 |
|
A question you want answered? Let us know and we will do our best to get you the answer. |
|
|
|
|
|
| Sound News | Press
Releases | Archives | Week In
Review | Editorials
| Articles | |
|
|
|
||||
|
|||||
|
|
FAQs: Unofficial Vortex 2 FAQ
Thanks to all the people who have either asked the questions, provide the foil or helped with the responses. Special thanks to Toni Schneider of Aureal Semiconductor for taking the time to work on many of the responses. Given that the Xitel Storm Platinum and the Turtle Beach Montego 2 are also in the Vortex 2 picture, its time for this FAQ to broaden its horizons. This will, as per typical FAQs, be a work in progress as we gather more information about all the cards. The answers to some of the questions still focus in on the MX300, especially in terms of software issues like the Zoran Soft DVD solution. Created: October 19, 1998 Index A3D 2.0 What other Languages Can I get this Information in? iXBT Hardware, one of Russia's leading hardware sites have translated our Unofficial Vortex 2 FAQ into Russian. What Manufacturer's Have Announced Vortex 2 Boards?
* Adobe Acrobat Reader Needed to view PDF file Click Here To Download * As we noted, much of the information in the FAQ has a MX300 slant to it. Untill we gather the same volume of card specific information to supplement the FAQ you can check out the following PDF file on the Xitel Storm Platinum* which is a copy of their product brochure with a picture and info. You can now also go to the Xitel site for information on the Storm Platinum. For information specific to the Turtle Beach Product we can offer the following information: Voyetra Turtle Beach is offering two versions of the Montego II. One is the direct version (two speaker out, AudioStation 32 software bundle) which is currenly an upgrade option on Dell systems which Turtle Beach has decided to offer throught their web site. The other is the full retail version. Details are not finalized as yet but we can offer the following from Turtle Beach. The quad-speaker retail version (w/ RCA digital out, gaming bundle), the Montego II Home Studio edition (w/ RCA and optical digital I/Os, quad-speaker support, a Roland GM/GS sample set, and Digital Orchestrator Pro music production software), and a separately-available breakout board (provides an upgrade to quad support and digital I/Os from two-speaker versions) are in the final stages of development, and were previewed @ COMDEX. We have a few questions on the go on this board, only one answer so far that came courtesy of ExtremeHardware. Here is the Q & A. Q: Will people who buy the 2 speaker OEM version of the Montego II have an upgrade option to the bracket board with RCA and optical digital I/Os and quad-speaker support? TB: Yes, owners of this card can upgrade with the bracket board I mentioned, adding RCA and optical digital outs as well as quad speaker support. That upgrade should cost somewhere between $70 and $99. Q: Can you tell me some more about the MIDI on the Montego II Home Studio edition? A: the MIDI chipset will be provided by Dream, and placed on a Turtle Beach-designed daughterboard. As noted before, it will contain Roland GS-compliant samples, but still no word on their specific make-up. Q: So what does this card look like? 3DSS: Aureal A3D Central has a picture of a vortex 2 card. We have borrowed Kirben's reference card picture and added to a pic of the MX300 here. SharkyExtreme has two pictures of the MX300. Check them out at Mx300 Shot 1 and MX300 Shot 2. We also have a pic of the Storm Platinum Board and a link to a pic of the Turtle Beach Montego II (without the bracket board) Q: So how big is this card anyway? 3DSS: The MX300 is 7 inches long. I want to start of with a correction to a statement I made in a post on Sunday October 18th. In it I stated that the Vortex 2 has a full reverb hardware engine. Toni Schneider of Aureal Semiconductor made the following comment to my post: Mark, you say "Vortex 2 has a full reverb hardware engine". This might be misleading to some people. Vortex 2 will handle EAX style reverb by doing the computationally intensive tasks of sample rate conversion and mixing in hardware, and the memory intensive task of running the reverb filters and delay lines on the host (using main memory and the host CPU). We think this is a great design that takes full advantage of the PCI bus and host memory (two things that you already have in your system), so we don't have to burden the sound card with additional cost for memory. Q: How much processing power will that
require on say a PII 300? Reverb Question #1 background: A reader quoted the following from Kert’s Wavetracing article: "Late order reflections will be handled with reverberation. The rationale is stated in the white paper: humans are capable of individually perceiving first order reflections while second and higher order reflections usually combine to form late field reflections, or reverberation. According to Toni, reverb works best for very large spaces that require long decay times. A good example is a cathedral or large cave that you move around in slowly, exploring as you go along, hearing everything echo for a long time. Late order reflection parameters constitute a part of environment properties: Late reflection (reverb) engine variables Input and output level of the late reflections. Reverb predelay. Decay time for late reflections. Brightness of late reflections. At the time of writing, late field reverb capabilities are unavailable and will be added via a driver upgrade, probably in A3D 2.1 " They then made the following statement To me it sounds like the reverb engine would only take care of late field reflections but Live's reverb engine (as I understand it) is an approximation of both early and late field reflections. If I'm right about that Aureal can't only use the reverb engine to match Live's since the first reflection is the most important (as I understand it). Q: Can you respond to this? Toni: Fixed early reflections as part of a reverb are pretty much standard in any reverb. Our reverb engine will model fixed early reflections just like Creative's when a game uses reverb only (via something like EAX). As we've described, the disadvantage of that approach is that things sound somewhat static and flat, because the reverb itself doesn’t change as you move around (the only thing that changes is the wet/dry mix between direct path and reverb). If the game uses A3D 2.0, the idea is to turn the reverb's fixed early reflections off, and substitute dynamically computed and individually rendered early reflections. The reverb can be left in there to do the late field rendering. That way you get it all, the interactivity of early reflections, and the long echoes of the reverb. Why the questions about EAX in a Vortex 2 FAQ? Well, its an open extension to DS3D and Diamond and Aureal have listed support for it in the comparison tables we have posted on 3DSS. In addition, Creative Labs has announced wide spread support for EAX in games coming this Christmas. A consideration for buyers will therefore be maximum compatibility with the games on the market in the near future. We know about plain DS3D, we know about A3D 1.x but we don't know about EAX. Q: Could A3D 2.0 be used to make the EAX calls saving the developer the work of coding 2 APIs? Toni: What we're currently doing to reduce the need to code to multiple APIs is the addition of DS3D support inside A3D 2.0. This allows a game use the A3D 2.0 API and still get the full benefits of DS3D. So in a system that doesn't have a Vortex 2, but has a DS3D accelerator, we fall back to generic DS3D mode. This benefits EAX platforms, since they use DS3D for positional 3D audio. More specifically, a game that uses A3D 2.0 as its only API will do the following to cover all types of hardware: On a Vortex 2: 3D direct path, reflections, occlusions, reverb. On a DS3D hardware accelerator: 3D direct path (same as you would get by using DS3D directly) On anything else: A2D emulation (similar to what you would get by using DS or soft DS3D) Q: Any estimate on timing for support of EAX on the Vortex 2? Are we talking about Direct X 7 time frame?
Q: Will the Vortex 2 provide the same or better reverb than the Live in EAX only games? Toni: Our reverbs will be high quality and sound great in games. Vortex 2 will combine that reverb with the highest quality 3D audio positioning filters we've ever done, and we think the result will be a highly competitive solution. Q: Will the Vortex 2 support EAX 2.0? Toni: The answer is yes. Vortex 2 can support EAX 2.0, and we intend to implement that support. From what I've seen of EAX 2.0 it is very similar to EAX 1.0. The new occlusion controls in EAX 2.0 are quite simple, and they are a small subset of A3D 2.0, so no problem for Vortex 2 to support them. Q: What can you tell us about the streaming capabilites of the Vortex 2? Toni: The stream routing in Vortex 2 is completely programmable. Vortex 2 features 96 DMA streams. These streams are audio data streams going from host memory down to the chip. 32 of those streams can additionally also go in the other direction, i.e. from the chip back up to host memory. Once an audio stream is inside the chip it is routed along the internal "Vortex Dataflow Bus" to go between the various core audio processing engines (sample rate conversion engine, mixer engine, 3D engine, wave table engine, etc.). The Vortex 2 driver software decides how to route the streams based on user settings and the real-time audio needs of the various applications running on your PC at any given time. Since no connections are hardwired, the chip has great flexibility that translates into optimal usage of hardware resources for any given task and the ability to re-program the data flow for new tasks. An example of the latter is multi-channel Dolby Digital streaming support which was recently integrated. This architecture also makes Vortex 2 a great fit for the upcoming WDM (Windows Driver Model), which is heavily oriented towards stream routing. As far as I know, the Vortex architecture is unique in terms of combining highly optimized core audio processing engines with complete topological flexibility to route sound streams between those engines. Having mastered the complexity of such a design, we are now reaping the benefits by having a solution that can deliver maximum quality at minimum silicon cost (BTW, both Vortex 1 and 2 are based on the same core architecture). Other PCI audio chips are usually either general purpose DSP based (for example the Cirrus Logic SoundFusion), hardwired (the Ensoniq AudioPCI?), or hybrids of hardwired Wavetable engines with audio effects DSPs on the side (I believe the EMU 10K1 and ESS Maestro II would fit this category). If you are looking at actual usage scenarios, things get complicated quickly. The basic rule is that Vortex 2 can handle any form of audio stream and multiple active streams can add up to 96 at any given time. Of the 96 streams, 4 are pretty much permanently allocated to overhead such as the DirectSound primary buffer, or line and mic input. That leaves 92 output streams. Of those, you can have a maximum of:
The different types of streams can be mixed and matched in various ways, but they always have to add up to a maximum of 96. There are other types of streams such as wavOut, or DOS legacy apps, of which you'd typically only ever have a few running in parallel at the most (Vortex 2 can handle them as well of course). A typical usage scenario for today's applications might be: You are playing an A3D game that uses 16 A3D channels for primary sound effects, 2 DirectSound channels for soundtrack streaming off a CD, 8 DirectSound channels for background effects. Additionally, you have e-mail running in the background which uses a wavOut channel to play an e-mail alert sound every once in a while. This would add up to about 32 streams (with overhead channels). Such a game could add another 32 A3D or DirectSound3D channels, and still be well within the limits of Vortex 2 (64 streams). This case would bring about any other PCI audio chip to its knees. Or for more advanced uses, you might imagine that a (future) game adds 32 3D reflections to the baseline described above, and has a DirectMusic soundtrack that maxes out at 32 voices. This would bring you to about 96 streams and take advantage of the entire Vortex 2 chip. I believe that a game like that would for sure go beyond what other PCI audio chips can handle in hardware. I am not going to re-invent the wheel here as there are some excellent resources available (even if some are not on 3DSS :) ) If there are specific questions that the resources do not answer, be sure to let me know and I will address them here. In the meantime please reference the following: A3D 2.0 Technical Brief at Aureal Semiconductor Kert's wavetracing article on Kert’s Page Our own wavetracing Q & A with Aureal's Skip McIlvaine. Q: How many wavetable voices does the Vortex 2 have. 3DSS: 64 hard + 256 soft Q: Does the Vortex II offer a connector for a MIDI daugher board? 3DSS: Yes Q: What is the general memory size of the Vortex II for downloaded instruments Toni: Samples are in host memory, so I think the only limit is the size of your host memory. Q: Will users be able to seleect preset reverbs and/or create reverb setting for playing MP3s or MIDI files? Toni: This ability already exists for MIDI files. We are planning on adding it for all audio types. Q: The MX300 supports both
a waveblaster header as well as a joystick midi port. Are these separately addressable
(appear as separate ports) to a sequencer? Or are they ganged together? Q: So can I use my external wavetable card and an external midi device at the same time? Is this a driver issue or does an external midi device have to be physically pluged into the MX300 for the option of a second midi port to be available? (like how it is able to detect which audio out jack is being used) Toni: The daughtercard is just like having an external synthesizer attached to the MIDI out port. It responds to the same events as the external synthesizer connected to the MIDI port. There is only one MIDI out port on the MX300 and it will drive the port as well as the daughtercard concurrently. As noted in the answer above, the only control the driver has is on the volume of the daughtercard. Q: I have a wavetable daughterboard. Since I won't be needing the MX300's midi capabilities, is there anyway to completely disable it and recover the memory by unloading the sample midi set? Toni: The wavetable memory is only paged into memory while the wavetable is active. When it is stopped the system memory used for the sample set is unlocked and can be paged to disk. There is no need to disable the wavetable to reclaim the system memory. Q: Will the Vortex 2 MIDI implementation support any kind of sysex? Toni: We currently don't support any sysex messages. Q: When I play MIDI files using the Vortex 2 wavetable I get four speaker playback but when I hook up a wavetable daughterboard the MIDI files only playback on 2 speakers. Is that normal?
Q: Are there any known issues? Toni: The Thrustmaster Rage 3D and some of the older Interact Joysticks have problems with our gameport (Vortex 1 & 2). The problem is easily resolved in their drivers but they have no desire at this time to put out an update due to these products being older and end of life. They are testing and making sure that their newer and future products will work correctly with our sound cards Q: Can you use a MIDI daughtercard and the MS feedback sticks at the same time or will there be conflicts? Toni: According to MS, the feedback sticks either use the Midi I/O, game port or Serial cable. Depending on the MS controller driver selection, there may be a conflict on the MIDI I/O port. To avoid this problem use the serial cable provided by MS by request. BTW- Just about every sound card will have this same problem since the internal Midi connection is identical to the Gameport Midi connection Editor's Note: Much of the concern around joysticks stem for the original Monster Sound's early problems with digital and force feedback controllers. These cards were not built around the Vortex 1 which has a virtually clean record with controllers (Thrustmaster Rage being one exception that comes to mind) Q: What titles near release will definitely support A3D 2.0 3DSS: Ones that are out right now are Half-Life, Heretic II, Motorhead and Sin. Since 2.0 is just starting to ship, 2.0 titles will begin to pick up soon. Out of the box the MX300 will use the Soft DVD player to decode AC3 and send it to the appropriate channel on the MX300. Q: If a user has a hardware DVD decoder, then they send the output directly their external DVD amp is this correct? Toni: Yes, using either one of the daughter cards, they can connect the digital AC-3 bitstream to their DVD amp, the same way you'd connect a consumer DVD Player. Q: At the MX300 launch you mentioned that there would be an option to use the SPDIF in the above situation with the MX300. Would that be simply as a digital pass through? Toni: Yes, the SPDIF out is simply the standard way to connect to consumer equipment like Dolby Digital amplifiers. Q: Is there any way to use a hardware DVD decoder and have the MX300 do the 5.1 output? Toni: I think the answer to the second question is no. Once the signal is sent to a hardware DVD decoder that decoder has to do the output. Q: If a user gets the upgrade card that allows full 5.1 output and if that user has an AC3 amp that has 6 discrete inputs, can they connect the 6 outputs from the MX300 to the AC3 Amp? Toni: Yes.
David: When you send normal audio out the S/PDIF, it will only send the front stereo signal. There is no way currently to send the signal going to both analog jacks. In order to send front and rear, we'd probably need to encode the streams on the fly into a Dolby standard for your decoder to decode. I seriously doubt that will happen considering the cost of such an endeavor (in terms of engineering time, money to Dolby, and CPU usage to encode). At the MX300 launch you mentioned that A3D 2.0 and the Vortex 2 have gone beyond 4 speaker panning and now use HRTFs. Q: Can you talk about the 4 speaker HRTF? Is the HRTF used on all 4 speakers? What about up/down virtualization? Toni: HRTFs are used on the front speakers. The fronts basically do similar processing as they do in 2 speaker mode until the sound moves in the region behind you. At that point the rear speakers kick in to fill in panning information for the behind locations. This gives you the best of both worlds: nice spread, immersion and some up/down by using HRTFs on the fronts, and strong behind cues from the rear speakers. Q: Is there a negative impact on the HRTFs from using 4 speakers? Toni: There is no negative impact I can think of. In the worst case, i.e. if you are sitting way outside the sweet spot, the system more or less sounds like a standard 4 speaker panning system. Q: What about side placements? To put this question in context here is a quote from Scot Willing of Qsound that I got from one of our readers. Can you respond to this? "For the creation of the side placement illusion, having real speakers and just using panning is actually a disadvantage. In the real world we're much better at judging side-to-side azimuth placement than front-to-back, because of the strong interaural time-delay and amplitude difference cues that are generated by actual sources to the sides. These approach zero for objects directly ahead and directly behind, so our front/back differentiation is not that good. Therefore, having real speakers behind really makes a difference for front/back differentiation, but actually works against the creation of convincing side placement without the use of 3D processing." In response the question as why Aureal didn't use 3D processing on both the front and rear speakers and Scott Willing answered "For one thing it doesn't fit the HRTF model easily... they'd actually have to come up with a new idea! (Sorry, cheap shot.) For another it would be very expensive to implement." Toni: I don't have any scientific data on this, but from my experience Scott's observations regarding weakened side placement on 4 speaker panning systems seems correct. That's one of the reasons why we decided to use HRTF 3D processing instead of simple panning. HRTFs make the side positioning more convincing (along with adding up/down cues). We don't use HRTFs on both fronts and backs because we found that the "HRTF in front, panning in rear" model works very well. Another reason is that people tend to set up their rear speakers in weird spots (because of physical space limitations), which makes HRTF in the back pretty wasteful (HRTFs are more sensitive top speaker placement than panning). Q: Will the add on card for the MX300 that allows for 5.1 support be used by A3D 2.0 for 6 speaker support? Toni: The main reason for the additonal analog out is to get to full 5.1. The Vortex 2 chip could theoretically also handle 6 (or even 8) discrete A3D outputs to that many speakers, but we are curently not planning to implement that support. The demand has been very low, especially compared to other features on the improvemetns list, such as 76 3D voices, reverb, etc. A lot of people have seen the emphasis to-date by us and others on four speaker support and have been asking about 2 speakers and headphones. The A3D alogrithms are optimized for 2 speakers and headphones as well as for four speaker support. Users by selecting their playback configuaration in the control panel will at the same time be selecting a custom algorithm. Q: Can the multiband equalizer be used discretely between the front and rear channels to compensate for the different speakers? Toni: No. The eq is a stereo eq, so the same eq curve will be applied to all four speakers. Q: What’s the ideal height for the 4 speakers relative to the listeners head? Toni: The fronts should be placed the same as in 2-speaker mode, i.e. to the sides of your monitor. The rears are not terribly sensitive to placement, as long as they are behind you, and spread apart from each other. My personal preference is to have them pretty far away from me (they sound more ambient that way because you hear more of the rear sounds reflecting around your room - you're also less likely to trip all over them if they are far away), and spread apart quite a bit. Q: Will the MX300 will allow you to play music CDs through four speakers? Toni: Yes. Even 10 band eq will work for CDs (can be disabled). Q: When I hook up my Vortex 2 based card to the Cambridge Soundworks FourPoint Surround System (and possibly some other 4 point systems with similar implementation) the subwoofer is muffled to the point of not working. Is their a fix for this? 3DSS: The next driver release from Aureal will look after this problem. In the mean time, if you download the Vortex 2 reference drivers and move the front/read slider one or two notches in either direction (we suggest to the front) then subwoofer will start to kick out the bass again. Q: The Sound Blaster Live is built around the EMU 10K1 which has a very powerfull 1,000 MIP DSP. The Vortex 2 is listed as having a 600 MIP hardware engine. Does this mean that the Vortex 2 is less powerfull than the EMU 10K1? Toni: Emu is quoting DSP MIPS, we are quoting hardware MIPS. The two are not the same. If we had to run all of Vortex 2 on a DSP, it would probably take 2-3 times as many MIPS (1200-1800 MIPS). Q: What kind of NT support can we expect for the Vortex 2
3DSS: The request for dual processor support in NT has been logged by Aureal but there is currenty no time line for when it might happen. Q: Will there be any Linux support for the Vortex 2? Toni: The only thing I know of are the 4Front Open Sound System guys. They are working on Vortex 1 drivers, and might move on the Vortex 2 when they are done. They are the leading Linux audio driver guys, but their stuff isn't free. Aureality has since posted the following update: Skip: We are working with a couple
folks (besides 4Front) to develop Linux drivers for Vortex. I'll let you know as soon as
we can announce the details Q: Will their be BeOS support for the Vortex 2? 3DSS: Yes, BeOS drivers are currently been developed for vortex 2 based sound cards.
Q: Will there be OS/2 support for the Vortex 2? Toni: Right now we only have OS/2 drivers for Vortex 1. I don't think there are any plans for Vortex 2 drivers for OS/2. As far as OS/2 functionality is concerned, Vortex 1 is almost equivalent to what Vortex 2 would be like (since there is no A3D in OS/2, the A3D capabilities of Vortex 2 would not be available anyway), Q: Are there issues with Vortex 2 boards and VIA MVP3 based super7 motherboards? David: Yes, there are
numerous issues with VIA MVP3 motherboards. Normally they are realized when you use any
Vortex card (and some other PCI cards as well) and an AGP video card. Later revisions of
motherboards tend to have the problem fixed as long as you download all the various VIA
patches from their web site. But still, some people never get it to work. It is a
motherboard issue, though, and we can't do anything about this.
All cards are shipping except the 4 speaker versions from Turtle Beach Turtle Beach's Montego 2 direct
version has bee shipping since early December. No date has been set for the
retail quad version.
|
|
|
|
|
symbol.
I will cycle them off every couple of days until nothing is new.
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.
