Creative Labs Soundcard Overview

Discussion in 'Links' started by S_Kinton, Mar 9, 2006.

  1. S_Kinton

    S_Kinton Ancient Guru

    Messages:
    1,884
    Likes Received:
    1
    GPU:
    512mb 8800GT & 22" Samsung LCD
    Sound Blaster
    The Sound Blaster family of sound cards was for many years the de facto standard for audio on the IBM PC compatible system platform, before PC audio became commoditized. The creator of Sound Blaster is the Singapore-based firm Creative Labs.

    The pre-Sound Blaster years
    The history of Creative Labs sound boards started with the release of the Creative Music System ("C/MS") board in August 1987. It contained two Philips SAA 1099 circuits, which, together, provided 12 voices of square-wave bee-in-a-box stereo sound plus some noise channels.

    It is interesting to note that these circuits were featured earlier in various popular electronics magazines around the world. For many years Creative tended to use off-the-shelf components and manufacturers' reference designs for their early products. The various integrated circuits had white or black paper sheets fully covering their top thus hiding their identity... On the C/MS board in particular, the Philips chips had white pieces of paper with a fantasy CMS-301 inscription on them; real Creative parts usually had consistent CT number references.

    Surprisingly, the board also contained a large 40-pin integrated circuit, bearing a CT 1302A CTPL 8708 serigraphed inscription and looking exactly like the DSP of the later Sound Blaster. Presumably, it could be used to automate some of the sound operations, like envelope control.

    A year later, in 1988, Creative marketed the C/MS via Radio Shack under the name Game Blaster. This card was identical in every way to the precursor C/MS hardware. Creative did not even bother to change any of the labeling or program names on the disks that came with the Game Blaster.

    First Sound Blasters: the right bundle
    The first board bearing the Sound Blaster name appeared in November 1989. In addition to Game Blaster features, it had a 11-voice FM synthesiser using the Yamaha YM3812 chip, also known as OPL2. It provided perfect compatibility with the competing Adlib sound card, which had gained support in PC games in the preceding years. Creative used the "DSP" acronym to designate the digital audio part of the Sound Blaster. This actually stood for Digital SOUND Processor, rather than for the more common digital signal processor meaning and was really a simple microcontroller, presumably from Intel. It could play back monaural sampled sound at up to 23 KHz sampling frequency (AM radio quality) and record at up to 12 KHz (slightly better than telephone quality). The sole DSP-like feature of the circuit was ADPCM compression and decompression. The card probably lacked an anti-aliasing filter, as it had a characteristic "metal junk" sound. Finally, it featured a joystick port and a proprietary MIDI interface. This interface lacked simultaneous input and output capabilities, so music software had to use eg. the FM synthesizer in order to play the input received from a MIDI keyboard.

    It is difficult to tell what microcontroller was used as "DSP" since not only did Creative stick a black label with a fantasy (C) COPYRIGHT 1989 CREATIVE LABS, INC. DSP-1321 inscription on the top, but also carefully scratched two thirds of the plastic surface underneath. Analysis of the device pinout suggests that it was an Intel 8051 microcontroller with a custom mask ROM. The labels on the FM synthesizer circuit and on the companion Yamaha 3014B digital-to-analog converter said FM1312 and FM1314 respectively, but luckily the manufacturer references remained intact below.

    In spite of these limitations, in less than a year, the Sound Blaster became the top-selling expansion card for the PC.

    The premature usage of the DSP word backfired at Creative when they finally included some real digital signal processing features in later Sound Blaster models and were obliged to coin a new term for them, ASP, for Advanced Signal Processing.

    Sound Blaster 1.5
    Released in 1990 dropped the "C/MS chips". They could be purchased separately from Creative and inserted into two sockets on the board. This change was probably related to Philips having discontinued the design, and to the lack of enthusiasm among users; the chips could be bought mail-order from Creative until 1993.

    Sound Blaster 2.0
    provided better support for multitasking operating systems, presumably thanks to the introduction of its own timer interrupt. It was the earliest Sound Blaster supported by OS/2.

    Sound Blaster MCV was a version created for IBM PS/2 model 50 and higher, which had a MicroChannel bus instead of the more traditional ISA one. It was little used.

    Improved quality: stereo and 16 bits
    Sound Blaster Pro

    The Sound Blaster Pro (May 1991) added stereo capabilities, but not yet at CD quality level, since it still had only 8-bit sampling. The first version of the Pro also used two YM3812 chips (one for left audio channel and the other one for the right one; both chips had to be programmed identically to get sound in the middle). Version 2.0 switched to the improved Yamaha YMF262 chip, also known as OPL3. MIDI support became full-duplex and offered timestamping features, but was not yet industry-standard MPU-401 compatible.

    Sound Blaster 16
    The next model, Sound Blaster 16 (June 1992) introduced 16-bit sampling to the Sound Blaster line. The cards also featured a connector for add-on daughterboards with wavetable synthesis capabilities complying to the General MIDI standard. Creative offered such daughterboards in their Wave Blaster line. Finally, the MIDI support now included MPU-401 emulation (in dumb UART mode only, but this was sufficient for most MIDI applications). The Wave Blaster was simply a MIDI peripheral internally connected to the MIDI port, so any PC sequencer software could use it.
    Sound Blasters with onboard wavetable synthesis

    Sound Blaster AWE32
    The Sound Blaster AWE32 (March 1994) introduced the EMU8000 processor which supported 30-channel wavetable synthesis. The AWE32 didn't use its General MIDI port to access the wavetable module—Creative decided to use a nonstandard port. As with the Gravis Ultrasound, software designers had to write special AWE32 support into their programs. To support older software, the AWE32 still featured OPL3 FM synthesis, and came with the AWEUTIL program which attempted to provide GM/MT32/GS redirection to the native AWE hardware, however the compatibility wasn't great and it used a lot of precious DOS conventional memory. Its usage on Windows was simplified by the fact that Windows 3.1 had drivers which made the FM synthesizer appear like just another MIDI peripheral, on its own MIDI interface.

    The Sound Blaster AWE32 was a full-length ISA card, measuring 14 inches (356 mm) in length. It needed to be this large because of the number of features included (the most available at the time), and the lack of an integrated controller (an ASIC). The AWE32 included two sound processors (the Creative controller and the EMU8000), an analog processing section with resonant filtering, a Panasonic/Sony/Mitsumi CD-ROM interface (for accessing old, non-ATAPI CD-ROM drives which were still in use at the time), 512 KB built-in memory, and two 30-pin SIMM slots (with their own memory controller) for adding sample memory. The AWE32 supported up to 16 MB of additional SIMM memory.

    Before the advent of ATAPI interfaces for CD-ROM drives, three main manufacturers—Panasonic/Matsu****a, Mitsumi, and Sony—each had their own interfaces with differing pin connections. Modern ATAPI drives are connected to the IDE controller (on the motherboard or an expansion controller), but these older drives had to be accessed via special interfaces which were typically included on sound cards at the time.

    Sound Blaster AWE64
    The AWE32's successor, the Sound Blaster AWE64 (November 1996), was significantly smaller, being a half-length ISA card (meaning it was only half the length of the AWE32). It offered similar features to the AWE32 (replacing the old CD-ROM interfaces with an IDE compatible one), but has a few notable improvements, including support for more midi voices simultaneously and a higher polyphony count. However, these additional voices were achieved via software emulation using host CPU resources, rather than being processed on the card. The main improvement was better compatibility with older SB models, and an improved SNR. The AWE64 came in 3 versions: A Value version (with 512KB of RAM), a Standard version (with 1MB of RAM), and a Gold version (with 4 MB of RAM and SPDIF ports). The SIMM slots were replaced with a proprietary memory format which could be purchased from Creative. A fourth version - a PCI version of the AWE64 - was introduced shortly after. It offered the features of the original ISA AWE64, but it had a PCI interface and was built around an ASIC so it had drastically fewer components on the board and ended up being much cheaper than its predecessors. This was followed by the AWE64D, which was a variant of the PCI AWE64 that was developed for OEMs. It offered the same features as the retail PCI AWE64, but has an architecture that was distinct enough to prevent the standard PCI AWE64 drivers from working with it.
    Multi-channel sound and F/X

    Sound Blaster PCI64 and PCI128
    The Sound Blaster PCI64 (April 1998) added four-speaker support (quadraphonic sound). Next in that series was the Sound Blaster PCI128 (July 1998). The PCI64/128 were lighter on features than the AWE series, but were basically one-chip cards that were quite inexpensive and attractive to most people who did not require the advanced MIDI features of an AWE card.
    Sound Blaster PCI512
    The Sound Blaster PCI512 was basically a lower-priced version of the Sound Blaster Live! Series. It also used the EMU10K1 processor but the firmware could not be flashed unlike the Live! Series to accommodate future EAX extensions.

    Sound Blaster Live!
    Sound Blaster Live! (August 1998) saw the introduction of the EMU10K1 processor, a 2.44 million transistor DSP. The EMU10K1 featured DirectSound acceleration, General MIDI wavetable output, EAX (environmental audio extensions, which competed with A3D before the demise of the latter), a high-quality 64-voice wavetable synthesizer, and the FX8010 DSP chip for real-time digital audio effects processing.

    A major design change from its predecessor (the EMU8000) was that the EMU10K1 used system memory over the PCI bus for the wavetable samples, rather than using expensive on-board memory. The FX8010 is actually a 32-bit programmable processor (with 1 KB of instruction memory). Effect algorithms are created by a development system that integrates into Microsoft Developer Studio. The effects are written in a language similar to C, and compiled into native FX8010 object code by its compiler, fxasm.

    The Sound Blaster Live!
    Features higher audio quality than previous Sound Blasters, as it processes the sound digitally at every stage. It has an internal fixed sample rate of 48 kHz, meaning that any recording done at lower sample rates (such as 44.1 kHz or 32 kHz) is first upsampled to 48 kHz and then downsampled. In a production environment with a Sound Blaster Live!, it is generally recommended to use 48 kHz sampling to maintain sound integrity. The later versions of the Sound Blaster Live! featured support for 5.1 surround sound.

    Sound Blaster Audigy
    The Sound Blaster Audigy (August 2001) features the Audigy processor and supports 16-bit play-back at up to 48 kHz (which is upsampled for the 96KHz 24bit DACs), and recording at up to 16-bit, 48 kHz. The Audigy can process up to 4 EAX environments simultaneously, and supports up to 5.1 speakers.

    Sound Blaster Audigy 2
    The Sound Blaster Audigy 2 (September 2002) features the Audigy 2 processor and supports 24-bit play-back at up to 192 kHz and recording at up to 96 kHz. The Audigy 2 supports up to 6.1 speakers and has improved signal-to-noise ratio (SNR) over the Audigy (106 vs. 100). It also features built-in Dolby Digital 5.1 EX (which is technically 7.1) decoding for improved DVD play-back.

    Sound Blaster Audigy 2 ZS, Sound Blaster Audigy 2 Value, and Sound Blaster Live! 24-bit
    The Sound Blaster Audigy 2 ZS, improves upon the Audigy 2 by having a slightly improved SNR (108 vs. 106 dB) and adds built-in DTS-ES (Extended Surround) for improved DVD play-back. The Audigy 2 ZS supports up to 7.1 speakers.

    The Sound Blaster Audigy 2 Value is a stripped down version of the Audigy 2 ZS, with an SNR of 106 dB, no GamePort, and no DTS-ES 6.1 playback.

    The Sound Blaster Live! 24-bit is not actually a member of the Sound Blaster Live! family, but is a further stripped down version of the Audigy 2 Value, with an SNR of 100 dB, no advanced resolution DVD-Audio Playback, and no Dolby Digital 5.1 or Dolby Digital EX 6.1 playback.

    The Soundblaster Audigy 2 NX is an external USB soundcard, supporting 24 bit playback.

    Sound Blaster Audigy 4 and Audigy 4 Pro
    The latest incarnation, the Sound Blaster Audigy 4 Pro improves on the Sound Blaster Audigy 2 ZS by improving the SNR to 113 dB. It serves as a stepping stone between the Audigy 2 ZS and the Zenith.

    Creative X-Fi - Xtreme Music, X-Fi Platinum, X-Fi Fatal1ty and X Fi- Elite Pro
    Based on standard PCI conncetion and has 10,340 MIPS of processing power (for reference, a 3.6 GHz Pentium IV has 10,224 MIPS). This is 24 times more power than offered by the Audigy. The 130 nm core operates at 400 Megahertz. With the X-FI's "Active Modal Architecture" (AMA), the user may choose three optimization modes: games, leisure, and creation. The X-FI will provide a function, called the "24 bit Crystalizer," to convert traditional audio sources into high-definition multi-channel audio. X-FI also includes 24 bit 3D MIDI.

    X-Fi still resamples 44.1KHz content (audio cds, lossy encoded music, lossless encoded music) to 48KHz internally just like its predecessors and still does not have a true 24bit effects engine just like its predecessors.


    For further information on some of the products above, see Creative's website!

    Taken from http://www.reference.com/browse/wiki/Sound_Blaster and then edited.

    Figured this could make a good sticky for people unsure about creative products and their general features and history
     
    Last edited: Mar 10, 2006
  2. S_Kinton

    S_Kinton Ancient Guru

    Messages:
    1,884
    Likes Received:
    1
    GPU:
    512mb 8800GT & 22" Samsung LCD
    Anyone want to add to the X-Fi section coz its a bit thin on the ground considering there are at least 4 models
     
  3. RejZoR

    RejZoR Ancient Guru

    Messages:
    4,211
    Likes Received:
    0
    GPU:
    Sapphire HD4870 Silent
    Some X-Fi info in your post is incorrect...
    The internal resampling and 24bit thingie...
     
  4. S_Kinton

    S_Kinton Ancient Guru

    Messages:
    1,884
    Likes Received:
    1
    GPU:
    512mb 8800GT & 22" Samsung LCD
    Care to explain?
     

  5. ROBSCIX

    ROBSCIX Ancient Guru

    Messages:
    16,241
    Likes Received:
    7
    GPU:
    22" LCD on GTX260 C216
    He's talking about the Band splitting paradigm. Meaning as you said the X-fi's still don't contain TRUE 24/96 DSP effects..so they resample, example: IF you record a 24/96 and use the DSP to process it it would be cut into "chunks" for processing at the DSP's internal frequecies. It is then reassembled into a 24/96 stream using SRC's Creative Labs states you cannot hear any artifacts or such, I have not tested this process myself. IN short while the card brags true 24/96 operation it in fact does not offer that if the DSP is used. Which is kind of a let down in my opinion. Creative has been using resampling DSP's for a long time. This is a theoretical way around the problem but it is still resampling none the less. Also you must consider your audio is being resampled from Near DVD or Dat quality down to CD quality then back up so you still have failry good quality, still another step in the chain. To audio purist it's a no,no.
     
    Last edited: Mar 9, 2006
  6. RejZoR

    RejZoR Ancient Guru

    Messages:
    4,211
    Likes Received:
    0
    GPU:
    Sapphire HD4870 Silent
    I'd say a "transparent SRC". It's still light years ahead Live! or Audigy 2...
     
  7. ROBSCIX

    ROBSCIX Ancient Guru

    Messages:
    16,241
    Likes Received:
    7
    GPU:
    22" LCD on GTX260 C216
    True, Your absolutley right, it is a great advancment for PC based audio. I was just answering a question. BTW the term "transparent SRC" is a contradiction in terms.
     
  8. izikog

    izikog Maha Guru

    Messages:
    1,126
    Likes Received:
    0
    GPU:
    Sapphire 6950@70 910/1400
    the x-fi is not based on PCI Express x1 but on standard PCI... as explained on creative site pci express is not good for audio hardware!
     
  9. ¬69er

    ¬69er Master Guru

    Messages:
    203
    Likes Received:
    0
    GPU:
    Ge5900Ultra
    My X-Fi does not resample :nerd:, not at CD playback. There is a clock option there, if set to 44.1 no resampling is involved.
    Read up here guys
    Is that true??? got the link?
     
    Last edited: Mar 10, 2006
  10. izikog

    izikog Maha Guru

    Messages:
    1,126
    Likes Received:
    0
    GPU:
    Sapphire 6950@70 910/1400
    I will find the article again but there is no need in your sistem spec I see that you have the x-fi so if you installed it in your pc you know that you installed it in the standard pci slot as I have mine


    EDIT:
    here is the link from guru and the text

    http://www.guru3d.com/newsitem.php?id=3005


    Last week we published our review of the SoundBlaster X-Fi Fatality soundcard from Creative Labs. One of the questions that remained to be answered was why Creative decide to go for a PCI only based soundcard, we found out that in the near future there will not be a PCI-Express version of the sound card. Steve Erickson, vice president for Creative's audio products was kind enough to answer this particular question specifically, here's his take:

    As far as PCI Express (PCIe) is concerned, which is the next bus, what we found is that the performance of PCIe is truly bad for audio. We are seeing four times degradation on the bus for audio.
     
    Last edited: Mar 10, 2006

  11. ROBSCIX

    ROBSCIX Ancient Guru

    Messages:
    16,241
    Likes Received:
    7
    GPU:
    22" LCD on GTX260 C216
    Good to know. I figured the wouldn't use PCI-X for the newer cards becuase they didn't need that amount of throughput...I guess there is other reasons.
     
  12. RejZoR

    RejZoR Ancient Guru

    Messages:
    4,211
    Likes Received:
    0
    GPU:
    Sapphire HD4870 Silent
    The info about X-Fi cards is still totally wrong...
     
  13. ROBSCIX

    ROBSCIX Ancient Guru

    Messages:
    16,241
    Likes Received:
    7
    GPU:
    22" LCD on GTX260 C216
    @RejZor, AFAIK the X-Fi does resample internally. They may have changed the name calling it this or that and the may have a advanced algo doing the job BUT it's just resampling. This is why 70% of the X-fi's power is for Sample Rate Conversion or SRC...
     
  14. RejZoR

    RejZoR Ancient Guru

    Messages:
    4,211
    Likes Received:
    0
    GPU:
    Sapphire HD4870 Silent
    That still doesn't change anything. The "history" list above still says something like "it still resamples and it's still crap" while thats far from true. God knows what other soundcards do with the sound and no one ever makes a big deal out of it even if they resample it from 44.1kHz to 8kHz and then back...
    Besides, even if X-Fi does resample, it was still light years ahead of any other soundcard in ANY test done to date. That does tell someting about how good it does at it's "resampling" doesn't it?
    And X-Fi also sports true 24/96 or 24/192 playback and true up to 24/96 recording. Only soundcard with crappified 24bit unit was Audigy 1 which is history anyway...
     
  15. ROBSCIX

    ROBSCIX Ancient Guru

    Messages:
    16,241
    Likes Received:
    7
    GPU:
    22" LCD on GTX260 C216
    Well no because Creative kinda double talks and ensure people that it's a good algo etc..either way it still does it. I don't think becasue people don't know about it then it's OK? This whole problem with CL cards is this internal resampling. Anyway I am not down on the X-Fi by any means.
     

  16. RejZoR

    RejZoR Ancient Guru

    Messages:
    4,211
    Likes Received:
    0
    GPU:
    Sapphire HD4870 Silent
    So, how sure can you be that X-Mystique doesn't do it too? Or X-Meridian? Maybe Realtek, VIA and Intel HD solutions? No one assures this and no one will admit it. Can you be sure? I don't know how info about Creative cards doing internal resampling "leaked out" but honestly if sound levels are so high as with X-Fi, i really don't care at all even if it resamples 50 times in between. And i also don't remember Creative admiting it or anyone else saying that X-Fi still resamples. 7000 mips dedicated to SRC is just a marketing talk to show how powerful it is, nothing else. Show me some real technical data or any kind of good prooof that it still resamples and i'll reconsider it. So far i haven't seen anyone claiming that X-Fi resamples. So most probably it doesn't. I'm sure people would make a big deal out of it since it was a big "problem" of all former Creative soundcards...
     
  17. ROBSCIX

    ROBSCIX Ancient Guru

    Messages:
    16,241
    Likes Received:
    7
    GPU:
    22" LCD on GTX260 C216
    Might aswell leav this one alone. Go out and do some research you should find the same info I did. Same as it claims true 24/192 effects when the DSP is only 44.1 or 48 so it's broken up 48 X4 and effects are applied to each part and then it's reintegrated and Creative calls this "True" . As I said I am not down on the X-fi not at all, just want you to know as it seems you do not. Why would it be marketing that 70% of the awsome power of the X-fi processor is used for SRC -sample rate conversion. Why would they waste all this power if the card doesn't resample? I am sure it's great and your not supposed to hear any added distortions or such. Just pointing something out. I considered buying a X-fi but I didn't becasue I just bought a Audigy 2 Value. I am not trying to down play your card Rej, not at all this is just the info I got. It is in no way "Leaked" the info was on the release sheets on creative site... The Band splitting pardigm etc...all boils down to 44.1 or 48KHz internal clock....meaning the X-fi's DSP only runs at these frequencies. some people get confused as there is alittle drop down box that allows other frequencies...this is just multiples of the origial 44.1 or 48khz..e.g.
    44.1X2 is 88.2Khz, 48Khz X2 is 96Khz. This doesn't mean the X-fi won't sound good in fact the opposite - these technologies are a new concept in digital audio and they were put into place for a few reasons according to Creative Labs. You can believe what you like...
     
  18. RejZoR

    RejZoR Ancient Guru

    Messages:
    4,211
    Likes Received:
    0
    GPU:
    Sapphire HD4870 Silent
    Actually after some reading i noticed that you HAVE to resample in order to synchronize different freq. sounds played simultaneously, otherwise you'd get borked up pitch if you'd want to play 8kHz sound at 48kHz rate or vice versa... It all goes down to how good you do that. Considering splitting of streams or not, the DACs, even the most high grade ones cannot handle the difference, let alone human ear noticing it.
    Besides, even if it does resamples, he should dedicate some more than just 3 lines for it... Saying that it does resamples but that it's so good at it that it doesn't really matter if it does that. But instead i'd get (as regular user) feeling that it's the same crap as SB Live! judging by the text... even though it's light years away from that fact...
     
  19. ROBSCIX

    ROBSCIX Ancient Guru

    Messages:
    16,241
    Likes Received:
    7
    GPU:
    22" LCD on GTX260 C216
    True, resampling may be needed in some circumstances BUT when it is being used when it is not needed that makes people worry about. If your playing 1:1 CD copies or DVD audio you do NOT want any messing with the streams. It goes through is resampled to 44.1 or 48Kz then resampled back up to 96hz or 192hz this introduces distrotion to the mix etc. It is totally frowned upon by the audio community it doesn't matter how good it is being done it should NOT be done in the first place. Sure for game audio such things are common and no big deal. This is why Creative cards have never been considered good for recording and any type of audiophile tasks even listening with high quality headphones or HT gear etc.. The idea that it doesn't really matter that it does resample is a matter of opinion and not a fact. Some people won't touch it with a 10 foot pole becasue of this. Creative puts nice spins on things to tyake your attention away from the fact. The X-fi's are great cards and the are top for what they are good at. Creative is notorious for changing specs and fudging things up and passing off hardware advertised to do one things when in fact it could never do it. Hey Rej, don't get me wrong man, I like the X-fi's but I also understand exactly what they are about. For games though they are great...BTW, Rej did you check your PM's?
     
    Last edited: Nov 26, 2006
  20. ¬69er

    ¬69er Master Guru

    Messages:
    203
    Likes Received:
    0
    GPU:
    Ge5900Ultra
    Resampling is necessary, needed and good when it is done properly. When it is not done properly it causes problems.

    With X-Fi resampling is done properly so you can't hear it.

    If you do not want to resample AT ALL you can always switch to AUDIO CREATION mode, tick the option CLOCK and set 44.1 or whatever your source is. That's so with X-Fi that I have.

    Saying that "X-Fi still resamples 44.1KHz content to 48KHz internally just like its predecessors" is completely wrong.

    Talking that X-Fi is bad cus it resamples is double :boob: wrong.
     
    Last edited: Nov 27, 2006

Share This Page