CloudBounce - CloudBounce 2.0

  • Published
    May 17, 2017
  • Released
    May 2017
  • Share
  • If you suggested the idea of an algorithmic, automated mastering service ten years ago, you would've been laughed out of the studio. Now it's a growing market with an increasing field of competitors. The likes of LANDR, Masterlizer, eMastered, ARIA and CloudBounce all promise fast, high quality services. Considering how new the technology is, the results are pretty good, especially when the process consists of simply paying a few bucks, dragging and dropping your track into a browser and twiddling your thumbs for a few minutes. These services have strengths but they pose little threat to actual mastering engineers. CloudBounce don't see themselves as being in competition with humans in studios. Indeed, they were developed in Abbey Road's own music tech incubator. While they're in the business of mastering music, it's a pretty different service to their human counterparts, offering different priorities, strengths and weaknesses. For instance, you cannot use an algorithmic master to cut vinyl. But you can get a serviceable master back in a few minutes for a fraction of the usual cost. If you're a professional producer with a release schedule, you'd at best use cloud mastering to polish up a demo you wanted to play out. And even then, you might prefer your own DIY mastering skills. This isn't to downplay what CloudBounce offers. The recently launched 2.0 version of their mastering algorithm comfortably covered my own (admittedly rudimentary) masters in these dance music-focused tests. Rather, it indicates that they're most useful for amateurs trying to take their craft to the next level—meaning, the vast majority of musicians and producers. So in that sense, they're not to be sniffed at. And it's certainly healthier for a musician to be thinking of creative questions rather than learning the ins and outs of such a demanding technical task. However, electronic dance music producers are generally more likely to toy with mastering their own music, given their role is closer to that of an engineer than a regular musician. But besides a good ear, good monitoring, a lot of concentration and a light touch, DIY mastering does require multiple, sometimes expensive, plug-ins. If you don't have the prerequisite saturators, compressors, EQs, multi-band compressors, limiters and whatnot, a one-off mastering fee of $4.90 is pretty attractive. Even an annual CloudBounce subscription with unlimited masters costs less than a suite of high quality plug-ins or an actual mastering session. If you do have those plug-ins and rate your DIY masters, the question then is, how does CloudBounce compare to your own mastering abilities? I would never boast about my own basic mastering but it's at least good enough to blend in a mix of professionally mastered material without sounding out of place. To use as an example, I bounced out a Soul Capsule rip-off peaking comfortably around -10db. I set aside the unmastered file for later use on CloudBounce and then applied my usual mastering chain. This consists of saturators, compressors, stereo width tools, some notch EQing and a limiter. Without going into too much detail, I only just touch the needles of the compressors and limiter, having seen professional mastering engineers do much the same. I dragged my unmastered file into CloudBounce and was greeted by a list of genre options. I found myself wondering what differentiated the algorithms of the Tech House, Electro House and EDM settings. There's also Techno, Deep House and Drum & Bass / Bass Music. It would be interesting to know what sort of processing is applied for each of these but the whole point of CloudBounce is that you shouldn't have to bother with such thoughts. To be safe, I chose Deep House. Then you choose from mastering options like Louder, Mid Dip, Warmer, Less Bass, Stereo Width and so on. I chose Warmer, largely because the other options sounded like corrective measures. You don't have to worry too much about these settings. You're subsequently greeted by an MP3 clip of your master wherein you can choose alternate settings and remaster the file to taste. Compared to my basic mastering, the CloudBounce master had noticeably softened the top end. The 808 hats were much softer on the ears. The mix had less superfluous information below 30hz or so and perhaps a touch more energy in the low mids. The overall soundstage seemed more compacted and focused, perhaps due to extra compression. Needless to say, you'll need decent headphones or monitors to discern the fine differences.
    Oddly enough, when I tried the Default setting with no mastering style selected, there was noticeable clipping between the transient of the kick and the body of the bass sound. Instead, I tried the Techno setting and Warmer. The clipping disappeared, so perhaps it's merely a case of needing to pick an appropriate genre. Compared to the Deep House master, the Techno setting had a touch less top end, more bass and more compression. It also had a higher perceived loudness measurement. I'd take either of these masters over my own effort.
    Using a Waves PAZ Analyser, my master was reading around -10db on the RMS scale while the Deep House setting was hitting -7.9db. The Techno master was even louder, hitting -6.1db. This is pretty loud in a context where vinyl is being played. It might be de rigueur on the radio or for commercial productions, but this is much hotter than I've seen respected mastering engineers treat music made for underground dance floors—unless when asked, of course. Digital files I've received from professional mastering sessions have clocked in between around -11db and -8.5 for more aggressive material. The point is that, in my electronic dance music-specific experience, CloudBounce 2.0 can master slightly hot for some contexts. Of course, this is what most people want, especially when you're sending out demos. And it's easy to forget that in this case I'm comparing a service that took a few days and cost hundreds of Euros to a process that took about ten minutes and cost a fiver. In any case, the majority of people using CloudBounce won't be checking the RMS levels of their master. They're just going to listen. And with a decent mix down and the appropriate mastering settings, they're probably going to like what they hear. Unless you've made a strong effort at becoming good at mastering, CloudBounce likely does a better job than you can. At $4.90 per track, it's a low stakes option. Just be sure to toy around with the different algorithm settings before downloading your final master.
RA