Audiobus/sequencing/recording tests – Beatmaker 2 vMidi Sequencing

Well, after my last blogpost (Audiobus sequencing post) i thought it only fair to investigate the options for supporting Audiobus with a currently available linear virtual midi sequencer running on iPad2.

Seeing as i’ve had Beatmaker2 sitting on my ipad for an eternity it’s the perfect choice for the test (being virtual midi enabled and linear in its recording/sequencing method). My initial concern regarding this app is all the additional features it has available and the amount of CPU this would grab, especially as the whole purpose of this experiment is to run as much synth power as possible and record the combined (and sequenced) output.

In Beatmaker (BM2 from here on), opening a new project and using 3 instances of the keyboard sampler instrument, I selected no sound or presets for the instruments which instead were set for virtual midi outputs on channels 4, 5 & 6 respectively.
Sunrizer, Sunrizer XS and Magellan were opened up and configured each to one of the vMidi channels, BM2 was checked to be sending to the iSynths and a simple sequence assigned to play on each synth.

So far so good. Everything sequenced is running through and playing back with 100% clarity. Now for recording…

Apps running under Audiobus

Apps running under Audiobus

Ran up Audiobus app, set the 3 synths up in the input slots (one restriction, 3 inputs maximum when recording) and set the output destination as the AudioShare app. My gut feeling is AudioShare has the smallest CPU drain of the current selection of output apps, plus the recordings can be ACP’d directly or even posted straight to soundcloud.

Ready to rock and roll – switch back to one of the synths, use the audiobus panel to start recording, fast switch to BM2, hit play… and… relax. Everthing fell in to place seamlessly. Switching mid recording over to Audioshare to watch to sequence fade and stop the recording caused no glitch, nothing. So we’ve got a successful test run recording the three synths, so what’d i do next?? Upload of course 🙂

So this got me thinking. We have been given a single effect slot on audiobus. And the NLog synths have the ability to fill this… then i wonder if the NLog could also be used as a forth sequenced synth? Interesting thought…

Killed off Audioshare for the time being, set up NLogPro in the Audiobus effects slot, vMidi channel 7. Added a new instrument in BM2 set up to suit and programmed an additional sequence for it. Noticed a bit of audio glitching, but nothing horrendous. Opened Audioshare and the glitching became far worse, unworkable.

Swapped out Magellan for its smaller iPhone counterpart, Magellan Jr. Also swapped NLogPro for NLogMini. Audio improved and playback was far far better.

Four sequenced synths

Four sequenced synths

This is where things become slightly confused. i tried this twice, identical setup and everything. The first time when i switched back to Audioshare, the audio glitching returned but only while recording. It took 1:30 to record the 45 second sequence, however undeterred i replayed the recording to find it perfect! My belief is the glitch comes from the overworked CPU which carries out both audio rendition and recording function simultaneously, hence the glitches are in time are synchronised across the board and therefore not apparent in the final recording. Interesting, still usable but a bit hit and miss.
Here is the second test result…

Take 2 of the second test came the following day. I expanded the sequence from 16 to 24 bars and reselected some of the sounds. Actual setup was no different whatsoever. This time there was no glitching either on playback ir recording, no issues at all. Perfect sound, perfect rendition. And here it is…

So, it is possible to run 4 iSynths, using the smallest CPU footprints possible, using BM2 as the sequencer for them and to record them using Audiobus.
Other iSynths that i did try during Audiobustest 1 but were immediately detrimental to the process were: Cassini, Samplr, PPG WaveGenerator and iPolySix. These all came out as being a bit too CPU heavy when it comes to multitrack recording, but they will pr.

Whilst carrying out these tests i’d been discussing my goings on on twitter, taking advice and ideas from others as the procedures progressed.
One interesting test carried out by Clif Johnston ( ) is that he setup and tried ‘#Audiobustest 1’ but used MultiTrack DAW as the output (recording) device to find each synth was recorded on a separate channel within the DAW! Something i was completely unaware of, and a very handy feature!

Prognosis (conclusion):
So it looks like the possibility of multiple synths with a linear sequencing package is already a reality, so long as you can put a bit of time in to BM2. Still having a hard time recording & manipulating controller movement. Some work, some dont.
Hardware/processing power is always going to be the hold-up factor with regard to what you might want to do, but certainly running the smaller synths it is possible to put together a backing track (and more) and record it in Multitrack, and then add additional live or sequenced parts, drums, whatever.
The process is not as impossible as i had originally believed, but there are still a few things in the way.
In these times of getting by I, like many of us, cant afford to keep upgrading my iPad, and i won’t ask for that to be thought of as any kind of urgent need at home. I will get by with what i have, use what is available to me and see how it can be abused and moulded into something useable. But i still don’t know if i’ll get along with BM2…

You may have noticed there are no drum apps included in the testing here. A while ago i tried testing Genome with as much as i could. One thing which that testing taught me was that drum apps seem very greedy when it comes to CPU hogging so for now i’ve ignored them. On a real track, i’d sequence the synths as usual, but would carry out the recording of the drums seperately, either from a sequenced track within BM2 or sequenced patterns within whichever drum app i was to use.

And a final set of Thanks…
Many thanks must go to Clif Johnston , Upright_Music , Sinapsya , Paul Collins , SmiteMatter and of course AudiobusApp AudiobusApp for encouragement.


~ by carbonishsound on 13/01/2013.

3 Responses to “Audiobus/sequencing/recording tests – Beatmaker 2 vMidi Sequencing”

  1. Hi interesting reading indeed and a good reference of the limits of what a ipad 2/3 can achieve with Audiobus.

    Software does seem to be moving quicker than the hardware that supports it very quickly in the last 6 months.

    At present I think BM 2, Genome, Cubasis and possibly Midi Pattern Sequencer are the only apps you can use on the outside of Audiobus to use as a virtual sequencer to multiple synths that gives you quantisation options which is great to get beats in time together. I don’t own the last 2 apps mentioned so can’t testify to that.

    Have you tried Little Midi Machine it can only control 2 synths at any one time but its quick at fast switching between multiple synths and easy to set up and use. I expect its CPU footprint is low as well. I know it hasn’t got all the bells and whistles like effects and recording that BM2 but its a option.

    I managed to control Grain Science and Megallan JR with it using Audiobus and also run BM 2 at the same time running NLog Pro and Sunrizer in Audiobus recording into audio share. Got very slight glitching I think and you have to be selective on what patches you choice especially in Sunrizer and Grain Science.

    Thanks again for the article and the mention!

    Paul Collins

  2. I too have been using BM to sequence synths and drums, but usually record them in-app or through Audiobus, and then paste the resulting audio files as new tracks in BM. You can also use BM to sequence drum apps, though if you’re using BM’s drum pads for this, you need to remap the MIDI notes that it sends. I’ve posted about how to do this, using MidiBridge, on the Intua support forum:

    • Very handy and useful forum post there! Thanks for that. At some point i’ll definitely be programming druns and that would have been a pain to work through! It’ll be interesting to see what the CPU can manage with different drum apps running. Its all a bit of a work around, but a worthwhile experiment. Cheers.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: