 |
|
Someone you want us to interview? Let us know at news@3dsoundsurge.com |
|
|
|
|
|
| Sound News | Press
Releases | Archives | Week In
Review | Editorials
| Articles | |
|
|
|
||||
|
|||||
|
|
The Sensaura Site Visit
The trip to Sensaura Every now and then, it’s nice to leave the sumptuous opulence of 3D SoundSurge Towers (Yeah, right) and go and visit the tireless individuals that slave away on the sound technology that we all love. So, packing my bags, I was off to London to visit Sensaura. All about Sensaura Now Sensaura are a little bit different to the rest of the soundcard world. They don’t make soundcards. They don’t make chips. They don’t do anything apart from research, develop, tweak and tune 3D sound algorithms. And play Unreal. But that’s work for them. Lucky blighters. Sensaura has grown out of CRL, the research and development arm of what was once the EMI/HMV giant. Much of Sensaura’s work is based on work done on recording classical music. The idea was to record music by having an artificial head with microphones where the ears are. This gives recorded ‘3d-sound’. But it’s like Dolby Surround, it’s only good for linear entertainment. So as computers developed, they wanted to apply this technique to computer sound. But while they were taking measurements for sound sources from different points around the head, they found that the ‘ears’ on the head were not accurate enough. So in the spirit of true pioneers, they built their own. It turned out that there were more than a few things that didn’t quite match up from the model to the real ear. So many refinements later, they now have a ‘digital ear’, which, when graphed, gives very accurate frequency readings for sound sources, much smoother than other models. From here, the basic sound model has evolved – HRTF’s for headphones, and HRTF’s with crosstalk cancellations for two speakers. And this is what people using Sensaura technology at the moment have. Sensaura has the technology Sensaura are justifiably proud of their achievements so far. The ‘digital ear’ modelling is far too complex to go in to here, but suffice it to say (and I’ll give the full gen on the newsgroups if anyone really wants to hear) that the ear models and crosstalk cancellations graph significantly smoother than ‘the competitions’. As you would expect, a company with a background of EMI isn’t just prepared to rest on it’s laurels, oh no siree. Sensaura have a few new ideas up their sleeve that are going to make the competition kick themselves for not thinking of it first. I put it to Sensaura that the sound card companies had moved ahead with environmental modelling while audio positioning still needed polish. To paraphrase the reply, Creative knew that they were somewhat behind the market – so they had to offer something new. And so because EAX is open, we’re implementing it (see later). But we have other things that we wanted to work on as well. So here’s the first – MacroFX. DirectSound3D uses volume as a distance cue. And for distances greater than 1 metre, that’s fine. And since the standard DS3D minimum distance is 1 metre anyway, most people don’t know there’s something that’s not as good as it should be. So Sensaura have invented MacroFX – a more accurate way of modelling close sounds, ie between your head and less that 1 metre. It is a combination of two things – a more accurate volume model that works better for near objects, and a modified ITD system. As we should all know, 3D sound uses the Interaural Time Delay – the difference it takes for a sound source to reach both ears – to help place sounds. Lets take a sound that’s two meters away, and at a compass angle of 85 degrees. The difference between the two will be slight. But when the sound is practically in your right ear, the difference will be huge. Nobody else’s sound model takes this into account. MacroFX does. And it does this with no extra instructions. Provided the programmer doesn’t set the minimum distance to 1 metre, MacroFX will work perfectly. The usual buzzing bees demo sounded good, especially for two speakers. I could tell the difference between on and off. But the best display came from Ye Game of Major Graphics and Sound – Unreal. A Skaarj was quickly summoned, and just as quickly dispatched with the Biorifle, my personal favourite. And I waited for the flies. Credit has to be given to Epic Megagames. There are at least 3 sound sources for the flies. All I can say is it sounded like I was in a swarm. They were everywhere. Even with my eyes closed it sounded so good I felt I could touch them. And on only two speakers. Yet this is on beta drivers that won’t be ready for a public demo for at least another month. You get the feeling that Sensaura really know what they’re doing. And four speakers too! Now, it has to be said that multiple speakers offer the most convincing 3d sound experience. The front/rear positioning is excellent. But as many people have found, it can seem like there is a sound ‘hole’ at either side. How bad this hole is varies on each chipset, but it can be said to be an acknowledged factor with them all. Sensaura’s answer is to use their 3d sound expertise to ‘fill-in’ the space in the configuration. This is MultiDrive. By using HRTF technology and some other ideas that Sensaura are keeping schtum about, MultiDrive aims to bring a more cohesive sound field to four speakers solutions. Again, the proof of the pudding is in the eating. Four Yamaha YST-M100’s, and a modified ESS Maestro-2E card were our bench system. MultiDrive is still being played with, and was only demonstrated to Videologic about a week before I went, but nevertheless, MultiDrive does what it sets out to do. Most people use mismatched speakers for 4-speaker audio (I know I do) and this doesn’t help. Quite a few people use the decidedly average 4-point surround speakers, and these just ain’t got the welly. So the rig I heard it on was definitely set up for a good sound. But even so, MultiDrive works. REM’s ‘Man on the Moon’ was set rotating in a circle, and I tried to pick holes. Not easy. A lot of four speaker rotating demos sound like ‘Here’s the front’s, and now the backs’. Some are better than others. MultiDrive just had this sound rotating in an eminently convincing circle. The side-to-side positioning isn’t completely perfect, but the difference between, say an SB Live!, is very noticeable. The difference between an MX300 and MultiDrive is less pronounced, but MultiDrive is still noticably better. What is even more convincing is when the sound moves away from the 90 or 180 degree and starts to move towards the speaker itself. This is superb. I was actually able to completely fool myself that the sound was coming from the suite door and not a Yamaha speaker. Spanking stuff. Bear in mind that this is a test rig, designed to give absolutely the best output. It will be interesting to see how well it performs in the home. The final ‘new’ technology had only been in development for three days when I got there. ZoomFX is again a refinement of the way 3d sound currently works. Imagine, if you will, a train. As in a big train, you know, engine and 8 carriages. Now model this in a game. You have your sound source. And the train zooms past the player. Hang on a second. 1 sound source? For a train? ZoomFX aims to correct this by ‘widening’ large objects as they ‘zoom’ past you, hence the name. At the moment, it’s really just R&D, but Sensaura hope to have it as a viable technology in not too many months time. The early ZoomFX work sounds right, but has some nasty phase effects in it. Unfortunately, ZoomFX will require extra work on the part of the programmer, and it will require up to 3 3d streams as the object gets closer. EAX Implementation Finally, we have Sensaura’s EAX implementation. This is nearly finished, and it sounds great. The team had just gone out and bought a truckload of software to test it with, so it wasn’t quite there, but I had a good play with Man in Field, and I have to say that their reverb algorithms are very good. It’s difficult to tell if there as good as Creative’s, but I think I can say in all honesty when you’re running out of ammo in Half-Life, you won’t really care. As for processor occupancy, well yes, this implementation is basically software. Despite the fact that the some of the chips they have hardware reverb, EAX is run on software lines to make it easily portable to any chip. Wrapping up Sensaura have only recently become an independent company. They have been a part of the massive CRL behemoth for many years. Given this background, you would expect to find a slightly conservative attitude to the company. Not so. Sensaura are unlike Creative or Aureal because they have very few market constraints. They have their deals in place, and so can basically work on pure research. Occasionally the market dictates the research that they do – EAX – but the emphasis is on constantly improving PC audio technology regardless of the hardware that it runs on. Consequently, the atmosphere is very laid back. Every member of the team has his or her unique speciality, and there’s a harmony to the work ethic that most companies could only dream of. As things moved towards a close, I asked them a couple of questions that I felt hadn’t been touched upon. 3DSS: Are you members of the IASIG? What’s it like? How much of a contribution do you make? Sensaura: Yes, we are. It’s basically a lot of like minded people trying to work together. It’s usually quite a good atmosphere – there’s no point arguing the toss in a "Mine’s better than your’s" style. However, there was one meeting – just after the Aureal lawsuit to Creative – and it was like us, Creative, Cirrus, Qsound and the like sat together, and Aureal at the other end of the table. It was quite funny. We get asked lot’s of questions about Aureal. They’re OK, and they’ve got a great marketing team. 3DSS: Of all the chips that you implement your algorithms on, which offers the least hardware acceleration? Sensaura: Hmm. It would probably be the Yamaha chipset for the 192 range. It’s doing an awful lot of MIDI work. 3DSS: On a similar vein, if E-Mu released a SDK for the 10K1, what would you implement? Given all that power from scratch? Sensaura: Good question. Obviously we’d implement everything we do now in software in hardware. I don’t know, actually. All the new ideas in hardware. That’s a really good question. We’ll have to get back to you on that one. Conclusions And with that, it was back to the real world and a late train back to London. Sensaura really know what they’re talking about, and have drive that you don’t tend to get from every company. There’s a dynamic air from the minute you walk into the research suite, and a genuine concern for tight and efficient code that doesn’t sap resources. I’ve been impressed by their standard 2-speaker algorithms (See SonicStorm Pro review, coming up soon) and MultiDrive is just that noticeably bit better than what is out there at the moment. All this, and without the constant marketing hyperbole. And there you have it. Sensaura. A little company taking on the big boys. And just to whet your appetites, you want to see what they’ve got planned for Dolby Digital. If you have any comments, questions or see any errors in
the article contact Simon@3dsoundsurge.com.
For information from Sensaura you can contact Mike
Clarke.
Related Information Our Sensaura Technology article Read our Gamer's Guide to DS3D
March 12, 1999 |
Articles A Gamer's Guide to Sensaura 3D Audio A Gamer's Guide to 3D Sound and Reverb APIs A Gamer's Guide to 3D Sound and Reverb Engines A Gamer's Guide to DS3D and A3D 1.x Virtual Surround Sound Technologies The Advantages of the Center and LFE channels Stereo to 4 or 5.1 Expansion Technologies |
|
|
|
All content, design and work is ©
2001 - 3D Sound Surge Please respect the copyrights of the articles and writers herein. All copyrights are enforced by 3DSS.