Throwing hardware at it!

I tend to keep with a policy of skipping each generation of hardware and since I dutifully skipped the 4th generation iPad, I broke down and bought an iPad Air. There was a $200 buyback that I took advantage of for my ancient and unused iPad first generation as well so it wasn't quite as much of a wallet shock. I went with the 64GB model because I do store tons of music samples on my iPad and the size does matter! I opted for the full size model instead of the Mini because again, with synth apps, the size matters.

Early reports have been very good related to music making on the Air and I can confirm that most of my performance concerns on CPU have been addressed. If you follow my tutorials or postings on Noatikl,  you might remember that I used mainly Sampletank due to its ability to give 4 MIDI channels without too much CPU overhead provided you use a light DAW along with Audiobus.

So, with the Air, I decided to try a piece with a heavy DAW and with 3 concurrent synths recording. The heavy DAW is my favorite, Auria and the synths are the Dxi synth, Thor and Alchemy Mobile. I drove all 3 synths with Noatikl which would usually bring my 3rd gen iPad to its knees. Without a hitch I was able to record all three concurrently into Auria, which is a very heavy DAW by itself.

Another recent addition to music making on the iPad is inter app audio as I blogged before and I added some additional tracks using the Arturia Oberheim SEM synth and my old standby Sampletank for the violin.



Throughout the piece, Auria was responsive and the tracks all recorded without issue. I think I can now add many more concurrent tracks with Noatikl and Auria is now a go-to DAW for me. I'll be posting more as I experiment further. Thanks for reading!

Noatikl hits the iPad! Generative music goes portable!

I'm just back from a wonderful vacation in Bermuda and was thrilled to see that one of my favorite desktop music tools is now available on the iPad - Intermorphic's Noatikl. If you're read my older posts, you know that Noatikl is basically a way to program generative music - you give your voices various rules to follow and it emits the MIDI events to play the way you instructed it too. Sort of like playing God, and who doesn't like that from time to time!

Noatikl comes with a built-in modular synth (Partikl) that you can use immediately without having to hook up virtual MIDI channels but to really get some great sounds, you will probably want to take advantage of Virtual MIDI and use your favorite synths or samplers on the iPad.

On the desktop version, its fairly easy to use up to 16 channels and the CPUs are usually up for the task. The first thing to get used to on the iPad is that you have to conserve the somewhat meager CPU on the iPad. Open up 3-4 different music apps at once and run them all in the background and you'll quickly find this a challenge!

As most initial releases go, this one has a few semi-serious bugs and niggles but I'm confident they will be addressed and there are workarounds for most of them - the worst I saw was that editing a pattern has to be done in "portrait mode" at the moment. Should be sorted in the next release.

Without going too deep into generative music, rules etc, I'll walk through what I did to create a nice chillout piece on the iPad.

First off, here it is:


I decided to use some "lighter" tools to create the sounds and record them so for this example, I avoided the excellent Auria DAW and stuck to the very light and usable MultiTrack DAW for recording.

In Noatikl, drum sounds all go to MIDI channel 10 and you create one voice for each "drum" and attach all of them to channel 10. To do this, you probably want to use rhythmic patterns instead of note patterns and you should make each voice use the GM "patch" you specify. Try to use drum kits that respect the General MIDI mappings.

In my case I used Drumjam - the "lower portion" where you can drag a kit to the pads. From Noatikl, I created a number of "fixed pattern" voices that play the exact rhythms you specify. The key to making things interesting is to create more than one pattern and give each one a "probability" of firing. So for the bass drum, I created a few patterns in Noatikl - one that played on beats 1 and 3, another with some syncopation and a third that only played once. Each had a weighted probability of 33 so that when Noatikl emits the MIDI, there will be some variation.

I did the same for the snare and high hats and then I put a very low probability on a beat 4 cymbal crash. This created a very nice laid back drumbeat on the song above. To further make things less "robotic", Noatikl lets you put "micro-delays" on each note and I varied each note from 2-4 microseconds to make the drumming a little more human sounding.

I used Virtual MIDI to hook up Drumjam to Noatikl via MIDI channel 10 so that when I start Noatikl playing, Drumjam puts out the sounds.

Next I wanted to add bass, guitar and an organ (on MIDI channels 1,2 and 3 respectively). To minimize the CPU usage, I chose to use SampleTank for the iPad. This is a good choice with Noatikl since just one instance of the app can take up to 4 MIDI channels in. This preserves CPU!

I created a simple rhythm for the bass, constrained it to a note range and told Noatikl to use a Pentatonic scale for the piece. I used a similar set of rhythmic patterns for the bassline again with a little randomness to keep it interesting.

For the Guitar, I did the same but then went into the "Chord" section of Noatikl and told the voice to use at least 2 notes and at most 5 - corresponds to how many strings. Also in the "Chord" section I put a  little delay between each note (1-2 microseconds) to give it a "strum" sound.

For the Organ, I programmed the voice to follow the guitar voice one beat behind and to harmonize with it in the Pentatonic scale. For notes I used between 1-3 notes at a time also with a slight random delay.

So with Sampletank, Drumjam and Noatikl open, I played around with the voices and tweaked until I got a nice evolving background sound. Once I liked it, it was time to fire up AudioBus and MultiTrack DAW to do some recording.

One confusing thing in Noatikl is that it can be a "source" for Audiobus, but this will only work if you are using the built-in Partikl synth. If you are doing Virtual MIDI, leave this unchecked and put the synths you are controlling into the inputs to Audiobus (in my case, SampleTank and Drumjam).

I then started MultiTrack DAW recording and pressed play in Noatikl. As the song was generating and recording, I occasionally muted and unmuted a few voices in Noatikl to bring things in and out.

After I got the recording in MultiTrack DAW, I decided to put a traditional "played" synth part as the lead. I closed Noatikl, put Nave into the Audiobus input and played back the recorded tracks while performing the lead part in Nave.

A few takes later, back to MultiTrack DAW to add some light effects, compression and fades and the song was complete.

I highly recommend that you check out Noatikl in more detail if you get the chance.

Beatmaker 2 and Bus news

  

I mentioned in my last post that I was using both Music Studio and BeatMaker 2 as Audiobus targets. Here is the piece I put together in BeatMaker 2. Of the two targets, I reported that I prefer BeatMaker 2 overall for its functionality despite its non-intuitive interface. 

For the piece above, I started by building up a groove using Loopseque - a really unique beat making app that recently added Audiobus support. At the time I started this piece, the Audiobus support was not there yet and I created the beat first using Audiocopy/paste to put it into BeatMaker 2. For drum beats I often prefer to use Audiocopy/paste rather than Audiobus since it is easier to get an exact size for the drum loops this way.

This is where Beatmaker's semi-intuitive interface first twarted me. The size of the sample exceeded the 32 MB default that beatmaker allocates for in-memory samples. BeatMaker 2 can either use samples "in memory" or can stream them right off the storage. For the life of me, I could not find out how to convert the "in memory" sample of Loopseque to a streaming version so I needed to expand the memory buffer for BeatMaker 2. Is there an option in the app? No, actually. You have to go into the Settings app on the iPad, select BeatMaker 2 and there set its buffer size. I'm sure there is SOME WAY to change from memory to streaming, but I didn't find it in my cursory forum mining on Intua's site. The good new is that the samples you bring in via Audiobus will by default be set to streaming. 

After getting the basic groove track in, I started to use Audiobus for the remaining tracks. I brought in a nice ambient sound from Animoog and recorded that directly into the app. I then added a DX7 sounding lead from Propellerhead's Figure which I "played" directly from the app into BeatMaker 2 via Audiobus. 

Having this basic melody established, I filled in the gaps with a few native MIDI tracks using BeatMaker 2's string samples and one of their synth pads. I put an LFO on the strings with a square wave to give a stuttering electronic feel to the track.

Finally I decided to give the some some background ambience with Mixtikl which was recorded via audiobus into the final track.

Overall, once you get used to Beatmaker's bizarre interface, it works very well. I applied some light compression to the overall mix to bring everything together.

As I was completing this, Garageband surprised me by suddenly supporting Audiobus as a target. This is very exciting and gives me yet another good option to use my many synths on the iPad into Garageband, add some parts from its powerful instruments and if I really want to go crazy on the mixing, I can upload it to my iMac and complete the effort in Logic 9 or Garageband on the iMac. This will be fun! 

In typical Apple fashion however, they are sticking to their "Roach Motel" standard - you can use Garageband as a target in Audiobus but not as a source. Music gets in, it doesn't get out (easily)! 

Bussing with Xewton music studio

I've had yet another few weeks with Audiobus and have been working a bit with Xewton Music Studio and Beatmaker 2.

In my last posts, I discussed the pros and cons of some of the other DAWs as Audiobus targets so here are 2 more. Auria and Multitrack DAW are both good targets but work only with audio files. The appeal of Beatmaker 2 and Music Studio are that in addition to acting as targets for Audiobus, they also provide MIDI tracks natively. This lets me combine their native sounds with other Audiobus inputs.

Both tools provide this but both also are imperfect and have their quirks. The quirks are more shortcomings in what they provide natively and not so much Audiobus. Beatmaker 2 is the more feature-rich of the two but suffers from its often counter-intuitive interface. It started its life as a low power iPhone app and that history shows. Many folks have used Beatmaker for years and for them, the issue is moot. For me, learning it is frustrating in many places. Most functionality is there and if you mine their forums you can figure things out.

Music Studio on the other hand has a beautiful interface and is fairly simple to use but has some glaring functional omissions. Fading out a song is almost impossible if you mix audio and MIDI tracks. The effects chain is also somewhat bizarre but if you are patient enough, you can get good results. For fading, you might do better to mix down and then copy into another app for posting.

My verdict for these two tools is to give Beatmaker 2 the nod. I need to get used to its weird interface, but functionality wise, it provides good audiobus support with decent MIDI capabilities. It's native samples are meh, but you can add your own.

Overall, MultiTrack DAW remains my favorite when I don't need MIDI. Auria is just too crash prone with audiobus (I'm on an iPad 3 with nothing else running but it is flakey with Auria).

Here is a track I put together primarily with Audiobus on Xewton Music Studio:


The beats and bass were created in the Korg iPolySix and it was easier to just mixdown and audiocopy/paste those into Music Studio - so no Audiobus there, though Korg does support it. The next few tracks were Audiobus from the Moog Animoog synth. I added a couple of native Music studio tracks for muted guitar, muted trumpet and flute. Then I added some ambient bells/drones via Audiobus from Mixtikl.

So, Xewton does work with Audiobus fairly well if you really want to use it. I think I may stick mainly to MultiTrack DAW with perhaps a little BeatMaker if I can ever wrap my head around its interface.