paulhazel.com

FourTrack, as its name suggests, is a 4-track recording application for the iPhone (and iPod Touch):

On the whole, the application works well. It’s very simple to set up and use (although I’m not mad on the clumsy navigation scheme). Recording quality is pretty damn good given the limitations of the microphones I used, which were the iPhone’s built-in mic and the one on the supplied headset (which are actually very similar in performance). The ability to transfer audio files onto a computer using wi-fi is very welcome and totally straightforward. To be honest, I haven’t done any serious multitracking myself, but Sonoma claim latency of less than 1mS so there shouldn’t be any problem in that department. You can get more technical detail here.

My first impressions were pretty favourable, then. Unfortunately, I’ve discovered a couple of quite serious problems with the program. These have been unearthed because of the (perhaps slightly unusual) way in which I was using the program. My first serious recordings were done at the SAND 2008 conference where I was trying to record a couple of the presentations to use as podcasts on this very blog.

So what I wanted was just a single track of audio, but lasting something around an hour (and the website claims that record times are “unlimited”). I recorded two separate presentations into two songs. What I found when I got them home and transferred onto my computer was that they would both play up to a certain point, and then iTunes would jump to the next track in the playlist.

The recording of the first presentation was about 46mins long, and it always ‘flipped out’ at exactly the same point, 32mins 5secs. The second recording was around an hour and 10mins, and would again always flip out at exactly the same point, but this time 38mins 1sec. I tried loading the files into Peak Express but it reported that both files were corrupt.

In work the next day I tried loading them into Pro Tools, with the same result. Eventually I got them to load into Sound Forge on a PC, and this is what I saw:

What it looks like is that FourTrack continued recording for the duration of the presentations, but in each case it lost the input signal at some point. However, I’m not sure it’s as simple as that: working on a rough memory usage of 5Mb a minute, the file size should be around 230Mb for a 46min recording. The file size is actually 162Mb, which is right if it stopped at 32mins. This suggests that FourTrack did actually stop recording where the audio signal drops out, but that somehow it logged it as continuing to record, which is presumably why I was getting a corrupted file message… Whatever: just to be sure, I recorded one of my own lectures the next day to see if I could repeat the problem and exactly the same thing happened.

Whilst going through this process, I also found what seems to be another quite serious bug in the program, insofar it seems to have a memory leak. Here is a screenshot from iTunes showing the memory of my iPhone with the two presentation recordings onboard:

And here’s the same thing with those two files deleted:

Now those two files only take up around 350Mb, whereas we can see from the above that there is a difference in the displayed application memory of approximately 1.25Gb! Whether this is actually a ‘memory leak’ I don’t know for sure. But I do know that memory management is one of the issues with Objective C, so I’m taking a semi-educated guess.

In summary, I’d say that FourTrack is basically a good program, but as yet it has some technical issues that need sorting out.

Soul Trapper is a very well-produced application for the iPhone that developers Realtime call an ‘audio adventure’: you experience it as if you were listening to a radio play, but at certain points you have to interact with it, giving it a game-like dimension as well.

The story, set in modern-day Los Angeles, plays out as a cross between the 1940s detective novels of Raymond Chandler and, say, Ghostbusters. This rather unlikely combination actually works rather well and after an initial acclimatization period I truly found myself getting involved with the characters and their Hellish plights. It’s more pulp than Chandler ever was, but the dialogue is littered with Marlowe-esque wisecracks and mannerisms and the locations are classic Chandler: missionary churches, surf-spattered coastlines, Cahuenga Boulevard, horseshoe-boothed bars. It’s a world where tough guys don’t drink their whisky, they inhale it.

There’s not much going on graphically. Each of the 23 chapters is simply represented by a single stark image. You’re listening. In fact, you find yourself listening really hard because many of the cues are quite subtle. In order to support this need for detailed listening the quality of the audio is very high throughout, and you’re best off with a decent set of headphones: many of the tasks would simply be unplayable over the iPhone’s speaker.

There are some great audio set pieces later in the game. Whilst in Hell (!) you play the hero swordfighting with a demon and you have to parry his strokes by listening to which side they’re coming from, and then very quickly parry them using onscreen buttons: very Luke Skywalker. Later, whilst recovering from this ordeal, you have to ‘centre your chakras’ by remixing synth tones in real-time. Brilliant, intuitive, fun.

It isn’t perfect: some of the voice acting is a bit cheesy; certain sections of the dialogue are merely functional; and in places the interaction isn’t all that meaningful or productive. But, overall, Soul Trapper is well worth the admission price and good value-for-money at £3.99.

So why am I reviewing Soul Trapper exactly? Well, here’s a couple of reasons:

Firstly, I am intrigued and fascinated by the idea of telling a story using only audio. In our Internet-driven world the default communications strategy privileges images and, in particular, moving images. It is a relief, therefore, to come across a developer willing to attempt something different.

As Marshall McLuhan has pointed out, media exist on a continuum between hot and cool (where by ‘hot’ he means high resolution, narrow bandwidth, requiring total concentration from the user, total involvement). By focusing on audio to tell their story, Realtime have exploited these characteristics of a hot medium to excellent effect.

Secondly, I am also intrigued and fascinated by the interactive narrative elements. The plot itself is not open to manipulation by the user: about the most you can do is effect the order conversations play out, or the way in which the protagonist moves around the limited maps. However, to make up for this, the story fair motors along, and you’re recompensed by some unusual interactive game-like elements (as mentioned above) that crop up in most chapters.

It really is quite an interesting and cost-effective solution to the problems presented by any type of interactive narrative. I shall be interested to see how Realtime develop these ideas in future releases.

To sum up: an excellent release for the iPhone. Highly entertaining and very interesting.

Bloom is a recent application for the iPhone developed by Brian Eno and Pete Chilvers. The press release describes it as “part instrument, part composition and part artwork”; in practice it’s an ambient music generator that allows the user to input notes via the touch screen. These notes are then a) displayed like ripples on a pond and b) taken up by the programme and variations are generated over time.

I’ve been playing with the thing all week. Despite being very simple to use, it’s very good at what it sets out to do. It’s hypnotic and relaxing, and does actually create convincing ambient music. There is only one sound available, a sort of cross between a piano and a harp, but there are subtly shifting drones that hover in the background as the melodies drift in and out… There are also a set of ‘moods’ that seem to change the scales used by the ‘pieces’:

There’s obviously a lot of very clever stuff going on behind the scenes. Presumably all the sounds are generated in real-time (i.e. no samples) which gives the music a very rich and warm sound. Knowing Eno, I’m guessing it uses an FM synth, probably built in Max/MSP or PD. And although the sequence generator seems to be little more than a delay line at first listening, if left alone the programme will generate endless variations on even the most simple of inputs.

I left it running today for about four hours, and it was still happily evolving when I turned it off. Running the programme this long did highlight one thing: it drained the battery in a couple of hours. Here it is in action:

I love it. It’s not a toy. It’s not a gimmick. Bloom actually turns the iPhone into a viable and meaningful instrument that allows you to produce some very listenable and sonically high-quality ambient music. I found it extremely satisfying to be able to tap out a quick sequence, let that evolve for a while while I went about my business, and then ‘add a new part’ just as I was passing by. Or shake it and start again. Whatever…

As with RjDj, it suggests a completely new type of relationship with both the music and with the technological device, and you find yourself operating somewhere between the seemingly incompatible realms of recorded and improvised musics.

- -

One final point: there has been a lot of comment about Bloom being a rip-off of Electroplankton (which has been available on the Nintendo DS for ages). This is just silly: I don’t particularly want to diss either Electroplankton or the DS, but, as this short video demonstrates, we’re talking completely different kettles of fish:

Released on October 10th, RjDj is one of the most interesting pieces of software that I’ve come across for some time. It’s difficult to describe what it is exactly, so you’d better watch the video (over 9 minutes but worth it, believe me):

I downloaded RjDj this afternoon and begrudgingly dragged my unloved and unused Apple headphone set out of the box—the software only works with this headset at the moment. It all worked perfectly first time, and within five minutes I was tapping, banging, clacking, and, yes, even singing along to “the soundtrack to my life.” Live and interactive: John Cage would have loved it.

Cooking the Sunday dinner became an experimental sound workshop: peeling potatoes, kicking open the flip-top bin, using a knife to create glissandi on the grill rack, whistling, thumping the worktop, running the tap, the clanking of saucepans, all became melded into some futuristic ambient-techno soundscape. Great fun!

At the moment, the number of scenes available is limited (5 only) but the website promises another 18 coming shortly. It could do with a way of exporting your recordings, and of course people posting comments on the RjDj site already want programmable delay times, use of better headsets, access to the individual audio channels, etc., etc.. Like a lot of iPhone applications, it borders on being a gimmick: something interesting and exciting for sure, but we’re not quite sure what to do with it…

BUT: what we’ve got here is an application that is sampling in real time, performing DSP on the input, playing that back and recording it at the same time. On a mobile phone. (In fact, RjDj makes the phrase ‘mobile phone’ suddenly seem redundant, out-of-date.)

Something important is happening here. It seems like one of those tipping-point moments, a paradigm shift. The gestural interface of the iPhone is exploited by RjDj in such a way that it allows not only a new way of making music, but a completely new way of experiencing music where our behaviour generates the events that become both the raw material and the gestures that shape our listening.

In fact, one could envisage a future where we no longer primarily bought music performed by other people. Instead we would buy new ’scenes’ and build up a library of software that would transform the music we listen to and the sounds we experience according to mood, behaviour, whim, or conscious control. All ‘recorded music’ would become permanently fluid, open to improvisation and gestural control.

RjDj costs £1.59.

Well, having wisely invested in an iPhone 3G, I signed up for MobileMe as well. As I already had 2 Macs at home—my ancient G4 Powerbook in the living room and my iMac in my office in the loft—some way of syncing these 3 devices together just seemed too good to pass up.

A couple of months in and things don’t look so good. There has been the well-documented climb-down over Apple’s misrepresentation of the service, the loss of emails and subscribers’ general unhappiness with the service (with Apple giving away another two months free usage in mid-August). So: is MobileMe any good or not?

The bad news: there’s just no avoiding the fact that, no, it isn’t very good. Firstly, it’s painfully slow. Secondly, the buttons don’t work properly: last night I looked at an email in one of my folders and when I tried to go back to my inbox all the navigation buttons were dead. And there’s the widely reported problem with the greyed out buttons when you open a new email: Apple do suggest a kludge for this but really they need to get it working properly in the first place. Thirdly, loading sets of images into the Gallery can be painful: they suggest you can’t load more than 1Gb at once—which is fair enough—but last night I left it loading only 23Mb and it still wasn’t finished by this morning! Truly pitiful.

[There we go, a perfect example. I was going to write something about not being able to create groups on the contacts page: better check I haven't overlooked a big 'Groups' button methinks. Go to MobileMe (already open on the Mail page) and click Contacts. Wait about two minutes for them to appear!]

Other gripes: no, you definitely can’t create groups of contacts. It handles images attached to emails very poorly. And no matter what service you use, it’s always losing contact with the server. These types of message come up all the time:

Good points? It’s a great idea. It looks good (compared to Gmail for example, which is butt-ugly). And, so far, it’s cheap.

Am I going to carry on paying for it? I would really like to, simply because of the functionality it allows me. But the biggest problem with MobileMe is that it is really, really, flaky. Unacceptably flaky, because unfortunately services like this need to be rock solid (which is something Google do very well).

Heaven only what Apple think they’re doing with MobileMe. They need to really sort the whole thing out once and for all: as it stands it’s a bit of an embarrassment. From a business point of view there is the short-term loss of revenue, but, in the long term, it will have serious implications for Apple’s standing as a web service provider and it will go on to harm iPhone sales as well (insofar as the compelling USP of the iPhone is its function as a web platform).

One can only hope that the relative lack of news on this front is a sign that, behind the scenes, they’re working frantically on a major upgrade to the service.

Back in January I had a good look at the iPhone and Jefferson Han’s work on multi-touch interfaces (here and here). And for those of you looking to gain some kind of contextualization on this fascinating and highly topical area of interface design I’ve recently come across Bill Buxton’s historical overview.

I casually mention these only by way of introduction to Microsoft’s (ahem) “new paradigm in computing”, the Surface computer. Here’s one of Microsoft’s own promotional videos:

For something a little more illuminating, and that briefly includes schematics showing the innards of the Surface:

And for the inner geek, here’s a full 18-minute test-drive of the thing:

Yes, it’s pretty impressive (although the thought of that bog-standard PC running Windows Vista hiding inside is a bit off-putting). Some thoughts:

1. It’s not clear how the security issues will work. I mean, have you ever transferred data from one device to another without generating security prompts? As these will be public devices it seems inconceivable that security will not be a huge issue, and yet not once do we see anyone even inputting a PIN number in any of the videos. As if!
2. Will all manufactured objects become ‘tagged’ in the near future to allow interfacing with surface computers?
3. If so, will there develop a universal tagging language that will be understood by all “surface-compatible” products?
4. Can we predict a new job description: Surface Designer?
5. I do think there is a certain inevitability about this type of product.
6. The ability of the Surface to act as a ‘docking station’ for mobile devices calls to mind one of McLuhan’s Laws of Media: if you push a technology to an extreme it flips over into it’s opposite. In other words, as mobile devices have gotten smaller and yet more powerful, the tendency for miniaturization flips over into single large device that many of them can simultaneously attach to like a Mother Ship.
7. Doesn’t the Microsoft Surface remind you of those black glass-topped gaming tables you used to find in pubs? Space Invaders, anyone?

Of course there are those who quite rightly question Microsoft’s presumptuous and overblown claims for their product: British multi-touch interface designer Andrew Fentem has a reasonable and well-argued critique of both Microsoft and Jefferson Han here. Fentem’s own Spaceman Technologies website is well worth checking out by multi-touch aficionados.

Finally, irresistibly, if only to puncture the corporate pomposity of Microsoft:

In a recent Alertbox column Jakob Nielsen looked at the way computer interfacing was represented in films and on TV. I’m not going to reiterate the content of his fine article here, but it did get me thinking about our preconceptions and expectations of the interfaces we use (or wish we used).

There are a whole series of interrelated problems, all based around the fundamental problem of too much information: too much to fit on screen, too much to manipulate, too much to think about. So - the theory goes - there must be some better way to manage this than the current QWERTY keyboard and mouse paradigm that dominates mainstream computing…

How about this:

Impressive huh? Or is it all smoke and mirrors set to a cool soundtrack? Looking at it closely, we can see that it’s simply a touch screen, but one that has multiple simultaneous sensing points. What’s more, there’s a shot of two people working on the machine together, so one can only guess that it recognizes somewhere around 16-20 sensing points at any one time. Apart from the nice-looking but probably-not-all-that-useful lava lamp blobs, examples of programs being used in the video include image manipulation software, word games, text entry, a Missile Command-type game, a real-time music program such as Max/MSP, and some kind of activity using molecular objects. There’s also evidence of a working toolbar. So, yes, impressive actually.

Evidence, then, of a functioning example of an innovative computer interface, albeit in a university research lab. Is it realistic to expect interfaces like this to appear in the near future? Is it Tomorrow’s World or Star Trek? Well:

Sorry, I’m not trying to jump on the “isn’t the iPhone fabulous” bandwagon but it’s interface - which you’ll note features a touch screen that recognizes at least two simultaneous sensing points - is certainly pretty exciting. And it’s on a hand-held device. And it’s coming to a shop near you, very, very soon.

Beam me up, Scotty!