Jump to content

TwoRocks

  • Posts

    26
  • Joined

  • Last visited

  • Country

    country-ZZ

Retained

  • Member Title
    Newbie

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. @ Juergen: I don't mean to drag this "hearing the least significant bit" discussion out, but there's a difference between measuring it and being able to hear it. I have no doubt that you've been able to measure the dither in the 24th bit. But hearing something that's at -144 dBFS when the SNR of the audio interface is only 112 dB (Tracker Pre), 117 dB (0404 USB), or 120 dB (1616m) respectively is a physical impossibility because the dither signal is 24 dB or more below the noise floor of the interface. I'm not aware of a single DAC that has a SNR of 144 dB (RMS) or better. Until such an interface hits the market, discussions about dither noise @ -144 dB are purely academic, IMO. I also don't see how the jitter you've measured with WaveOut and DirectSound could degrade the playback quality of the analog signal at the output of the E-MU DACs, as the audio data stream gets buffered (= asynchronous operation) in those E-MU USB audio interfaces, so the only jitter that can have an effect on the sound quality is the jitter of the unit's internal clock, and that clock is -- as you've measured and posted here yourself -- very stable with jitter below 200 ps. And that same internal crystal clock circuitry triggers the D/A conversion all the time, no matter whether the audio data gets sent over ASIO or through the WDM audio stack. There may be jitter "on the way to the buffer", but at that point it gets "thrown out". That's the main argument for asynchronous operation, not just for these "beer budget" interfaces from E-MU, but for those from the more expensive audiophile companies, as well. Re. DxD: not to go too far off topic here, but I share your excitement about that recording technology. ) It's "the best of both worlds" (PCM and DSD). While DSD may have gotten very close to analog reproduction in straight tape-to-DSD transfers or direct recordings (transients look unbelievably good), DSD cannot be used in the mixing and mastering process (because of its single-bit nature), and having to convert things to PCM for the mixing and mastering and then back to DSD for the final SACD leads to audible artifacts so that end results are sometimes poorer in quality than "pure PCM" all the way. DxD, on the other hand, is multi-bit, very much like traditional PCM, but at a sample rate of 352.8 kHz! (8x that of Redbook CDs!) I've looked at some graphs of transient signals recorded with DxD, and the results are practically as good as what DSD delivers in that regard, but without any of the downsides I just mentioned. Finally we are getting "full analog quality" in the digital realm (without any of the shortcomings of analog, like tape hiss and such...)! Yay! Just like one of those Pavlovian dogs, I start to salivate every time I dream of E-MU's "next big audio interface" being a DxD (and PCM) model. But AFAIK, none of the "top dog" DAC chip manufacturers (like Cirrus Logic and AKM) has released a DxD IC, yet, and that unit from that Danish company that's built with discrete components is just too friggin' expensive for somebody on a lunch money budget. (Yes, I've already exhausted my beer budget for the next 12 months, so I'm down to my lunch money... LOL) Okay, back on topic, folks... LOL
  2. @ CharlyD: It was good to see that most of the responders to this thread have a hard time distinguishing if the bits are going through WDM or ASIO. Looks like Microsoft did a pretty good job with their sample rate coverter. Nope. The Microsoft sample rate converter does a pretty lousy job. The trick when using WDM drivers is to give KMixer nothing to do, keep it from doing sample rate conversion and let it pass the audio straight through to the audio device's WDM driver. In order for that to happen, the audio interface must be set (at least in the case of those E-MU USB audio interfaces) to the native sample rate of the track(s) being played before starting playback, plus the master volume in the Windows mixer must be all the way up, plus the balance setting must be smack-dab in the center, plus there must not be any other audio streams. (Use the "mute" checkbox in the Windows mixer for everything but "Wave".) While not all of that is necessary to prevent sample rate conversion, all of it is necessary if you want to keep the original "audio bits" mostly unaltered (except for some minor alterations that happen in the least significant bit of each sample when using WDM, anyway, despite giving KMixer nothing to do). Then, and only then, does the original material reach the converters of the audio interface almost unaltered via WDM (except for that dithering in that least significant bit that Juergen identified in his tests). @ vortecjr: While there are different versions of WAV files (more about that in a moment), the trouble you are encountering with your player most likely has to do with the sample rate and/or bit depth of the material, not with the WAV format itself. I refer you to my previous post and my experience with those AIFF files with different resolutions. The same applies, in principle, to WAV files: your player may have no problems with 44.1 kHz material @ 16-bit resolution, but may be unable to properly play WAV files with 192 ksamples/second and 24-bit resolution. Now re. the actual WAV "format": WAV is actually a so-called "container format", and it can "contain" -- hence the term "container format" -- a whole bunch of different audio formats internally. Most of the time when people refer to WAV files, they are talking about "pulse code modulated" (PCM) audio material in stereo or mono with varying sample rates and bit depths. However, multichannel (e.g. 5.1) PCM WAV files exist, as well. If you try to play those in a media player that is not equipped to handle such a multichannel audio data stream, all you get is -- if you are lucky -- white noise. I've tested that with a six-channel WAV file in Windows media player. The same track played fine in full surround beauty in VLC player. (Again: nifty, capable little player...) The audio data inside a WAV container can even be DTS-encoded. That's essentially the same as what you get with those few DTS-encoded CDs, like the Eagles' "Hell Freezes Over". Re. your S/PDIF question: none of the E-MU USB audio interfaces (and barely any other audio interfaces on the market) can handle more than 96/24 stereo material over S/PDIF. E-MU's current top-of-the-line PCI (not USB) audio interface, the 1616(m), supports up to 192/24 stereo material over S/PDIF, but it does not do the auto-switching of sample rates that E-MU's USB audio interfaces are capable of ( because of a different architecture = built-in hardware mixer). One more nifty feature of the E-MU 0404 USB is that it supports AC-3 and DTS pass-through over S/PDIF, so if you have any material encoded in those formats on your computer's hard drive, you can send it to your A/V receiver's S/PDIF input (or an external multichannel DAC with a S/PDIF input, if such a thing exists). 192/24 output only works over the analog outputs of E-MU's USB audio interfaces, and only when connected to a PC. On a Mac, playback is currently limited to 96/24 over the analog outputs and to 48/24 over S/PDIF.
  3. I hereby declare victory... ) Got those excellent Kent Poon recordings (thanks again for the link, Joel), and even @ 192/24 resolution, they are playing perfectly in WimAmp via the Japanese ASIO driver -- no crapples. (For those who haven't read my previous posts: that's short for "crackling, clicks, pops" and other assorted nastinesses.) However, I can only second Joel's warning to CharlyD: when dealing with new material/formats and/or players, use headphones and/or low volume, first. Seeing that those Kent Poon recordings were AIFF files, not WAV, and I hadn't played any AIFF files in WinAmp before, I turned the volume down, down, down when I played the first track with 44.1 sample rate, and it played fine. I thought: "Great! Another format that WinAmp can handle natively!" So I turned the volume up again (thankfully not all the way), loaded the same tune at 192/24 resolution, and I practically jumped out of my armchair... WHITE NOISE! (Or was it pink? Lavender? LOL) I'm just greatful that the player wasn't spitting out brutal transients, otherwise I would be shopping for new speakers now... Then I thought: that evil Joel ;-) probably hasn't revoked his jinx spell of my WinAmp player, yet... But the far less dramatic reality is that WinAmp simply does not like AIFF files with 96/24 or 192/24 resolution. Not one bit. (Pun intended.) Not via ASIO, not via WDM. (Windows Media Player fails here, too, but the VLC player can handle the format, albeit via WDM... amazing little player, that VLC...) I got around the problem by loading the AIFF files into Audacity (free audio editor, very useful little tool -- even runs from a USB stick) an re-saving them (no sample rate conversion or resolution change) as FLAC files. And those play fine in WinAmp, all the way up to 192/24. (Audacity can handle 24-bit WAV files, too, if your player does not like FLAC files.) Playlists with mixed resolutions (44.1/96/192) work perfectly, too -- the ASIO plug-in makes the E-MU switch bit rates just fine. So I'm having my cake, and I'm eating it, too! ) And as that still seems to make me the only man on the planet who's had success with this, maybe somebody should call Guinness. ;-) (Okay... somebody can just buy me a Guinness... LOL) I doubt that my success has anything to do with using a different E-MU USB audio device (the Tracker Pre). My "money" is on a lean, clean OS configuration. Turning all processes and Windows services off that are non-essential for audio playback can make the difference between a computer that has a chronic case of "the crapples" and one that doesn't. Maybe my powerful CPU (Core 2 Duo @ 3.6 GHz) helps, too, but I kinda doubt that, as well, as CPU utilization is only around 8%, even when playing those fine 192/24 tracks. I have, however, set the buffers in that Japanese ASIO plug-in to the maximum value allowed (63), and thread priority to "time critical". Maybe that will help some people here to have success with WinAmp and ASIO, too... Sooo... who's got more links to free hi-res downloads? BTW: those Kent Poon recordings really show what even Redbook audio @ 44.1 kHz sample rate and 16-bit resolution is capable of, when done right. While I can hear a difference between 44.1/16 and 96/24 resolutions of the same track -- a teensy bit more detail / clarity --, the differences are minute, and the recordings sound exquisite even @ 44.1/16, which goes to show how lousy most audio is being produced today. (And don't get me started on the "loudness war" topic... wha!) Oh, one last thing, for the record: I could claim now that I suddenly have the hearing of a bat all the way up to 90+ kHz (yes: I'm the bat man! LOL), but I honestly cannot distinguish between the 192/24 and 96/24 resolutions of those fine tunes, and not between WDM and ASIO playback at 192/24, either. That dithering noise in the 24th bit @ -144dBFS just keeps eluding me... )
  4. Joel, thanks for the download link to those free 192/24 tracks. I'm gonna download them now and will test my setup with them in a little while. Just don't jinx it in advance, okay? ;-) LOL It may just work fine! I insist. LOL And thanks for sharing your story of auditioning and testing the 0404 at your friend's place (again.) (Man! If this thread weren't so fraggin' long, I'd read through it from start to finish... but there are only 24 hours in a day, and I've already added the night to that, so I'm up to 48 hours, and it's still not enough time... ;-) ) 400k audiophile system! Great maker! Impressive that the 0404 stood its ground in comparison to the Apogee gear. Not surprising to me, though. Now some may ask: why? Because the senior analog and digital circuit designer at E-MU who designed some (or all) of their audio interfaces, including their top-of-the-line 1616m model, used to work for Apogee before he joined E-MU... That's why you get Apogee quality at E-MU prices with those mighty fine "beer budget" interfaces... :~) I think I can see some expensive audiophile gear makers start trembling in their boots... LOL
  5. @ Harald: Thanks for testing. I didn't think of going the FFT test route. Good thinking. The lack of any high frequency content when playing the 88.2/24 transfer of the Beck album when using WDM confirms it for me: the 0404 USB locks the sample rate to what's being shown in the display of its control applet. So it's confirmed then: no auto-switching of the E-MU USB audio interfaces with WDM. ( TwoRocks needs therapy now. Cries in pillows. Not a pretty picture... LOL Joking aside, though... would you do one final test for us, set the sample rate in the E-MU control applet to 88.2 kHz before launching your media player, then play that same track from the Beck album again using the WDM driver and do the FFT test, once more? I'm pretty dang sure that you will get the HF content this time around... In the same context: As I see it, the 0404USB driver locks on/to the first samplerate it receives (on a stream off a player like foobar or, maybe MediaMonkey), and the rest - even if at different samplerates - get converted to this "first setting". No. As your FFT test has shown quite clearly, the E-MU USB audio interfaces don't do auto-switching of the sample rate when the WDM drivers are being used. But the devices don't sync to the sample rate of the first track they get to play, either. They simply play at the sample rate that was set / showing in the control applet's display before launching the media player and starting playback. That's why I asked you above if you would do that one final test at a setting of 88.2 kHz. That also means, of course, that KMixer is always doing its horrible, horrible sample rate conversions to the sample rate that's locked in for the E-MU audio interface if the locked sample rate is not the same as the native sample rate of the respective track. Which means for me: going back to WinAmp with that Japanese ASIO plug-in... The horror! Run for the hills! The world is coming to an end! ;-) LOL One last thing for now, regarding not being able to send anything over the S/PDIF output of the 0404 when using WDM for playback: I would be surprised if that were not possible, at all. I believe a number of people here have used the 0404's S/PDIF outputs to route audio material from their computers to an external DAC, not needing / using ASIO drivers for that... In the words of my favorite "defective detective", Adrian Monk: "I may be wrong now. But I don't think so." ;-) Anyhoo... thanks for all your testing, Harald Gruß ins schöne Frankenland! )
  6. Hi Harald, thanks for your offer to take some measurements of the 0404's analog outputs. I doubt, however, that such a test would tell us what the actual sample rate is at which the 0404 operates... The easiest test method would be using some dedicated test device (maybe like the Audio Precision unit that Juergen has) that just gets the feed from the 0404's S/PDIF out and shows on a convenient display: running @ 44.1/16... running @ 96/24... running @ 384/32... LOL And while we're waiting for conclusive results, we can "bless" this-here fine forum with some "Germlish"... Jawohl!
  7. Juergen, thanks for elaborating further regarding "what gets done to those samples" when using DirectSound and WaveOut. That -178 dBFS number is helpful, as it means that the noise "added" by using WaveOut (and the dithering that goes with it) gets totally "buried", so I'm not concerned about some minute extra noise that's far below the noise floor of the overall interface. But I think we should probably stop digging deeper into the technical side of things here, as we might otherwise give some people here a headache... LOL Now, re.: for me (non native English) it is a little bit difficult to describe with words Geht auch auf Deutsch... :-) Aber wohl besser nicht hier ;-) Okay, back to English... I totally agree with you re. the 1616m... I love that thing, too. It's really my "audio interface of choice" and the "cream of the crop" audio interface for professional recording (and audiophile listening) -- and so stunningly affordable, considering the quality of its circuitry. Even those very expensive DigiDesign ProTools HD 192 I/O units don't have better ADCs and DACs. I would have long gotten myself a 1616m if the price differences between the U.S. and Europe (Germany) hadn't been so ludicrous: if I had still been living in the U.S. some 1.5 years ago (I moved back to Germany 2 years ago), I could have (and would have) gotten myself one for $350, but the price in Germany at that time was the equivalent of about $700! Nuts! So I got myself the Tracker Pre as a decent interim solution until the European E-MU prices for their top end audio interfaces "come down from the Stratosphere", relatively speaking. Now E-MU has announced a new PCI Express version of all their current PCI audio interfaces, something that promises to eliminate some -- or all -- of the PCI-related problems people encountered with those interfaces in the past. Once those hit the market, I'll probably buy one in the U.S. and have it reshipped to me through a friend. Now back to the topic of auto-switching sample rates. There's a fundamental difference between E-MU's PCI and USB audio interfaces. All PCI products come with that powerful digital signal processor and the associated "PatchMix" application -- essential a "patchbay and a mixer" on a chip. That's very useful for professional audio recording, but actually gets in the way of "lazy audiophile listening", as auto-switching of sample rates does not work with any of E-MU's PCI products, at all -- not with WDM, not with ASIO. All of E-MU's USB audio interface auto-switch sample rates when using ASIO. The big remaining "mystery" at the moment is: does the display of their little control applet "tell the truth" (= sample rate is locked, which would violate standard WDM driver behavior), or do we get auto-switching in accordance with common WDM driver design guidelines / standard behavior and is the display just not getting updated accordingly? Stay tuned for our next episode... )
  8. Joel, while the Tracker Pre can be USB bus-powered, it doesn't have to be. It has a 5V DC input, so if you feel the power drain is too much for your notebook computer, you can just use an external 5V DC adapter, instead. It doesn't come with one, but any ole decent unit from the local Electronics Barn will do. (No, wait... I meant to say: Computer Shed... no, that ain't it, either... LOL) Now for the 0202 USB: if you have no need for either digital I/O, or powering the audio interface via USB, the 0202 USB is very similar to the Tracker Pre, purely from an audiophile playback standpoint. The DAC chips are exactly the same as in the Tracker Pre, SNR is identical, and the clock is as rock solid as that of the Tracker Pre: < 100 ps. And it's even cheaper than the Tracker Pre: audiophile listening pleasure for a hundred bucks! (Or: "on a lunch money budget", if you prefer... almost a scary thought, ain't it? LOL) Re. the clicks, pops and crackles (I just call them "the crapples", for short... LOL) -- I had a lot of stuff running on my machine yesterday, including a dozen open browser windows, and a bunch of other apps, when I ran the test and suddenly got "a case of the crapples". I'll run another test in a few days, with just the basic OS (XP 32) and the player (WinAmp) running. It would be nice if I could also put that Japanese ASIO driver through its paces with some 192/24 material, as I'm currently "maxing out" at 96/24. Can anybody point me to some good, free 192/24 downloads? Doesn't matter whether it's mainstream stuff, Jazz, or Classical music. There must be some online stores out there, catering to the audiophile crowd, offering free samples to whet people's appetite for their catalog... (In a pinch, 176.4/24 will do... LOL) @ CharlyD Re. that loopback test to determine the actual sample rate: setting the clock to external won't work. When set to external, the 0404 tries to sync to an external clock signal, like from the S/PDIF output of an upconverting CD / DVD player or something like that. When you do the loopback thingie and set sync to external, the unit has to try to sync externally to its own internal clock... nice challenge! LOL For the WDM auto-switching test of the sample rate, I suggest you set the clock to internal, leave the sample rate setting in the control panel applet at it's standard 44.1 kHz, then play a 44.1/16 tune first, and then some 48/24 material. That should suffice to determine whether the unit does switch sample rates automatically when using WDM drivers, or not. If it does, we all can forget about that Japanese ASIO plug-in for WinAmp, once and for all, and run WinAmp henceforth via DirectSound or WaveOut with all the conveniences that come with it... unless your ears can hear the difference in that least significant bit of a sample that gets altered, and that would spoil your listening experience... ;-) BTW: The fact that those bundled apps all require ASIO does not get in the way of loopback testing, as you should be able to do the playback with WinAmp via WDM and the recording in, for example, Wavelab Lite via ASIO. That should be totally possible. What I don't know is whether Wavelab auto-detects the sample rate of the S/PDIF input channel it's set to record from... ooops... Let us know how it goes!
  9. Have you ever compared that Tracker Pre to the 0404? Looks like the Tracker Pre can do everything the 0404 can, just some different chipsets.... Purely used as a playback device (=DAC) for a PC (or Mac), the Tracker Pre and the 0404 USB are almost on par. The differences between the DAC chips are small (SNR of those in the 0404 is around 5 dB better), and the clock circuit of the Tracker Pre has even lower jitter than the already excellent < 200 ps of the 0404 USB that Juergen reported: it's < 100 ps! That, sort of, makes up for the inferior SNR... Furthermore, the Tracker Pre draws quite bit less power than the 0404 USB, so it does not require a separate DC adapter and can be USB-powered -- which is neat when you're out and about with a notebook computer to record nature sounds like croaking toads... LOL But, of course, you could also be doing some audiophile listening with said notebook computer, the Tracker Pre and some quality headphones on a long intercontinental flight... ) Just make sure the airline is not "Oceanic" -- otherwise you might get "LOST" ;-) On the other hand, the 0404 USB has a bunch of features the Tracker Pre doesn't have. For example, the Tracker Pre does not have any digitial I/O, no optical or coaxial S/PDIF. That's why you can't use it as a DAC for an external component like a CD or DVD player with S/PDIF out. IOW: you can't use the Tracker Pre as a stand-alone DAC. It always has to be hooked up to a computer to be used as a DAC. The total absence of digital I/O is also the reason why I have no way of testing / measuring at what sample rate the Tracker Pre is actually running. As I mentioned before, with the 0404 USB, one can always grab its S/PDIF output with another device (or possibly even do a local loopback of S/PDIF out to S/PDIF in) and determine the sample rate that way, or test for bit accuracy. That's not possible, at all, with the Tracker Pre. It also does not have that convenient post-DAC "analog master volume knob" that the 0404 has. It does have a knobbie for adjusting the headphone volume (and a very nice headphone preamp to go with it!), though... Oh, and the Tracker Pre has no MIDI ports, but that's probably not really an issue for the audiophile crowd... That about covers the differences between the two units... As far as those troubles with the 0.67 ASIO plug-in go: as you're not having problems with DPC latencies (I really cannot recommend testing computers meant for audio playback with that tool enough), try cranking up the buffer setting of that Japanese ASIO dll all the way to 63. BTW: I had a "funny" experience after my last post: I was first playing some 44.1/16 material with WinAmp and that plug-in, and everything was fine, and then I loaded a 96/24 track, and all of a sudden I was getting the worst distortion and crackling... you must have jinxed it, Joel... ;-) Well, I did crank the buffer up all the way to 63, and now all is well again in playback land... I frankly cannot say whether the buffer setting had gotten screwed up before I made my last post, and "7" would always have been too low, or whether some conditions on my machine have changed that have necessitated the higher buffer setting. I may have to do some "housekeeping" on my machine... Anyway: try a buffer setting of "63" and see how that works for you...
  10. Hey Joel, re.: Two Rocks----you are the first person in the history of mankind I have seen have success with the Japanese ASIO driver! Call Guinness! Now I feel like the first man on the moon... LOL Seriously, though: I have zero problems with it. I can understand why you find neither Foobar with ASIO nor JRiver acceptable solutions for your record collection and playback needs. I wouldn't want to convert everything to FLAC, either, or having to reset ASIO all the time. Annoying! I have FLAC, APE, WAV, WMV and MP3 files in my collection, with different sample rates and bit depths, and they all play just fine with WinAmp and the 0.67 Japanese ASIO plug-in "talking to" the native ASIO driver of my E-MU Tracker Pre, properly switching sample rates the whole time. I can just lean back and listen to everything completely "unmessed-with", hands-free. No pops, no clicks, or any other audio artifacts. I do, however, have quite a "beefy" system with a 3.6 GHz Core 2 Duo CPU (an OC'd E8400), and I've also optimized my machine for streaming playback. I will be posting such an "optimization guide" for digital audio workstations (and the same things apply to any PC where uninterrupted audio and video streaming is the goal) pretty soon at one of my sites (where I've already posted that "bit accuracy setup guide" I mentioned in an earlier post on this thread) at: http://daw.kickassproject.com/ So Joel (and everybody else): if your WinAmp player did not freeze up completely every time you used that Japanese ASIO plug-in, there may be hope for you, yet, to get it working flawlessly, just like it's working on my system. I wouldn't get into the Guinness Book of World Records no more, as the only man on the planet who managed to pull that off, but I'm willing to sacrifice that... LOL As a starting point, before the guide gets posted, here are some of my system specs and some initial pointers: Asus motherboard with Core 2 Duo CPU @ 3.6 GHz, 2 GB DDR2 RAM @ 800 MHz, several big 7200 RPM SATA-2 hard drives, Windows XP SP3 (32-bit), WinAmp 5.551, Japanese ASIO plug-in 0.67, E-MU Tracker Pre with native ASIO driver. (Don't use ASIO4All -- bad idea in general, and especially if your audio interface already comes with its own ASIO driver...) To eliminate as many sources as possible that could interrupt the audio stream (resulting in those nasty pops, clicks, and what-have-you-nots), nuke all non-vital Windows background processes and services, including virus scanners and the like, and make sure that DPC latencies -- a major source of pops & clicks -- are not an issue. You can get the DPC latency checker here: http://www.thesycon.de/deu/latency_check.shtml If that thing shows anything in the yellow, or even in the red, flawless playback via ASIO is unlikely until you've eliminated the DPC latency trouble makers. See, the thing is: WDM works with much larger buffers than ASIO (at least at default settings), so huge DPC latencies don't cause the same trouble with WDM as they do with ASIO. You can, however, increase the ASIO buffer size of that Japanese ASIO plug-in, by selecting it in the WinAmp options / preferences and then clicking on "configure". I have my buffer set to "7", and it's working fine, but then again: my system is "lean and mean"... LOL Whew... long post... Just one last thing, to get back to that sample rate switching with WDM: I still suspect that the E-MU WDM driver might auto-adjust sample rates just fine and simply not reflect those changes in the display window of its little control applet. IOW: I don't trust what the control applet shows in its display. Why? Because locking the sample rate would go against how negotiation of capabilities and switching sample rates usually works within the WDM audio stack, and I find it hard to believe that E-MU would intentionally go against common driver design guidelines. That's why I'm still hoping that somebody like Juergen can confirm that it either auto-switches -- or doesn't -- with the WDM drivers by feeding the E-MU source material of different sample rates (say: 44.1/16, 48/16, and 96/24), configuring it to output things over S/PDIF and then test with an external unit whether the sample rate at the E-MU's S/PDIF out remains locked at the rate the control panel applet displays, or whether it changes... (Yeah, I'm persistent, I know... but that's the engineer in me: "Don't trust. Verify." That's my mantra... well, one of them...)
  11. Man, Juergen, we seem to be posting practically simultaneously here at the moment! LOL Thanks for your input regarding sample rate switching when using WDM. How did you verify that the E-MU uses one fixed sample rate and does not switch when the player program sends audio data with a different sample rate to KMixer? Also by checking via the 0404 USB's S/PDIF output? I'm asking because that contradicts Microsoft's documentation regarding how KMixer works. (Not saying you're wrong, just that there seems to be a contradiction.) KMixer, according to MSDN, queries the audio interface's device driver every time a new sample rate "comes into play" to find out whether the device can play back that material, and unless E-MU programmed their WDM device driver in a non-standard fashion, it should affirm all sample rates it can play and switch things accordingly, even if that change is not being reflected in the display of that little control applet. When I did my A/B listening tests of DirectSound vs ASIO a while ago, I manually set my Tracker Pre to the proper sample rate of the source material (96/24) before launching WinAmp (when using DirectSound). Not seeing a change in the E-MU control applet display when the sample rate of the source material changed when using WDM, but seeing the change properly reflected when using ASIO made me pick ASIO over DirectSound (because of the "extra convenience"), especially because I did not and do not have any issues with that Japanese ASIO plug-in for WinAmp. However, others might not be so lucky with that ASIO plug-in, so being 100% certain about whether the 0404 USB auto-switches sample rates, or not, when using the WDM driver seems like a vital piece of information for quite a few (potential) 0404 USB users.
  12. Thanks for the additional info, Juergen. I don't want to take this into a "territory" that is too technical for the fine audiophile crowd here, either, so to wrap things up and just confirm that I understood you correctly, it's something like this, yes? Original sample: 111111111111111100000000 Same (16-bit) sample via DirectSound: 1111111111111111x0000000 Same sample via WaveOut: 11111111111111110000000x (with x indicating the bit that's getting changed) If that's how it is, I'm not surprised I'm not hearing any difference between DirectSound vs. ASIO. I can totally live with that least significant bit being different... not that I would have to, as I'm using ASIO, anyway... ) P.S. One little extra nugget of information for those who might not be aware of this: the DACs of all E-MU USB audio devices always operate at 24-bit resolution. If the audio data stream only provides 16 bits of resolution, bits 17 through 24 (the eight least significant bits) of a sample are simply set to zero.
  13. @ CharlyD Sorry to hear you have problems with the ASIO plug-in for WinAmp. If you downloaded the 0.70 version, you might want to try 0.67. That's what I'm using, and it works without a hickup for me. I just copied the .dll file into WinAmp's plug-in folder (no .exe) and then selected it in WimAmp's output settings. My answer to your question "Is ASIO needed?" is: "Most likely not." LOL I've looked up the KMixer documentation on MSDN again, and I believe your understanding of how KMixer operates and how you describe it in your post is 100% correct. As I stated in my little guide: as long as you give KMixer nothing to do, the output bit stream of WinAmp should get passed straight through to the E-MU USB audio device. The one vital point that E-MU has not confirmed, so far, at least not anywhere I've been looking and lurking, is: The 0404 DAC will then automatically lock to whatever sample rate it is presented but does not report any changes to the USB Contol Panel. My ears -- via Mackie active studio monitors and/or Sennheiser HD 280 Pro headphones -- tell me: yes, it automatically adjusts to the sample rate of the input audio stream. I can't hear a difference between ASIO playback and WDM playback, either, but I can hear a difference between different resolutions of the exact same studio source material. (Such stuff is hard to come by, but I got it in the form of the special two disc edition of John Mellencamp's latest album "Life, Death, Love And Freedom". The standard CD obviously contains the tracks in standard 44.1/16 format. The "bonus" DVD contains the exact same tracks as 96/24 wav files -- no changes to the mix, just higher resolution. The only way to compare "apples and apples", in my book...) That leads me to conclude that source material of different formats is automatically played back at its respective native sample rate by E-MU's USB audio devices, no matter whether ASIO or WDM is being used, and that in the case of using WDM, only the device's control applet display does not reflect such changes. I would still like confirmation for that from E-MU, though, so I've posed that question to E-MU's developers at the "unofficial E-MU forum" (where I hang out a lot... LOL): http://www.productionforums.com/viewtopic.php?f=126&t=10641 It will be interesting to hear what they'll have to say in this matter. (And I'll bug them until I get an official answer... LOL) So, long story short: you should be able to get bit-accurate playback with the 0404 USB (and E-MU's other USB audio devices) using either WDM or ASIO. The one advantage that ASIO offers is that you don't have to "jump through quite as many hoops" as with WDM to eliminate things that might mess up the original bit stream. But if you take all the precautions to give KMixer nothing to do, the audio should sound equally pristine, no matter what type of driver you use. In both cases, it's important to turn everything off in the media player (e.g. WinAmp) that might alter the source material. BTW: thanks for the tip to uncheck "Enable Volume Control". I'll incorporate that in my little guide. ASIO is definitely the preferred choice, though, when doing professional audio recording, as the buffers of the WDM audio stack cause considerable playback delays that can make it difficult for performers to record a new track in sync with something that's being played back. Using the ASIO drivers, instead, can bring those delays (ASIO "latencies" is the actual technical term) down to 2 ms with a powerful enough digital audio workstation. That's roughly an order of magnitude better than with WDM... ) Update: I just caught Juergen's latest post right after making this post, and even though I can't hear any differences between using WinAmp with DirectSound and ASIO, auditioning 96/24 material, I trust his long experience in this matter that ASIO is the only way to get bit-accurate playback. How subtle those differences between DirectSound and ASIO may be and whether they are audible, at all, when dealing with 24-bit samples is another matter all together...
  14. @ Juergen First of all, thanks for pointing me to "Specifying the Jitter Performance in Audio Components". Got it. Another "fun read" for tonight to put me to sleep... LOL Also thanks for pointing out that DirectSound does "mess with" the original sample, even when the volume is all the way up. That was news to me. One follow-up question(s) re. DirectSound, if you don't mind. You said that... The Direct Sound driver in Windows is also not Bit True and shifting the signal one bit down (even when having the volume all up) Would you give us a concrete example? Because it can't be the usual binary bit shift -- e.g. all bits "one bit to the left" or "one bit to the right" -- as that would either cut the volume in half or double it. Example: Original sample (MSB to LSB): 111111111111111100000000 Bit-shifted sample (one to the right / "down"): 011111111111111110000000 So how does DirectSound alter that sample with the volume cranked all the way up? I would imagine that it "only" changes (something in) the least significant bit(s) and does not make a bit shift of the whole sample... Unfortunately, I cannot test these things myself at the moment, as I only have the 0404 USB's "little brother" -- the Tracker Pre -- which does not have a S/DIF output the signal of which I could feed into the S/PDIF input of another device to record the bit stream to compare it with the original... ( P.S. I have had no issues whatsoever with that Japanese ASIO plug-in 0.67 for WinAmp, and I've been using it for quite some time now. Guess I'm one of the "lucky ones"... ) P.P.S. To avoid possible confusion: when Juergen compared WDM and DirectSound, I'm sure he meant to say "WaveOut" vs. "DirectSound". Both are part of the WDM (Windows Driver Model) architecture, and both use the WDM "miniport" driver of the E-MU USB audio device at "the bottom of" the audio stack. "WaveOut" and "DirectSound" can and do, however, behave differently, as Juergen pointed out. But that's something that happens in the top layers of the audio stack, right at the output of the player application. Here's a diagram to illustrate things: http://msdn.microsoft.com/en-us/library/ms790062.aspx Now you can see why it's a good thing if the playback application happens to support ASIO -- like Juergen's fine JRiver thingie natively, and Foobar or WinAmp with an ASIO plug-in -- as the audio bitstream goes from the player program "through the ASIO tunnel" straight to the audio device's ASIO driver, bypassing all those "WMAud" and "KMixer" layers (and possibly more) in between that you can see in the diagram.
  15. Yes! I whole-heartedly agree with you, Juergen: I wish people would specify more clearly what kind of jitter they are talking about. Clock jitter? Line jitter? CD drive seek jitter? The jitter they got from drinking too much coffee in the morning? LOL Juergen, I presume you were referring to this neat PDF on the Wolfson website explaining jitter in the context of the S/PDIF interface: http://www.wolfsonmicro.com/uploads/documents/en/SPDIF_Paper_v1.2%20May%202007.pdf If not, would you kindly provide a link to the document(s) you were referring to? And for those Computer Audiophiles here who are bold enough to dig into some more technical papers about jitter, Julian Dunn's papers at the Nanophon website are another fun read (yeah, I know: I have a strange definition of fun... LOL): http://www.nanophon.com/audio/index.htm But if that's more than you ever wanted to know about jitter, you might feel content just reading the relevant excerpts from the Wikipedia entry: http://en.wikipedia.org/wiki/Jitter In any case: jitter under 200 ps for the internal clock of the 0404 USB means that those excellent ADCs and DACs can really "show off" what they are capable of. I know of no other audio interface in that price segment (except for the two "little brothers" of the 0404 USB from E-MU's product line) that come even close to the stability of the clock and the quality of its ADCs and DACs. And the fact that it also has an excellent headphone amp and can be used as a stand-alone DAC, as well, doesn't hurt, either... One last thing: 2 ns on the eye diagram sounds a lot more like it -- those 8 ns John Atkinson came up with just seem way to high. That's why I suspect improper termination and signal reflexion issues in his test setup. Thanks again, Juergen, for putting things even more in perspective.
×
×
  • Create New...