New Toys

It seems like it has been hardly any time at all, but my trusty Zoom H4n was simply no longer cutting the mustard for my portable recording wants and needs. So the time came to upgrade! (...Before the end of financial year.)

So after much deliberation, I chose a Sony PCM-D50.

These have actually been on the market for a while, and don't appear to be in production anymore (?), which is a crying shame because from the looking around I've done this is still the best handheld recorder on the market. When I bought the Zoom it was both a great deal cheaper and had XLR inputs, which tipped the boat in its favour at the time, but aside from the occasional planned field recording trip the XLR inputs didn't get much use and my growing dissatisfaction with it prompted me to go searching for an upgrade.

I'm absolutely in love with it. Having both the monitor and record levels on such smooth-turning dials was the first thing I noticed and appreciated. The buttons as well are fantastic and near-silent - one thing that was always very annoying with the Zoom was how every button press was massively amplified, and how incredibly sensitive it was to any sort of handling noise. The start up time is a smidge faster than the Zoom, and the menus themselves nicer to navigate given the button layout. It also feels sturdier, despite weighing roughly the same as the Zoom with the batteries in. The benefits of an aluminium casing instead of plastic.

In terms of size it's about the same, although half the width, which saves space. It's about the same size as a Nintendo 3DS XL actually, so if you're in need of a case, they're cheap and easy to find. It also has 4GB of in-built memory which the Zoom lacks, although the tradeoff for this is being stuck with the Sony Pro Duo memory cards which are a lot pricier than the standard SDs.

The real value, of course, is in how much better the Sony's microphones sound compared to the Zoom's. The first and most obvious improvement being the significantly lower noise floor, which alone is worth the upgrade.  The microphone quality is really great bang for buck here, you would have to pay a heck of a lot more to get something better.

This is definitely supplanting the Zoom H4n in my bag. I'm hoping to at some point in the future to purchase another portable recorder with at least four XLR inputs for the more heavy-duty planned field recording sessions, at which point I'll retire the Zoom for good. Super excited to go do some spur-of-the-moment field recording with this thing!

Pitch Correction Shenanigans

Since I've specialised in sound design more so than music, I've actually not needed to do much in the way of pitch correction in the past, since most of the time of any music mixing or composition I do tackle is usually of the synth and midi-sequenced variety, so tuning concerns and pitch problems very rarely come up.  The occasional sound effect has been nudged into key but that's usually been something fairly simple and easily accomplished by ear.

However, recently I wanted to sample a steel tongue drum, with one of the goals to make a VSTi of it.

(This magnificent bastard)

It's a lovely instrument, a slightly lavish Christmas present to myself last year. It's fun and easy to play, and quite meditative, but as you might expect, its pitch varies dramatically according to ambient temperature, by up to as much as a semi-tone sometimes.

This is okay so when you're just jamming for fun solo, but for the purposes of a VSTi some pitch correction is necessary - especially if you ever want to throw it into a composition with any other instruments. Which is where my adventure into the world of pitch-correcting plugins has come in.

Here's a small snippet of the session recording, to get a feel for the sort of sound a steel tongue drum produces:


For the steel tongue drum, the vibrato and overtones from the resonance - which is a huge part of the drum's sound - rule out pretty much anything other than Melodyne, which after some fairly exhaustive research appears to be the only polyphonic-capable pitch correction plugin floating around. It's tough to swallow, though - $300 is a big investment for a plugin, especially one that you would have need of infrequently. It would definitely be a must-have for anyone more heavily involved in recording and mixing music though.

This is probably to the surprise of absolutely nobody in audio - Melodyne is fairly well known as the gold-standard, standing head and shoulders over Auto-Tune. But in looking for alternatives, I was pleasantly surprised by the quality of WavesTune! It can't handle polyphonic, but based on my admittedly limited experimentation with it, with vocals and a lot of other instruments the result is even smoother and more natural-sounding than Melodyne!

VariAudio in Nuendo (and Cubase too, presumably) wasn't too shabby either, although its analysis was somewhat spotty and required a great deal of intervention. Still, considering it comes with the program, I was also pleasantly surprised by it. By controlling how much you want to flatten the pitch on a segment-by-segment basis it was very easy to preserve the vibrato, which is something a lot of the pitch-correction plugins stomp all over.

I gave Audition's pitch correction tools a whirl too, for curiosity's sake. The Auto Pitch Corrector actually wasn't too shabby with basic tasks, but like any automatic process is prone to blowing up in your face horribly and spazzed out at the slightest hint of vibrato.  The manual pitch correction is a bit better and more versatile, but nice as the interface controls were, you would need to either have perfect pitch or pair it with some hardcore analysis, and its tolerance before it starts sounding horribly processed is extremely narrow. Still, it's a feature that could eventually catch up to Steinberg's VariAudio... if Adobe spends any time on it. The pitch manipulation tools in Audition took a massive step back after 3.0 and have never quite recovered.

For now, though, Melodyne and WavesTune get my votes of confidence! Urgh, now to decide how much use I'll get out of Melodyne in the future... $300 isn't horrific for a plugin of that calibre, but when the money is tight... maybe I should have bought it instead of the drum!

Recording Workshop in Melbourne

Had the opportunity to head down to Melbourne for a sound recording workshop a week or so ago held by Stephan Schutze.  Stephan hired out a studio in Docklands for the day to record some things that aren't easy to record normally, like flaming torches and arrows and dry ice-on-metal, etc.

This is a quick picture I snapped of the space - impressively large!  Once the doors were closed it was wonderfully insulated from outside noise, but of course that massive expanse of concrete made for some incredible reflections!  I'll be curious to see how many of the final recordings that compromise.  Although the reflections sounded fairly epic in themselves, I kind of wanted to record an impulse response of it!

A lot of the workshop wound up covering familiar ground, but it was definitely a lot of fun just listening to someone be so enthusiastic about sound recording and share some of his recording anecdotes.  I think if the same workshop turned up three or four years ago it might have saved me a lot of lessons learnt the hard way!

There was far too much material covered for me to remark on at length, but the biggest points of interest I took away from it were:

1. Dry ice on metal is a very cool combination, and I have to find a way to play with this myself sometime.

2. He showed off a neat hack using a pair of lapel mics and a drilled-out manikin head as a cheap binaural microphone solution.  I'm extremely curious as to what the recordings of that one will sound like and how effective it is versus an actual binaural microphone.

3. Safety first!  Made me think twice about smashing up that vintage TV that I've been lugging around since University days.  Vaccuum tubes involve some fairly toxic chemicals, who knew!  Also when it comes to miking up a car, taking not only moving parts but heat into consideration, as well as wind noise not just from the car's movement but from within the motor itself.

4. For field recording, specialist groups will get you further than more generic approaches.  An interesting angle I hadn't considered before, but obvious in retrospect!

5. Splitting signals, having one channel running hot and the other at a much lower volume for safety.  This is something I normally try to do when recording with multiple microphones, but the idea of multiplying out the signal was an interesting one I'll have to explore at some point.

Perhaps I'll have more food for thought once the videos and recordings come out, but overall, it was a good experience (and also great to catch up with some of the other audio people present).  It's pretty rare to find any sort of audio workshop happening in Australia - plenty for music production but precious little dedicated to other audio disciplines!  So major props to Stephan for going out on a limb and organising it.

Recording all the fun and weird things

Been feeling a little bit scatterbrained lately!  That's my only excuse for not updating this blog for so long.  It's a poor excuse, I know, and likely to be repeated.

Part of my professional development goals at Sega is to record something at least once a week, and it's been really awesome so far. Back when I was with KMMGames I did make an effort to record when I could, but it was largely a mixture of opportunism and need.  The audio room was in the less-than-ideal position of being wedged between train tracks, a construction site, and the QA room, which meant I was limited to either waiting for a quiet moment or only recording very loud sounds.

No such issues at Sega!  The rooms are very well insulated which makes it extremely convenient to set up a quick recording session, even if only a single mic for five or ten minutes.  It's paid off very well so far, with a rapidly growing library of sounds, many of which have already been useful.

Here's a picture of just last week's recording props:

A bit of soft discarded plastic from a popped plastic ball, a wind up toy fan, and a bunch dried clay bits that were leftover from a sculpture I made when I was in an arts-and-crafts kind of mood after cutting through Lincraft one day (incidentally, the source of all my arts-and-crafts moods).  They sounded kind of interesting when I was gathering them up to throw them out, so instead stored them in a bag and put them with the rest of the junk I packrat for potential recording.

Useful discovery - bits of solid clay actually make very good rattling bone sounds.  With a little bit of processing and layering I think you could make a pretty good sound set for a skeleton out of these.

One advantage to recording your own sounds where possible which has come up with this is not only how many more variations you can record - because however good the libraries are, it's rare you can get more than a couple of decent matching variations from the same sound set - but also the potential to play some of them like instruments.  To drag the sounds out, or make them short or punchy, soft or violent, or use them on or with a wide variety of surfaces.  One thing I've been learning to do better is to not just record the obvious sound a prop makes, but to wring every single possible sound out of it.

Just some of the things we've recorded in-studio lately have been rubber and leather gloves, squeaky toys, paper and plastic bags, balloons, silly putty, soft furry slippers, a vampire cape, and a jaw harp.

It's been a lot of fun, too!  If I were to come up with an art analogy, recording sound is like working directly with paint and clay, as opposed to working in photoshop or 3DSMax.  You can get really great results either way, but there's something very satisfying about dealing with the raw materials, even if it's often more clumsy and uncontrollable.

Succeeding Quickly VS Failing Quickly

An idea I was introduced to recently was the concept of 'Succeeding quickly' as opposed to the more popularly spruiked 'Failing quickly'.

'Failing quickly' is one way to explain a popular method in the creative side of the sound design process - for those troublesome sounds which you might not already have a very clear idea for.  It can also be described as 'throwing things at the wall to see what sticks'.  It's like rapid-fire brainstorming in sound design terms.

'Succeeding quickly', on the other hand, is more a case of working on a series of sound effects quickly, dumping the ones you get stuck on in favour of moving on to new ones.  The idea is that rather than spend a day on a sound you just can't seem to nail, you move on to the next sound, and then the next sound, some of which you will probably get first go.

I'd always thought the former method the best approach.  After chewing on the idea for a while though, I think there's a great deal of merit to the latter approach.  Without putting a name to it, this is what I will naturally do on any project which I have free range on.  Certainly this was my approach on Happy Feet 2, where audio milestones were few and non-descriptive.

The advantage of this was that time didn't get wasted troublesome sounds.  I could move on, take care of the low-hanging fruit, and often when I revisited the sound effect the next day I would have much better luck.

Some of it was simply due to being creatively refreshed - lots of artists or writers will tell you how some days they can spend hours on a piece and then the next morning do the same work in minutes (a phenomenon I've noticed that diminishes with experience, but can still occur to the best of us.)  Another reason might be because in the intervening work I've found some inspiration or fallen it into a creative groove.  Either way, at the end of the day, the result is you have more bang to show for your buck (literally, as the case may sometimes be) and a lot less frustration.

Of course, this is more useful is the earlier stages of a project, when there a long list of sound assets that need to be made.  Later on you don't have the luxury of moving on to another sound, because the sound that is giving you trouble needs to be designed yesterday.  That's when the 'Failing quickly' approach can come in handy again.

Anyhow, this is one of those things which is fairly obvious when you stop to think about it, but I'd never actually stopped to think about it before, and putting it down in words helps in figuring out the why and how.  And with that, you can decide which approach best fits your current needs.

Happy Feet

Happy Feet 2 (the videogame) has been out for a while now. It was the first project I ever tackled completely solo (from a sound design point of view of course).  For a budget game in the openly reviled sector of kid's licensed games, it has reviewed decently, although considering the last time I checked most of those reviews seemed to be in foreign languages it's hard to tell for sure.

I had planned to make a big blog post about the challenges involved with mixing a rhythm game with an adventure game, or the rationale behind some of the design decisions, some of the localisation challenges, or even a post-mortem on all the things I'd wished I'd done differently with fresh ears and a little bit more distance and wisdom (and more than a few facepalming moments). Now that I've reached this point, however, this picture is all that comes to mind:

I hope I never have to use maths to sync sounds to an animation ever again.

Risk in Sound Design

Caught up with one of my old colleagues a week ago and we inevitably got to talking shop.  The sound design scene in Australia feels very small sometimes so any opportunity to exchange thoughts and tips is always eagerly seized upon.

One thing we were talking about which gave me some good food for thought was the role of risk in creating stand-out sound design.

The ultimate challenge for a sound designer is not merely being a good sound designer, but a great sound designer.  And the thing that separates the two is more often that not the one who takes the risk.  There is the way that you know will sound 'right' - the sound and mix that you are certain will work, because it's tried and true and audiences have been trained to accept it.  I think there's a lot of merit to this approach, because sometimes the audio isn't there to be the star, and all you need is something safe and functional.  There's a common saying in sound design, which is 'if you've done your work right nobody will notice it'.  Also known as 'bad drummer syndrome' - you only notice the drummer in a band if they're bad.

However, if this is all you do, you run the risk of falling into being a 'good' sound designer instead of perhaps a 'great' one.  This is where the risk comes into play.  It's the work which does something different that is most often remembered.  The one which chose a novel approach to the mix, which maybe chose unusual sound effects for mundane objects, which pushed the envelope of originality.  Creativity and originality are traits which are often lauded in the sound design profession, but I think it's far better to describe these traits as 'risk-taking'.

Thinking about it in regards to game audio, I think this can manifest most strongly in implementation.  When technical factors become involved, it's very easy to choose the easier solution, the one you know will be implemented correctly and will work and sound right.  But in doing so you often throw away the opportunity to go one step further and do something really impressive.  It's a trap I've often fallen into in the past, especially on projects when I've been frustrated with a lack of code support and so will reach for the basics first and only then try and reach for more ambitious implementation.  While this is a good approach in theory - you have the safety net of at least being able to do the bare necessities - in practice it's more likely the most basic option is the one you will eventually end up with.  It's a far better bargaining position if you reach for the most ambitious option first, and if necessary then work your way down.

Of course, the biggest issue with taking risks, is that sometimes you will fail.  You have to be prepared for that, too.  But just like investment, the biggest risks can also pay the biggest dividends when they work out.

Anyhow, there's a bit of what's been occupying the stream-of-consciousness the past week.