HDCD

[Andreas]

Steve's Audio Fidelity Gold HDCDs did not show any clippings, as far as I remember.

Wouldn't that simply mean he didn't use dynamic range extension?

Probably. Does "dynamic range extension" necessarily result in clipped samples when ripped to a PC?

[Crummy Old Label Avatar]

Why not use audio capture software (like Audio Hijack for OS X) to digitally record a HDCD file from Windows Media Player; then rip the same track off the CD - and compare the results. No need to make an analog capture.

That will tell you immediately if Windows Media Player really does decode HDCD. It would also be interesting to see what differences would be revealed.

I'd do it myself -- I do have Audio Hijack -- but there's no way in hell I'll reinstall Windows Media Player on my machine (and does the Mac version even have the HDCD feature? I have no idea).

[lukpac]

Probably. Does "dynamic range extension" necessarily result in clipped samples when ripped to a PC?

If the peaks would otherwise be 0dB, I'd assume so. If the peaks "should" be -6dB, then no.

[Andreas]

You mean that the missing 6 db are encoded in the least significant bit? And why should this increase the dynamic range?

(You win one bit at the top, you lose one bit at the bottom ==> nothing is gained.)

[lukpac]

You mean that the missing 6 db are encoded in the least significant bit?

That was my understanding. I could be wrong.

And why should this increase the dynamic range?

(You win one bit at the top, you lose one bit at the bottom ==> nothing is gained.)

You're asking the wrong person...

[krabapple]

[quote="Crummy Old Label Avatar"]Why not use audio capture software (like Audio Hijack for OS X) to digitally record a HDCD file from Windows Media Player; then rip the same track off the CD - and compare the results. No need to make an analog capture. [QUOTE]

That's what I've wanted to do...but I haven't got the Windows version o fcapture software (I forget the name of the software) to work. I haven't tried that hard either, though.

[krabapple]

Probably. Does "dynamic range extension" necessarily result in clipped samples when ripped to a PC?

If the peaks would otherwise be 0dB, I'd assume so. If the peaks "should" be -6dB, then no.

Of course, in the latter case, normalizing the track would give you the same effect, rather more cheaply than using an HDCD rig.

[Patrick M]

That's what I've wanted to do...but I haven't got the Windows version o fcapture software (I forget the name of the software) to work. I haven't tried that hard either, though.

Total Recorder?

[krabapple]

That's the one.

[Andreas]

You probably know this, but it is worth repeating.
http://www.hdcd.com/partners/proaudio/m ... ating16bit

Peak Extension
Peak Extension is a restorable (with HDCD decoding) soft peak limiter that allows peak levels of up to 6 dB above standard full scale peak level (+6 dBfs) on 16-bit HDCD-encoded recordings without generating "overs." The limiter has a carefully crafted "easy over" curve, designed to mimic the sound of analog tape saturation, which operates over an input signal level range of -3 dBfs to +6 dBfs. This squeezes the top 9 dB of the input signal's range into the top 3 dB of the 16-bit recording.

During HDCD-decoded playback, the peak limiting is undone by the HDCD decoder chip using a precisely mapped inverse of the limiting curve controlled by a hidden code. The dynamics of the original material are restored up to +6 dBfs, thus extending dynamic range by up to 6 dB. During undecoded playback, the effect of Peak Extension is that of a very high-quality standard limiter. Use of Peak Extension is optional in the 16-bit HDCD encoding process.

What I find interesting is that they describe the peak extension method as an invertible compression. In other words, it works like a compressor, squeezing the top 9db of the recording into the top 3db, but with the lost information stored in those least significant bits. There should be no clipped peaks on the .wav, if I understand this correctly.

[lukpac]

Sure, if you simply change the gain digitally. But I'm not sure if you'll see the same flat tops by doing a D/A -> A/D conversion. I actually did this a bit recently. I should see if I still have the files and check it out.

I took a look at some Who's Next comparisons I did recently (Japanese CD next to 3 different dubs of said CD). That one has some mild flattops. I was wrong - the D/A -> A/D copy seems to have the same flat tops. What's interesting is that I wasn't able to record that quite as hot as the original - my thought was some of those flattops were being "fixed" when played back, causing higher peaks, but that doesn't seem to be the case. Strange.