Jump to content

Hanz

  • Posts

    10
  • Joined

  • Last visited

  • Country

    country-ZZ

Retained

  • Member Title
    Newbie
  1. Hi Buch, There is no improvement in sound quality, as far as I can see. I am now using async USB which sounds better than SPDIF (through Halide Bridge) that I used on the original, but through the Bridge the updated PK90 does not sound better than the original one. Auraliti was reluctant to update (so they stated somewhere here on this forum) because it most probably would not have any effect on sound quality. And after all you could primarily consider this an audio device and not a computer, for which updates are a normal thing. I did not expect sound quality improvements. I was only after functional improvements. My original PK90 had MPD v0.16.0 and could not play DSD, no. I haven't yet tested DSD playback on the updated PK90 because I have not been able to rip my SACDs. I was planning to do this already months ago, but I have not found a suitable Playstation 3 yet. Your problem with the slider in MPAD when playing DSD is an MPAD issue I suspect. As I mentioned above, I programmed my own MPD client, and I remember that the slider required quite some time and effort to get it working because of the tweaking of the communication with MPD. It was rather critical and had to be just right. Try another client would be my advice. Hanz
  2. Just a message for all you guys who have bought an Auraliti PK90 and want to have a software upgrade (and have contacted Auraliti for this in vain). I bought one about 1.5 years ago. There were connection problems then with my NAS that were solved by SSH-ing into it (Auraliti provided me login data) and coding a workaround. This product obviously was not 100% ready for market then, as some of the utility scripts from the device 'home page' also did not work. As far back as then Auraliti promised a software upgrade. It was going to be made available to all customers for download from the website. They were 'nearly ready', just wanted to make sure it was rock solid, so nobody was going to end up with a dead PK90. After some months of not seeing anything appear, I mailed them to enquire about the status of the upgrade. Not ready yet. I have mailed them every few months or so again. For more than a year I have not received any answer anymore from them at all. Just complete silence. In the mean time more reasons, beside a proper solution for my NAS, to want to upgrade popped up. I wanted to have a newer version of MPD, among other reasons for DSD playback and I wanted to connect a DAC that has a Tenor TE8802 USB chip inside, for async USB, that was not working with the PK90 - I had reasons to think later versions of Voyage MPD would support this chip. So I got really sick and tired of this silence, even on mails asking for purchase of a hardware upgrade (a new SSD), that they had offered me before, and that I started asking for, believing that the downloadable upgrade would never appear. So, finally then, I decided to try to upgrade the PK90 myself. I opened it up, checked out what was inside, connected a keyboard and a display to watch what was happening when it booted up... And soon I discovered that there was really nothing special about it. Standard hardware, standard software. Nothing of their special manufacturing I mean. It should be possible to upgrade to a later version of Voyage MPD without problems. The only special thing about the PK90, the SoTM USB card, should work with Voyage MPD out of the box. And to come straight to the point: I was right, you really can upgrade. With hardly any trouble. When I finally upgraded (to Voyage MPD 0.9.1) it took me about half an hour to have it up and running, with my NAS connected, MPD set up, and, best of all, my TE8802 DAC working. Here's what you have to do: When you connect a keyboard and display to it and boot, you have to press F2 and keep it pressed when you see the Auraliti splash. You then fall into the BIOS setup screen. There you discover that Auraliti has fixed the boot sequence. You cannot select any other boot device than the PATA SSD that holds Voyage MPD. The reason for this is that the BIOS was changed with Intel Integrator Toolkit, a piece of software to create an OEM BIOS version out of a standard BIOS. BUT: these changes can be undone! Not by removing the CMOS battery, not by changing the BIOS jumper setting, the usual remedies for BIOS trouble. No, you have to run a program ITOOLKIT.EXE, which is a DOS (yes, DOS...) utility that comes with the Integrator Toolkit (you can freely download integrator toolkit from the Intel website). The only problem is: how are you going to run this ITOOLKIT.EXE on the PK90 ? Well, you have to make an old 2,5" HDD with IDE interface into a DOS bootable disk, put ITOOLKIT.EXE on it and boot the PK90 up with it. To create this bootable HDD you create a DOS bootable USB stick first. Make sure fdisk is on it. Then boot a computer with this stick, your HDD also attached. Then you make the HDD DOS bootable with fdisk that's on the stick. You can find details of this procedure on the web. The only thing you cannot do this way is making the HDD DOS partition active, the last step in the process. For that you need to use a program called diskpart, that is present in Windows. To boot the PK90 with the HDD you have to replace the SSD (the 2GB Transcend device) with it. When the Auraliti has booted up in DOS, type 'itoolkit remove', that's all. The Auraliti splash is still there, but the BIOS now gives you the opportunity to change the boot sequence. Now remove the HDD and put back the SSD, attach a USB stick or a USB CDROM drive with an .iso of the latest Voyage MPD LiveCD, boot from that, and install. Of course you can also use a 3.5" HDD, but then you need an adapter (for a few bucks), since the connector on the main board is for 2,5". When you have reinstalled Voyage MPD, the first thing you have to do is make the power button functional. It does not do anything out of the box, and you can only shut down from the command line. To make the power button functional you have to install acpid, and create an event file (I will not go into details here, you can find these on the web). Next you create a subdirectory of the MPD music directory where you want to mount your NAS, and add an entry in fstab. Next you configure ALSA for USB audio, to be precise, make USB audio the default card. Then configure MPD to use the USB device that you have attached, as shown by aplay -l. Done! Maybe I forgot something (for instance, how about all these ever growing logfiles, like the MPD logfile, considering this tiny storage space...) but it has been running here for a few weeks now without any problems whatsoever. Of course, the 'home page' the PK90 normally spits out, which also gives you access to some utility scripts, is gone. But it was useless anyway and the scripts partly did not work. There also was an MPD client installed that is not anymore there now. But I never used that as it was way, way too simple and I have programmed my own special client (to be able to play and control my extensive and specially tagged collection of classical and jazz with complete ease of selection, and to prevent the utter mess that can arise if you start putting all your music on HDD, things you hardly hear anybody about... But those are different stories.) And now you have a cleanly installed machine! I say this because what I saw in the original PK90 install seemed to be remnants of a whole bunch of tinkering and trying. Maybe because they used a single firmware for both PK90 and PK100 or maybe because tinkering and trying was still neccessary then to get everything working, but anyway, it did not leave an impression of much 'cleanliness'... Of course you need to have some knowledge of hardware, DOS and Linux. You don't have to be an expert but do not try this upgrade if you are not confident about your skills. And beware, you lose your warranty on the device when you do this. Auraliti is not going to help you out if you mess up. The SSD Auraliti has put in is an odd device, put on the market by Transcend a few years ago, to enable people to give their old laptop a second chance by offering a faster replacement of their IDE HDD. But these devices are actually quite expensive now because they are becoming obsolete (maybe that is the secret reason why Auriliti did not answer my requests for a hardware upgrade). You can buy a much spacier and faster SATA III SSD for the same money as these 2GB PATA thingies. Actually there is a whole series of them differing in storage space. But they are all expensive, ridiculously expensive if you look at the price per GB. The PK90 main board has an SATA interface, so a regular SSD should not be a problem. Haven't tried yet (but I will soon) to run Voyage MPD off a real SATA III SSD, it will surely speed things up. Though you don't need the ocean of space that a real SSD offers: even the smallest are 30/32 or 60/64 GB. The only challenge will be to get it inside the small PK90 cabinet while not disturbing the passive processor and chipset cooling. Maybe the newer PK90s have different hardware, but at least some portions of this description will still be useful. I have not tested exhaustively, though, as most of my music is 16/44. So far I have only tested FLAC playback, and up to 24/176 it works fine. To wrap up, it is not my intention to bash Auraliti stuff. But there IS something wrong with their customer support: it seems to be non-existent. It is a very small and probably very busy company, yes. But come on, ignoring customer emails ... (I have seen more of these complaints on this forum). Good luck! Hanz
  3. Well, I decided to give the bridge some more burn-in time before proceeding to PK90 tweaks and measurements. I let the two run continuously for about 100 hours. Boy am I surprised! The sound signature has completely transformed. It now greatly surpasses my CD transport in every respect: low level detail, separation, naturalness, sound stage... I have never before experienced burn-in effects THIS extreme. Don't know which of the two needed the burn-in, the PK90 or the bridge, or both. I'll try the PK90 on the USB input of my DAC again, without the bridge. Later. For now I'm completely involved in happy listening!!
  4. @audiozorro: just read the thread from the top and you'll understand. Furthermore, it was not my intention to start some kind of discussion about CD transports vs. computer transports. I did not make any general comparative statement about the one or the other.
  5. @phofman: I was comparing PK90 + Halide bridge to SBT/CDP through coax. So all SPDIF. You probably got confused by my remark about the lows with and without the bridge. By that I meant that the bridge did not improve SQ in all respects. @barrows: well, what can I say... curious, yes. I surely didnt't expect this either. But I love my Audio-GD ref7 DAC, I'm completely satisfied with the sound quality it provides through SPDIF. I don't need to switch to USB (I bought the PK90 to be more 'future-proof', as I said). I only want my network based music to sound at least as good as my CD transport before throwing out my CDs. And I was nearly there with the SBT, but controlling it was not up to my standards. MPD is much better in this respect I think. As far as the PS is concerned: I am using the PK90 with a low noise linear PS from Aurality. As far as tweaking of Voyage MPD is concerned: people ARE doing this... there must be some reason... And I haven't seen any signs in the PK90 that this kind of tweaking has been done already (I can SSH into it). You say you get better results with PK90 through USB than with your CDP through SPDIF on the same DAC... I don't consider that hard evidence in favour of the PK90. Who knows, maybe your DAC has an SPDIF implementation that's not that good... Still, if so many people rave about the PK90... maybe I just have a bad one... I would really like to test first if the PK90 is outputting bit perfect. Anybody here who knows how to do that ? Preferably at the output ... connecting another computer and re-assembling the streamed file at the other end ? (so not by tee-ing ALSA in the PK90)
  6. So, I finally got my Halide bridge last week. I'll come straight down to business: the PK90 with the Bridge does sounds somewhat better (highs, definition and soundstage) but still not as good as my CD transport. I have a tweaked Squeezebox Touch (all the Soundcheck software tweaks and a super low noise power supply I built myself) that I planned to be replaced by the PK90, primarily because I don't like the way you control the Touch. The sound quality I get from the Touch is very, very good. I compared it with PK90 + Bridge: the Touch wins on all fronts. And that for roughly 25% of the cost of the PK90 + power supply + Bridge... So, what to make of all this (apart from the fact that the Touch is really very good value for money...) ? My gut feeling says it's the PK90. With all the things I changed while experimenting, the sound quality difference between the PK90 and my CD transport stayed more or less the same, qualitatively. My experience with changing things in my stereo (cables, electronic components like output coupling capacitors, power supplies, whatever) is that usually all kinds of subtle shifts in sound quality are the result. Often you cannot even say the change has made it better or worse, just different. It's not always that easy to give an analytic, quantifying judgement when comparing sound qualities. But in this case it consistently was easy: the CD transport SQ has wider soundstage, better definition, the layers of the music are separated better, the music is more detached from the speakers. And the only thing that stayed the same in all the experiments in this case is the PK90. I don't think it's the PK90 hardware, I think it's software. ALSA and/or MPD. The bits aren't coming out the right way, somehow. It has to be, the difference between my CD transport and the PK90 are just too big to be caused by clock jitter alone. And the clock jitter specs of the Bridge are better than those of the low jitter clock I built into my CD transport, so I don't believe it's the Bridge messing things up either (although the lows sound less firm with the Bridge than without it, which makes me somewhat suspicious of it). I remember from tweaking my Touch that the software tweaks had much more effect than hardware tweaks. Especially changing the scheduling priority of the audio streaming processes had a dramatic effect. And as I said above, there are people doing this with voyage MPD too (the PK90 runs on this). For instance, take a look here: http://www.symphonic-net.com/kubotayo/articles/voyagempd.html#patch (Why this improves sound quality is beyond me. I can understand that the processes taking care of streaming the bits out should have a high priority. But that giving them higher priority improves sound quality can in my view only be because that cures buffering problems. What else could it be ? But then again, buffering problems ??? Surely, such an issue that is so utterly basic to streaming high quality audio should have been addressed first in any design, I would imagine...) Tweaking the PK90 software it shall be then!! (I'm not giving up that easily)
  7. My CD/SACD player does not play DVDs. So I cannot compare 48KHz. It would be nice to know if the PLL/packet alternation is causing this, as a purely technical matter. But even if the SQ differences between CD player and PK90 would be significantly smaller at 48, or disappear, it wouldn't help me in any way solve my 'problem', as upsampling from 44 to 48 is not an option, for obvious reasons. And 99% of my music is 16/44. The Bridge will arrive in a week or so. I have good hopes.
  8. Well, since I don't want to lose my Ref7 (I really love it), and I don't want to ditch the PK90 either, I decided to order a Halide bridge. As I said, it is a gamble, because maybe the PK90 is already that good that the bridge cannot improve much, and certainly not get the signal better than what my CD transport delivers. After all, it was made for a standard computer with jittery USB data and dirty USB power lines, not that hard to improve something there. But on the other hand, the SOtM card USB power is already very clean, and my experience with (low jitter) clocks is: the cleaner the power supply, the more stable they get. The Halide only goes up to 24/96 but so does the Ref7. And it is still affordable; it would be rather ridiculous to go and spend a fortune on a top-of-the-line async USB-SPDIF converter under the circumstances. I usually don't buy audio stuff without listening first, but as the PK90, the Halide isn't available in any stores here (I live in Europe), so I had no choice than to go by reviews. I'll post back the results (not very interesting actually, but who knows somebody can benefit from this topic).
  9. Hi all. In order to do away with my CDs and switch over to network based audio I recently got myself an Auraliti PK90, thinking that was going to provide me with extremely good USB playback. My Audio-GD Ref7 DAC is not an USB DAC in the true sense, its USB port was actually optional, and it uses adaptive mode USB. But since there is an ultra low jitter clock on the dedicated USB card (SOtM) inside the PK90, I thought that would eliminate the need for asynchronous USB. I could have gone for Auraliti's SPDIF out model, the PK100, but thought a USB player would be more future-proof. But alas, although the PK90 sounds good, it does not sound quite as good as my CD transport and I do not want to take a step back in sound quality switching over to network based audio. The CD transport has a wider soundstage, better highs, everything is just located better and more accurately defined. All in all simply more involving. Not a night and day difference, but you notice immediately when switching between the two sources (listening to the same music, same passage, on CD and accurately ripped on NAS. And it's not a volume difference thing). I know these differences quite well to be caused by jitter. Since the SOtM board inside the PK90 has such good specs, I really didn't expect this. And my CD transport (actually a SACD player used as transport) isn't even that esoteric in quality; it's a Marantz SA15S-1, although with a heavily modified power supply and a TentLabs low jitter clock. And the music sources I am comparing are not that advanced in quality either: all 16/44, although excellent recordings. I varied some things, to try to get to the deeper cause of this phenomenon: - Hooked up a mains filter before the PK90 power supply (a low noise power supply from Auraliti) --> made no difference - Replaced my el cheapo USB 2.0 cable with an audiophile grade one (Furutech) --> made no difference - Maybe the network and/or the NAS had some influence. So I put my test music on an USB stick, plugged it in the PK90 and compared playing from that with my CD transport. --> same sound quality difference as before - I played the PK90 through my Audio-GD USB-SPDIF converter ('Digital Interface'), and compared again (CD transport attached to BNC SPDIF of my Ref7, Digital Interface to RCA SPDIF) --> same sound quality difference as before. (The Audio-GD USB-SPDIF converter also uses adaptive USB, and probably has mostly the same electronics at its USB input as the Ref7, so this last test was not very useful, I admit) Assuming the USB port on the Ref7 is not messing things up (in fact, I mailed Kingwa, Audio-GD's chief, about this and he stated that SPDIF in and USB in should sound the same, if the music source is the same, jitterwise, and the cabling is ok, of course. Possibly only minor differences), it seemed the PK90 was the cause of this. So I got into this adaptive vs. asynchronous USB thing a bit deeper, reading all kind of stuff on the web, and found out that async is actually better by design, not by implementation. That is, even when you implement adaptive really well, as in my case, with the SOtM card having this very accurate clock, async is still better because the clock in adaptive mode is synthesized (it needs to have different frequencies, based on the data rate) while the clock in async mode has a fixed frequency. Maybe that explains the sound quality difference in my case too: although the SOtM clocking is quite good, it cannot beat my CD transport fixed frequency low jitter clock. (There could also be some unresolved real time software issues in the PK90, its OS is Voyage MPD Linux, and although this OS is supposed to be optimised for audio streaming, there are quite a few websites dealing with tuning Voyage MPD's real time settings for best sound quality. Once I tweaked a Squeezebox Touch, it had this issue too, among others. Tweaking the priority scheduling of the audio processes made a huge difference to sound quality. But the guys at Auraliti seem to know what they are doing, so I guess they tackled these issues.) So, what next ? I'm out of options besides getting myself a Halide bridge (or something like it). It's a bit of a gamble but the bridge would certainly benefit from the clean USB output power that the SOtM card is providing. It would mean that the accurate clocking on the card is quite useless, though. Hell, I'm even beginning to doubt the whole use of this PK90/SOtM approach for high-end audio. Or are my vues on/my understanding of these matters too simplistic ? Anyway, I'm tempted to say: if you have a adaptive USB DAC, and are thinking of buying an Auraliti PK90, don't expect any miracles to happen... Any thoughts anyone ?
×
×
  • Create New...