|
cum in my meowth posted:This one's killing me 'cause it's so simple but I just can't reach it. It's the blip synth that starts at 3:22 and just goes EBG http://www.youtube.com/watch?v=xtK2ApmBUw4 It sounds like detuned squarewaves but my attempts with stacked subtractors are off. Lets break it down into steps. First lets start with what we know: 1) It is almost certainly derived from a single square wave or a combination of square waves in a high register. They are not tuned far apart (doesn't sound dischordant) and if there are multiple squares they are all tuned to the same octave. 2) It is monophonic and there is non linear portmento on it. 3) Sound is staccato with a fast attack and the sound stops abruptly. Therefore it probably has a short amp decay, extremely short amp release and extremely short amp attack. 4) Filter/Envelope is definitely non linear. First port of call - to make this sort of sound we want a synth that has really fast envelopes. Fast filter modulation is the important part of this sound since the timbre is very easy to make on any synth that can generate multiple square waves. I picked a minimoog emulation called Minimonsta. It has really really fast envelopes and the best thing? You can recursively modulate the filter -> envelope! This means you can change the 'shape' of its attack by having it modulate itself. I'll cut to the chase. Here is my best attempt after 15 minutes of dicking around: http://media.putfile.com/temp-9 Theres a touch of tape delay (Voxengo Analog Flux) and a stereo convolution reverb on it (Pristine Space using 2 impulses freely available on noisevault. These are 'Lexicon PCM91 Studio A - Direct Guitar.wav' and 'Lexicon PCM91 Plates - Dark Plate-Classic Lexicon!.wav) Because of the way Minimonsta's routing works you can't see everything thats going on in this screenshot so heres the fxp which I think you can use in the demo. Either way, you should probably buy it because its a great synth.
|
#
¿
Oct 6, 2008 22:13
|
|
|
|
| # ¿ Apr 26, 2024 02:01 |
|
|
As a general tip, I would discourage 'stacking' and 'layering' synth sounds to create complex sounds at first. This is just something I see alot of because many people do it with samples (especially one shot drum samples). The reason why I discourage it is because you don't really learn the mechanics of your synth by doing it and theres only so much sound shaping you can do with layering. Furthermore, layering can also introduce many issues like phasing problems and if you aren't paying attention then it can be hard to figure out where the problem is. Focus on making good sounding single patches before you get into ensemble patches. Even after 3 years I find myself rarely making ensemble patches. Even on my Xpander and Virus B, I rarely ever use multi mode and a I rarely ever layer synth sounds in FL Studio. Other than that, think it through logically. Familiarise yourself with linear and non linear changes in pitch, envelope times, filter cutoff etc so you know roughly what they sound like. I do this with an LFO set to control pitch or filter cutoff or whatever and you should be able to hear it clearly. After a while, you will begin to tell when portmento is linear or not or when filter -> amp envelope attack is non linear. You may not be able to tell the extent of it, but it gives you a place to start experimenting which is closer to where you want to go. Good luck.
|
|
#
¿
Oct 6, 2008 22:37
|
|
|
To make convincing vowel sounds out of a subtractive synth you need at least 3 highly resonant filters (and preferably alot more). You can do it with two but the range will be really limited and the articulation wont be that clear. I wouldn't even bother trying to make choir sounds with a subtractive synth that has only 1 resonant filter unless that synth has some proprietary gimmick up its sleave that helps make convincing vocal sounds. Mostly it seems impossible because many people try to make choir sounds on synths that really don't have the tools to be able to do it properly. If you have the right tools its not very hard. I have made a few choir patches on Virus B with two filters so I'll post clipits when I get home from work if theres any interest in this sort of thing. They only sound like choirs in 1 octave though so it is a somewhat frail illusion and may well illustrate why you need a filter bank to do it properly. Check out this thread for some samples of string/vocal sounds made on a Nord Modular. You can use any synth as long it has alot of resonant filters and it helps to have a real choir sample that you can use to model from. WanderingKid fucked around with this message at 11:42 on Oct 8, 2008 |
|
#
¿
Oct 8, 2008 11:37
|
|
|
wayfinder posted:Wouldn't a square wave be the better bet to simulate a human voice? It doesn't really matter what the raw material is. The most important thing is to model the formants right. To get basic vowel sounds you really need at least 3x highly resonant filters, preferably more. You don't strictly even need a signal since you can just drive all of the filters into self oscillation and that will do fine. The rest of it is to do with correct modulation of the filter and amplifier. The easiest way to do this is with a vocoder or talk box. You can manually do it with LFOs and ADSR envelopes if you have to but the results will be choir like at best. You won't be able to form intelligible words and talky sounds will mostly be gibberish because you don't have the necessary control - not like using your mouth/throat as the modulator. I'll post some vocalizations I made on a Virus B when I get home from work. Virus of course has only 2 resonant filters so its not quite there but you can get choral-ish sounds out of it for an octave or two around middle C. If you go to the extremes of the keyboard it loses the 'vocaloid' quality completely. Virus B doesn't have self oscillating filter models so I did have a signal dialled in. If I remember correctly they were usually sine waves mixed with wavetables. One of them I know for sure is derived from some kind of church organ and the others could be organs or dance pianos. Back in the Virus B days the wavetables weren't named so its hard to tell where they came from. The harmonic content of the signal is more or less irrelevant in the sense that all you really want is a shrill fundamental. The most important thing was to set the filter center frequencies/osc sine frequencies to the right places and drive up the resonance then start playing keys in the right register. Its also capable of talking gibberish but I've never made such a patch because that kind of thing is mostly unusable. If you don't have 3 or more filters you will get better results with a 4 to 6 op FM synth. If you have 2 filters you can get very limited results. If you have a synth with only 1 filter then forget about it. WanderingKid fucked around with this message at 13:14 on Jun 29, 2010 |
|
#
¿
Jun 29, 2010 12:33
|
|