ALive! - features - Dave Rossum

Music

One of the most anticipated upgrade for the music community is support for the SF2.1 format (SoundFont version 2.1). Can you tell us the estimated date for this update and whether we have to pay for it?

There’s a joke in the design community that the product engineering is all done, what’s left is "simply a matter of software." Of course, software has become so complex that the hard part of just about everything is "simply a matter of software" and that’s certainly true here. It turns out that in order to efficiently support SF2.1, there’s a whole new driver architecture being put in place. The engineers involved would say that it will be in customer’s hands in the 3rd quarter of 1999; the more experienced managers would give a more conservative estimate of merely the second half of 1999. So I think you can count on it in ’99, and of course we want to bring it out as soon as we can. As far as paying for it, that’s what we call a "business decision" which means we engineers don’t want to deal with it. So I can’t give you an answer either way on that.

What happens after SF2.1 is implemented, what will new versions of the SF format be like? A lot of musicians would be very interested to know about Creative/Emu's plans, as well as to participate in the design.

In 1998, Creative participated in an industry-wide initiative to standardize on a file format for downloadable sounds. The end result is a new standard called DLS-2 which was ratified by the MIDI Manufacturers Association. The DLS-2 standard was also adopted by the ISO committee as part of a new international standard known as MPEG-4. Microsoft has also announced support for DLS-2 as part of their new DirectMusic interactive music API.

We took the best part of SoundFont, the creative expression of the E-mu synthesizer, and the best part of the DLS standard, a flexible file format for describing synthesis, and we combined them into a single, powerful standard. The exciting thing for musicians is that they can now write all their content to a single standard that will have broad support in the industry. And even though there are few tools or content available for DLS-2 today, you can develop your content using SoundFont, and know that it will convert easily to DLS-2 in the future.

Internally, we’ve called DLS-2 "SoundFont 3", but whether that will be an official name or not I can’t say. We’re certainly happy that the standard has been adopted so broadly, and also that the key portions of our original SoundFont technology that allow for further innovation remain within the standard.

It's not very likely that Creative will release the EMU10K1 programming specs to the public domain. Will Creative consider releasing it to other companies under a NDA (Non-disclosure Agreement) to develop add-ons for the SBLive in terms of effects, alternative synth-engines etc?

Of course, any such arrangement would need to take into account a great many factors, but I would say that Creative would consider working with other companies in this way. Presently, we are engaging third parties to provide support for the E-mu Audio Production Studio under BeOS, MacOS and Solaris, so clearly it can be done. But supporting another company is a burden - it means that you’ve got to provide external support for all your development tools, and if you want their software to work with ours you need to write down standards and establish tests, etc. So it isn’t something we’d enter lightly. However, if the benefit clearly exceeds the costs, I’m sure we’d do it.

One other feature that should be simple to implement is (software) support for a second wave-output. The SB Live! already have the connector (Rear speakers). Why can't we use this output for a total of four mono channels of output (or two stereo channels)? Does Creative have any plan to add on this feature?

We’ve received this request from a number of users, and have looked into how it might be accomplished. The bottleneck is inside the architecture of DirectSound itself. DirectSound currently can only address one audio card with one wave out device at a time. What’s needed is support for a "virtualized" wave device. We will be working with Microsoft in the future to try to remedy this, if the DirectSound architecture can be adapted in a compatible manner.

contents


Back To Top	h t t p : // a l i v e . s i n g n e t . c o m . s g	Copyright © 1998-2003 ALive! and its writers.