HDCD ripping

[325xi]

Yes, it's that silly "loudness race"!

http://en.wikipedia.org/wiki/Loudness_war

P.S: Check out the wav clips of "One of Us" and the pathetic result after adjusting the new version using replay gain, compared to the original version.

Wow, it also says about HDCD: "When played back on a normal CD player the sound will be compressed and "loud". When played back on a CD player equipped with HDCD decoding, a dynamic range expander is brought into play, reversing the compression applied at mastering."

Which means I'm better to avoid HDCD CDs altogether unless I have HDCD-enabled player!

What happens when I rip HDCD disk? Am I getting HDCD info, or just the base comressed recording?

[mudlark]

Hi,
I don't think you have it correct about HDCD. My understanding is the tracks are identical, but an area of the disc is used to store extra info that can be decoded with the correct hardware.

check the Wiki for HDCD

"HDCD is a proprietary process, and no accurate technical description has been released to the public. Hence, none of its claims have any evidence. In fact, since the technology does not involve compression, and still allows an audio CD to hold the normal 80 minutes of audio within 700MB, the additional 4 bits of bit depth it claims to add to the recording cannot physically exist.

HDCD encoding places a control signal in the least significant bit of the 16-bit Red Book audio samples (a technique known as in-band signaling). The HDCD decoder in the consumer's CD player, if present, responds to the signal. If no decoder is present, the disc will be played as a regular CD.

In itself, the use of the least significant bit does little to degrade sound quality on a non-HDCD player (only decreasing the signal-to-noise ratio by a minuscule amount)."

Regards,
M.

[325xi]

Then we have an interesting mismatch between your reference and this http://en.wikipedia.org/wiki/Loudness_war



What may be is that HDCD itself does not involve compression, but allows to do mastering with more compression, assuming that HDCD will make it up.

Anyways, what about ripping?

[mudlark]

The whole point of HDCD is to improve the sound so why would the engineers use compression?

The only reason engineers use compression is to reduce the dynamic range of a track so that loud sounds are quieter and quiet sounds are louder.

I have it on faultless authority that it is possible to rip HDCD discs, but all you would gain is, hopefully, a good engineer who didn't use compression or set levels too high. The extra info is only available in the hardware and doesn't as far as i know exit via the digital connections. This is similar to dvd-A and SACD.

[Skunk]

Then we have an interesting mismatch between your reference and this http://en.wikipedia.org/wiki/Loudness_war



What may be is that HDCD itself does not involve compression, but allows to do mastering with more compression, assuming that HDCD will make it up.

Anyways, what about ripping?

A bit perfect rip with FLAC files automatically contains the HDCD data in the LSB, so your best bet is use the WMP workaround to maximize the quality of the tracks. There is a thread about it but I don't know the details or if it truly works.

Altering the LSB with digital volume control, or an imperfect rip (i assume), will ruin the HDCD data- but also leave the reduced SNR- i.e. no gain from not having the data (which may be useful later, should you obtain necessary hardware).

I need to figure out the workaround, because I have a ton of HDCD discs and no decoder. Some HDCDs do sound very compressed, especially Neil Young albums. Others, like Garcia's 'The Pizza Tapes' and Mickey Hart's 'Supralingua' sound great without the decoder, but not any better than a well mastered redbook cd I guess.

I'd also be curious how much bit error HDCD can tolerate, and assume that WMP must be as good as EAC in this regard...

[325xi]

A bit perfect rip with FLAC files automatically contains the HDCD data in the LSB, so your best bet is use the WMP workaround to maximize the quality of the tracks. There is a thread about it but I don't know the details or if it truly works.

Altering the LSB with digital volume control, or an imperfect rip (i assume), will ruin the HDCD data- but also leave the reduced SNR- i.e. no gain from not having the data (which may be useful later, should you obtain necessary hardware).

I need to figure out the workaround, because I have a ton of HDCD discs and no decoder. Some HDCDs do sound very compressed, especially Neil Young albums. Others, like Garcia's 'The Pizza Tapes' and Mickey Hart's 'Supralingua' sound great without the decoder, but not any better than a well mastered redbook cd I guess.

I'd also be curious how much bit error HDCD can tolerate, and assume that WMP must be as good as EAC in this regard...

Good point. That actually makes sense after reading that wiki entry - if for some reason labels are interested in producing louder sounding albums, HDCD does offer another opportunity for that. So instead using it to increase SNR, they can use it to produce louder but more compressed album, where compression partially compensated sonically by HDCD. And since not every label feels that urge to produce the loudest ever album, we have that difference - some HDCD sound good on non-HDCD player, and some - compressed. Now we just need to find out which labels do that bad thing.

HDCD is a deviation of Red Book - are you sure normal DAE (to wav files) using regular optical drive will pick those bits in a form that can be decoded later on pc?

[mudlark]

I've read some more which suggests that HDCD does use compression. The stuff is recorded at 20 bit and compressed to 16 bit for ordinary cds and the decoder uses info on the disc to expand the 16 bits back to 20 bit allowing a gentler filter.

Can someone give me the actual details or has the info all dissappeared into Microsoft's clutches.

[Skunk]

Good point. That actually makes sense after reading that wiki entry - if for some reason labels are interested in producing louder sounding albums, HDCD does offer another opportunity for that.

HDCD is a deviation of Red Book - are you sure normal DAE (to wav files) using regular optical drive will pick those bits in a form that can be decoded later on pc?

I should note that the Neil Young discs don't sound louder than the others, just less good. I went through about ten HDCD FLACs in audacity and didn't see any 0dB points. The waveform of the Neil Young disc looked about the same as others, so maybe it's mastering that's the difference.

I'm pretty sure regular DAE picks it up. People report that EAC>FLAC does light the indicator in capable receivers/DACs.

[Skunk]

I have it on faultless authority that it is possible to rip HDCD discs, but all you would gain is, hopefully, a good engineer who didn't use compression or set levels too high. The extra info is only available in the hardware and doesn't as far as i know exit via the digital connections. This is similar to dvd-A and SACD.

Not only possible, but unavoidable if you get anything near perfect rips.

It is not similar because you can rip neither of those other formats, and it does exit the digital out, or wouldn't be decoded by receivers.

[325xi]

Are there software decoders - to get "normal" wav/flac that can be sent to non-HDCD DAC in full?

[P Floding]

Wow, it also says about HDCD: "When played back on a normal CD player the sound will be compressed and "loud". When played back on a CD player equipped with HDCD decoding, a dynamic range expander is brought into play, reversing the compression applied at mastering."

Which means I'm better to avoid HDCD CDs altogether unless I have HDCD-enabled player!

What happens when I rip HDCD disk? Am I getting HDCD info, or just the base comressed recording?

Since HDCD is a way to squeeze better-than-CD audiophile sound out of CD's, hence getting top-notch production treatment, I doubt very much those CDs would sound bad without HDCD decoding. The HDCDs that I have sound just fine.

If I remember correctly HDCD swithes between different anti-aliasing filters dynamically depending on the required transient behaviour, so as to avoid time smearing. There might be something about dynamic range too... Nevertheless, they sound generally good even without HDCD decoding.

BTW, this was quite funny from the Wiki:

"HDCD is a proprietary process, and no accurate technical description has been released to the public. Hence, none of its claims have any evidence. In fact, since the technology does not involve compression, and still allows an audio CD to hold the normal 80 minutes of audio within 700MB, the additional 4 bits of bit depth it claims to add to the recording cannot physically exist."

http://en.wikipedia.org/wiki/HDCD

The author has obviously never heard of other coding schemes than linear PCM.. I bet one could use, for example, DBX to put a lot of dynamic range onto a 16/44 CD. Perhaps he confuses data compression with dynamic range compression.

I had a Krell CD player at one stage which did HDCD, and HDCD CDs sounded very good indeed. Very startling transients -which is where I feel CD often falls short.

[Skunk]

Are there software decoders - to get "normal" wav/flac that can be sent to non-HDCD DAC in full?

http://www.head-fi.org/forums/showthread.php?p=2680596

http://www.head-fi.org/forums/showthread.php?t=221104

Shows the change in dynamic range using a 24 bit 44.1kHz files compared to 16 bit, using WMP + plugin* to create a wav/FLAC (that could be played on Sb3). It's not proven this is equal to what a hardware decoder can do, and HDCD capable external digital converters are flooding the used market right now. Contrary to that idea, however, the Levinson 360 is one commonly available DAC that does do the decoding in a DSP chip (software), so I suppose WMP could very well be getting it right.

*http://www.chronotron.com/content.php?page=downloads

[DCtoDaylight]

Somebody should edit the wiki entry, because it's really wrong...

The details of HDCD have been released, and because it's a patented process, the details are available to anybody with the desire to get a copy of the patent!

In a nutshell, HDCD is a dynamic compressor, in the same vein as DBX or (to a slightly lesser extent) Dolby. The basic idea is the same, compress the dynamic range of the music for the recording media, then expand it during playback. The net result is lower distortion at amplitude extreme's, and a better signal to noise ratio.

The biggest difference between HDCD and things like DBX, is that the de-compressor (or decoder) is controlled by a seperate control signal, not the amplitude of the signal itself. This was always DBX and Dolby's problem. If the playback amplitudes weren't properly calibrated, the whole system was thrown off. In the case of HDCD, the control signal is buried into the noise floor created by the dithering algorithm used when converting from 20 bits to 16. That's why a properly ripped HDCD retains the HDCD encoding. The control signal is buried in the LSB of the noise floor, you won't hear it, but a decoder that's looking for it's specific signature, will read it.

The compression curves of HDCD were chosen as a compromise, to maximize the improvement in sound quality when properly decoded, but NOT to sound overly compressed when played back without proper decoding. Inspite of this, some producers like the compressed sound, and use the encoders hoping no one will actually use the decoder!

Happy listening!
Dave

[Skunk]

Can someone give me the actual details or has the info all dissappeared into Microsoft's clutches.

Thanks to Gerald Jansen at www.diyaudio.com , and google of course: http://www.fine21.net/hdcd/data.htm

[crooner]

I'm going to try an interesting experiment to see if HDCD data can be successfully recovered from the SB3.

I'm going to play a FLAC file created from a HDCD disc ripped using EAC.

Since I don't have an HDCD compatible DAC, I'll take the digital out from the SB3 and feed a Philips standalone CD-R machine.

The resulting CD-R should enable HDCD decoding in Windows Media Player.