I know most people are concerned with the audible noise a power supply emits, but which power supplies give the components in your computer the cleanest power? I know its all relative to how clean the power the PSU is being fed. I ask because I'm very interested in creating an audiophile headphone/speaker setup and using the computer as the transport/source. Clean power is a must in this situation, as important as having the PC be silent. Anyone got any takes on this? The seasonics surely interest me, but how clean is the power?
-Mag
Electrically Quiet Silent PSU
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee, Devonavar
Already considered. I already have a DAC and amp(s) in mind. I just want everything in the chain to be as clean as possible. Including the transport.sthayashi wrote:If you're interested in an Audiophile setup, then you're far better off using your separate DACs and Amps (with their respectively seperate power sources). The reason is that even though the PSU could be clean as a whistle, the other components in your computer may not be as good.
-Mag
Power for digital out doesn't have to be nearly as clean as analog amplification. Any of the power supplies listed will be sufficient in this regard. I would be more concerned about whether your transport method can sustain enough bandwidth for your audio.MagusG wrote:Already considered. I already have a DAC and amp(s) in mind. I just want everything in the chain to be as clean as possible. Including the transport.
-Mag
-
Edward Ng
- SPCR Reviewer
- Posts: 2696
- Joined: Thu Dec 11, 2003 9:53 pm
- Location: Scarsdale, NY
- Contact:
You're going to have an extremely difficult time getting a positive answer on this one because it's something that few to no people have done any specific research on. I am also unsure as to who, if any members here have the necessary equipment (oscilliscopes, anyone?) to check this. Not only that, but remember that the cleanliness of the power from the PSU isn't just dependent on the AC being fed to it, but also the other components in the computer. Remember that there are all sorts of things on the 3.3 and 5volt lines being driven at once from the mainboard, not just the PCI slots, so even if you picked a PSU that tested out as very quiet on those rails, something else you install may affect that; at this rate, considering the power must past through the mainboard before reaching the PCI slots and/or drive controllers, you need to also focus on how clean the signals within the mainboards are. I don't know if anybody's tested signal noise at the bus level, either!
I've had great success building a satisfactory sounding platform myself, and I didn't pay particular attention to PSU or mainboard noise--the most important part is the quality of the sound card. If you want to see the system I built as a music platform, just see this link and/or this one as well.
As a matter of fact, more important than the noise level of the PSU or mainboard is your software configuration! To ensure that you are getting the most uncolored, most natural output, you must make sure that the music source has never been resampled (before sending out to the DAC), never been put through any form of filter, DSP module or gain and never been encoded to a lossy format. In other words, you want to be sure that the original data, 0s and 1s, makes it absolutely, positively 100% intact, unmodified out to your DAC. The best part about using a PC as a transport is that, by doing a direct rip of the audio from a CD in digital will avoid jitter completely; this is something the majority of standard compact disc transports simply cannot do. The only place where jitter can rear its ugly head is from the point where it sends the signal out to the DAC, and this, too depends more on the sound card than the PSU or the mainboard.
What I found was that there are a few places where your sound will get resampled and/or filtered before reaching the output to DAC...
First of all, it is highly recommended that you stick to lossless compression codecs if you plan to get an absolute 1:1 digital signal sent to your hardware, based upon the original source material. Many people will tell you that some of the best lossy codecs, when set to the highest possible quality settings, will produce sound that is not audibly different, but I personally believe that if you have the space to spare, lossless codecs guarantee an identical sound, regardless of your hearing abilities (it may be possible that the best lossy codecs are not audibly different, I am not going to say they are or aren't noticeably different, that's an argument left for some other place, some other time; I just know true lossless is a guarantee of no change in quality, period).
EDIT: I just wanted to add that I personally recommend FLAC and Monkey's Audio as good lossless codecs; FLAC is even supported by a couple portable audio players.
Second of all, Windows 2000/XP itself has a software layer known as Kmixer that automatically adjusts the gain and resamples the audio before sending it on to the hardware for processing; you must bypass this to ensure an unadulterated signal to the hardware. There are a couple different ways to do this; the best way is to use a good media player that can output via kernel streaming or ASIO directly to the hardware layer; this completely bypasses Kmixer all together, and also reduces latencies, which is critical for recording artists and musicians. You will also need to be sure that your sound card has drivers that support ASIO or kernel streaming input, whether from the original manufacturer or from a third party. Your alternative option, and this is much harder to find, is a sound card that has drivers capable of bypassing Kmixer while receiving material via Waveout. It just so happens that the sound card I chose can do this, if I want--it's the Echo Audio Mia MIDI, with its proprietary PureWave Kmixer bypass mode.
EDIT: For playback software, I also recommend foobar2000, which has the highest degree of control and the full install package already includes the necessary kernel streaming output module. WinAmp also has a third party kernel streaming output plug-in, but when I tried it, it didn't have the degree of control that foobar2000 does (iIrc, bitrate was one function that, back when I tried it, wasn't an available option in WinAmp).
The third place is in the sound card itself. When I first built this machine, I had a Creative Labs sound card based on the EMU10K1 engine; the problem with cards of this sort (for example, Live!, Audigy, Audigy2 and Audigy2 ZS) is that the sound engine itself only operates internally at 48KHz. Any sound going in or coming out that's not 48KHz will get resampled at least once; if it's not 48KHz going in, it will get resampled to 48Khz. If your DAC doesn't support 48KHz, it may or may not even be able to understand the signal from the Creative Labs card; I know this as I tried an Audio Note DAC on my Audigy2 ZS Platinum Pro and it refused to have anything to do with the signal from the card; something about the PCM transport format from that card just didn't agree with the Audio Note. I chose the Echo Audio Mia MIDI because it supports full manual sample rate lock in the driver panel. For all my lossless, Kmixer-bypassed material that goes to the sound card, it is 44.1KHz sample rate locked to guarantee zero resampling of the data before the DAC (I only have CD Audio; don't know of software that lets me rip DVD Audio 1:1 directly to the computer, or any non-proprietary software for DVD Audio playback on PC). The Mia MIDI even supports two different transport modes from the coax; standard consumer level S/PDIF, or professional AES/EBU output to your DAC!
Needless to say, I am more than satisfied with my PC as an audiophile platform. If I were to even consider touching my sound system, it would certainly not be at the source, but somewhere after it, beginning with the DAC. Luckily, I am more than satisfied with my sound system.
-Ed
I've had great success building a satisfactory sounding platform myself, and I didn't pay particular attention to PSU or mainboard noise--the most important part is the quality of the sound card. If you want to see the system I built as a music platform, just see this link and/or this one as well.
As a matter of fact, more important than the noise level of the PSU or mainboard is your software configuration! To ensure that you are getting the most uncolored, most natural output, you must make sure that the music source has never been resampled (before sending out to the DAC), never been put through any form of filter, DSP module or gain and never been encoded to a lossy format. In other words, you want to be sure that the original data, 0s and 1s, makes it absolutely, positively 100% intact, unmodified out to your DAC. The best part about using a PC as a transport is that, by doing a direct rip of the audio from a CD in digital will avoid jitter completely; this is something the majority of standard compact disc transports simply cannot do. The only place where jitter can rear its ugly head is from the point where it sends the signal out to the DAC, and this, too depends more on the sound card than the PSU or the mainboard.
What I found was that there are a few places where your sound will get resampled and/or filtered before reaching the output to DAC...
First of all, it is highly recommended that you stick to lossless compression codecs if you plan to get an absolute 1:1 digital signal sent to your hardware, based upon the original source material. Many people will tell you that some of the best lossy codecs, when set to the highest possible quality settings, will produce sound that is not audibly different, but I personally believe that if you have the space to spare, lossless codecs guarantee an identical sound, regardless of your hearing abilities (it may be possible that the best lossy codecs are not audibly different, I am not going to say they are or aren't noticeably different, that's an argument left for some other place, some other time; I just know true lossless is a guarantee of no change in quality, period).
EDIT: I just wanted to add that I personally recommend FLAC and Monkey's Audio as good lossless codecs; FLAC is even supported by a couple portable audio players.
Second of all, Windows 2000/XP itself has a software layer known as Kmixer that automatically adjusts the gain and resamples the audio before sending it on to the hardware for processing; you must bypass this to ensure an unadulterated signal to the hardware. There are a couple different ways to do this; the best way is to use a good media player that can output via kernel streaming or ASIO directly to the hardware layer; this completely bypasses Kmixer all together, and also reduces latencies, which is critical for recording artists and musicians. You will also need to be sure that your sound card has drivers that support ASIO or kernel streaming input, whether from the original manufacturer or from a third party. Your alternative option, and this is much harder to find, is a sound card that has drivers capable of bypassing Kmixer while receiving material via Waveout. It just so happens that the sound card I chose can do this, if I want--it's the Echo Audio Mia MIDI, with its proprietary PureWave Kmixer bypass mode.
EDIT: For playback software, I also recommend foobar2000, which has the highest degree of control and the full install package already includes the necessary kernel streaming output module. WinAmp also has a third party kernel streaming output plug-in, but when I tried it, it didn't have the degree of control that foobar2000 does (iIrc, bitrate was one function that, back when I tried it, wasn't an available option in WinAmp).
The third place is in the sound card itself. When I first built this machine, I had a Creative Labs sound card based on the EMU10K1 engine; the problem with cards of this sort (for example, Live!, Audigy, Audigy2 and Audigy2 ZS) is that the sound engine itself only operates internally at 48KHz. Any sound going in or coming out that's not 48KHz will get resampled at least once; if it's not 48KHz going in, it will get resampled to 48Khz. If your DAC doesn't support 48KHz, it may or may not even be able to understand the signal from the Creative Labs card; I know this as I tried an Audio Note DAC on my Audigy2 ZS Platinum Pro and it refused to have anything to do with the signal from the card; something about the PCM transport format from that card just didn't agree with the Audio Note. I chose the Echo Audio Mia MIDI because it supports full manual sample rate lock in the driver panel. For all my lossless, Kmixer-bypassed material that goes to the sound card, it is 44.1KHz sample rate locked to guarantee zero resampling of the data before the DAC (I only have CD Audio; don't know of software that lets me rip DVD Audio 1:1 directly to the computer, or any non-proprietary software for DVD Audio playback on PC). The Mia MIDI even supports two different transport modes from the coax; standard consumer level S/PDIF, or professional AES/EBU output to your DAC!
Needless to say, I am more than satisfied with my PC as an audiophile platform. If I were to even consider touching my sound system, it would certainly not be at the source, but somewhere after it, beginning with the DAC. Luckily, I am more than satisfied with my sound system.
-Ed