When is Hi-Res not really Hi-Res?

[wgscott]

Why not just spot-check, and then focus on those that look suspicious?

 

[Bruce B]

Why not just spot-check, and then focus on those that look suspgicious?

Because even doing that, some have slipped through the cracks.

 

[rbbert]

there is no excuse for digital clipping. Seeing some of the recent graphs in this thread are just sickening. there is an awakening taking place and thats good.

I'm not sure what you mean? This has been going on for over a decade, and while overall I think newer releases overall may be a little better in this regard than those from a couple of years ago, it's probably not from any kind of "awakening", it's more likely that even the most tin-eared listeners realized that music with a dynamic range of only 10-15 dB just isn't very pleasant to listen to

 

[Julf]

I'm not sure what you mean? This has been going on for over a decade, and while overall I think newer releases overall may be a little better in this regard than those from a couple of years ago, it's probably not from any kind of "awakening", it's more likely that even the most tin-eared listeners realized that music with a dynamic range of only 10-15 dB just isn't very pleasant to listen to

Well, the loudness war has been going on for at least 20 years, and I don't see that getting much better, but hi-res downloads are a fairly recent thing, and along with that there has been the temptation to offer quick-and-dirty "hi-res" recordings that are made by up/resampling and zero-padding CD/redbook material, or in best case, resampling sacd material.

I think the awakening tomelex is talking about is the rising awareness of this practice, and the increasing amount of enthusiasts who don't just accept the vendor clams of "hi-res", but actually do at least rudimentary checks using spectrum analysis software and things like the FLAC encoder to detect zero-padding. Some hi-fi magazines (such as "Hi-fi News and Record Review" in the UK) have actually started to do the same as part of their reviews.

 

[rbbert]

Unfortunately there are quite a few "true hi-res" (in the sense of genuine 24/96 or even 24/192) recordings offered for sale which are still compressed and/or peak limited all to hell. There might be increased awareness of this, but I personally haven't noticed much change in the marketplace offerings. I hope I'm wrong.

 

[Julf]

Unfortunately there are quite a few "true hi-res" (in the sense of genuine 24/96 or even 24/192) recordings offered for sale which are still compressed and/or peak limited all to hell. There might be increased awareness of this, but I personally haven't noticed much change in the marketplace offerings. I hope I'm wrong.

Compression / peak limiting is at least an "artistic choice" by someone - artist, engineer, producer, label executive - while I would hesitate calling upsampling or zero-padding an artistic choice

 

[Orb]

Just as an example of high rez situation; 3 of the 5 high rez download albums reviewed by Hifi News this month when analysed by them were not true-native high rez and had flaws....
Cheers
Orb

 

[rbbert]

Compression / peak limiting is at least an "artistic choice" by someone - artist, engineer, producer, label executive - while I would hesitate calling upsampling or zero-padding an artistic choice

But Tom's post specifically mentioned "digital clipping", not fake hi-res.

 

[Julf]

But Tom's post specifically mentioned "digital clipping", not fake hi-res.

Right. And it was to that part that I referred to with "the loudness war has been going on for at least 20 years, and I don't see that getting much better".

 

[rbbert]

Right. And it was to that part that I referred to with "the loudness war has been going on for at least 20 years, and I don't see that getting much better".

Which was the whole point of my post; Tom apparently thinks there is increased awareness (and therefore improvement), whereas I'm not so sure...

 

[Julf]

Which was the whole point of my post; Tom apparently thinks there is increased awareness (and therefore improvement), whereas I'm not so sure...

And I don't disagree - as long as 99% of music buyers don't care, the record companies won't either.

 

[Orb]

In agreement with you guys, because the jazz and classical "lite" releases seem more often than not to suffer more these days from loudness wars, this shows it has spread beyond the rock/pop culture of studio recordings, and this is being seen even with hirez recordings in these categories.

Quoting what I mentioned in a different thread this month:

Just to add a modern example of a "lite" genre recording.
High rez version of Diana Krall's Glad Rag Doll has most of its spectra content up to 1khz between -10dBFS and -30dBFS, this is from an artist who is recognised to having great talent and fits closer to the audiophile niche.

I did emphasise 0dBFS, but should be noted my concern in that post also for up to -20dBFS, which shows some performance considerations for certain DAC architectures and implementation.
Cheers
Orb

So two trend headaches still around with hirez, and one of those common to both standard CD and hirez, a good recording is at least 50% better in terms of dBFS figures and usually over twice as good (around -50dBFS to -60dBFS for 1kHz related music spectra as reported by Hifi News - they show all FR range but I just focused on lower as it fitted with other thread)
Cheers
Orb

 

[sasully]

Bruce,

When you say 24 bit compressed versions of the original 16 bit CD can still be "hi-rez", it's distinction without a difference.

The whole point of offering 24 bit files is that they allow for substantially greater dynamic range than 16 bit files. When anyone offers a 24 bit file claiming "hi-rez" with 1/2 the dynamic range of the original 16 bit CD, that's ugly.

Redbook offers 96dB of dynamic range; with dither and noise shaping it offers well over 100 dB of dynamic range. There is actually little audible point in offering 24 bit consumer format files (though there's good reason to do recording and production at 24 bits) . Mainly they sell it because they can, and because people think it guarantees better sound, which of course it does not. A better point (though it's mainly in the category of 'foolproofing') would be that consumer AVRs typically process input audio at 24bits anyway, so might as well offer it at 24 in the first place.

 

[sasully]

Why do you want to do another conversion? Every change degrades the sound.

Nonsense.

*So much* nonsense on this thread, it's mind boggling.

 

[sasully]

Well, the loudness war has been going on for at least 20 years, and I don't see that getting much better, but hi-res downloads are a fairly recent thing, and along with that there has been the temptation to offer quick-and-dirty "hi-res" recordings that are made by up/resampling and zero-padding CD/redbook material, or in best case, resampling sacd material.

.

Since early days of DVDA and SACD, "high rez' releases have included releases where the dynamic range has been limited, and/or sourced from standard-rate sources. The label 'high resolution' has *never* been a guarantee that all the 'benefits' of the formats were exploited (any more that Redbook's 96db of dynamic range have been).

 

[Bruce B]

Nonsense.

*So much* nonsense on this thread, it's mind boggling.

Everytime you do a sample rate conversion it degrades the sound. This is a fact.

 

[sasully]

Everytime you do a sample rate conversion it degrades the sound. This is a fact.

1) you wrote *every change* degrades the sound. That's nonsense.

2) and more fundamentally, you need to define 'degrade'. If I sample at 192 kHz and carefully filter out the content above 44 kHz later (i.e., to resample at 88kHz) , i've removed content that no one will ever hear in the first place. That's decreasing the bandwidth but is it 'degrading' the *sound*'?

 

[Andre Marc]

1) you wrote *every change* degrades the sound. That's nonsense.

2) and more fundamentally, you need to define 'degrade'. If I sample at 192 kHz and carefully filter out the content above 44 kHz later (i.e., to resample at 88kHz) , i've removed content that no one will ever hear in the first place. That's decreasing the bandwidth but is it 'degrading' the *sound*'?

Sasully:

You seem to be very knowledgeable in the area of computer audio with some strong opinions...can you give details of your set up/system
and if you have any industry affiliation?

 

[RBFC]

It would be nice if both sides would clarify their talking points. Why does a change in sampling rate change the sound, or why doesn't it. To merely make declamatory statements does not provide any educational or discussion-based value!

Lee

 

[Bruce B]

1) you wrote *every change* degrades the sound. That's nonsense.

If you had been reading the thread, I was making reference to sample rate and bit depth

2) and more fundamentally, you need to define 'degrade'. If I sample at 192 kHz and carefully filter out the content above 44 kHz later (i.e., to resample at 88kHz) , i've removed content that no one will ever hear in the first place. That's decreasing the bandwidth but is it 'degrading' the *sound*'?

Yes, it is degrading the sound. And to answer Lee... If you change the sample rate from 192 to 88.2, the sample rate change does not just happen from 44.1kHz to 96kHz... .it affects the whole bandwidth, even in the audible range. Even if you "carefully" apply a filter, that filter is affecting the audible range. Let's say the filter slopes 6dB-12dB/octave. Well one octive is from 20kHz to 40kHz. The slope starts back in the audible range.