Can Mac/PC compete with High End CDT??

I recently built a music server based upon the cics Memory Player (cMP2). It has been both very rewarding and eye/ear opening to say the least.

In the past, I have tried using my daily office computer to stream uncompressed audio to my Playback Design MPS-5, but I always found the sound quality was boring and lifeless compared to discs played directly on the player. I quickly dismissed computer transports as a serious audiophile alternative, but a few months ago I came across a discussion on cMP2 and decided that I would see what all the fuss was about.

The philosophy behind cMP2 is to have a barebones computer configuration with an audio player whose goal is to maximize sound quality at the expense of convenience and features. It is not for everyone as it requires users to be a somewhat tech savvy and the user interface is a very basic and clunky compared to standard Win/Mac applications.

cMP2 consists of a recipe for putting together and tweaking the computer hardware and bios settings. It comprises two distinct applications: cMP and cPlay. cMP is meant to replace the standard Windows shell and cPlay is the audio player application. Both cMP and cPlay can run under Windows XP/Vista/7 without any of the recommended system optimizations; however, the sound quality will be no better than using something like Windows Media Player.

When I first loaded the OS and the cMP & cPlay applications onto my new computer, I played some audio tracks just to ensure that all the components were working and that I was getting some sound. The sound quality was the usual dull, murky, lifeless audio that I have always associated with computer based transports. So I started making changes to the bios to severely under-clock and under-volt the system. Also disabled all extraneous on-board peripherals that were not required for a music server (eg. power management, floppy, firewire, LAN, etc ...). Immediately, you could hear a big improvement in sound quality. The music came alive with greater detail and transparency ... but it was still a far way off from the sound I get from disc. Next procedure was the Windows optimization steps by disabling all windows services that are not required for our music server. Note that these steps will result in loss of regular windows functionality, but that is of little significance for our intended goals. Again, sound quality takes a significant leap forward and system latency starts to really improve without all the extraneous services running in the background (optimized system around 3-5 usec versus 70-100 usec on my daily computer).

The best way to run cMP is to have it configured as the replacement shell for XP/Vista/7. This avoids extra overhead and processes that are loaded with the standard Windows shell. I don't really see any advantage of using cMP if you do not use it in this manner. cPlay is an ASIO player that loads the song into memory before playback so that access to the disk I/O is not required during playback. It allows for up to 24/192 playback with either 145db/121db SNR or SoX upsampler. You can also configure cMP to use an alternative player to cPlay if you choose.

I was originally using digital coax to connect the music server to the MPS-5 and what a big mistake that was. Computer systems are inherently noisy and the coax connection was causing detrimental effects on the overall sound of my system. I never realized how bad it was until I turned off my music server and noticed how much better everything sounded. I have always refrained from using Toslink based on unfavourable comments by many people, but was I ever wrong with this assumption. The galvanic isolation of toslink resulted in superior performance as compared to a pricey Transparent Ref Digital coax in this particular instance. Proximity of the computer server to the audio gear is also very important as the CPU and nasty switching ATX power supply generate significant amounts of RFI.

I can happily say that my little experiment has been a great success. The SQ is getting very close to that of playing discs directly on the MPS-5. I am still slowly tweaking the bios settings as well as the windows configuration trying to remove the last bit of extraneous windows processes without causing the system to freeze. I also believe my next significant step will be the replacement of the switching power supply to a robust linear power supply that I am having Gilbert Yeung of Blue Circle design and build for me. I'm sure once that has been done, the sound quality will be the same if not better than playing discs on the MPS-5.

If you are an audiophile that loves to tweak, then you will really enjoy the self satisfaction of building one of these machines. If you do decide to assemble one of these servers, then I would recommend configuring your system for dual boot so that you can switch between a full service Windows OS and the heavily streamlined cMP2 OS that can do very little but play beautiful music.

System Specs:
Intel Core 2 Quad Q8300
Gigabyte GA-EG45M-UD2H m-ATX motherboard
Kingston DDR-2 PC2-8500 2GB Single Channel
Antec 550W Earthwatts
WD 2TB Green 3.5" HD
RME HDSP 9623 Audio Card
Samsung 22x DVD
Themaltake ISGC-400 CPU cooler
Zalman HD160XT Plus HTPC

Windows XP
cics cMP & cPlay

Kijanki, if the component resonates on soft feet, you have the component and its wires moving in the very magnetic field that their circuit produces with some delay. This is why isolation can so improve a component. Nothing stupid about this.

If electronics resonates on soft feet it can have (if any) unpredictable results. Not only that soft feet can provide nice isolation from vibration (as Vibrapods do) but there is no way of telling how it might affect the sound (if at all).

As I said before - keep footers under you SS gear wet to have more liquid sound or hang electronics in the air to have airy sound. What the heck - move it upstairs to have better highs or to basement for better lows. Logical - isn't it?

Kijanki, no it is illogical. Everything has a resonant frequency below which all hell breaks lose.

I believe so however, there are so many combinations I would imagine it's impossible to come to a consensus. I would certainly be concerned if anyone made absolute statements regarding either approach, now or in the future. The number of possible combinations in the computer, DAC and audio equipment world are too vast.

For my part, I was using a Mark Levinson No. 37 transport with a 360S DAC. I wasn't listening to it much anymore having dedicated more and more time and resources to analog so I sold the Mark Levinson gear. I personally liked the sound of this set-up but it's lack of use made me think I should move on and try something else. I like to try different things and see what works for me and I also like trying new technology, even if I ultimately determine that it's not for me.

In deciding to experiment with computer audio, I selected an older MacBook Pro (something I already had on-hand), connected it to a Weiss DAC-2 via Firewire and used the Amarra playback software with iTunes. I've found the experiment quite interesting and very satisfying. I've since added the Saracon up-sampling software from Weiss and I'm considering a Weiss DAC-1 now that they have added a Firewire input to it.

I appreciated the Weiss approach to computer audio as it uses asymmetrical Firewire allowing the DAC to control the computer. To me, this eliminates a certain number of variables that could theoretically degrade sound quality and it made sense to try and minimize the amount of audio hardware inside the computer. Other manufacturers use this approach using asymmetrical USB but I have no experience with that. It may work well too.

I find the Apple/Weiss/Amarra/Saracon combination to be very relaxed and musical, doing everything the Mark Levinson did well while adding a significant amount of detail. This detail did not come through in a cold, clinical, analytical way nor did it come through in a warm, dark, "heavy" way. It seems to me that it lives right in the middle giving a good dose of musical honesty.

Please remember that I'm describing what I hear, in my system. It's quite a startling difference at first but the more I listen to it the more I realize the Mark Levinson was either missing this information or perhaps masking it.

For a little context I'm using an Atma-Sphere MP-1 preamp with Atma-Sphere MA-2 monoblocks driving SoundLab A1-PX panels. My analog front end consists of an SME 30 with Graham Phantom II tonearm and a Benz LP and an Ampex ATR-102 with ARIA electronics. I value musicality and accuracy equally and prefer to live in the neutral zone or perhaps a tad on the warm side when required. I think this is why I liked the Mark Levinson gear in my system. It added some warmth and weight to CD's which I felt was required for my taste.

So, returning to the sound of my Apple/Weiss/Amarra/Saracon combination - On high resolution file formats or up-sampled formats you can really start to hear a positive difference, especially with the Reference Recordings HRX disks. The higher the resolution, the more detail, spacial information, dynamics and accuracy you get all while increasing the relaxed nature of the presentation. Obviously starting with the highest resolution possible is a great improvement over CD. It becomes more analog-like while giving you dynamics and impact that analog just can't quite do.

For me the Mac/Weiss/Amarra/Saracon combination is relatively easy to use. I understand Macs fairly well having used them in some form or another since the early 80's and the Weiss information on their web site combined with the instructions on "How to Make a Music Server" on the Amarra web site go a long way to getting the best sound out of this combination.

I would state that it does require some technical know-how and a bit of fiddling with software to get everything right. There are operating system issues that need to be addressed, dongles that need to be attached, drivers that need to be installed and things that need to be turned off in order to get it to work well. It's just not as simple as buying a CD player and plugging it in however, as long as you like experimenting and have some basic knowledge of the computer it's fun and you will get some really satisfying results.

There is certainly far greater flexibility with regards to the final sound from the computer approach than the CD player approach. In other words, you can alter the flavor to suit your personal taste in a way that is simply impossible with a CD player. It's like having multiple cartridge and arm combinations on a turntable. The Amarra software for example, has a digital EQ that can make nasty sounding CDs sound listenable and even enjoyable. The Saracon software has various options for up-sampling your files yielding subtle differences in the final digital file.

In conclusion, it has been my experience that computer audio can compete with high end CD transports while offering the opportunity to listen to high resolution file formats, up-sample any resolution (to 196 kHz in my set-up), adjust EQ in the digital domain if required or desired and the convenience of a music server. These are things that a CD player can't do. It does require some experimenting so don't expect to simply drop in a computer and expect to come to a conclusion in a few hours of listening. I've been woking on this and enjoying it for several months - and I'm not finished yet - but I do believe that computer audio can yield some very satisfying results.

Have fun!