[jaddie]

Quote:

Interesting explanations! So if stereo sounds unnatural, how do binaural recordings fit in there since they are stereo per se, but the recording technique makes it sound more three-dimensional?

Binaural is a very special case.  Two microphones, ideally in either real ears or an artificial head with simulated ears, is used to record all information that a real human set of ears would.  Then the recordings are played back on headphones, usually open, on ear.  One of the more developed and refined processes is called Holophonics by Hugo Zuccarelli.  The 3D effect can be startling, encompassing directionality from everywhere around the head (download the holophonics demo, you'll see what I mean -- http://www.holophonic.ch/archivio/testaudio/Cereni%20-%20Holophonic.mp3), or somewhat mild, but in any case there's no good way to play that recording on a set of speakers.  The stereo effects are very lack-luster on anything but headphones, and completely collapses on speakers.  
 
Quote:

Mmk that's good to know then. I'm not sure what artifacts sound like though. They can produce sound above 20 kHz?

Artifacts are undesired by-products.  They should not be audible in a good system.  Yes, high rate systems can produce signals above 20KHz.
 
 
Quote:

I don't even know what oversampling is; is it the same as up-sampling? 

 
Kinda, the goal is to up the sampling frequency so the reconstruction filters don't have to be quite as severe.
Quote:

 If 24-bit is only really advantageous for post-processing, then why do "normal people" buy such music and don't use DSPs?

Some people have to have what they believe is "the best", even if the best is imaginary.  Do the comparisons for yourself, then if you still really want to shop for high-rate files, see if you can find any you want that are confirmed to be 24/88, 24/96 or 24/192 all the way from the original master to your files.  If you can't confirm the file's heritage, you may want to hold off.  
 
I hate it when people post stuff they aren't sure about, so I'm not posting a link for this, but I recall hearing that AIX Records produces both high-rate and surround recordings, and if I recall, they produce the surround recordings in both audience and band perspectives.  Sorry if I'm wrong, just can't find it on their site quickly, perhaps someone with more time to dig can correct me.

 

[stv014]

Quote:

Kinda, the goal is to up the sampling frequency so the reconstruction filters don't have to be quite as severe.

 
The requirements on the overall filter response do not actually change, it is only easier to implement a digital filter with a very fast roll-off than an analog one. In other words, the advantage is that the analog filters do not need to be as severe.
 
Oversampling vs. upsampling in the context of DACs is usually audiophile terminology to differentiate between using a typical modern oversampling DAC chip only, and upsampling the signal first with an external DSP chip (or even software) and then feeding it to the - possibly old style non-oversampling - DAC. However, upsampling will not magically avoid the trade-offs involved in designing a reconstruction filter. On the other hand, it allows for more control over the filtering. But the filters built into a decent oversampling DAC chip are generally good enough to be transparent at 44.1 kHz anyway, so the "upsampling" approach basically just adds more complexity and cost without real audible benefits (which is not unusual for audiophile products).

 

[stv014]

Quote:

Originally Posted by miceblue /img/forum/go_quote.gif
 
Mmk that's good to know then. I'm not sure what artifacts sound like though.

 
They are not supposed to be audible at 44.1 kHz with good filtering (which is not that hard to achieve), that is the point. The differences (ringing etc.) between the filters can be audible at low sample rates like 22.05 or 11.025 kHz.

 

[kkcc]

I generally love technological progression but with music, too many variables and bottlenecks exist in the production or post process that would have contributed to a much larger difference than HD vs CD formats. 
 
I would rather buy some nicely produced CDs than some lousy HD files.  I've learned to only consider purchasing HD version for an album that has excellent reviews on its production.  Even tho I've HD daps I really ain't a believer.  Good music always trump good format.

 

[jaddie]

Quote:

 
The requirements on the overall filter response do not actually change, it is only easier to implement a digital filter with a very fast roll-off than an analog one. In other words, the advantage is that the analog filters do not need to be as severe.

Thanks for fixing my poorly written statement...I over-simplified.  What I meant is this:
http://www.analog.com/static/imported-files/tutorials/MT-017.pdf

 

[StudioSound]

On the subject of HD vs CD quality, can anyone explain these?

There's clearly nothing about 40kHz here, or on any of the other tracks from this album. Why is this a 192kHz file? Is there any benefit to it, or is it simply paying more for literally zero benefit over a 96kHz track?

Similar situation here - very little above 40kHz. But what is up with those high energy bands across the track? Shouldn't those be filtered out?

Isn't a -20dB spike that goes on for about 20 seconds potentially damaging to your equipment? Same thing is happening here around 30kHz as well, and above that it's just noise. (clearly sourced from DSD)

A high energy band around 18-20kHz, or around 30kHz seems to be common in a lot of "HD" tracks. It just seems like all you're paying for is more interference/noise in most cases?

If my hearing only extends to 18kHz (which is very quiet) shouldn't I be filtering out everything above that?

 

[miceblue]

Yeah I'm wondering how to interpret 24/192 or 24/96 music too. What is all of the 'blue stuff' in the spectrograms? Noise? Some of my music also has the 'random high frequency line' in the spectrum as well.

 

[jcx]

narrow, constant flat high frequency lines are probably "spurs" - spurious tones that are coming from something in the system - not the music

 

[miceblue]

I sent an e-mail inquiry to a vendor that sells digital HD music and the respondent replied:
 
Quote:

Thank you for your question. We receive the most up to date masters from our record label partners. Each record label has their own recording/mastering process when it comes to high res. We are not a record company. We don't do anything to music provided to us by record companies. If something needs to be change or revised it is sent back to label for the process to be taken care of. We are just a retailer. We don't downsample or upsample. 
 
Whenever possible we try to provide as much information as possible about the mastering process that is giving to us by the labels. 
 
Older recordings when remastered may present qualities that originally were unnoticeable when first released or in previous formats. We only hope that our label partners have taken note of that. When it is an issue we do our best to have them correct it.
 
In general with regards to our hi-res content, we DO test them in-house as well as by an independent 3rd party, at various time points within each track in the album. What we look for is a gradual rolloff in frequencies up to the appropriate point for 96/24, 192/24, etc, and reject anything with an obvious brickwall cutoff. Note that depending on where in a track a sample is taken, the rolloff may be at a slightly different frequency, which is natural with dynamic music.
 
I hope this information is helpful.

 

[jaddie]

I sent an e-mail inquiry to a vendor that sells digital HD music and the respondent replied:

Meh. Sounds quite familiar

 

[miceblue]

Sorry to bring this thread up again, but what exactly is the point of having a studio master and a CD master? i.e. why not just have a studio master down-sampled to 16/44.1 and burned to a CD?

 

[bumeibuhuo]

I do purchase the occasional HD album for kicks but I don't seriously expect it to be any better than a Redbook album.

 

[xnor]

What you're effectively paying for is the sometimes less destructive mastering. That's it.
 
(Other times it's the same mastering, sometimes even just 44.1/16 resampled (!), ...)

 

[miceblue]

So bit depth affects the noise floor of a recording.

Hypothesis: If HD music sampling rates (> 48 kHz) make a difference in audio quality, then downsampling HD music with signals above maximum frequency obtained from the Nyquist frequency defined by the Red Book standard (22.05 kHz) would degrade the audio quality because the natural ultrasonic frequencies that only HD sampling rates could capture would be eliminated.

Either this ABX test was a complete failure, or I was identifying a pattern and answering the wrong answer. What are the chances of getting 0/10 on a coin flip?

Spoiler: abx results

foo_abx 1.3.4 report
foobar2000 v1.3.3
2014/08/24 13:27:15

File A: C:\users\Michelle\Desktop\LessLoss ABX\LessLoss - Drums Duet No. 6.wav
File B: C:\users\Michelle\Desktop\LessLoss ABX\LessLoss - Drums Duet No. 6(1).wav

13:27:15 : Test started.
13:30:32 : 00/01 100.0%
13:31:26 : 01/02 75.0%
13:32:34 : 01/03 87.5%
13:32:50 : 01/04 93.8%
13:33:30 : 01/05 96.9%
13:34:19 : 01/06 98.4%
13:34:52 : 01/07 99.2%
13:35:17 : 01/08 99.6%
13:35:36 : 01/09 99.8%
13:35:59 : 01/10 99.9%
13:37:06 : 01/11 100.0%
13:37:21 : Test finished.

 ---------- 
Total: 1/11 (100.0%)

Setup was Foobar2000 via Wine, Chord Hugo, HiFiMAN HE-560 (15-50,000 Hz frequency response)
Track A is the original 24/96 WAV recording. File B is a downasmpled 24/48 version.

The passage I was listening to definitely has ultrasonic frequencies with the original recording


I didn't downsample to 44.1 kHz because I don't know what the downsampler does (I'm using XLD for OS X). At least from 96 khz to 48 kHz, you sample exactly half as much. Still though, theoretically the file would be degraded because there are clearly natural frequencies above the 24 kHz maximum frequency obtained from a 48 kHz sampling rate.


Apparently 0/10 correct answers in a coin flip is 0.09%, so I'm pretty sure I was just clicking on the wrong answer.


Track of choice, "LessLoss - Drums duet #6," was obtained legally for free from:
http://www.lessloss.com/drums-drums-drums-p-203.html

 

[bigshot]

Hypothesis: If HD music sampling rates (> 48 kHz) make a difference in audio quality, then downsampling HD music with signals above maximum frequency obtained from the Nyquist frequency defined by the Red Book standard (22.05 kHz) would degrade the audio quality because the natural ultrasonic frequencies that only HD sampling rates could capture would be eliminated.

 
The problem with that theory is that tests by the AES indicate that frequencies outside the range of 16/44.1 recording add absolutely nothing to the perception of sound quality of music. If we can't hear them, they don't matter to anyone but bats.
 
However a high sampling rate can wreak havoc in some home audio components that aren't designed to deal with super audible frequencies, creating artifacting and distortion down in the audible spectrum. In these cases, having the super audible frequencies in the recording makes it poorer quality, not better.
 
So basically, it's a no win situation. If the inaudible frequencies are there, it can only hurt the sound quality. It can't improve it.
 
Downsampling simply removes the information that isn't able to be contained in high sampling rate/high bitrate files. It doesn't affect the frequencies and dynamic range that humans can hear at all. There is no more resolution in 20Hz to 20kHz and 90dB of dynamic range in 24/96 than there is in 16/44.1. The difference is all outside our ability to hear. Useful for mixing and mastering, pointless for listening to music in the home.