The X-Fi Audio Ring: Powerful and Flexible

Putting aside marketing buzz words and the slightly tacky Xtreme Fidelity moniker, the new generation of sound cards from Creative Labs is quite impressive. The technological advancements come from the inclusion of four key new features: the ring based architecture, a high quality sample rate converter, a powerful DSP, and on-board RAM. First on the chopping block is the Audio Ring.

Current sound card architectures are based on a linear flow of data. For straight up audio listening or recording, this is not a problem - there isn't any need to deviate from the norm in these cases. And that's why we haven't seen a fundamental change in audio architecture until now. It is when we want to intensively multiplex these components and perform operations on large numbers of audio streams that we run into problems. The backbone of X-Fi is its ability to process any audio stream on any component in the audio chain in any order, and any number of times without having to leave and come back in.

The ring supports up to 4096 internal audio channels that can all be taking different paths through the hardware. Don't confuse these internal channels with the number of voices that the card can handle. The X-Fi is maximally capable of playing up to 127 simultaneous voices. These 4096 channels include channels necessary for advanced filtering and effects processing. The bus is time division multiplexed rather than interrupt driven or otherwise mastered. Time division multiplexing involves handing out the entire bandwidth of the bus to a single channel on a set time interval. The result is that each channel has a deterministic bandwidth and latency - aspects important to effective audio processing. For example, in a household with two teenage daughters, we could time division phone bandwidth by letting one daughter talk during even hours of the day and the other during odd hours. This way, the entire bandwidth of the phone (or audio ring) is fully allocated and you would always know which daughter (or channel) is using the bandwidth at a given time.

Another reason why Creative may have chosen to support so many internal channels is so that it can handle algorithms that call for splitting or duplicating an audio stream for multiple different types of processing and analysis. This could range from using CMSS-3D (Creative's new surround upmixing feature) to efficient implementations of complicated high order N-tap and feedback filters. Creative doesn't talk much about the uses of these internal channels, but many of their new features include detecting aspects of the audio being played in order to enhance the sound. It seems likely to us that the algorithms that implement their new features would rely on the high number of channels and the flexibility of the ring architecture to get as much done at one time as possible in order to keep from introducing an unacceptable amount of latency.

Now that we know how the ring works, here's what each node on the ring does:

Transport: This handles moving audio streams in and out of local RAM as well as over the PCI bus to system RAM. The Transport engine manages over 1000 DMA channels to both the PCI bus and local SDRAM. Being able to manage so many DMA requests at a time is important for dealing with the latency of PCI and effectively managing the volume of data with which the new solution is capable of working.

Tank: Its name is based on the tanks that held mercury used to create high quality analog delay lines. The tank engine is capable of fractional and modulated delay lines (delays that don't line up with the sample pattern and delays that change over time). The tank engine supports up to 1024 accesses per sample. This means that the tank engine can assist in many types of effects including: chorus, reverb, reflections, and interaural time delays (for positional audio effects). These effects are employed quite a bit in many of the audio processing features of the X-Fi, so avoiding the use of a DSP for the creation of a delay line is very helpful in spreading out processing power.

SRC: The Sample Rate Converter seamlessly transforms any audio stream or channel to any other supported sample rate with very low ripple and THD+N. We will cover this node in detail as it is one of the key features of the new architecture, allowing Creative to convert the sample rate of all audio multiple times independently of any other audio with little to no discernable loss in quality.

Filter: The filter engine implements 512 floating point 2nd order IIR filters. This is the fundamental building block of sound synthesis, 3D spatialization, equalizers, speaker calibration, and a host of other features. There are 13 filter types implemented in hardware from direct and parametric 5 parameter EQs to notch and peak filters. Complex (higher order) filtering and synthesis can be achieved by looping through the filter engine multiple times until the desire result is achieved.

Mixer: Consisting of 256 audio summers, 1024 parameter combiners, and 4096 single segment parameter rampers, the mixer is where internal audio channels come together. Parameter mixing and ramping are used to control effects combining for things like 3D audio effects in game. Instead of single segment, the parameter rampers can also support either 1536 multi segment rampers or 4 multi segment shapes (possible shapes are linear, pseudo exponential, and pseudo logarithmic).

DSP: The X-Fi Quartet DSP is so named because it supports 4 hardware threads. Each thread has access to two SIMD (single instruction, multiple data) units for easy stereo and complex math processing. NVIDIA has called a multiple SIMD processing unit MIMD (multiple instruction multiple data), but Creative has dubbed the overall architecture TIMD (Thread Interleaved Multiple Data) due to the hardware threading support alongside traditional methods. The Quartet DSP is central to many of the X-Fi features and we will cover this hardware at length as with the SRC engine.

Audio I/O: This node handles moving the final processed audio streams to output either digitally or through a DAC, and can also acquire input from each of the source options on the hardware. Audio sources in memory do not enter the path through this node, and likewise for audio written to memory (or a file). This block manages all the physical I/O ports on the X-Fi card.

Having easy access to any of these structures at any time during audio processing will greatly simplify the process and increase the complexity of operations possible on X-Fi hardware. In order to further optimize the architecture, Creative has introduced three distinct modes in X-Fi. These modes outline the basic type of path that audio streams will take through the ring. For instance, professional audio programs require very low latency audio from the sound card when recording. In order to accommodate this, the Creation mode limits the types of processing done to less complex (and faster) algorithms. Creation mode also allows synchronized audio streams for proper integration into a studio setting. These features aren't necessary when in Gaming or Entertainment modes, as the focus of these modes is on generating or heavily processing audio before input or output.

Architecting the card like this isn't useful unless the nodes on the ring are powerful enough to exploit the potential. Creative has spent quite a bit of time in making sure that this is the case. Their components are high quality and introduce as little distortion and noise as possible in order to make heavy processing of audio on a consumer sound card a reality. (Whether this is really a good thing or not is still up to the end user.) Over the next couple of sections, we will cover the most important components of the audio ring, as well as explore what is possible when all of this hardware sings in unison.

Index X-Fi Processing Elements: Exploring the SRC
Comments Locked

110 Comments

View All Comments

  • SDA - Tuesday, August 30, 2005 - link

    >>Why do you post opinion like that and give no example? What is the point? >>
    Well, gee, I don't know. Why do you post your own opinion? Because we both like to post our opinions, of course. Everyone here does, else their posts wouldn't be there.

    I agree that it's only fair to provide examples, so I have done so in my post. Beyond that, some general rules before buying ANYTHING:

    - Determine what measurements and features you should be looking for. For example, with a CPU or GPU, you'd know you'd want to look for benchmarks in applications that you find yourself using most often, and what features you wanted or needed for those applications (SSE3? Multiple cores?). With a PSU, you would look for power quality (as determined by rail sag under load, AC ripple on all rails under load, etc.), ability to supply a given amount of power over a long period, whatever features you needed (modular cables? no fan?), component quality, and preferably a large amount of samples (there is often a great difference in quality from PSU to PSU from the same manufacturer).

    - See if you can obtain some background knowledge in that field. It's a pain, I know, but it can come in really handy if you don't know what sources to trust or what features to look for. For power supplies, that means getting some basic power-specific EE knowledge (or asking an EE, heh); for CPUs or GPUs, that means trying to understand a little about how these processors work, what specific attributes are valuable in the applications you use, and how performance can be affected by the rest of the system.

    - Find resources that you trust. You can identify these by whether or not the reviews or analyses in these resources take into account relevant measurements and features, and by how knowledgeable the reviewer seems in a specific area. A video card review that only does AquaMark and 3DMark should not be trusted (and the site or person that does it should be regarded with some suspicion in that field), and neither should a power supply review where the reviewer does not once mention ripple or do anything more than check the rails under light load with a Windows utility.

    - Consider bias. This is not relevant at AT, where I at least trust the editors, but it is very important with some guy that spent a big chunk of money buying something. Of course someone who spent $500 on the latest and greatest processor is going to say that they like it, unless something is horribly wrong. How often do you hear people say that they think their stereos sound bad, that their computers are overpriced, that the PSU that they bought doesn't use the components they would expect at its price point? Now, I grant you, rational people do buy things BECAUSE they feel they are good choices, but there is almost always a significant degree of personal bias with some personal item.
    (bias is also relevant when considering smaller or less trustworthy sites that may rely on manufacturers' goodwill, including but not limited to THG.)


    A little long, I know, but maybe what little I know can help you with future purchases. If not, you don't have to read it, heh.
  • Eskimooo - Wednesday, August 31, 2005 - link

    thanks for that. Much appreciated. Well, I did quite a research and virtually all reviews I came across including digit-life appreciate technology behind X-Fi. Most reviewers like even this much doubted Crystallizer feature, which is not obligatory and adjustable anyway.I do not believe Creative is so powerfull to exercise pressure on all the reviewers. So if both RMAA etc tests and subjective tests give positive results this card is probobly good and I think I am going to buy it
  • Eskimooo - Tuesday, August 30, 2005 - link

    previous one posted to early, sorry. So my final question to all who listened to music using the X-Fi card and can compare it to other sound cards: Is there anything that would give more better features at the same quality level and at the same price? Please do not reply on Audigy or Live because X-Fi is a completely different card.
  • DerekWilson - Tuesday, August 30, 2005 - link

    It depends on what you are looking for, but in most cases I will have to disagree with you.

    The Gina 3G sells for $350 and does not off the quality of the X-Fi Elite Pro. Features include hardware 3D audio buffers, filters, and effects. Again Pro cards fall short (especially for gaming).

    In fact, the only thing that I would give as an advantage of similarly priced professional audio solutions is balanced audio. This is a big advantage if the user intends to setup a completely balanced audio chain.

    If I had to guess at a product that would best the X-Fi in terms of quality, I would have to point to the LynxTWO-B ... For which you would easily spend more than twice what the X-Fi Elite Pro costs. If you know of a pro card that you feel fits your claim, please point it out to us and we will do a comparison.

    From what we have seen, your prescott/athlon64 comparison doesn't make much sense. If you take away the non-pro qualifier, we would still maintain that the X-Fi Elite Pro is the best sound card for any application at it's price point (unless you need balanced I/O). If you include no upper limit on price, then we will have to wait for our LynxTwo comparison before we shoot our mouths off too much.

    So to recap, pro audio cards at the same price point or below offer fewer features, equivalent or lower sound quality, and similar recording capabilities (with the exception of balanced I/O).
  • SDA - Tuesday, August 30, 2005 - link

    >>The Gina 3G sells for $350 and does not off the quality of the X-Fi Elite Pro. Features include hardware 3D audio buffers, filters, and effects. Again Pro cards fall short (especially for gaming). >>
    I never said anything about gaming, unless you're interested in recording gaming sessions. Anyway, you're relying on a logical fallacy-- specifically, a hasty generalization. More specifically, that the Gina 3G (a small sample) is representative of all pro audio cards in its price bracket.


    >>If you know of a pro card that you feel fits your claim, please point it out to us and we will do a comparison. >>
    Well, the E-MU 1212m is a rather capable card at half that price. The 1820 and 1820m are also capable pro cards. (Yes, I'm aware what E-MU is. The issue is capability for the price, not who owns what.) The Delta 1010LT is also an option, I guess, though it's a little dated now. There are also external options... Edirol equipment was fairly popular here last I checked (prefer separates, so I'm not entirely up-to-date). You're welcome to do your own research, of course-- that would be a good idea if you're going to make statements about what's the best.


    >>From what we have seen, your prescott/athlon64 comparison doesn't make much sense. If you take away the non-pro qualifier, we would still maintain that the X-Fi Elite Pro is the best sound card for any application at it's price point (unless you need balanced I/O). If you include no upper limit on price, then we will have to wait for our LynxTwo comparison before we shoot our mouths off too much.

    So to recap, pro audio cards at the same price point or below offer fewer features, equivalent or lower sound quality, and similar recording capabilities (with the exception of balanced I/O).>>
    The comparison was solely to illustrate the issue of specialized comparisons. If you are willing to remove the "non-pro" qualifier, it is irrelevant.

    I appreciate that the fine folks at AnandTech often have experience with as many as three or four different sound cards, but I must ask: exactly how many pro sound cards have you tested and reviewed? I don't mean "read the specs of", I don't even mean "measured the noise floor of", I mean given a full, comprehensive test. I sincerely doubt it is a very high number, and that you have and are utilizing the proper measurement equipment (and again, simple audio quality measurements say nothing but that everything out there today has obscenely good measurements).

    Why do I doubt that? Well, ffs, you didn't even recognize the "24-bit Crystallizer" for what it was, relying instead on Creative's description that it clears up the high end and adds dynamic range. Someone who can't recognize a crude equalization's effect on the sound and who then regards said equalizer as a noteworthy feature is not someone I am inclined to trust for statements about professional audio.

    If you're curious now about the Crystallizer, you might want to read Digit-Life's significantly better review, linked below.
    http://www.digit-life.com/articles2/multimedia/cre...">http://www.digit-life.com/articles2/multimedia/cre...
  • yacoub - Wednesday, August 31, 2005 - link

    "I appreciate that the fine folks at AnandTech often have experience with as many as three or four different sound cards, but I must ask: exactly how many pro sound cards have you tested and reviewed? I don't mean "read the specs of", I don't even mean "measured the noise floor of", I mean given a full, comprehensive test. I sincerely doubt it is a very high number, and that you have and are utilizing the proper measurement equipment (and again, simple audio quality measurements say nothing but that everything out there today has obscenely good measurements).

    Why do I doubt that? Well, ffs, you didn't even recognize the "24-bit Crystallizer" for what it was, relying instead on Creative's description that it clears up the high end and adds dynamic range. Someone who can't recognize a crude equalization's effect on the sound and who then regards said equalizer as a noteworthy feature is not someone I am inclined to trust for statements about professional audio. "

    Well this is what happens when you have a PC hardware review site in the day and age where the hardware is running out of 'hot new features' and starting to overlap with professional grade equipment in order to justify the pricing.

    It's also what happens when you get a review site reaching out into random other areas like digital cameras - realms where there are other sites that exist solely to do such reviews and will obviously do them to a much finer degree. As you get into the realm of professional audio products with the features of PC hardware audio products, it necessitates one of two things: Delving into the field of pro audio equipment to a large degree, or not making comparisons to items in a field one isn't highly knowledgeable about. :)
  • yacoub - Wednesday, August 31, 2005 - link

    Ah, I see you say as much here:

    "Derek - The concern I have is that your review, as it stands, is a ringing endorsement for a product in a market you do not fully understand. The users who rely upon Anandtech as their only source for this type of reccomendation are likely to purchase something like this, even though there are a wide variety of competitive solutions out there for a quarter of the price. I feel you should at the least post a disclaimer that your audio review process is a work in progress and make it very clear that you do not fully understand the market that the X-Fi is being marketed to, nor have adequate experience with competitive audio solutions. "

    Very well put!
  • DerekWilson - Wednesday, August 31, 2005 - link

    We've played with a few different cards extensively, in cluding the TerraTec EWS88 MT the M-Audio Delta 1010 and the EMU 1212 ... We didn't run any RMAA tests on them, as they were used for listening and recording. From our experience, I can say I'd rather have the X-Fi Elite Pro for a non-balanced setup. Currently I have an external ADC for converting balanced line-ins to ADAT and run digital audio to the computer and I use the Gina 3G for recording the ADAT signal.

    Regaurdless of our experience, we haven't done indepth electrical analysis of these parts yet. It is our intention to test other consumer, semi-pro and pro cards and compare them. We drew our assessment of the X-Fi from our experience with hardware and the in-depth tests we have done so far.

    As far as the Crystallizer goes, we are talking about more than one effect if Digit-Life is correct. Yes, there's equalization, but that's not all. I wasn't expecting multiband compression to be added which (as digit-life points out) decreases the dynamic range of the original signal. We will certainly be speaking with Creative about this (there are other things that could be going on that an RMAA test doesn't determine), and if our analysis leads us to the same conclusion we will certainly have some things to say about it. We certainly noticed the general increase in dB level though ... If I didn't mention it, when comparing crystallizer audio with unmodified audio we decreased the volume to match.

    On top of that, the word noteworthy carries a positive connotation. If you mean only to say that we noted the Crystallizer as a feature and tested it, that's fine. Beyond that, our assessment was that the crystallizer had limited application to some audio sources that were already subpar. I would say we were not wrong.

    At the same time, we are still finding our way in PC audio. We recognize we have some room for improvement and appreciate any direction and help we can get. We will expand our coverage to include indepth signal analysis on features like the Crystallizer. There are difficulties in going down this route, and it is hard to correctly talk about the value of psychoacoustic functions or environmental effects from such analysis. We haven't gone there yet because it is like opening a can of worms. Any suggestions are welcome.

    Thanks very much for taking the time to bring all of these issues to our attention.

    Derek Wilson
  • SDA - Wednesday, August 31, 2005 - link

    Why, jooc, would you rather have the Elite Pro? Just what listening equipment are you using (I'd really like to know this part), and what tests have you performed with them (other than introductory spec measurement)?

    Noteworthy does carry positive connotations, but then the fact that your review sees it as a feature that can be useful for some does mean to me that you've classified it as a noteworthy feature when, in reality, it is utterly useless (anything that can benefit from it can benefit from a better, simpler software EQ that you can carry with you from card to card).

    I tend to agree with Reflex here. I understand and appreciate that you guys are trying, and you'll notice I haven't made any stupid accusations about bias, but the fact is that this is an unqualified endorsement. If you're still finding your way in PC audio, you should be a little more cautious about making sweeping general statements.
  • DerekWilson - Wednesday, August 31, 2005 - link

    I would rather have an Elite Pro because I also like to play games and movies as well as record and listen to music. I do to perfer my external ADC for recording, and would probably rely on it instead of the soundcard itself.

    For listening, I generally stick with my Sony MDR-7509 headset. I also have the Sennheiser HD580 that I use to audition sound hardware. The reason I stick with the Sony gear for general listening is due to my environment. I can turn everything off when I need to do a noise test or listen closely to something, but the lab with all the computers and workstations running is not a quiet environment. I realize that open air headests will reproduce (especially) the low end in a more appropraite manner, so I do listen with them, but I know the sound a little better on the MDR-7509s as I've been using them for a long time.

    I am planning on picking up the HD650 as I've heard great things about them.

    The tests I've performed with them include recording and editing audio on them. I've used each of the solutions I mentioned in my home rig before I moved to lightpipe. I'm definitely not saying they don't get the job done well. And if that's all you want a sound card for, then I'd certainly go with one of those solutions over the X-Fi. Likewise, if all you are doing is stereo audio listening then one of the other solutions is still a better choice.

    The advantage of the X-Fi (and I'm sorry if I didn't make this clear in the article) is that it is capable of high quality recording, high quality playback, EAX 2+ with 127 voices, has lots of analog and digital I/O, and generally meets any requirement anyone could have from a consumer or semi-pro sound card (except, again, balanced I/O). But $400 is still too much to pay for this solutioin.

    Doing something like a double blind subjective study on audio is difficult. People that don't know how to listen won't be any help because even if they hear a difference they won't know how to describe it very well. People who do know what they are talking about are hard to come by in volume. Don't get me wrong, we'd love to do something like this. But we just don't have any idea how to work out the logistics. Suggestions are welcome.

    Also, I appreciate the suggestion to avoid general statements about the goodness of something. It is a good suggestion even in cases where we know everything about everything in detail. There are always surprises and erring on the side of caution is the best way to go. We will be more careful in the future.

    Derek Wilson

Log in

Don't have an account? Sign up now