Synth Creation

[Deleted User]

@haulin_notes said:

@Fruitbat1919 said:
I would like to see a modern take on the D50 theme: Take the separate OSCs you were thinking of, but make them segmented, so one for the attack segment and one for the sustain and release segments. Now make all aspects of these modulate by a great array of multi point and loopable envelopes and complex LFO units - the modern take bit. I’m not talking basic LFOs either, I’m talking MATHs style multi use units. Think of something new and extensive for the modulation matrix.

D50 meets Matrix 12

Yes! Envelopes seem to be an afterthought, the neglected stepchild sometimes. Loopable and tempo sync-able envelopes (at the very least) add another dimension of pulsing, evolving sounds. The envelopes on the Sub 37 are incredible. I haven't figured out MATHs yet, but it is great. Gotta like modules that are smarter than you, and surprise you with the combination of modulations.

Also, at least bi-timbral. Need at least two layers imho. A bassline and a pad or lead, etc. And maybe an option to filter them separately or together.

Not asking alot... Ya know, just come out with Zebra 3 before Urs does! On iPad. For $9

Yeah maybe we are asking a bit much for a first time project lol

haulin_notes

@Fruitbat1919 said:

@haulin_notes said:

@Fruitbat1919 said:
I would like to see a modern take on the D50 theme: Take the separate OSCs you were thinking of, but make them segmented, so one for the attack segment and one for the sustain and release segments. Now make all aspects of these modulate by a great array of multi point and loopable envelopes and complex LFO units - the modern take bit. I’m not talking basic LFOs either, I’m talking MATHs style multi use units. Think of something new and extensive for the modulation matrix.

D50 meets Matrix 12

Yes! Envelopes seem to be an afterthought, the neglected stepchild sometimes. Loopable and tempo sync-able envelopes (at the very least) add another dimension of pulsing, evolving sounds. The envelopes on the Sub 37 are incredible. I haven't figured out MATHs yet, but it is great. Gotta like modules that are smarter than you, and surprise you with the combination of modulations.

Also, at least bi-timbral. Need at least two layers imho. A bassline and a pad or lead, etc. And maybe an option to filter them separately or together.

Not asking alot... Ya know, just come out with Zebra 3 before Urs does! On iPad. For $9

Yeah maybe we are asking a bit much for a first time project lol

Haha! True! Good to start simply and do one thing extremely well. But also while leaving room to grow and expand with updates. Lots of great and useful ideas in this thread for any synth developer.

haulin_notes

Like several people mentioned above: can't go wrong with oscilloscope(s) and AU. AU is gold! (little elemental table joke).

And, if there is a phrase sequencer or arpeggiator.. a transposable one is soooo very useful. The Korg Monologue has one that is wonderful. Wish all synths had it. A walking bassline is often is good thing!

JohnnyGoodyear

@Processaurus said:
I think the original poster has an idea for the synthesizer, he's just asking what features iOS'ers like to see? As far as features for a synth, most people here would be only be interested in a synth that can be connected to other audio apps, Audiobus 3 is many of our favorite way. IAA is widely used as well.

Ableton Link is a great way to pass tempo information between apps, but IAA has its own way of doing tempo that works well. Midi clock is the messiest, but is very handy to people with outboard midi gear, or sequencers, or a computer.

One could sidestep all of this by making the synth an AU plugin, then it is up to the host to do everything. Some talented developers believe this is the future.

People with midi controllers would like to be able to play the synth with a midi keyboard, at a bare minimum. It is nice to be able to hide the onscreen keyboard, when you have the midi keyboard, for more real estate. Good iOS synths have midi learn, for the knobs, the best communicate their knob positions back to the midi controller (Animoog, Model 15). If not MIDI learn, at least a static CC# for each knob, and a dedicated fan of the app could knuckle down and create a template for their midi controller.

Just a personal opinion, there may be plenty of people here that would disagree with the design philosophy, but after a couple years of diving into iOS music making, I enjoy the clever, but simpler concept apps the most, the music toys, because terribly deep apps take an investment of patience that often doesn't happen, especially when you already have a bunch of very good, complicated synthesizer apps. Like, why would I learn a new complicated synth, when I haven't figured out the PPG wavetable synths? Something that makes an interesting, new flavor in a way that is easy to grasp are my favorite. A good example is Laplace, it is a single page for the synth, almost everything visible about the sound on that page, and it sounds different, and interesting. The simpler apps that each do something different lend themselves well to a collection. The synths that are pages and pages of endless, exhaustive possibilities not so much, because you have to strap on your thinking cap, unless you use it regularly, or semi-exclusively.

Very good.

Samu

What ever you cook will be interesting to check out

One thing to keep in the thoughts is to keep all applicable parameters as both modulation sources and destinations.
Custom-Waveforms(in practice a 'Value/Step-Sequence' with a 'smoothness' parameter for stepped or interpolated values) for modulations are ofter overlooked.

Other things that are also often overlooked are the creative possibilities that are opened when given full control over start-phase and polarity & bias of the oscillators as well as doing some calculations between oscillators (add, subtract & multiply) as well as doing 'quantise'(both bit-depth and sample-rate) on the output prior to feeding them to the next step in the signal chain.

Not even the all options with wave-table synthesis have been fully explored yet (ie. use them as both oscillators and modulation sources and and do some crazy maths between them) or change the wave-table content while it's being played back.

Keeping the modulation rates fairly high can also lead to some interesting results...
(I still recall when I first heard my first 'multi speed' tune on the C64 that was changing the SID parameters at 200hz and wen't like 'The SID can't do this!' or when the first samples were played using a glitch in the volume register).

I think the trick here is to keep it as simple as possible on the surface but deep when digging into it...

[Deleted User]

@haulin_notes said:
Like several people mentioned above: can't go wrong with oscilloscope(s) and AU. AU is gold! (little elemental table joke).

And, if there is a phrase sequencer or arpeggiator.. a transposable one is soooo very useful. The Korg Monologue has one that is wonderful. Wish all synths had it. A walking bassline is often is good thing!

Personally I would keep this in mind for the project. Think of the simple things. Look at what Bram did with Od...oops sorry ehem....Rozeta. Control AUs are a great thing. So back to what I said - how about a complex envelope generator Midi AU?

anickt

@JohnnyGoodyear said:

@Processaurus said:
I think the original poster has an idea for the synthesizer, he's just asking what features iOS'ers like to see? As far as features for a synth, most people here would be only be interested in a synth that can be connected to other audio apps, Audiobus 3 is many of our favorite way. IAA is widely used as well.

Ableton Link is a great way to pass tempo information between apps, but IAA has its own way of doing tempo that works well. Midi clock is the messiest, but is very handy to people with outboard midi gear, or sequencers, or a computer.

One could sidestep all of this by making the synth an AU plugin, then it is up to the host to do everything. Some talented developers believe this is the future.

People with midi controllers would like to be able to play the synth with a midi keyboard, at a bare minimum. It is nice to be able to hide the onscreen keyboard, when you have the midi keyboard, for more real estate. Good iOS synths have midi learn, for the knobs, the best communicate their knob positions back to the midi controller (Animoog, Model 15). If not MIDI learn, at least a static CC# for each knob, and a dedicated fan of the app could knuckle down and create a template for their midi controller.

Just a personal opinion, there may be plenty of people here that would disagree with the design philosophy, but after a couple years of diving into iOS music making, I enjoy the clever, but simpler concept apps the most, the music toys, because terribly deep apps take an investment of patience that often doesn't happen, especially when you already have a bunch of very good, complicated synthesizer apps. Like, why would I learn a new complicated synth, when I haven't figured out the PPG wavetable synths? Something that makes an interesting, new flavor in a way that is easy to grasp are my favorite. A good example is Laplace, it is a single page for the synth, almost everything visible about the sound on that page, and it sounds different, and interesting. The simpler apps that each do something different lend themselves well to a collection. The synths that are pages and pages of endless, exhaustive possibilities not so much, because you have to strap on your thinking cap, unless you use it regularly, or semi-exclusively.

Very good.

The last paragraph says it all for me. I just don't have the time any more for a lot of tweaking. Give me a bunch of presets, an interesting sound engine. a flexible way to modify it to meet my needs, put it on one screen (or maybe a primary and a secondary) and I'm good.

ryanjanik

@brambos said:
Free advice: spend the holiday break researching, coming up and sketching out a solid idea/vision that feels right in your mind. Something that you yourself would really like to own. You should be able to sum up its essence (or its differentiating quality) in a just few words.

Don’t start your project as a design by committee, so get the solid fundaments down first, before using this helpful community to give you further insights. Otherwise it will be very difficult to give focus to your project.

Good luck.. making music software is lots of fun!

Bram thank you for this solid advice I admire your work

ryanjanik

Thank you to everyone for the advice, and keep it coming! I’ll be posting a thank you present for you guys soon

Will

One of the things I really really love about the Loopy app: It has MIDI learn per port. That means if you're using USB Device A, you can have CC3 connected to controlX and if you're using USB Device B, you can have CC25 connected to controlX. In a world where lots of users have multiple controllers in use a different times, this is really helpful stuff.

On the flip side: Loopy has no facility to export the mappings. So if you want to transfer them to a new device, you've got to rebuild them from scratch.

Semi-related: go ahead and setup/publish a default MIDI mapping for everything. Users can change it as needed via MIDI Learn if they can't set their controller to align but having published defaults means that apps that ship with CC maps (Quantum, Modstep, Rozetta...) can take advantage of them.

Will

Sorry if someone already mentioned this one: support for standard MIDI Program Change messages is a big deal for some of us. Particularly for live scenarios or for using the app with sequencers that support them.

u0421793

There should also be a button marked “hit”.

rezidue

Kind of related but I’m stuck in analysis paralysis trying to figure out how to get started. I have a few ideas for AU instruments/effects. Do I learn objc? Swift? Both? AudioKit+Swift? Juce + c++? Juce seems great for multi platform but I really just want to make IOS apps for now. I felt like I would start with something simple that routes midi and then move onto effects and eventually sound generation? Any advice? I have a background in ruby and python.

iOSounddesign

What i would really love to see on iOS would be a comprehensive Lo-Fi synth/effect, with special sonic degradation abilities. Some combination of tape warble/saturation, old school DAC emulation. On the desktop i have the http://www.psychicmodulation.com/ stuff, could be a nice niche to get into...

ryanjanik

@u0421793 said:
There should also be a button marked “hit”.

Hmm what’s this about?

ryanjanik

@rezidue said:
Kind of related but I’m stuck in analysis paralysis trying to figure out how to get started. I have a few ideas for AU instruments/effects. Do I learn objc? Swift? Both? AudioKit+Swift? Juce + c++? Juce seems great for multi platform but I really just want to make IOS apps for now. I felt like I would start with something simple that routes midi and then move onto effects and eventually sound generation? Any advice? I have a background in ruby and python.

Wow my background is also Ruby and Python as well as Java. I’m going with Swift and AudioKit for now and see where that takes me. If anyone else has advice on this it would be great

ryanjanik

@iOSounddesign said:
What i would really love to see on iOS would be a comprehensive Lo-Fi synth/effect, with special sonic degradation abilities. Some combination of tape warble/saturation, old school DAC emulation. On the desktop i have the http://www.psychicmodulation.com/ stuff, could be a nice niche to get into...

That echomelt looks pretty awesome

Littlewoodg

@ryanjanik said:

@u0421793 said:
There should also be a button marked “hit”.

Hmm what’s this about?

i think “hit song” is the parameter indicated

ryanjanik

@Littlewoodg said:

@ryanjanik said:

@u0421793 said:
There should also be a button marked “hit”.

Hmm what’s this about?

i think “hit song” is the parameter indicated

Of course

rezidue

@ryanjanik said:

@rezidue said:
Kind of related but I’m stuck in analysis paralysis trying to figure out how to get started. I have a few ideas for AU instruments/effects. Do I learn objc? Swift? Both? AudioKit+Swift? Juce + c++? Juce seems great for multi platform but I really just want to make IOS apps for now. I felt like I would start with something simple that routes midi and then move onto effects and eventually sound generation? Any advice? I have a background in ruby and python.

Wow my background is also Ruby and Python as well as Java. I’m going with Swift and AudioKit for now and see where that takes me. If anyone else has advice on this it would be great

I think I’ll do the same. If I end up in objc so be it but either are probably more helpful than jumping into c++ right now when I just want to do iOS dev.

Littlewoodg

@ryanjanik said:

@Littlewoodg said:

@ryanjanik said:

@u0421793 said:
There should also be a button marked “hit”.

Hmm what’s this about?

i think “hit song” is the parameter indicated

Of course

ryanjanik

@rezidue said:

@ryanjanik said:

@rezidue said:
Kind of related but I’m stuck in analysis paralysis trying to figure out how to get started. I have a few ideas for AU instruments/effects. Do I learn objc? Swift? Both? AudioKit+Swift? Juce + c++? Juce seems great for multi platform but I really just want to make IOS apps for now. I felt like I would start with something simple that routes midi and then move onto effects and eventually sound generation? Any advice? I have a background in ruby and python.

Wow my background is also Ruby and Python as well as Java. I’m going with Swift and AudioKit for now and see where that takes me. If anyone else has advice on this it would be great

I think I’ll do the same. If I end up in objc so be it but either are probably more helpful than jumping into c++ right now when I just want to do iOS dev.

I agree. I learned to play music by emulating people I loved, not by focusing on music theory, yet I learned a lot of it along the way. In that same way I’ve learned programming by making stuff I want to make, and was forced to learn the programming languages along the way, lol.

ryanjanik

@rezidue said:

@ryanjanik said:

@rezidue said:
Kind of related but I’m stuck in analysis paralysis trying to figure out how to get started. I have a few ideas for AU instruments/effects. Do I learn objc? Swift? Both? AudioKit+Swift? Juce + c++? Juce seems great for multi platform but I really just want to make IOS apps for now. I felt like I would start with something simple that routes midi and then move onto effects and eventually sound generation? Any advice? I have a background in ruby and python.

Wow my background is also Ruby and Python as well as Java. I’m going with Swift and AudioKit for now and see where that takes me. If anyone else has advice on this it would be great

I think I’ll do the same. If I end up in objc so be it but either are probably more helpful than jumping into c++ right now when I just want to do iOS dev.

I learned music by emulating my favorite music not starting with music theory, but ended up learning a lot of it along the way. It’s been the same with programming with me that I want to make cool stuff and end up learning a lot of programming in spite of myself lol

analog_matt

@ryanjanik -

Awesome, sounds great. Good luck, and keep us posted!

Lessons I've learned the hard way over the years-

• Pick an idea you love. That way, if it's not commercially successful, you'll still have an app you are proud of and like to use yourself. (The same thing can be applied to music-making in general, I suppose )

• Narrow down the features you want. Then, cut them in half. If you can find the restraint, only focus on the 20% of features that provides your app with 80% of its functionality.

• My first rule of audio programming: Don't wear headphones when writing new sound generation code. Don't want to lose your hearing if things don't go as planned and the code results in loud screeching noise.

• They say people are happiest when they're solving problems. By that logic, Audio developers should be among the happiest people in the world.

• Audio programming can be challenging. Unexpected roadblocks, undocumented functionality, and errors can seem insurmountable. There will be times you feel like giving up. You have to keep going. It's like climbing Mt Everest. One step at a time. Keep pushing forward.

• Focus on something simple and ship it. There's nothing quite like the confidence boost and feeling of getting an app out the door.

• Get the app out to Beta testers as soon as possible. Things you suspect will be incredibly important, may not matter at all to real users. And, real users will discover crucial things that may not have even crossed your mind.

• Stick to your beliefs and core-dreams for the app, even if you get some pushback. Henry Ford use to say "If I listened to what people thought they wanted, I'd be raising faster horses instead of building cars"

analog_matt

@rezidue said:
I felt like I would start with something simple that routes midi and then move onto effects and eventually sound generation? Any advice? I have a background in ruby and python.

Routing MIDI: Cem has graciously open-sourced much of the code in his new ChordBud app

Which is built with Swift and AudioKit:

https://github.com/cemolcay/MIDISequencer

https://github.com/cemolcay/MIDITimeTableView

A good intro to getting started with AudioKit is the Ray Wenderlich tutorial:
https://www.raywenderlich.com/145770/audiokit-tutorial-getting-started

The Ray Wenderlich videos are really great for learning Swift. I'm a subscriber and often reference them.

rezidue

@analog_matt said:

@rezidue said:
I felt like I would start with something simple that routes midi and then move onto effects and eventually sound generation? Any advice? I have a background in ruby and python.

Routing MIDI: Cem has graciously open-sourced much of the code in his new ChordBud app

Which is built with Swift and AudioKit:

https://github.com/cemolcay/MIDISequencer

https://github.com/cemolcay/MIDITimeTableView

A good intro to getting started with AudioKit is the Ray Wenderlich tutorial:
https://www.raywenderlich.com/145770/audiokit-tutorial-getting-started

The Ray Wenderlich videos are really great for learning Swift. I'm a subscriber and often reference them.

Thank you so much for this. I will dive in.

ryanjanik

@analog_matt said:
@ryanjanik -

Awesome, sounds great. Good luck, and keep us posted!

Lessons I've learned the hard way over the years-

• Pick an idea you love. That way, if it's not commercially successful, you'll still have an app you are proud of and like to use yourself. (The same thing can be applied to music-making in general, I suppose )

• Narrow down the features you want. Then, cut them in half. If you can find the restraint, only focus on the 20% of features that provides your app with 80% of its functionality.

• My first rule of audio programming: Don't wear headphones when writing new sound generation code. Don't want to lose your hearing if things don't go as planned and the code results in loud screeching noise.

• They say people are happiest when they're solving problems. By that logic, Audio developers should be among the happiest people in the world.

• Audio programming can be challenging. Unexpected roadblocks, undocumented functionality, and errors can seem insurmountable. There will be times you feel like giving up. You have to keep going. It's like climbing Mt Everest. One step at a time. Keep pushing forward.

• Focus on something simple and ship it. There's nothing quite like the confidence boost and feeling of getting an app out the door.

• Get the app out to Beta testers as soon as possible. Things you suspect will be incredibly important, may not matter at all to real users. And, real users will discover crucial things that may not have even crossed your mind.

• Stick to your beliefs and core-dreams for the app, even if you get some pushback. Henry Ford use to say "If I listened to what people thought they wanted, I'd be raising faster horses instead of building cars"

Wonderful advice thank you so much