POST
Host: slashdot.org
Content-Type: application/x-www-form-urlencoded
Cache-Control: no-cache
Accept: */*
Origin: http://slashdot.org/
X-Requested-With: XMLHttpRequest
User-Agent: Mozilla/omnichad like Gecko
Referer: http://slashdot.org/comments.p...
Content-Length: 46
postersubj=Re:BAH!&postercomment=how+very+meta
Must be hard finding an HTTP/1.1 compliant web browser that supports the Host header.
Since there's no ransom demand, the only way to make money is if ALM was publicly traded - they're not. Otherwise, you short sell some stock before making your demands public and hope to make money on the plummeting stock price.
"True" AI isn't artificial at all. If you make a human brain out of circuitry, you have "real" intelligence. Either the human mind is a non-deterministic machine that can't function without a soul, or "artificial" intelligence is possible.
Stuff that matters.
JPEG was only an analogy - editing and effects introduce artifiacts is the point. The noise floor drops directly from the higher bit depth and it better handles effects added, but you also get noise/junk audio when you layer multiple audio streams (drums, guitar, vocals, etc) and try to combine them into one waveform. That is improved by having the higher sampling rate - Nyquist has nothing to do with this. If it wasn't an improvement, then no engineer would waste the disk space recording at a higher sampling rate when the final output is only CD quality.
I disagree that FM is that bad. It's probably around 96Kbps in quality if that were really apples to apples. FM doesn't screw up sounds that much - it just flattens the waveforms horribly to help keep the SNR down.
Pandora's not exactly half as popular anymore because of it.
If you want to have your cats enjoy it, you're better with the virtual low-pass of quantizing it down to around 20,000Hz. If the engineer can't hear much higher, they can't fix noise or harmonics or distortion in those ranges. So it's much better if you cut them out.
Now that digital services (all of them) are streaming 128kbps or better
Pandora's free service is only 64K. Pure garbage.
Apple streams at 256. Others stream at 320 (e.g. Spotify). Free Pandora is the bottom of the barrel for streaming quality at 64K. We're no longer in the dial-up days.
The Dolby tech on cassettes isn't for compression. It's to defeat SNR loss on lossy magnetic media. If you don't play it on a Dolby player that reverses the "compression" it will sound a bit muddy like listening to an FM broadcast.
Digital studio equipment records in and keeps audio in a format at a higher bit-depth AND sample rate. Every edit and every layer of multiple tracks is like resaving a lossy jpeg. Your final result may contain all those extra bits, but they need to be averaged out and resampled away because the detail is already lost by that point. That's why you record in 192KHz / 24-bit and then master to 44.1KHz / 16-bit. GP post is 100% correct.
The noise floor they are talking about is introduced by the editing - not inherent to the recording.
The idea of a directional ethernet cable makes sense, but I'm sure it's probably snake oil. If the shielding was grounded on only one side (the source end), it's less likely to introduce noise into the circuitry on the DAC side. But probably at a level that would make coil whine out to be a jet engine.
Get hold of portable property. -- Charles Dickens, "Great Expectations"
