Slashdot Mirror
Build High-End Audio System w/ Hard Drive Storage?
nganju asks: "Hard Drives have finally reached the size where I can rip down 1000 CDs directly to WAV files, and skip the compression step (read: headache) altogether, ensuring that the audio playback is what the original CD author intended. Now the question is, how do I get that WAV data off the computer and into the amplifier with minimal distortion? Are there D/A PCI cards comparable to high-end CD players? Or is the best solution some direct digital output card (SPDIF) and a standalone D/A converter component? Specific model names would be greatly appreciated."
Just use optical output at the soundcard such as sb audigy.
The signal will be in totally digital form, until your amplifier D/A converts it back to audio.
That way your computer/soundcard won't affect the sound quality at all.
There are no atheists when recovering from tape backup.
Have you compared the prices of 200GB drives with 400GB drives? The higher capacity drives cost more than double. Now if you happen to have more storage space than you know what to do with right now, then fine, but sooner or later you'll find that running a script overnight to compress your files is less painful than the cost and effort of a hardware upgrade.
And if you're doing backup by syncing over the network with another system, then you cut in half the storage cost on the other system, as well as the network time.
If you have enough computing power to manage a file system on a hard drive, then you have enough power to uncompress a file as you play it--that's why they put MP3 players into DVD players; it didn't increase the hardware cost.
Slashdot: Failed Car Analogies. Amateur Lawyering. Anecdote Battles.
Lossless compression does not discard any data; when you decompress the data, you get back exactly the bits you compressed. The trick here is to represent common values with short bit strings, so you can represent the same data in less space.
Lossy data compression discards data that (theoretically) won't be noticed. The theory is that people tend not to notice the missing data, because the human brain is very good at interpolation. If you don't trust this theory, then you should avoid lossy compression. But not lossless compression.
Wikipedia, as usual, has a lot of good material.
Another solution is to get a card with S/PDIF digital output and an amp which supports it. This is a good solution if you already have such a soundcard, but the soundcard upgrade can be expensive. On the whole, if you're starting from scratch, I'd go with USB.
That's it. I'm no longer part of Team Sanity.
Do an ABX test.
Anecdotal stories do nothing to convince me.
Besides: A 256Kbit CBR MP3 is crap. The --alt preset settings are very high quality CQ VBR settings.
Comparing video artifacts to audio ones is apples to oranges.
Simply do an ABX test and _then_ let's talk.
Take a 10 Mb text file and zip it. You'll get a much smaller file. Unzip it and you get exactly the same file. Now zip a .wav file and you'll get a file about 20-40% smaller. Unzip and you get the same .wav file. Now heavily optimize zipping specifically for audio data you'll get a file 30-70% of the originial size. Now optimize that zipping routine to allow unpacking on the fly for convenience. Now unzip the FLAC audio on the fly giving you the original .wav data and then send that audio stream to the hardware exactly as you would have if you had just sent the .wav file. (This is very simplified just to get the gist of it. For a more comprehensive overview try this.)