Everyday Genius (/Crazy) Ideas

Have an urge to learn, or a calling to teach? Want to share some useful Youtube videos? Do it here!
RobC
Posts: 676
Joined: 10 Mar 2018

13 Jun 2018

Since I have a lot of random ideas, which either do or do not work, I thought it would be a good idea to keep it in one place, so if folks aren't interested, they easily can avoid it; and so it doesn't clutter threads up either.

If I come up with something, I'll just add it in a reply.

"You're dumb! - Am not!!"

Image

Starter:

As a start, I was thinking a lot about what would sound best for a synthesized sound when it comes to its frequency spectrum.
When creating a kick drum for example, we could try to copy a recorded kick drum's sound by more or less equalizing it that way; but why would we do that when it comes to a fictional sound?
Needless to say, it's very easy to get lost while creating a sound from scratch. It might appear to sound great in that moment, but later we might realize that the sub bass turned out very weak. Now, since every person's hearing can be very different, I feel we could tops try to balance that frequency spectrum slightly to a simple standard.
See, for example if we listen to a sine wave sweep, we might not even realize that we hear that constant power with varied audio levels - some frequencies louder, some softer.
Today I thought, if I split the frequency spectrum of a sound into just larger stages (example 20-80 Hz for sub bass), then listen to a sine wave that generates a center frequency (50 Hz in for sub bass) or if that doesn't work well, then an up-down looped sweep (20-80 Hz), and then nudge the given frequency band so it sounds more or less as loud as the sine wave with the constant power, then at least I can avoid that my synthesized sound ends up with a completely unbalanced frequency spectrum.
Maybe I could try the same with more detailed bands, but I'm afraid the more detailed, the more the dynamics would flatten out, too.

I'm sure we all toyed around with a multi-octave sine wave and heard, that it sounds pretty good, actually, so this might be worth a try.
Last edited by RobC on 15 Jun 2018, edited 1 time in total.

User avatar
rcbuse
Posts: 689
Joined: 16 Jan 2015
Location: SR388
Contact:

13 Jun 2018

You certainly do have a lot of very technical ideas about sound generation and production. Have you ever thought about using something like Reaktor, or getting into DSP where you could flesh them out?

User avatar
Loque
Posts: 3112
Joined: 28 Dec 2015

13 Jun 2018

Sounds like you want to create a fundamental for your sine sweep. Why not!? A kick often is not only 1 thing, often a sub, a base, a top and click, all with different decays. Top and click can be added by FM with 1 or 2 short decay as an example. Lots of different ways to achieve this.

And you are right, that you might think how a kick drum must sound today, is boring and outdated tomorrow.

If you are looking for more experiments, try different "attack" modulations, try a saw instead of a sine, which you filter-sweep out or try to move the filter-self-oscillating-resonance according your drum-sound. Shapers, amps, saturations, EQs, transient shapers, gaters and compressors are always nice for additional shaping - and ofc this could be done in different frequencies with different timings.

IMO, most of this is too much work, you get lost in sound design and nobody really will get the differencies - except, you want to create something really new - THAN you NEED to go NEW ways...
:reason: 10, Win10 64Bit.

RobC
Posts: 676
Joined: 10 Mar 2018

13 Jun 2018

rcbuse wrote:
13 Jun 2018
You certainly do have a lot of very technical ideas about sound generation and production. Have you ever thought about using something like Reaktor, or getting into DSP where you could flesh them out?
I'm preparing for something in-between ~ the Rack Extension field - I'm willing to spend time learning anyway.

RobC
Posts: 676
Joined: 10 Mar 2018

13 Jun 2018

Loque wrote:
13 Jun 2018
Sounds like you want to create a fundamental for your sine sweep. Why not!? A kick often is not only 1 thing, often a sub, a base, a top and click, all with different decays. Top and click can be added by FM with 1 or 2 short decay as an example. Lots of different ways to achieve this.

And you are right, that you might think how a kick drum must sound today, is boring and outdated tomorrow.

If you are looking for more experiments, try different "attack" modulations, try a saw instead of a sine, which you filter-sweep out or try to move the filter-self-oscillating-resonance according your drum-sound. Shapers, amps, saturations, EQs, transient shapers, gaters and compressors are always nice for additional shaping - and ofc this could be done in different frequencies with different timings.

IMO, most of this is too much work, you get lost in sound design and nobody really will get the differencies - except, you want to create something really new - THAN you NEED to go NEW ways...
True, sounds are quite complex, though I'm looking for ways to be able to limit how many sounds I need to create ~ inspires a lot of interesting solutions. My mentality was that sounds have to be unique.

Now, taking frequency bands apart for a kick drum, is an interesting one, since I've been thinking of that, too, but I was quite worried how it would sound in the end ~ especially if taking 10 bands, designing them until each sound punchy; ~ well, I only theorized about it so I can't say anything yet, but that's yet again a very interesting part of sound design with not necessarily predictable end results.

The hard work, and creating something new comes also with the benefit that - in case things go well - there's a chance that at the mixing stage we can rather focus on yet again creativity and expression (macro dynamics), instead of having to worry about getting everything in place.

User avatar
Loque
Posts: 3112
Joined: 28 Dec 2015

13 Jun 2018

Checking how a kick might sound, if i have different partials with different decays is still on my todo... Spectra might be suitable for this job. I still cannot really get my head around, what a partial is !?!? So i'd rather stick on waves and frequencies - so a pitch-shifting solution with an envelope shaper might be a solution too.
:reason: 10, Win10 64Bit.

RobC
Posts: 676
Joined: 10 Mar 2018

13 Jun 2018

Loque wrote:
13 Jun 2018
Checking how a kick might sound, if i have different partials with different decays is still on my todo... Spectra might be suitable for this job. I still cannot really get my head around, what a partial is !?!? So i'd rather stick on waves and frequencies - so a pitch-shifting solution with an envelope shaper might be a solution too.
When it comes to multiple frequency bands, we're practically layering, so it's a new world where we can set the sample start (I forgot to say that I'd split bands with Linear Phase FIR filters and square shaped "crossovers" and then sample them all), change pitch, time stretch, offset the sample in time, etc. So many possibilities ~ that can be super cool, or totally horrid. xD

Here comes another idea to mind. We start off with an oscillator (which is as detailed as vector graphics), then we start applying any basic effect, like FM with another oscillator. Here comes the question that wouldn't it be possible (or maybe it already exists) that the instrument "samples" a "loop" of the detailed waveform, where it stores every mathematical information needed to recreate it with all the fine details (like a vector) without the need of turning the signal into a wave file? Or is Europa our best bet for that? Maybe graining, etc. Those come with artifacts, though. I mean, if the synthesizer "knows" how things were calculated, it should be possible to "learn" the created waveform, and then start again with the new, complex waveform.
So practically, it's like sampling, but (while sampling has it's interesting destructive effects) this way it would have maximum detail. It would be possible to easily apply FM to it again, since all the details are present for modulating pitch at high frequencies.

User avatar
Loque
Posts: 3112
Joined: 28 Dec 2015

13 Jun 2018

Guess you mean resynthesizing. Spectra does this and I think Europa does this to, maybe Grain also in some way. There are others out there in VST world that can resynthesize as well.
The math is FFT analysis with some additional stuff to get the partials. From the you can recreate the sound. Most synths cannot resynthesize the envelopes of each partial.
:reason: 10, Win10 64Bit.

jimmyklane
Posts: 540
Joined: 16 Apr 2018

13 Jun 2018

RobC wrote:
13 Jun 2018
Now, taking frequency bands apart for a kick drum, is an interesting one, since I've been thinking of that, too, but I was quite worried how it would sound in the end ~ especially if taking 10 bands, designing them until each sound punchy; ~
In all likelihood you will have phase cancellation problems. Let me explain: try and take two kick drums and combine them. They MAY work together, most likely the bass disappears. Why? Phase cancellation. So, you are on the right track to perhaps take a beater sound (the “click”) and combine it with another sound “the thump”....this has been going on for many years in the hip hop crowd, where they HP one kick, LP another and combine the two on one pad of an MPC....or Kong in your case. That’s why the sampler has four slots in Kong.

If you attempt 10, you’re most likely going to end up with one of two things:

A: a wimpy sounding mess
B: a kick drum that takes up too much of the spectrum.

My advice is to stick with 2-3.
DAW: Reason 10,

SAMPLERS: Akai MPC 2000, E-mu SP1200, E-Mu e5000Ultra, Ensoniq EPS 16+, Akai S950, Maschine

SYNTHS: Mostly classic Polysynths and more modern Monosynths. All are mostly food for my samplers!

www.soundcloud.com/jimmyklane

RobC
Posts: 676
Joined: 10 Mar 2018

14 Jun 2018

Loque wrote:
13 Jun 2018
Guess you mean resynthesizing. Spectra does this and I think Europa does this to, maybe Grain also in some way. There are others out there in VST world that can resynthesize as well.
The math is FFT analysis with some additional stuff to get the partials. From the you can recreate the sound. Most synths cannot resynthesize the envelopes of each partial.
The problem with that is, that it comes with artifacts. I would prefer a completely clean method, where the synthesizer creates a calculation to exactly draw out the same waveform. For a very simple example: if we start out with a saw shape, and we get a waveform that is a saw then a triangle (for simplicity's sake), then it should be able to analyze and create an internal calculation that draws exactly a saw and triangle shape next to each-other, just like vector graphics perfection.
That said, I would already be more than happy if a synthesizer could capture and export a perfect loop/cycle as a wave file.

This reminds me of yet another idea: While messing around with a generated static sound, that I loaded into NN-XT, I could create very interesting sounds from that just by changing pitch, and setting up a very short part to loop. Of course, it would be easier to be able to sweep through the sample with that few millisecond looping to see if there's something more interesting coming out from other parts. I have to check if I can do that with Grain in a clean-sample way without artifacts.
I think, the result is more close to a sample for wavetables; though if I prefer the clean sound, I can't just play it as is, cause it tends to fall apart, so I need some time-stretching for each note. Though Reason's pitch shifting in the sequencer seems to do a pretty good job.

RobC
Posts: 676
Joined: 10 Mar 2018

14 Jun 2018

jimmyklane wrote:
13 Jun 2018
RobC wrote:
13 Jun 2018
Now, taking frequency bands apart for a kick drum, is an interesting one, since I've been thinking of that, too, but I was quite worried how it would sound in the end ~ especially if taking 10 bands, designing them until each sound punchy; ~
In all likelihood you will have phase cancellation problems. Let me explain: try and take two kick drums and combine them. They MAY work together, most likely the bass disappears. Why? Phase cancellation. So, you are on the right track to perhaps take a beater sound (the “click”) and combine it with another sound “the thump”....this has been going on for many years in the hip hop crowd, where they HP one kick, LP another and combine the two on one pad of an MPC....or Kong in your case. That’s why the sampler has four slots in Kong.

If you attempt 10, you’re most likely going to end up with one of two things:

A: a wimpy sounding mess
B: a kick drum that takes up too much of the spectrum.

My advice is to stick with 2-3.
True, but phase issues can be dealt with by offsetting in time, changing pitch, etc. - for the most part; and what can't be helped, may not sound that bad either, since there are some nice sounding comb filtering effects.
That said, it's not a problem if we take a look at these issues, even if we're familiar with them, cause we can come up with new ideas on the way.
Now, there might be less problems if I just talk about a simple sine-kick. The good thing is, that there's more room to bring out essential parts with multi-band work with tons of possibilities if we sample all the bands. From there, we practically can create a new sound with new harmonics with the help of distortion, setting sample-start to get a click, etc. Heck even what infra or ultra frequencies may appear can yet again be played around with, like in the other thread, changing its pitch to an audible range, whatsoever. Although I'd do it with Linear Phase filtering, adding dynamics there can result with artifacts, so there's probably no more or less clean way. (...and we can take even those apart and then do a bit of mix-match and add together what sounds best.)

The beauty here is that even when we created a very colorful kick, we still can equalize and filter it again.

That said, for design purposes, I'm more likely to just take some main bands apart, and not for every octave (or even third octaves) from 20 Hz.

RobC
Posts: 676
Joined: 10 Mar 2018

15 Jun 2018

Just checked Grain and it certainly can do that 'zooming on a snippet and looping it', then we can scroll through the sample. Exactly what I was looking for when it comes to generating sound with a synthesizer, then sampling it. It's the (seemingly) simple (very useful) tools that I look for, and not necessarily the 'impressive' ones. I mean, it's not exactly the main use of Grain, I get that, but it's a fantastic little sub-feature.

Practically, Grain is what I was missing from NN-XT. Also, kind of like an offspring of Malström and NN-XT / NN-19 (DNA test needed? xD).

User avatar
normen
Posts: 3251
Joined: 16 Jan 2015

15 Jun 2018

This is Reason. Would be much simpler if you just made combinators so a) people know what you mean b) you can find out some things simply don‘t work like that yourself and c) people can help you improve your ideas if they work.

RobC
Posts: 676
Joined: 10 Mar 2018

15 Jun 2018

normen wrote:
15 Jun 2018
This is Reason. Would be much simpler if you just made combinators so a) people know what you mean b) you can find out some things simply don‘t work like that yourself and c) people can help you improve your ideas if they work.
There's more to it in my case, since when I start typing a basic idea out, it inspires more and more complex/useful (to me at least) ones sometimes. Likewise when others say something, it can bring something else out of me again.

I did the silent method for a decade, and I had lots of ideas which I was afraid to share, but I realized, it's worth it cause it puts my brain on a roll. Communication is very important.

Also, in case there is some confusion, by the time we understand each-other with people, even there new ideas pop out.

I understand that it can be bothersome, but that's why I decided to make 1 topic, so it can be avoided in that case. Even if I just "blog" ideas here out and nobody comments, the whole posting process and writing the comment/post is a type of multitasking with a bit of pressure, that makes my (probably) ADHD brain work better. There's lots of potential in this type of dumbness xD which I barely made use of in the past decade, since I had no clue about it, let alone how to make the most of it.

But yeah, I made examples before, but people don't really check them ~ can't blame anyone, cause reading and typing can be quicker and easier. I might do some anyway.

User avatar
normen
Posts: 3251
Joined: 16 Jan 2015

15 Jun 2018

Yeah, most of what you said rather supports having a .rsn file to grow and go along with it. That way theres some kind of result for people looking at that thread later instead of having to go through the whole back and forth.

RobC
Posts: 676
Joined: 10 Mar 2018

16 Jun 2018

normen wrote:
15 Jun 2018
Yeah, most of what you said rather supports having a .rsn file to grow and go along with it. That way theres some kind of result for people looking at that thread later instead of having to go through the whole back and forth.
These ideas are pretty varied, though, so not really straight forward. Plus I want to get into Rack Extension development, even if at some point I have to sell a kidney to be able to hire programmers. xD

User avatar
normen
Posts: 3251
Joined: 16 Jan 2015

16 Jun 2018

RobC wrote:
16 Jun 2018
These ideas are pretty varied, though, so not really straight forward. Plus I want to get into Rack Extension development, even if at some point I have to sell a kidney to be able to hire programmers. xD
Two more reasons to add .rsn files to the posts.

RobC
Posts: 676
Joined: 10 Mar 2018

19 Jun 2018

normen wrote:
16 Jun 2018
RobC wrote:
16 Jun 2018
These ideas are pretty varied, though, so not really straight forward. Plus I want to get into Rack Extension development, even if at some point I have to sell a kidney to be able to hire programmers. xD
Two more reasons to add .rsn files to the posts.
What do I do with ideas like wanting an easily programmable keyboard tracking module where we can set keys and ranges just like with combinator. If we go very simple, setting from key X to key Y how it should change panning from hard left to hard right, or the other way around. Practically generating CV values via MIDI at a set range, or programming specific values for individual keys - where even single notes can be programmed ~ Combinator Programmer -style.

I do want to make example files with whatever I can, but I had a lot of farming work to do ~ sad truth is, the taste of home grown vegetables can't be compared not even to what they bring to those market places, let alone canned stuff or neon-light-ripened artificial things at supermarkets. (...and no, not because the sold stuff would be much better. xD They never bring any of the good stuff there.) The terrible time is when I run out of my own food, and then return to the sold stuff and realize just how bad they are.

User avatar
normen
Posts: 3251
Joined: 16 Jan 2015

19 Jun 2018

RobC wrote:
19 Jun 2018
What do I do with ideas like wanting an easily programmable keyboard tracking module where we can set keys and ranges just like with combinator. If we go very simple, setting from key X to key Y how it should change panning from hard left to hard right, or the other way around. Practically generating CV values via MIDI at a set range, or programming specific values for individual keys - where even single notes can be programmed ~ Combinator Programmer -style.
For those ideas the corresponding .rsn file would illustrate the functionality you mean in the combinator and you'd only have to explain the UI.
RobC wrote:
19 Jun 2018
I do want to make example files with whatever I can, but I had a lot of farming work to do ~ sad truth is, the taste of home grown vegetables can't be compared not even to what they bring to those market places, let alone canned stuff or neon-light-ripened artificial things at supermarkets. (...and no, not because the sold stuff would be much better. xD They never bring any of the good stuff there.) The terrible time is when I run out of my own food, and then return to the sold stuff and realize just how bad they are.
Yeah, sometimes I feel bad for being able to pretty much look out of the window to see the cow I'm drinking the milk of - like literally ;) But then I'm thinking about how I could have opted for money instead :)

RobC
Posts: 676
Joined: 10 Mar 2018

19 Jun 2018

normen wrote:
19 Jun 2018
RobC wrote:
19 Jun 2018
What do I do with ideas like wanting an easily programmable keyboard tracking module where we can set keys and ranges just like with combinator. If we go very simple, setting from key X to key Y how it should change panning from hard left to hard right, or the other way around. Practically generating CV values via MIDI at a set range, or programming specific values for individual keys - where even single notes can be programmed ~ Combinator Programmer -style.
For those ideas the corresponding .rsn file would illustrate the functionality you mean in the combinator and you'd only have to explain the UI.
RobC wrote:
19 Jun 2018
I do want to make example files with whatever I can, but I had a lot of farming work to do ~ sad truth is, the taste of home grown vegetables can't be compared not even to what they bring to those market places, let alone canned stuff or neon-light-ripened artificial things at supermarkets. (...and no, not because the sold stuff would be much better. xD They never bring any of the good stuff there.) The terrible time is when I run out of my own food, and then return to the sold stuff and realize just how bad they are.
Yeah, sometimes I feel bad for being able to pretty much look out of the window to see the cow I'm drinking the milk of - like literally ;) But then I'm thinking about how I could have opted for money instead :)
I guess I could mimic something close to a more flexible keyboard tracker - though let's not forget, when communication starts, and there's some misunderstanding, new ideas come up!

A cow multiplies her own value (in one way or another xD). They are evil for real, though.

RobC
Posts: 676
Joined: 10 Mar 2018

20 Jun 2018

This one's more generic:

There's this thing I call "No Shit Edit".

It originated from when in the past, on youtube I had statistics where people listened to songs half way on average. They don't give you much time and chances for you and really dislike repetition, it seems.
Therefore, I first thought, it would be a good idea to take your final arrangement, and wipe out everything that you don't like. But then again, now that I think about it, that's rather about personal taste. You might end up trashing what some people like, and keep what they don't.
In the end, I thought it might be best to keep varying parts, and even build the arrangement as varied as possible, so people can hear something new or noticeably different change with every 4 bar at least and even tone common chorus repetition down. If everything is really good, you're more likely to not only get full plays, but even repeat listens.
After, you could upload the original, full mix to other places, if there still are any that don't make a big deal if people leave your upload half way.

Finally, when it comes to a No Shit version, song length is the least thing you should worry about. Sound good and entertain all the way.

User avatar
normen
Posts: 3251
Joined: 16 Jan 2015

20 Jun 2018

RobC wrote:
20 Jun 2018
This one's more generic:

There's this thing I call "No Shit Edit".

It originated from when in the past, on youtube I had statistics where people listened to songs half way on average. They don't give you much time and chances for you and really dislike repetition, it seems.
Therefore, I first thought, it would be a good idea to take your final arrangement, and wipe out everything that you don't like. But then again, now that I think about it, that's rather about personal taste. You might end up trashing what some people like, and keep what they don't.
In the end, I thought it might be best to keep varying parts, and even build the arrangement as varied as possible, so people can hear something new or noticeably different change with every 4 bar at least and even tone common chorus repetition down. If everything is really good, you're more likely to not only get full plays, but even repeat listens.
After, you could upload the original, full mix to other places, if there still are any that don't make a big deal if people leave your upload half way.

Finally, when it comes to a No Shit version, song length is the least thing you should worry about. Sound good and entertain all the way.
Lol Rob, you really SHOULD make something tangible each time you do one of your wheel reinventions :) I mean who can say he's got a collection of new kinds of wheels? :D

Edit: Yes, this implies that "cutting the shit away" is one of the most common art creation techniques :)

RobC
Posts: 676
Joined: 10 Mar 2018

20 Jun 2018

normen wrote:
20 Jun 2018
RobC wrote:
20 Jun 2018
This one's more generic:

There's this thing I call "No Shit Edit".

It originated from when in the past, on youtube I had statistics where people listened to songs half way on average. They don't give you much time and chances for you and really dislike repetition, it seems.
Therefore, I first thought, it would be a good idea to take your final arrangement, and wipe out everything that you don't like. But then again, now that I think about it, that's rather about personal taste. You might end up trashing what some people like, and keep what they don't.
In the end, I thought it might be best to keep varying parts, and even build the arrangement as varied as possible, so people can hear something new or noticeably different change with every 4 bar at least and even tone common chorus repetition down. If everything is really good, you're more likely to not only get full plays, but even repeat listens.
After, you could upload the original, full mix to other places, if there still are any that don't make a big deal if people leave your upload half way.

Finally, when it comes to a No Shit version, song length is the least thing you should worry about. Sound good and entertain all the way.
Lol Rob, you really SHOULD make something tangible each time you do one of your wheel reinventions :) I mean who can say he's got a collection of new kinds of wheels? :D

Edit: Yes, this implies that "cutting the shit away" is one of the most common art creation techniques :)
I mean, when your track is polished, and you still cut stuff out. Should I call it Body Part Removal then? This would practically play the chorus only once for example. So, really ripped.

User avatar
normen
Posts: 3251
Joined: 16 Jan 2015

20 Jun 2018

RobC wrote:
20 Jun 2018
I mean, when your track is polished, and you still cut stuff out. Should I call it Body Part Removal then? This would practically play the chorus only once for example. So, really ripped.
Yes, yes, doing everything properly but still knowing you'll kick a lot of it. I mean that's what Funk is, songs without all the music ;)

RobC
Posts: 676
Joined: 10 Mar 2018

20 Jun 2018

normen wrote:
20 Jun 2018
RobC wrote:
20 Jun 2018
I mean, when your track is polished, and you still cut stuff out. Should I call it Body Part Removal then? This would practically play the chorus only once for example. So, really ripped.
Yes, yes, doing everything properly but still knowing you'll kick a lot of it. I mean that's what Funk is, songs without all the music ;)
But there would still be a full version, this is just something for the unforgiving internet.

Post Reply
  • Information
  • Who is online

    Users browsing this forum: No registered users and 3 guests