Getting High-Quality Audio From a PC

audiophile writes "Just because it's a PC doesn't mean it can't output good-sounding audio. In the same vein as specialty A/V products, you can find PC-based A/V systems with extensive audio processing and step-up performance specifications, including Signal-to-Noise ratio, which can make a significant difference when using the analog outputs. Media center manufacturer Niveus shares tips for getting high-quality audio from a PC."

Comments

Re:Actually it's pretty easy..

[mrjb]

Audigy 2 and other EMU10k1 chipsets are locked to 48khz internally, which has caused me a lot of grief when wanting to play back stuff at other rates. If you're playing back 48k exclusively this is fine, otherwise better get a soundcard that supports the different sample rates of your choice natively.

Re:My answer

[Lumpy]

unfortunately your answer is wrong though. Most high end sound cards are PCI. 24 channel recording cards for pro tools do a GREAT job at rejecting and shielding against PC noises.

these guys have been doing this for quite a while now. and if you only want "good audio" from a pc for your hom theater, get something with spdif and toslink and call it done. Let a high end reciever/processor do the work converting the AC3 track to analog instead of the garbage consumer level PC audio cards.

Re:MP3

[dal20402]

99% of music is indistinguishable from CD in 256kbps AAC (I don't have many 256k MP3s).

But some waveforms are just too hard to compress. In particular, harpsichords, solo classical string instruments, and solo electric guitar (through some filters) start to sound strange even at 256k.

A good but not foolproof way to figure out what is going to be troublesome to compress is to compress it losslessly using FLAC or ALAC and look at the resulting mean bitrate. Most stuff that compresses to between 400-600kbps, which is most music, will be fine at 256k. Some of my music, though, exceeds 900kbps lossless, and I even have a couple tracks over 1000kbps (where uncompressed PCM = 1411kbps). In all cases this stuff sounds like crap compressed to 256k. The harpsichords, in particular, sound harsh and flat, since the exceedingly complex waveform they make just can't all fit.

For me, it doesn't matter in the end, since I rip everything losslessly and then compress it for the car or the iPod where sound quality really doesn't matter anyway. But some people may not want to use hundreds of GBs of disk space or may have more music. For them, strategic ripping is in order.

Re:Age considerations?

[Kadin2048]

To be perfectly honest, except in cars that are really well-insulated from the outside world, most of the quality on a CD or a decently-encoded digital file is going to be lost.

If your threshold for music listening is around 95 dB (which is only safe for 4 hrs/day anyway), and the road noise in the car at highway speeds is 60-75dB (the latter is allegedly the figure for my car as-built, a VW GTI, from an auto magazine) then you only have at most 35dB between the noise floor and max while driving.

That's already worse than most cassette tapes, particularly decent ones with Dolby NR. (I don't think they even sell the metal tapes anymore, but my recollection is that they were 30+ dB right out of the package without any noise reduction at all.)

In a way, it explains why so much popular music and FM radio is compressed: there's no reason to offer more than 40dB of dynamic range, because (assuming your listeners don't turn it up to unsafe/painful volumes) they'll never hear the quiet parts because of all the noise in the listening environment.

Frankly, I think the biggest single thing you can do, in terms of improving the sound system in your car, is to install a lot of Dynamat or other sound-insulating material. Most car stereos, even the factory ones, have far more resolution than you'll hear except when sitting in a parking lot. If you can drop that noise floor even 5dB, you'll get that much more "loudness" out of your stereo without upping the volume to dangerous levels.