switch the EQ processing type to FIR, the reverb/ringing gets cut off since the code internally windows the IR, basically "quantizing" the EQ into a coarser response (fewer DFFT bins) that cannot represent it (obviously it's an EQ plugin, so the developers are assuming you aren't trying to build a reverb with it and that long ringing is not actually intended!)
Asahi Lina (朝日リナ) /
npub14w…zuequ
2025-03-21 08:54:02
in reply to nevent1q…7tau
Author Public Key
npub14w78207els8vs5fxduhhval0r9zgujpf2khcqyfuhmkt2tlyvcyq2zuequPublished at
2025-03-21 08:54:02Event JSON
{
"id": "7c6d3f581c13c93dfb8e2309a2e4187f5356df14dba0e7b4a38669b2632417f1",
"pubkey": "abbc753fd9fc0ec851266f2f7677ef19448e482955af80113cbeecb52fe46608",
"created_at": 1742547242,
"kind": 1,
"tags": [
[
"p",
"ace0f3c148f697cf31419997c0823a88d6e506359197c1afd5e7ba1025b7f616",
"wss://relay.mostr.pub"
],
[
"p",
"28c204cfbb6d4709485339c4e17e7485bb4450ab64112cf3092dd7cd7ae15d5b",
"wss://relay.mostr.pub"
],
[
"e",
"56510e682fbcd663b2dedf61166988661f69b141f0bace5ba9be9739b0096579",
"wss://relay.mostr.pub",
"reply"
],
[
"proxy",
"https://vt.social/users/lina/statuses/114199576053395514",
"activitypub"
]
],
"content": "nostr:nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpq4ns08s2g76tu7v2pnxtupq363rtw2p34jxturt74u7apqfdh7ctqu2mc3s Here's a relevant video. People call this effect \"filter ringing\", but it's just a natural consequence of applying a phase change.\n\nhttps://youtu.be/H3ZyaXgTqjM?si=Gvz1HEuomidFuMon\n\nThe DFFT relationship is interesting to think about. The length of the DFFT spectrum is equal to the length of the equivalent FIR. That means that the more DFFT bins you have (the more \"graphic EQ bands\" you have), the longer the resonance can be. For example, with 1000 \"EQ bands\" you can create a reverb that lasts up to 1000 samples.\n\nIn my example above I use a single EQ filter, but it's a complicated IIR filter so it doesn't fit neatly into few DFFT bins. You'd need a lot more EQ bands in a \"graphic EQ\" to replicate it. That's why the \"reverb\" (or \"filter ringing\" if you like) is audible. If I switch the EQ processing type to FIR, the reverb/ringing gets cut off since the code internally windows the IR, basically \"quantizing\" the EQ into a coarser response (fewer DFFT bins) that cannot represent it (obviously it's an EQ plugin, so the developers are assuming you aren't trying to build a reverb with it and that long ringing is not actually intended!)",
"sig": "f52e821b8ad020d3bd81aed73c58bbef964d7e272a4acf31fe1b97bb6ab9b2c667f147b583e6c56850356126c58366cd1e111dccd29fee672fc811fa1dbd1a0a"
}