 |
3DsoundSurge Special Feature
Aureal vs Qsound - The Great
Debate
Qsound:
Scott Willing
Aureal: Toni
Shneider
Date Started: June
22, 1998 Date Ended: July 8, 1998
Scott Points Out He Isn't Insulting Aureal
Toni has suggested that I've
been insulting and that I'm spreading misinformation. Pretty serious charges. Let's see
here...
In response to specific statements from Toni, I've challenged some Aureal marketing
strategies and claims.
I'm the guy who yells at the TV when some commercial tries to imply that my life will be a
bed of roses if I buy some product. (Jooky syndrome.) It doesn't mean that I think the
product is bad compared to competing alternatives.
It only means that I object to the way it's marketed, and I reserve
my right of free speech to call marketing claims and inferences
into public examination.
I've been very careful to identify my opinions as such and invite correction in case I'm
wrong on any point of fact. Toni has been free to express his company's position, so we
stand on equal footing here.
I'm not rabidly frothing "A3D sucks!" and for anyone who cares to look, you'll
in fact find that I have respectfully recognized Aureal's technical achievements in print
more than once in this debate alone. I've also made numerous attempts to encourage the
reader to make up their own mind, which is what free speech is supposed to be about.
"Weaving insults?"
I didn't set out on an anti-Aureal campaign. This whole thing started when I copied
3DSoundSurge and 3DAI on an edition of QSound's developer newsletter. This was for their
background information as 3D audio journalists--more or less an FYI and I told them so. I
didn't ask for it to be posted.
Subsequently I was asked for permission to quote a few sections of my newsletter on 3DSS,
and I granted it.
Aureal chose to turn this into a debate. Obviously I'm not the sort to shy away from an
opportunity to bring forth issues for public scrutiny.
Aureal has taken particular exception to my statement that the developers I've spoken to
are overwhelmingly supporting general DS3D in current and future 3D audio titles. This is
true, and remains true. I have the notes and email documentation to back that up.
Why is this such a scary statement?
Perhaps there has been a misunderstanding from the start, so just in case, let me put this
down for the record: I never made any claim, nor intended to imply, that the widespread
support of DS3D excluded the possibility that these same developers might also support
other features of Aureal hardware via the A3D API, or indeed of anybody else's hardware,
released or otherwise, by whatever means. OK?
Such a misconception might explain the vigor of Aureal's attack on a simple statement of
fact. I'm not "pitting" DS3D against A3D; the two are not mutually exclusive!
To defend this evidently dangerous little statement of mine, I have been called upon to
list titles that will support general DS3D. I have refused to talk about specific projects
because it puts individual developers in the middle of this debate. I suspect they'd
rather just get on with their work.
All I did was generalize to some e-reporters about an emerging trend for their background
reference. I still say it's their job to check this directly with developers as a service
to their readers--that's what good journalists do.
If my refusal to quote specific examples appears to damage my "case" in the
short run, or if it leaves me open to accusations of "copping out" I'm not
going to lose any sleep over it. The truth will become evident over time, so no worries
here.
Aureal's reaction to this report is confusing. On one hand Toni is saying that Aureal
welcomes DS3D titles because they support A3D hardware. So far so good; that makes sense.
On the other hand my report seems to have caused great discomfort. If the former is true,
why the latter? Is it Aureal's position that in fact the majority of 3D titles will only
support Aureal algorithms? Nowhere have they made this statement, so what's the message
and/or problem here?
Now to the issue of discouraging support of alternate hardware.
Toni argues that what I heard from Aureal staff was a recommendation for a host emulation
strategy that had nothing to do with competing hardware. In fairness I must concede that
this is one possible interpretation of the advice I heard.
I will readily admit that what constitutes "discouragement" or
"encouragement" is potentially quite subjective. Two people could look at the
same set of facts and draw different conclusions.
All I can recommend is that the reader consider the overall actions of a company and ask
themselves if the mantle of support for open standards and free competition fits or not.
I asked about A3D 2.0 support for competitive hardware, and got this response:
"A game written to the A3D 2.0 API will run automatically on those platforms with no
additional work for the developer."
This doesn't give anyone enough information to draw a conclusion, because the critical
term "run" is undefined. This could mean:
a) Aureal algorithms (A2D?)
run on the host, and the competing
hardware is merely used as an output device (a waste of its
capabilities) or,
b) it could mean that DS3D
commands are passed via A3D 2.0 to
the competing rendering engine, such that it can render the
sound scene using its own algorithms.
If A3D will actually pass DS3D commands to a competing hardware card, that would be great
news! If this is true, then I am truly mistaken about the intent of A3D 2.0.
However, if the former is true and the owner of a competing DS3D card is treated to
something like A2D instead of the native algorithms of their hardware, then A3D 2.0
doesn't support the DS3D hardware in any real or useful sense.
I leave it to Toni to clarify this point for our illumination.
For the reader's reference:
When QSound says that the QMDX (or QMixer) API supports all 3D hardware, we mean quite
literally that standard Microsoft DirectSound3D positional commands will be passed to the
hardware (unless the developer specificially decides otherwise.) The actual rendering of
the most important sounds will normally take place on the hardware using its native
algorithms, whether for headphones, 2 speakers, 4 speakers or anything else provided by
the card/codec/whatever. In other words, if you have 3D hardware you will hear what you
bought.
The support of universal standards
============================================
Universal open standards make it easier for developers to code their games for maximum
effectiveness across multiple vendor hardware. That's dead simple. I think it stands to
reason that a technology vendor who wouldn't want to discourage the support of competing
hardware would participate actively and willingly in the process of creating open
standards... but that's just another opinion.<
An example for the reader's consideration:
QSound and Aureal both recognized that resource management was a problem in DS3D.
Aureal developed a resource manager for their hardware, accessed through proprietary
commands. To be fair, this preceded the advent of the property set mechanism in
DirectSound - which is a universal means for extending the API. However, this mechanism
has been available for a year, and a protocol for resource management has now been defined
for all to use. Toni indicates that Aureal may support if he sees evidence that it is...
um, a standard. (?)
Is this proactive support of open standards?
On the other hand, QSound also developed a resource manager for their hardware, but it is
accessed through the DirectSound mechanism of property sets. It was different than
Aureal's, and QSound could see that developers now had to support both separately. If a
new card came out after a title was released and it used yet another management scheme,
the developer would have to put out a patch to support it. We thought this sucked.
Subsequently QSound offered the protocol they had developed to Microsoft for consideration
as the basis of a standard property set protocol that all vendors could adopt. This sort
of thing doesn't happen overnight, but Microsoft has made changes as they saw fit and are
actively implementing what they now call the "Voice Manager" for the benefit of
developers, OEM's, and ultimately consumers.
QSound is of course committed to supporting this standard for our OEM licensees.
Furthermore, as part of our survey, QSound is attempting to ensure that developers are
aware of the Voice Manager and its significance. (We also tell them about the option of
using QMDX which has its own manager, useful for any hardware.)
I must assume that it would be trivially easy for Aureal to support this protocol
(alongside their own perhaps, for backwards compatibility) to provide access to their own
resource manager. Like the evolving EAX-based environmental extension, this provides the
possibility of universal base functionality across all platforms.
It does not restrict in any way a vendor's ability to offer proprietary extensions via
property sets or even proprietary API's.
However, in both cases Toni has said that IF these standards become significant THEN
Aureal will support them. There's a chicken and egg thing happening here.
Open standards are developed when competitors work in cooperation to make life easier for
content developers and end-users. You can roll up your shirtsleeves and help make it
happen, or hang out on the sidelines and stall until the market forces you to comply.
I stand by my assertion that with respect to open standards and easy cross-vendor
development support, there is a clear difference in corporate philosophy in evidence...
but of course, that's just my
opinion.
Moving along...
In the discussion of what constitutes a standard in terms of market penetration, Toni
sweeps aside the significance of the four-speaker DS3D cards that dwarf his company's
sales. Not much of an issue he says; few people plug in the extra speakers so it really
doesn't count. Interesting.
Then, under Marketing vs. Reality, Toni reprints a bunch of quotes to deflect my
invitation to prove that A3D technology meets the promise of its own marketing. For the
zillionth time, my point wasn't whether or not people think it's good or valuable. (I
could go pull quotes from QSound press releases too.)
The question was: does it work as advertised? By extension, should we expect the stuff
that isn't even here yet to work as advertised, or should we wait and listen and see how
easy it is to use and then form an opinion? (Geez, another dangerous suggestion.)
On the issue of 3D audio effectiveness in various output scenarios, here's a summary that
doesn't tell the whole story, but is certainly
> Headphones:
> Advantages: best overall rendering of 3D parameters, especially
> up/down positioning when compared to speakers. Disadvantage: you
> have to wear them.
>
> 2 Speakers:
> Advantages: good average on all 3D parameters, best ergonomics.
> Disadvantage: you have to sit in the sweet spot.
>
> 4 Speakers:
> Advantages: best front/back rendering, larger listening sweet spot
> than 2 speakers. Disadvantage: no up/down rendering, you have to
> have enough room for them.
>
> In other words, no output device is perfect and there is no
> universal preference among users.
These are sane words indeed. They are excerpted from a
FAQ written by Toni and available (for the moment anyway)
at http://3daudio.pair.com/text/finalword4.txt
Headphones, 2-speaker and multi-speaker 3D rendering all have have their strengths and
weaknesses--i.e. none of them is perfect--and they also have their rabid fans and rabid
detractors. I suppose I should have just pulled this quote before and saved myself the
trouble of trying to make the point myself.
Toni's own admission (incomplete, but a good start) of the strengths and weaknesses of
delivery systems merely underscores the wisdom of developers supporting all the options.
The way to do this is by using an API that assumes no particular strengths or limitations
of the hardware, and letting consumers choose the rendering product and configuration that
makes sense to them.
I've said the following elsewhere; might as well say it here.
1. No currently available
"3D" technology is perfect.
2. Despite imperfections, all
available 3D technologies can add
incredible value in appropriate applications.
3. Skilled creative use can
dramatically enhance the perceived
effectiveness of any illusion, especially when it comes to
sound. The results are highly subjective.
4. I think Aureal's HRTF stuff
is as good as any HRTF stuff, I
just don't think it's as good as their marketing claims it is.
If that is considered an insult, I can only point out that Toni's own quite reasonable FAQ
skewers the claims to which I object with greater authority than I can bring to this
discussion, vis-a-vis the possibility of my own bias.
With all due respect, I close with the same message as always:
Don't just take my word, or Toni's word, as gospel. Listen, think,
(BTW, if any readers would like to express their opinions in private,
I invite your mail at smw@shaw.wave.ca, pro or con.
P.S. I won't quote
you without your permission.)
Scott Willing
QSL
|
|
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.