Here is a link to a White Paper.
In the world of high-end audio it is quite easy to fall prey to simple assumptions. A fan-less computer is NOT noiseless. It cools the CPU (and partially cools the motherboard) via a heat-transfer device similar to the ones we use. These heat-sinks and heat pipes move a lot of heat to the surfaces of the exterior cooling fins. Heat moves air across these fins and this moving air can be detected by a decent digital sound pressure meter. Thus, the fan-less servers are not silent (whether or not you can hear such noise). Using our good, digital sound pressure meter, we find that our hybrid cooling systems – heat sink + heat pipes + Noctua™ fan on the bottom of the computer set at a very low speed – generates the same level of decibels as simply having no fan.
More importantly, the measured noise at 3 feet from the computer are no higher than the readings for the audio system on standby in the absence of the computer. We have measured the noise on several audio systems on standby, and they all generate relatively high measurable noise because of the heat being generated. The Noctua™ fans generate 7 db of noise in an anechoic chamber, while we have yet to measure the stand-by noise at less than 20db in any listening room, including in our own systems with all household appliances turned off – and our facility is on a dirt road 3 miles from the nearest black-top in a small town in Montana.
Most importantly, all fanless servers simply do not produce low enough temperatures on the critical motherboard to suit us (and none of these fan-less companies publish these temps). This is probably not too important for these server makers, because none of those servers do all the things a Baetis computer can do. For example, if you are transcoding multi-channel DSD files on the fly to PCM, motherboard temperatures can rise to dramatically high levels after several hours of play. Such temps are NOT good for longevity.
First, you should understand that no music server can operate without some “screen” that shows the choices to be made. Most servers have fluorescent 5” screens that are pretty much useless unless you have your nose right in front of the screen. Second, the computer’s advantages over simple music servers is that the software can provide not only the best audio but also the widest flexibility in sources and formats, as well as critical ripping capability. This complexity requires some training and some “housekeeping”. Thus, YES you must have a monitor, however small, for those times such as
- initial set-up;
- our Mastering Digital Audio™ training sessions; and
- whenever there is any problem with the hardware or software.
Most of the time you will be using at your listening position a small iPad or Android tablet for everyday use (choosing albums, playing or pausing, etc). But you must keep the monitor itself handy, just in case. Please see our paper on Where to keep the Monitor – usually it is just a question of not being able to immediately see how it can be fit onto your equipment rack (especially those tall, narrow racks that dominate 2-channel audiophile rooms). Please read this paper then let us know if this is still an obstacle to achieving the best possible digital audio quality. The usual solution is a small stand right next to the main equipment rack; but please read about all the other options here.
The answer is NO. This question is usually asked by someone whose “computer expert” friend suggests that Linux is more “stable” than Windows. It is not, because we do NOT set up Windows in the usual fashion (see next question). More importantly, the main benefit of the computer with a Windows system is that it is totally “OPEN” in nature. We do not have to worry about special software that might be less flexible within Linux than in Windows. Of course, we cannot use Mac operating systems, because the Mac is inherently closed from a hardware point of view. We simply could not have developed our special daughterboard for a Mac computer, nor used the best version of JRiver for a Mac computer. Yes, that’s right – the Mac is quite limited as far as audio quality is concerned. It simply is not competitive to our system, as a-b comparison after a-b comparison has confirmed.
Finally, the main reason for using the ubiquitous Windows system is that the software makers for highest end audio quality are numerous and JRiver may eventually lose its edge. This includes Digital Room Correction software that, right now, is best within a Windows environment (talk with us about that). We want to be able to switch ripping and playing software at any point in the future with the least confusion, the least cost, and the least chance of requiring new hardware – for us, this means a Windows environment.
This is true. Of course, our computers can indeed play 176.4kHz via either USB or HDMI, and for certain DACs, via Ethernet. But, native 176.4 up-sampled to 192 or down-sampled to 88.2 will sound better, via the Baetis AES/SPDIF, in our opinion, than via ANY other media server’s output of any kind. Why is 176.4 such a big deal? Because some audio engineers think that it is important to play certain native 176.4 files (e.g., those by Reference Recordings™) at their native sample rate. We would certainly agree that this is optimal, ceterus paribus, for such files. But other things are never equal. First, the reduction in common mode noise within the Baetis computer for AES far outweighs the beneficial aspect of playing these files in their native sampling rate. Second, in the years since Reference Recordings started to record interesting music in a 176.4 sampling rate, the up-sampling and down-sampling algorithms have improved quite dramatically (for example, JRiver did not use the SOX resampling sub-routine even a couple of years ago.) We are not saying these engineers who laud the 176.4 sampling rate are “wrong” in theory. We are only saying that playing a computer file in its “original” sampling rate is far less important than it was a decade ago. Our respect for these pioneers is unbounded, and we understand the so-called Nyquist Theorem, or as some call it, “the LAW of multiples of 44.1”. We would not be in this business, quite literally, had these gentlemen not been conducting their research way before us.
But to take this theory and apply it in a manner that says – “buy Brand X of media computer, because it DOES 176.4 via AES” is where theory breaks down, badly. You must listen to the Baetis AES output playing a 176.4 file at, say, 192 using the SOX resampler and compare this to any other music server playing the same file via AES at 176.4.
We do several things to minimize the chance of the OS crashing. First, we disable several Windows features that are troublesome for stability and/or might harm audio quality. For example, we disable Windows Media Center so that ONLY the JRiver Media Center software can be used for playing. We also don’t allow you to add any other software without our express permission, or the warranty is void.
Second, we do not install and do not permit you to install any e-mail handling program, nor use any web-based e-mail handler. Third, we automate the Windows IMPORTANT update process while stopping various optional updates.
Fourth, our training sessions teach most users, for the very first time, how to use a Windows computer. It is amazing how little the average computer user knows about SAFE operations – some users have never been taught how to properly disconnect a drive or even turn off the computer. No matter how “computer-savvy” you are, we can pretty much guarantee that you will learn something from our training sessions. If you take our training seriously you will not only hear the very best audio quality but also should NEVER suffer a Windows crash. If you do, then our support people will fix it quickly and you’ll learn something in the process. You may have to devote a few hours of your time to the learning process, but as the economists say, “there ain’t no such thing as a free lunch…TANSTAAFL”. Everything worthwhile takes effort and this is what Baetis Audio quality is all about. Our hardware and support, taken together, are unbeatable.
If you are considering selling your Baetis server, please contact us directly so that we can instruct you on how to proceed with the recertification and transfer process. This is the only way the new owner will get our unparalleled support services and peace of mind. Conversely, do not purchase a Baetis on, say, Audiogon if the computer has not been recertified by us. Otherwise, the warranty is invalid, and the software may be corrupted.
How can your computer compare with a $20,000 music server that is built from the ground up for audio quality and is very highly reviewed?
You can only find out by auditioning each “server”. You can do this in one of our small group of highest end dealers, or you can ask the dealer for a reasonable trial period. Technically, however, you should keep in mind that the manufacturer’s claim of “built from the ground up to put out the best audio” is not really to the point. The modern Intel Core iX CPU is so much more powerful than the specialized chip used in “music servers” that there is no technical reason why it can’t beat the specialized chip at doing JUST AUDIO.
And then there is the question of the software. Yes, the music server software is specially designed for audio. But again, there is no reason why the quality of the music server’s software can’t be beaten by the JRiver® software that also does a list of things that can’t be done by “music servers” — rip Blu-rays and DVDs, provide for the playback of Blu-ray files and any other audio/video file, playback of DSD music either in native format, DoP, or in PCM depending on your DAC.
Remember our servers are NOT for everyone. If you have an inexpensive DAC or receiver, it is highly unlikely our server will improve the sound of your system. If you own a truly good DAC or pre/pro then, yes, you will be pleased.