Audiobus: Use your music apps together.
What is Audiobus? — Audiobus is an award-winning music app for iPhone and iPad which lets you use your other music apps together. Chain effects on your favourite synth, run the output of apps or Audio Units into an app like GarageBand or Loopy, or select a different audio interface output for each app. Route MIDI between apps — drive a synth from a MIDI sequencer, or add an arpeggiator to your MIDI keyboard — or sync with your external MIDI gear. And control your entire setup from a MIDI controller.
Download on the App StoreAudiobus is the app that makes the rest of your setup better.
Comments
Did Filtatron really go AUv3?
I wish. Seems possible, though. But not yet, anyway.
well, I clearly wrote that I've checked all the examples posted (and a lot more on YT and SC) and I'm not exactly unexperienced with synths in general.
I have the software predecessor of John Bowen's hardware Solaris and in fact it was John Bowen himself who turned the afforementioned modular patch into a standalone synth.
(just to be clear it wasn't about some KVR freebies)
You may call me whatever names you like, just don't question my ears or call me naive.
My setup consists of 2 different hardware DSP environments, 2 Windoze DAWs and IOS.
It's no problem to disagree about sound preferences, but imho it's a weak response to turn a fact based comment (and answer to a question posted) into a supposed attack.
I just followed what the thread's title suggested - and obviously this version is not my cup of tea. That's about it, no need to extend this further.
It seems to me there is a big difference between a claim about taste and this claim you made:
"you'll never be able to build any 'reasonable' synth in such an environment.
The interactivity on the circuit level steals way too much cycles from the CPU."
Hubba hubba! (Doesn’t add anything to the discussion but had to be said)
indeed: it's the difference between the bottomline in post 2 about what the developer(s) told me - and an answer to a question in post 1, based on a real world example
It seems to me that the quality of synthesis comes from the design of the oscillators, filters, envelopes, etc. The point of this thread is to highlight the benefits of delving more deeply into those designs. In order to do that you need to understand how sound works -- how logical expressions form organized tools that interact in complimentary ways to produce rich sound. Of course, the more knob modulation you get going, the more it processing it will eat up -- you will find a similar case with Native Instruments' Reaktor Blocks, or VCV Rack, and you will also find this happens when you start automating parameters in a DAW. Right now a lot of the limitations (like with video games) comes from the hardware available. So, we used to have black and white iPods, now we have the iPhone X but as far as modelling physical chips, even the iPhone X quickly limits how much algorithmic math you can get going at once. One advantage with Audulus in this case is the ability to optimize logical patterns while streamlining how you achieve it.
In a sense it appears that you are actually criticizing the Audulus community for not being able to make synthesizers that are as good as the other iOS synthesizer apps around.
Right now there are a few bright people working at very foundational synthesis techniques, being inspired by academic research papers, innovations in Eurorack designs, etc. As I mentioned earlier in the thread, the work being done on Rob Hordijk's designs -- which can be found here -- include just a few examples what I find to be captivating regarding the philosophy of synthesis.
You said that you would "appreciate if the developer(s) would focus more on [special tasks], instead of the current 'high complexity - sacrifice tone' approach.
From what I have taken the time to share above, not only is the community focusing on tone, but there is also alot of beautiful work being done on "special tasks," like quantization, for example, or sequencing for that matter.
Is it just me that finds this freaking exciting?
Nope, but as a fairly regular SnuXov user I'm waiting with baited breath to hear how it works out providing AU plugins to the rest of the system before diving in. Now that could get me more than a bit fidgety.
why would I 'accuse' anyone ? let alone a community...
I only mentioned some flaws I hear in the engine's sound, but that's no more than one personal opinion.
Regarding the design aspect I'd never discourage anyone to dive into modular to develope ideas or try them out. I know the dude from Relab-DK (aka Warp69) built and tested a lot of his early Lexicon 480 reverb algorithm within a modular system for ease of use.
The same applies to 'my' synth example which had it's roots in a modular patch and has been significantly modified/enhanced while being turned into a standalone plugin.
The final result would be inoperable in a modular environment, but on the other hand the synth wouldn't exist without the modular idea as it's seed.
I don't judge, but communicate experiences - no less, no more.
One of those experiences is that the load with sophisticated modules can max out any system with ease - and all modular approaches had a significantly higher CPU load.
In my case it's easy to compare because synths and modular modules use the exact same (basic) code fragments.
>
I wrote that my personal use of modular synthesis is for 'special tasks' (with a rather low number of modules), opposed to complete synths like a (say) Jupiter-8 or similiar.
So I'd appreciate focus on individual module quality (for obvious reasons) but I'd never expect readymade solutions, which would be quite contradictary to the modular approach.
Here is a very minimal 3 operator synth in the FM style. Low CPU. Had fun building it today in the sun.
^ very tidy, nice work! I like that exponential decay on the envelope, I hadnt seen that module
I love that MG low pass filter! Doesn’t sound like a “sacrifice” of tone to me! My first foray into Audulus was just a single oscillator into that sucker, with an ADSR to control it and then into an amp and ADSR. The resulting bass patch sounded as good as some of Model D and Model 15 basses. I was shocked.
I’m never going to be an Audulus gearhead, (don’t have the patience) but no way I agree with anyone who thinks its “sound engine” (whatever the hell that is) is inferior. That doesn’t even make sense in this context where any formula can be used to create any sound in the same way that an app developer does, but without having to know any programming language, or even have a development PC.
It’s like saying a gourmet chef with access to every ingredient on the planet will never be able to make a decent meal because his pots are the wrong color.
@Telefunky is right about CPU loads. The patch-ability of Audulus makes it much more demanding for FM style synthesis. Above I have a 3 operator, 1 envelope monophonic patch. KQ Dixie is 6 operator 6 envelope 16 voice polyphony, for example, and it very easy on the processor. At the same time, for me, it quickly becomes a dead interface because I can't really see what it's doing what like I can in Audulus. The 3 operator patch above isn't very sonically interesting at this point. So I am trying to learn how I can improve that.
When he criticizes the sound engine I think some things need to be teased out. The quality in Audulus is definitely there. It sounds great. The machines you build in it though can be more demanding, so things need to be thought through in different ways. I personally just found his take way too heavy handed, but there is definitely validity in what he says.
I would basically have to agree with @Telefunky — modular audio environments are very interesting, and Audulus is a great implementation, both UI/UX-wise and technically — of one.
However, trying to do “normal” music production in such an environment is a bit like building your own car from scratch each time you want to go to the supermarket.
I‘ve used Audulus (and SunVox) countless times for specialized, one-off “DSP tasks” (noise generator, spectral manipulation / inversion, specialized EQ or Delay, looper, etc.), and they’re amazing for that.
Another very sensible use case for modularity is the MIXING and EFFECTS stage of a production.
So, as soon as Audulus gets AUv3 hosting and can be used to host “real” synths and effects, with totally free routing between them — now THAT will be amazing and I’m sure @Telefunky will agree
Meh... It depends. I built two custom patches for my Keith McMillen Instrument's BopPad and I am quite happy with the sound, CPU load and versatility.
You can make yourself little tools, interesting oscillators, etc. Because you can delve deeply into how everything works, you end up getting more control out of things. I wouldn't say it "is a bit like building your own car from scratch each time you want to go to the supermarket." You can make generative patches, but you can also make synths that work in productions.
This winter I drew some schematics, did a board layout, soldered all my components and mounted a couple of guitar pedals into their respective enclosures. I could have just bought some pedals, but now understand how those circuits work. So if I go to use something complex like a Chase Bliss Audio Pedal, having done some basic circuit board work on my own, I will get way more out of them -- I will understand how the dip switch routing is shaping the signals.
The other thing is I treasure those pedals now, just like I treasure the things I build for myself in Audulus. That's priceless.
@futureaztec yes, I understand the “meta-usefulness” of knowing how all the stuff works... I still think that for real music production, ready-made synths (including their optimized UIs) are more practical.
But I’m also probably biased because I used to do music production commercially, often for clients, and so it’s a bit like plumbing or being a carpenter or whatever... “what gets the job done, fast”.
but I fully appreciate the more idealistic view, of course.
@Telefunky I guess my main criticism of your criticism is that you're not specific about what actually needs to happen. You say we need to spend more time on making individual modules better, but without explaining how or why. What's wrong with what's there right now?
If you create a modular synth patch in Audulus, all you need to do to make it a standalone synth is to attach some knobs to the relevant controls, select everything, and press Group. This is essentially how modules are made, but you can also make a fully-voiced standalone module.
Also, the engine that's used by Audulus is pretty much shared by all digital modular synths. The tone of a normal subtractive synth comes mostly from its filter and VCA. It doesn't seem like you've spent much time with the recent additions to the library. I did an A/B of our Moog filter against a Eurorack Moog clone filter and no one could tell the difference. And that's not even my favorite filter of the bunch that's in there!
Furthermore @sevensystems too - Audulus allows you to create all-in-on synthesizers that you can just pull up and use without treating it like a modular synth. Check out this simple example for one:
I am designing more of these, as well as clones of actual synthesizers so that people can jump in and feel more comfortable with Audulus from the get go.
So in short, Audulus doesn't "force" you to work at a low level where you have to "build your own car from scratch." At more it's like when you make music, you have to plug in your guitar to your pedal board and your pedal board into your amp. It's neither rocket science nor car mechanics.
But really, if anyone thinks modular synthesis is hard - respectfully, you are wrong. It looks complicated, but 90% of the time all people are doing are hooking up a normal Minimoog-like subtractive synth patch VCO->VCF->VCA with a sequencer or keyboard control and some effects.
Deadmouse uses almost exclusively modular synthesizers, no? Not to mention basically all sound designers use modular synths because the flexibility of the routing you get is better.
Once you understand modular synthesis and get used to the workflows, its a pretty quick way to make music, especially vs. a lot of the menu diving you have to do on normal standalone synths that aren't 1 knob per function.
I agree with this. For music production tasks, having to build your tools may be a big distraction or timesink.
The real advantage to knowing how to use Audulus is that you can build custom things you need for your own purposes. Things that may not exist in pre-built app form. I am not a dev, so this is a very cool thing for me. Also, you may save a lot of money since you may not have to always buy your tools when you can make one easily yourself.
I think embracing both workflow strategies is the best. Use the good pre-built apps when they suit the purpose, and focus on music production. Build a custom tool when needed, or even invent something that doesn’t quite exist yet.
Really, I can relate to both sides of the Audulus argument.
@cracklepot - I feel like it's more of a time management argument. Why not play with Audulus on a fundamental level building stuff from scratch when you don't have a deadline or aren't feeling inspired musically, and then have a bunch of "presets" lying around for when you are looking to get to work? If the object is only to make music as quickly and efficiently as possible, then sure just get a synth with a lot of presets and blast away. It worked for Depeche Mode at first, right? But there's no getting around the fact that learning more about your instrument will make it possible to get better sounds and ultimately free your creativity more - even in those assembly line situations where you need to make a fat bass patch but don't actually know what goes into making a bass patch fat.
For me personally, I love to tinker with Audulus and build simple stuff as well. I am not under any time constraints and enjoy learning all of this digital music tech at a fundamental level. It is the most effective way to learn these things.
I know others may feel they don’t have the time, patience, or mental fortitude to learn Audulus, for whatever reason. They could still benefit from the freely available modules for now, and take some time to learn more about building stuff later.
I think that it doesn’t have to be either/or is all I was trying to get across. When the AUv3 update hits, this will be the case even more. Use it all.
I slightly reworked the 3 Operator PM Synth I posted earlier going off of the suggestions of the module author.
Here it is again. a little more aggressive, by modulating the course frequency ratio knobs.
As mentioned I can only judge about what I hear (trying to pickup recent stuff and leave out oldies) and the result was a rather tinny sound. Considering the sheer number of elements in those patches my conclusion was that 'infamous' sacrifice assumption.
I totally get your point that most building blocks CAN be extended ad infinitum, but that wasn't very obvious in any of the examples. And of course that there may be new stuff.
But I consider myself experienced enough (18 years with synths) to tell by listen if something is an interesting candidate to own.
Wouldn't even bother to get Audulus 3 for those few bucks if it wouldn't require a new iPad.
(sorry, but I won't part with IOS-9 on my Air-2 and an iCA4+)
not sure if you refer to content or presentation here, but the videos are solid.
No mumbling words and back-forward-back-clicky-dee-click and a nice presenter's voice.
That's correct, it's not complicated at all and the exploring thing is adventurous and inspirational - as long as you have the time.
Or someone is interested in his personal signature soundset. I know some great sounding Psytrance from Israel that's heavily based on modular.
(as a sidenote: I don't think programming (languages) aren't complicated either. Once you get behind the 'programming paradigm', the language doesn't matter. The choice is more a thing of convenience and efficiency)
But (as mentioned by @wim ) the algorithmic part isn't that simple, even if all math operations are be available.
As the designer of the Access Virus once mentioned (a favourite quote of mine):
it's easy to pick a filter formula from a math book and transform it into code, but it takes a lot of experience to choose the proper one - and the one that sounds best may not even be the one that's the most precise in technical terms.
There also is a strict time frame limiting calculations, you HAVE TO get your stuff ready in that amount of time. Those specialized DSP chips (or cpu instructions) do exist for a reason.
(this is a general hint and in no way questioning that Audulus isn't setup properly in that context)
A sine wave is a sine wave no? Maybe I am getting this wrong, but digital sound doesn't have a character like tinnyness. The only thing I can think of is that you have sampling quality and all of those details, but this doesn't factor here. For example, here is a member who posted an envelope today. The funny thing is that the so called "devs" are the community members because Audulus is both a modular synthesis environment as well as an open source programming interface -- like Reaktor.
This is why @wim pointed out that your criticism is "like saying a gourmet chef with access to every ingredient on the planet will never be able to make a decent meal because his pots are the wrong color."
Mean one is tasty! Thanks for these
SnuXov user (lover) here too, though I don’t think it’s analogous really with Audulus (in PC/MAC, SnuXov is to Psycle or Jeskola Buzz as Audulus is to Bidule or Reaktor maybe) the better iOS analogy is probably zMors modular which hosts AUv3 already. Personally I just buy and use them all...
digital sound can have any character the calculation processes.
There's few doubt that the guitar amp sims by Yonac and Positive Grid don't sound identical.
If all would just be numbers, this could hardly be the case.
That's what's called an 'engine' - the basic internal sound processing of an app.
Some stuff may be common (shared by libraries), some may be custom built - but not all are created equal.
FluxFx sounds crap in my ears (but it's perfectly ok for live performance), the Holderness FX are on par with Strymon pedals (imh ears).
well, that's stretching 'Open Source' quite a bit
But I'm actually quite interested in Audulus, because it partly seems to apply the strategy of the former Creamware (now SonicCore) Scope system.
This programming style was quite hip in the late 90s, I remember the Danish Koblo system, but likely there were more. I'm familiar with such stuff since 2000 and I'm a developer myself - but in the database/kowledge management domain, not audio.
@Telefunky I started two write something but between you and a couple of other people today I am worn out. If someone here has the energy to explain the difference between an amp simulation and a digital FM synthesizer please do.
Save your time and don't bother.
There is no fundamental difference for an amp will oscillate when feedback is high enough. An amp contains filter stages, too.
In the end almost all digital sound processing breaks down to a simple delay and feedback, but there's a million different ways to implement that. Another reason why there's variance in different sound processors.
ps: I may refer once again to the Access Virus designer (Mr. Kemper), who's also known for his Profiling Guitar Amp. Which quickly became the defacto standard for modelling amps, including the (very) big stages. Metallica etc. The amp model inherited it's excellent sound processing from the synth experience of the company.
This is all fairly annoying because the point was not between an amp and a frequency generator, it was between amp modelling and digital synthesis. Most of the content of what you say falls into the pattern of "I know smart people so I am right," or "I have owned synthesizers for 15 years, so my ears are special."
There is also the issue that there are multiple levels of development. There are people who design the nodes for the people who design modules for the people who design synthesizers or sequencers with them.
The source of all of this annoyance is that this thread was started to share some of the things that I find inspiring and fascinating about Audulus that I thought others would too and you have come along to claim that "you'll never be able to build any 'reasonable' synth in such an environment." Then you proceeded to write these long confusing half baked paragraphs that seemed to be fairly mixed up, probably because they don't appear to be coming from a place of interest, but from some sort of hell bent attempt to defend a position that was falling apart.
But, then, for me I feel like this might just be one of those things I have to learn to walk away from and continue to work on what I was enjoying and forget about trying to point out some flawed reasoning, when, to me the sound in Audulus is fantastic, so what do I care what you think?
This programming style was quite hip in the late 90s, I remember the Danish Koblo system, but likely there were more. I'm familiar with such stuff since 2000 and I'm a developer myself - but in the database/kowledge management domain, not audio.
I do do audio/DSP programming and you need to stop now.