Thursday, August 23, 2012 by Tim Webb
NLog Synth not only "supports" Audiobus, it already works! Tempo Rubato's Rolf Wöhrmann tweeted out this picture today.
In the picture above you see SoundPrism Pro and NLog Synth Pro acting as the sound sources. These are both being fed to an instance of Loopy HD. All of which are running on the same iPad. Alone this is great news; another app to play with on Audiobus!
Things got even more interesting in his next tweet. In this setup, pictured left, SoundPrism Pro's audio is first going through NLog Synth Pro and then into Loopy, all through the power of Audiobus and science. NLog is now a filter or effect for SoundPrism!
Apparently all that talk about Audiobus being easy for developers to work with is true. Rolf has already added a whole new audio path to NLog this morning! Seriously. I was talking with Michael Tyson about the Loopy update, when he mentioned he got a call from Rolf early in the morning with this filter idea. A couple hours later Rolf already had it working and compiled for testing!
I spoke with Rolf about the process of adding Audiobus to NLog, trying to get a sense for how challenging it will be for other developers to implement it. He mentioned how easy it is to set up an app to be an Audiobus destination, being the output in the flow shown here. In just 1 or 2 hours he had NLog set up so that it could be a recording destination for Audiobus.
Fitting it in between two apps as the Filter insert was only slightly more difficult, taking about half a day. Keep in mind that he is the guinea pig beta testing the SDK documentation, so some time is added there. Rolf said it took about the same amount of time and effort as implementing Sonoma's Audio Copy/Paste. I asked him how he felt less experienced iOS developers would fare:
Michael made it very easy. He calls it a 30-minute approach to implment it. To be fair this only works if you had constructed your app, let's say, in a reasonable, standard, way. However, if you have coded your app in a strange way, then of course it won't work in 30 minutes.
It's so quick, I mean no fear. If your app is well designed you are done in a day.
He added a bit of advice for developers who plan to implement Audiobus in the future.
All of this Audiobusing things together is only fun if you have low latency. Every point in the chain adds latency. So what Audiobus requires is that your app is tuned to work with low latency buffer sizes.
Audiobus works best with 256 buffer size, and this is the lowest apps traditionally use. However, I made an informal review and I would say about a third already handle this. For another third it is probably not hard to adapt. For the rest it may take a few days.
Audanika's Sebastian Dittman, co-developer of Audiobus with Michael, was so excited he did a little demo. First you hear SoundPrism Pro, then you hear the same instrument being run through NLog's filter and effects, all recorded into Loopy.