How to verify HD audio?

[6speed]

With so many sources for HD audio (>16/44) now, what methods are available for verifying a difference or benefit. What software is needed and what does one look for?

1. Increase bandwidth: Any content >20kHz.
2. Increased dynamic range: The higher the better, and there are several algorithms for calculating dynamic range.
3. Decreased noise floor: Is there any way to calculate this or visualize it vs frequency?
4. Peak level: I assume 0dbFS is bad and may be an indicator of clipping?
5. Clipping: How do we calculate this programmatically without visually reviewing the waveform?
6. Playback artifacts: Given different playback bandwidths, filtering algorithms, etc is there anything to investigate within the source material or is that exclusively a function of the playback chain?

 

[Blumlein 88]

Looking for signal beyond 22 khz is one thing. Audacity can show that in either FFT or spectrogram form. So can other softwares like Adobe Audition or Musicscope or some others. Even the various output displays in Foobar.

Dynamic range has nothing to do with it. In fact many modern remasterings have poor dynamic range. No recordings use even 16 bits anyway.

Actually rather than answering the rest checking for high frequency content is about the only one. HD can have clipping, high or low noise floors, high or low dynamic range, various filtering types, etc etc. HD merely means it is all in a high sampling rate. And even that could be artfully faked with ease. The HD idea implies nothing you can count on beyond a greater than 44 khz sample rate.

 

[Vincent Kars]

I use MusicScope
Tells you a lot about the content
http://www.thewelltemperedcomputer.com/SW/AudioTools/Spectrum.htm

A 24 bit 192 kHz recording made by upsampling a CD

 

[6speed]

My goal was to compare different releases for the same albums, to see if the HD or remastered versions were better or worse. This was an objective comparison, not a listening test.

I found a few free apps that can make spectrograms, which is a convenient way of comparing different releases of the same track. However, what I settled on is simply comparing crest and R128 DR figures in JRMC. This confirms some commonly held assumptions:

1. Many remasters have a lower DR than the original CDs.
2. HD versions offer no additional DR.
3. HD versions often provide negligible content >20kHz.

What is surprising is when an album has been remastered several times and the latest HD version happens to have the highest DR.