|
Ok, part two of the pissing contest, why not, I've nothing better to do (God forbid I go and make some music when I can flame with other randoms on a buliten board). I probably should apologize for arrogance towards you but it was somewhat caused by your arrogance towards me and others in this discussion (by the attitude and post count, I assume you're among the forum's noblety, right?).
| quote: | Originally posted by DigiNut What the hell... proprietary? Do you even know what that word means? Here you are correcting my terminology and you can't even string together a coherent sentence.
First of all, it doesn't matter whether you call it granular synthesis or granular resynthesis - neither is actually correct, since it's really just an effect. However, it is almost universally known as granular synthesis - go ahead and search on google, you are probably going to find 100 times as many results for the former term. |
Well, I'm not natively english speaking, the term "proprietary" I used as "at one's own will", I cannot always remember the right english term in heat of discussion (and in this case it should have been "any wanted position" or whatever) however semantic difference between synthesis (generation) and resynthesis (re-generaton from an existing source, in this case the original audio) is the same in any european language.
It really doesen't matter which term is more oftenly misused, if one of them is semantically more correct.
| quote: | Originally posted by DigiNut
Granular synthesis is a glorified gate. The only differences are:
1. It uses slices of usually less than 50 ms (which is about the cutoff point where the human ear can hear discrete events);
2. It has a linear and ramped attack and release, unlike a gate which is simply a pulse, and the attack and release are equal; and
3. Some GS plugins allow you to perform additional processing on the individual slices. |
Wrong. If it's "glorified" anything it's a "glorified" digital delay line (it buffers and replays audio), but that's only a tad bit closer than comparing it to a gate.
If it's a sampler doing granular resynthesis, it does what I said, "scans" through the sound and plays small parts of it, not nessesarily in the exact order they're in the original part. Smoothing isn't even required, it's still granular even without smoothing.
If it is an effect, then it needs to buffer audio so it can repeat it, rearrange it etc.
Gates don't buffer. What you refer to above as "gate" is a popular "trance gate effect" more precisely a sidechained gate processor i.e. controlling a gate's amplifier with ANOTHER signal (or controller value) to silence or radically decrease the volume of a sound. The traditional dynamic gate consists of two parts, a level detector and a controlled amplifier. Once the level detector detects input signal's amplitude (or RMS) falling under certain treshold, it then makes the amplifier silence the output. When the input signal returns above treshold it restores the volume. It doesen't buffer.
Or a more graphical example. This is our original sound, each letter representing, say, 25ms of audio and lowercse meaning lower volume:
ABcDEfGHiJKLmNOP
This is what sidechained gate set to cut off every 100ms and hold 50ms would do:
AB..Ef..iJ..mN..
This is what dynamic gate with treshold set to "lowercsse" would do:
AB.DE.GH.JKL.NOP
This is what granular resynthesis set to spread 50ms chunks (sampled at every 25ms) at 75ms intervals would do:
AB.Bc.cD.DE.Ef.fG.GH etc.
With the spread less than chunk size we get overlaps and phase-canceling which results in a flanger-like sound. I've never heard any gate produce that flanger-like effect.
| quote: | Originally posted by DigiNut
Stuttering by definition is repeating the same sound. |
Yes this bit is (the only one that is) totally and indisputably correct, and Fruity Granulizer has a control (hold) that will make it repeat the same audio slice indefinetely (or until the hold control is toggled off). It's a sampler, it doesen't even have to buffer audio (like effect granulizers need to).
Your idea of F. Granulizer is very strange. From your assumption that it's a processing plugin, and not a generating one (it's basically a sampler) I asume you haven't really used it, or haven't really used it more than just a peak at what it does.
Or you're just being headstrong insisting that you were right (which is ok with me, since I can be like that, except now I'm just right  )
| quote: | Originally posted by DigiNut
Using MIDI-triggered gates for stuttering is nearly impossible, since MIDI timing is about as accurate as a sundial and nowhere near precise enough to handle the miniscule slices that are usually used in a proper stutter.You need plugins with sample-accurate timing, and you have to tell them exactly when to start buffering and when to start repeating the buffer - which is essentially the exact same thing as doing the edits by hand, except it's LESS flexible, because you're limited to whatever effects the plugin can produce as opposed to your entire array of effect plugins if you did it by hand! |
You really didn't understand what I was saying. I was not talking about gating (what is it with you tranceheads and gating?), I was talking about simply retriggering a sample, where you've previously cut the audio at places where you want stutters, and loading the pieces in a sampler program mapping them to different keys.
I fail to see how on earth could you fet the impression I was talking about gating.
And btw, at the rate of 700+ events per quarter note (ppq or tpq) you can be pretty darn precise. Syncing anything you do in MIDI to some freely streaming audio would be a real bitch, but I wasn't suggesting anything like that here (I was suggesting it above, with dblue Glitch, but I wasn't suggesting being anal retentive).
In real life, no matter what his nano-highness said, you don't really need to be that anal retentive and sample acurate. In real life human hearing doesn't have that temporal precission so missing a plosive would sound different then hitting a plosive, but not at "event" level, but at "timbre" level. We get that good old Stockhausen''s transition between events and timbre.
Even more so, stutters can sound interesting even without being "plosive-aware" as you suggested. Neither BT nor Hybrid invented them, there is no single authority in this world that decides what a "good" stutter is.
I've been doing stutters for a long time (from using 9xx and E9x effects in FT II back in '97 to slicing audio, retriggering and fucking about with IDM-ish plugins today), so I think I know pretty much what I'm talking about.
No wonder a lot of trance sounds sameish whit that "only do it the right way" mentality. (It's a joke, a bad one, so keep you knickers on, most dance music sounds sameish, yes, I'm aware, it's the same anti-experimental copycat mentality that's causing it too)
| quote: | Originally posted by DigiNut
FYI, the "low-quality timestretch" on older D'n'B tracks is just linear resampling, unless there's some particular effect you're talking about that I'm not aware of. |
You probably don't know what effect I'm talking about, since no one can mistake that flangy stuttery sound for resampling. It's a timestretching effect on old Akai samplers and low quality stretching algos in software (Cool Edit 96 anyone?) that works on the almost exatcly the same principle as F. Granulizer. It would take the original ABCDEFGH sample and repeat slices in the ABCBCDCDEDEFEFGFGH fashion with some smoothing envelopes in between (or no envelope at all).
This discussion is in desperate need of some audio clips, since now I'm pretty sure a lot of this negativity is caused simply by missunderstanding.
| quote: | Originally posted by DigiNut
And as for the FL native plugins, functionally speaking they are as versatile as many other plugins out there; however, it's the sound quality on them that sucks. If you think it's simply the users that suck, then feel free to put your money where your mouth is and show us what you can do with your precious Fruity Granulizer. |
Sure, that I will most gladly do. However, what will stop you from going all "golden ears", "sensual, full buqet, but not too ripe, a breath of cherry" on me and claim that you hear some "lack of depth and clarity", us mere mortals cannot hear?
Granular resynthes (in the form needed for stuttery effects or better said, artifacts) is not in great need of very smart DSP. It's a pretty simple effect, there are no quality issues that could arise.
But for the sake of the original poster and others interested in the matter, I will provide both zipped FLPs and mp3 samples to illustrate my points.
|
HEY HEY HEY PEOPLE!!!!
