We are beginning to build up quite a collection of creative audio processing tools under iOS and these had had a couple of notable additions in recent weeks including the brilliant Flux: FX. Another recent release that falls into this ‘creative’ category is apeSoft’s Sparkle.
I’ve reviewed a number of apeSoft’s (and, indeed, Amazing Noises; the two development teams are linked) apps in the past, including apeFilter, Reverb FDN and Limiter, and they always brings something with a touch of ‘left-field’ to the table. Based on the app’s blurb on the App Store, Sparkle – or, to give it its full title, Sparkle Advanced Cross-Synthesis – looks like it is made from the same kind of mold. So, if you are into sound transformation and sound design, is Sparkle something to add to your iOS music app collection?
Sparkle is currently priced at UK£4.99, is a universal app, requires iOS7.0 or later and is a modest 19MB download. The underlying idea of the app is that it will take temporal elements from one sound (referred to as the source sound) and will then apply those features to a second sound (known as the target sound).
The app’s description – and the in-app manual – say little more than this about the exact nature of the processing but, in principle at least, this is not totally removed from what’s achieved by a vocoder, where a vocal sound is blended with a synth sound and can be ‘played’ as a sort of synthetic vocal instrument.
Sparkle can be used in two configurations. First, in Sampler mode, both the source and target sounds are based upon pre-recorded audio files stored on your iOS device. There are a number of suitable files included with the app but you can, of course, add your own (for example, via iTunes File Sharing).
Second, the source sound can be a ‘live’ audio input via the mic, Audiobus or IAA. You can, therefore, feed Sparkle with audio from another app (there are actually various ways to set this up depending upon the Audiobus or IAA route you take), let it apply its processing to the target audio (that still has to be an audio file held within Sparkle itself; more on this below) and then the output can be recorded into your DAW of choice (although Sparkle includes its own internal recording feature if you prefer that option).
If you have read the recent Reverb FDN review, then certain design elements of Sparkle will seem familiar. The screenshots shown here refer to the iPad layout. Things are, of course, somewhat different on the iPhone but the same controls are available even if they are arranged differently. The upper strip of controls that provide access to the ‘play’ (transport) control, the audio files stored within the app (this is where you load either the ‘source’ or ‘target’ audio files from), the preset system and, to the right, toggles for the accelerator and LFO options (these can be linked to parameters within Sparkle for real-time adjust of the effects) and the Tools menu (where you can set up those real-time controls or access the manual).
Along the base of the display you can access the various post-processing audio effects – a parametric EQ, a high-shelf filter, reverb and a compressor. This row of options is completed by the internal recorder and the output volume controls.
The bulk of the screen is taken up by two circular displays that house the waveforms for the source and target audio and the various controls that dictate just how the target audio is processed based upon the source audio’s sonic/temporal properties. Easy to miss amongst this is that the two large circles are actually input gain controls for the source and target respectively.
Within each circle – and located top-center – are mode toggle buttons. For the source audio (on the left) these allow you to switch between sampler (internal audio file) and mic/AudiobusIAA inputs, while for the target audio (on the right) you can toggle between normal, robot and whispering options. These three different modes obviously apply some different filtering to the audio output but the names are suggestive of the effect that is produced.
Perhaps the key control is the Hybridization slider located just above the two circles. This essentially controls just how much of the temporal properties of the source audio are imprinted upon the target audio, with higher values producing a more complete transformation. At lower settings, you can hear the target audio in a more ‘normal’ fashion but, depending upon the other settings, it may not be unaltered (i.e. unprocessed).
Underneath both waveforms are the ‘ratio’ controls. These control the relative playback speed of both the source and target audio flies. A value of 1 means that they are played back at their original tempos but you can speed up, slow down or even reverse playback to create all sorts of additional interesting effects or to match the tempo of any rhythmic elements in Sparkle’s output to other elements in a wider workflow. Obviously, when using a mic/Audiobus/IAA source input, the ratio control for the Source audio is disabled.
In between the two circles is the Pitch Shift control and a small virtual keyboard icon. With the former, you can set a static pitch shift for the audio output. Again, this might be useful if you are trying to match the pitch to some other audio elsewhere in another app.
However, if you tap on the keyboard icon, a virtual keyboard appears. From here you can ‘play’ the pitch-shift amount in real-time. You can also do this via a suitable MIDI keyboard and I had no problems hooking up my CME Xkey and triggering the pitch-shift via this route. Depending upon exactly what audio you are processing within the source/target combination, you could create something melodic via this route, although, given the sorts of processing results Sparkle generates, ‘conventional’ melodic is not perhaps what you should expect. It is, however, a lot of fun to play with.
There are a few other controls within each of the main circles. In the source audio circle, you also get an adjustable filter. Here, using a combination of finger gestures, you can change the gain and width of the filter slope and this changes the tonal character of the source audio that, in turn, will influence how its properties are imposed upon the target audio. Within the target audio circle, you get Envelope and Denoise controls. The former influences the tonal preservation of the target audio while the latter selectively removes spectral components. Higher values of both take you further from your original audio.
Whatever the precise algorithm sat behind Sparkle’s hybridization process, I have to say that the experience of exploring it is a very interesting one. Unlike something such as Turnado or Flux: FX, which offer a number of very different effects options, Sparkle is pretty much a ‘one trick only’ effect.
However, because the results of that effect are dependent upon the combination of source and target audio choice, the end product is always something different. You might have to work a bit harder to use this kind of effect but, if you are prepared to experiment, you will be rewarded with some genuinely inspirational – and often wonderfully quirky – audio results.
Some combinations didn’t work particularly well but if helps if the source audio choice has some rhythmic element to it (drum loops and vocals can work well here) while the target audio is best if it has something going on most of the time (vocals can work well here also but synth pads, mixes of instruments or even full mixes make good options).
For example, with a drum beat source and a vocal target, the end result is a sort of rhythmic electronic vocal effect (depending, of course, on exactly how you dial in the various settings). The more complex the rhythmic elements within the drum loop, the more interesting the vocal effect becomes. As a special vocal effect added to an electronic dance music production, there is plenty of ear candy potential here.
From a practical perspective, there were a couple of extra things I might have liked to see in the main screen. First, while you can audition the audio files when selecting them in the File Manager (and so anticipate what sorts of effects you might create), it might be nice to also have an ‘audition’ button within each of the ‘circles’ just so you can remind yourself what the raw material sounds like.
Second, a bypass button, so you could just switch off the processing easily, would also be nice to have. Third, some means to blend the source (unprocessed), target (unprocessed) and the processed signal (I kind of smart wet/dry mix control) would also be good to see and open up some further options.
These are, of course, the sorts of details that might easily be added and I’m sure apeSoft probably have a number of other possibilities they are considering on this front. Fingers crossed they can keep the ape moving forward with some regular updates.
Used as a stand-alone app – with both source and target audio based upon pre-existing audio files – I had no issues using the app at all.
Used under Audiobus, I was able to place Sparkle in the Audiobus Effects slot and feed it live audio as the ‘source’ signal from the Input slot (for example, using Funkbox), use that to process a pre-recorded audio file used as my ‘target’ audio, and then feed the result out to Cubasis sat in the Audiobus Output slot. This worked great and, providing you can match the tempo of any rhythmic element of the source audio to that contained within your Output slot project, you can get as creative as you want.
Things were perhaps less straightforward under IAA, although this is not through any technical issues or bugs I encountered (indeed, Sparkle performed very well) but simply because of limitations with how audio can be routed through the app to and from any IAA host app.
For example, I tried inserting Sparkle as an insert effect on an audio-based drum track. In this position, the audio on my Cubasis track acts as the Source while, as usual, the Target audio is based upon whatever audio file I chose to open in Sparkle. At one level this works fine; the Cubasis drum track could be used to modulate the sound of the Sparkle audio source. However, there is then no way to record the processed audio output from Sparkle directly back into Cubasis (or, at least, not an easy way I could find; solutions welcome if you have one). Equally, while using my Cubasis drum track as the Source audio for Sparkle, I couldn’t then hear it ‘dry’ within my Cubasis project.
The solution for this latter issue was simple enough; either duplicate my drum track so I can dedicate one instance of it for driving Sparkle or use Sparkle as a send effect rather than an insert effect. Neither of these allowed me to record Sparkles processed audio directly back into Cubasis though so the most straightforward option was simply to use Sparkle’s internal audio recording capability and indulge in a little old school audio file copying/pasting; a bit of a pain but it does get you there.
I have to say that this is not really just down to Sparkle; it is simply that audio routing is not yet as flexible in your average iOS DAW via IAA as it might be in a desktop DAW. Hopefully, that is something that will change with time.
The other option that would have be great to see is the ability to use Sparkle’s processing in a situation where both the Source audio and the Target audio are already present as audio tracks within your DAW. This would require you to be able to send both audio tracks to Sparkle at the same time, for it do then do its thing, and then for Sparkle to route the processed signal back to a third track within the DAW. I’ve no idea if that’s even technically possible under iOS (or whether you could create some sort of workaround via Audiobus in some way) but it would be great to see (hear!).
I really like what Sparkle can conjure up. Providing you are prepared to spend some time marrying a suitable source with a suitable target, your efforts can be rewarded with some really interesting results – from rhythmic robotic vocals through to quirky, pulsing drones – whether it’s a musical project or a sound design project, Sparkle has something to offer. The MIDI controlled pitch shifting is also a lot of fun with the right combination of source/target audio.
At UK£4.99, Sparkle is hardly likely to lead anyone into bankruptcy court, and for many may well represent a casual, curiosity-based, purchase. That said, it probably is a bit of a niche app. If you’re into cutting edge sound design or glitchy electronic music styles, Sparkle is well worth a look though. It requires some work to find useful combinations of source/target but, when you do, the results can be great.