Programming builds, swooshes and creating drama

S

STASIS

Guest
I just wanted to start a thread discussing techniques used for designing dramatic sounds, ie wooshes, swooshes and noise buildups.

I realized this is one of the biggest things lacking from my music, the ability to expertly build drama by designing cresendos and builds that make certain parts climax, etc. All well done electronic music has at least some of this and I think its an element well worth learning.

So, how do you guys approach this, speaking in synth programming terms, you can be specific to the software you work in or speak simply theoretically.....I personally have decided to thouroughly learn absynth, as the envelopes, AR pulses, sampling and oscillator modes provide the perfect formula for creating these sounds.

Anyone else successfully design sounds like this using Absynth? I want knowledge on this subject, please share!
 
Massive topic. But for starters. The most common method of creating cresendos on instrument sounds as opposed to drum loops, is to have the filter open and the gain increase together. Standard techniques involve assigning key pressure or better still using the mod wheel to open the filter and raise the gain.
You can also assign an lfo to pan the sound , so as you move the mod wheel up(+) the lfo moves faster and faster(cycles) and pans left/right intermittently. For rigid left right movement, pick a pulse or square lfo, for fluid left to right movement pick a sine lfo.
Another nice little trick is to have the release of the note create a gain, this can be quite dramatic and fluid when playing a sequence of notes.
 
Last edited:
Interesting...

How would you go about creating a gain with the release of a note? What feature must a synth have to do something like this? Do you mean self-oscillation? About assigning the mod-wheel to several such paramaters- good idea.

Why haven't I thought of that yet? I am trying to draw in envelopes to accomplish the automation I need, and in many cases this is necessary, but live automation in this aspect would prove much more useful.

Tell me more Samplecraze- when creating swooshy builds and such, do you just use noise? Do u use multiple oscillators, each generating noise at a different tuning?

Do you add harmonic content just depending on the context of the track you're working on? I think you know many of the techniques I'm after- so fill me in if you would be so kind.

I checked out your site- very accomplished/professional. I liked it indeed. Thanks-
 
Thnx Statis:cheers:

You need VenvRel-velocity envelope release
A patchcord that would increase gain on release would look like this:

Draw in the velocity as linear on a graph, if you have that facility, so 0-127.

Patchcord:

VenvRel<ampVol 100%
Vel<VenRel +50%

For mod wheel:

Mwhl<FilFreq 100%
Vel<FilFreq 50
Fenv+<FilFreq -50

Draw the filter curve as linear 0-127
Filter be a 4 pole set at 10000Hz and Q at 60

When you move the Mwhl up or positive it opens up the filter to it's maximum(100%)

For Lfos:
Use a sine(shape) Lfo1 with a setting of 6-10(rate)

Patchcord:
Lfo1+<amppan 100%

The Lfo1 will pan the sound left/right at the speed of the rate.
 
Only on certain sounds, but the way the filter opens and closes gives the perception of resonance. If you need to apply resonance use the function called FilRes, Filter Resonance. This is very cool on lead lines that have the filter tracking the note, very cool and ear piercing as you go up the scales. It's also cool on snares and kiks if used with velocity. You set the resonance to peak at the highest velocity when you hit the snare, it gives it that snap[ and crack effect.

You can also use resonance musically, for pad sounds that you want tyo sizzle either on pressure or over time. You can also use it for some wacky effects , resonance is a cool parameter.
sko04.gif
 
Samplecraze said:
Thnx Statis:cheers:

You need VenvRel-velocity envelope release
A patchcord that would increase gain on release would look like this:

Draw in the velocity as linear on a graph, if you have that facility, so 0-127.

Patchcord:

VenvRel<ampVol 100%
Vel<VenRel +50%

For mod wheel:

Mwhl<FilFreq 100%
Vel<FilFreq 50
Fenv+<FilFreq -50

Draw the filter curve as linear 0-127
Filter be a 4 pole set at 10000Hz and Q at 60

When you move the Mwhl up or positive it opens up the filter to it's maximum(100%)

For Lfos:
Use a sine(shape) Lfo1 with a setting of 6-10(rate)

Patchcord:
Lfo1+<amppan 100%

The Lfo1 will pan the sound left/right at the speed of the rate.
Click to expand...

awesome!!

i just searched forever to find this post!!

thanks!!
 
excellent thread... im taking notes as we speak. big up samplecraze
 
I just recieved my copy of Spectrasonics Atmosphere and let me tell you, this is the answer (if you have the money). You get about 3 Gigs of the sounds you are talking about and they are all the highest quality possible.

You can try tweaking as much as you want, but I doubt that you will recieve the kind of sound that Atmosphere will give you.
 
Shenlong, true that the Atmosphere is excellent but you forget a few important facts here about tweaking away. Eric Persing, who owns Spectrasonics, was a sound designer for Roland back in the 80s and his products are very well designed fonts that start from waveform up, the same procedure that all us SF developers practice. To create fonts , you need a decent sampler with a strong mod matrix, a tone generator , a tuner and bucketloads of experience and talent.
You can either sample or create from waveform up. We create from waveform up, simply because we are confined to strict parameters when it comes to SF development and cannot use a sample from a competitors product as this would infringe on the licencing agreement. The second reason, and more importantly, is memory. How do you think we cram 512 individual and unique presets onto a 32 meg board that you then buy and install into your module/sampler etc...? It takes a great deal of experience and skill to try to create a multilayered piano preset that is less than 500k. The other reason we use waveforms is simply because we can shape the waveform or waveforms into any sound that we desire. The maths behind waveforms can be complicated and the waveforms must be perfectly tuned, looped at full integers and well presented with no noise or artifacts. In essence, an oscillator is a waveform and shaping the osc is shaping the waveform.
Today we have the luxury of no memory limitations so to create a piano font that is 20megs is not a problem for the host (vsti or sampler or SF Player) and we can then concentrate more on creating better and richer textures. It is important for all programmers to be able to create any sound they so wish and sharing the knowledge of SF development is a selfless act and one that I enjoy tremendously. And if I can help anyone who wishes to create their own sounds using the tools that they have at their disposal then I will as it is truly a great education and a wonderful way for someone to understand then use this knowledge to programme their synths/samplers etc..
Eric is an excellent SF developer and creates top class content but you could too, so long as you know how and have the right tools. This is how libraries are created, this is how Eric and all the rest of us started, from nowhere, armed with knowledge we create most of the sounds that you use in the synths of today.
I encourage everyone to try and programme their own sounds. Why buy a piece of gear for a grand and only be limited to use it's own presets? Make your own custom soundsets and you will create an identity for yourself.
 
Samplecraze said:
Shenlong, true that the Atmosphere is excellent but you forget a few important facts here about tweaking away. Eric Persing, who owns Spectrasonics, was a sound designer for Roland back in the 80s and his products are very well designed fonts that start from waveform up, the same procedure that all us SF developers practice. To create fonts , you need a decent sampler with a strong mod matrix, a tone generator , a tuner and bucketloads of experience and talent.
You can either sample or create from waveform up. We create from waveform up, simply because we are confined to strict parameters when it comes to SF development and cannot use a sample from a competitors product as this would infringe on the licencing agreement. The second reason, and more importantly, is memory. How do you think we cram 512 individual and unique presets onto a 32 meg board that you then buy and install into your module/sampler etc...? It takes a great deal of experience and skill to try to create a multilayered piano preset that is less than 500k. The other reason we use waveforms is simply because we can shape the waveform or waveforms into any sound that we desire. The maths behind waveforms can be complicated and the waveforms must be perfectly tuned, looped at full integers and well presented with no noise or artifacts. In essence, an oscillator is a waveform and shaping the osc is shaping the waveform.
Today we have the luxury of no memory limitations so to create a piano font that is 20megs is not a problem for the host (vsti or sampler or SF Player) and we can then concentrate more on creating better and richer textures. It is important for all programmers to be able to create any sound they so wish and sharing the knowledge of SF development is a selfless act and one that I enjoy tremendously. And if I can help anyone who wishes to create their own sounds using the tools that they have at their disposal then I will as it is truly a great education and a wonderful way for someone to understand then use this knowledge to programme their synths/samplers etc..
Eric is an excellent SF developer and creates top class content but you could too, so long as you know how and have the right tools. This is how libraries are created, this is how Eric and all the rest of us started, from nowhere, armed with knowledge we create most of the sounds that you use in the synths of today.
I encourage everyone to try and programme their own sounds. Why buy a piece of gear for a grand and only be limited to use it's own presets? Make your own custom soundsets and you will create an identity for yourself.
Click to expand...

humm. any reccomendations for any hardware/software to help with the sound creation process???



oh.. and big ups for the knowledge
 
Good point sample craze. I might consider it later on.

Also, would you care to expand on what you said about license agreement?
 
Sure.
Alias, like I said a sampler , a tone generator and a tuner. The sampler can be software or hardware, software nowadays is better because it can actually import soundfonts and has multisave functions. The tone generator can be the TC Spark XL which has a generator built in but any synth that has oscillators is a tone generator and even easier Soundforge or similar editor has in their menu option(tools-synthesis) a synthesis engine that can create tones. In the case of Soundfonge, you can have pure waveform tones, a menu appears whereby you select whether you want a sine, saw, triangle etc...you then select the frequency ( get a chart, makes life easier, they are everywhere on the net-Google) then you select the length of the waveform, so 1 sec and then loop the waveform( sample), this should then be perfectly tuned and looped for use later when you want to programme the waveforms to create presets or instruments.Remember to create at least 10 waveforms for each preset so you have 2 waveforms on an octave for a 5 octave keyboard. Trust me, it takes no time at all once you get the hang of it. Soundforge also has an FM tone generator as well and this is quite cool because you can select eg: a sine with 4 operators(maximum on Soundforge) and configure the operators in any order and draw in the graph the attack. decay ,sustain and release of the waveform (ADSR). You can create countless possibilities and it has some presets of it's own you can use that can create horns, basses, leads etc...

Regarding licencing:

When a font is created for a manufacturer, they in turn own all the licences on that font, so they can , in effect, resell the font in any format they choose or package it any way they want. A font also has what we call a digital signature that is encoded into the font when we create it , this is done so as to avoid someone using a font that is distinctly yours or the manufacturers. A copyrght if you will. The licence is an agreement between you and the manufacturer that buys the fonts off you. They then own the licence and can use it how they want. If the licencing arrangement is one of royalties, then everytime your font is used and sold , you receive a royalty payment, if it is on a fee basis, then they 'buy' the licence off you and own the font. Example: I have created a number of soundsets for Emu, predominantly the Audity2000, XL1 and XL7 along with all the arp and BTS templates. They are licenced to Emu on a fee and royalty basis. At the moment I have created 2 roms for Emu and they were to be burnt as rom boards for the modules and command stations but since the release of the EmulatorX and a new product on the horizon (NDAs do not allow me to talk about it) the roms have been reformatted to be sold as libraries for these s/wares. In effect, the fonts have gone from Eos based architecture and format for the modules/samplers to a virtual(software) format. I own the material, they own the licence. This allows them the luxury of knowing that I cannot then go and sell the fonts to a competitor, it is theirs exclusively. The difference between these 2 products and the soundsets I created for them before is that I have granted them a licence but held onto the ownership rights, so my name has to appear alongside the products whereas before they bought the fonts off me outright and used them as thiers.
Although a font is a font, we do have ways of insuring our copyright is not infringed upon so we all either digitally sign our material or add a parameter in there that no one knows about so as to safeguard our work.
In the case of my sample cds and download packs , when you buy the cd or pack, they are Licence Free, which entitles the user to use my material in released material or in any commercial capacity, except a duplictaion and resell, and no fees have to be paid to me but I do instist oon creditation. Creditation is my marketting, you can use my samples but name the source when you release the material. This is now a major headache for distributors as the big boys sell their sample cds Licenced, which means if you use the samples in a release or commercial project you then have to pay an additional fee to 'clear the samples'. A ripoff by all standards, why should you pay 100 bucks for a sample cd to then have to pay for each and every sample that you use in your commercial projects?
This debate has been going on for years, so I always advise buyers of sample cds or cd roms or download soundfonts to first read the disclaimer and copyright terms on the package and if the product or it's contents are Licence Free or if fees need to be paid.

Sorry about the long ramble but I feel the content is important , especially for newbies, as they get roped into buying sounds off the net and then find out that they still have to pay additional fees to the holder of the licence.
I'm knackered now.
Peace.
 
I just use either Ambient effects, or a really dramatic piano.
 
You must log in or register to reply here.
Back
Top