HDTracks Offers 352/8/24 Downloads

[Boris75]

What's the issue again???

 

The issue is that all this noise wastes bits and, more importantly, can generate distortions into the audible band if not filtered properly.

 

By the way, one question mark at the end of the sentence is enough to get people to understand that a question is being asked.

[Kees de Visser]

If anything, it might be preferable to have everything above say 24kHz completely filtered out.

As soon as you figure out a way to do that perfectly without negatively affecting sound quality, please let all the digital audio engineers who agreed 8x oversampling was better than a brickwall filter in the CD player know how it's done.

There are very good SRC sample rate converters these days that come very close to perfection. My favorite is iZotope. I recently did some testing by downloading a DXD demo from 2L, converting it myself to 32/44.1, converting it back to DXD and doing a "null test" (subtracting one signal from the other). The difference below 20 kHz is very low level, far below 24 bit dither level. To me this indicates that up to 20 kHz these files are as identical as it gets.

IF there are audible differences between the two rates, it is most likely caused by the high frequency content.

[Jud]

There are very good SRC sample rate converters these days that come very close to perfection. My favorite is iZotope. I recently did some testing by downloading a DXD demo from 2L, converting it myself to 32/44.1, converting it back to DXD and doing a "null test" (subtracting one signal from the other). The difference below 20 kHz is very low level, far below 24 bit dither level. To me this indicates that up to 20 kHz these files are as identical as it gets.

IF there are audible differences between the two rates, it is most likely caused by the high frequency content.

 

Yes, but what does the fact that there are excellent SRC applications using non-brickwalled filtering have to do with the fact that brickwall filtering at 24kHz wouldn't sound very good?

[Boris75]

Yes, but what does the fact that there are excellent SRC applications using non-brickwalled filtering have to do with the fact that brickwall filtering at 24kHz wouldn't sound very good?

 

Who is suggesting that brickwall filtering would sound good?

[One and a half]

The issue is that all this noise wastes bits and, more importantly, can generate distortions into the audible band if not filtered properly.

 

By the way, one question mark at the end of the sentence is enough to get people to understand that a question is being asked.

 

With DSD, the sound quality from the same music compared with even hires PCM is, in broad terms, far superior. It is often noted on pages at CA that brickwall filters close to 20kHz have detrimental effects on sound quality, and those that are pushed higher up the band can have a slower sloping filter avoiding this issue.

 

Wasted bits due to the noise, a compromise, to attain the quality of DSD, this is what the technology is, if there was a solution for the wasted bits, it would have been thought of long ago, but it's here for now. Have you heard this noise during playback?

 

DXD and DSD are treated equally as far as SQ is concerned, DSD has a lot more efficiency.

 

More question marks are an emphasis only, I could have used ?!. If that's not grammatically correct, then a swear word would be closer to grammatical truth but not appropriate for these pages.

[Jud]

Yes, but what does the fact that there are excellent SRC applications using non-brickwalled filtering have to do with the fact that brickwall filtering at 24kHz wouldn't sound very good?

 

Who is suggesting that brickwall filtering would sound good?

 

"It might be preferable to have everything above say 24kHz completely filtered out."

[Boris75]

"It might be preferable to have everything above say 24kHz completely filtered out."

 

Thanks. I missed that remark. I for one have always been suggesting on this thread that this file could harmlessly be converted into a 24bit 96kHz file, which allows capturing frequencies up to 48kHz and thus gives more than ample space for gentle filtering. I agree that going to 48kHz sampling complicates filtering with potential sonic consequences.

[Boris75]

With DSD, the sound quality from the same music compared with even hires PCM is, in broad terms, far superior.

 

I generally enjoy DSD, but in the one instance when I could compare DSD and PCM head-to-head (2L's La Voie Triomphale in 24bit 192kHz PCM and DSD on my Benchmark DAC2), PCM won by a slender margin. The 24bit 192kHz PCM offered a bit more detail and presence than DSD, which sounded a bit laid back (though still extremely good).

 

It is often noted on pages at CA that brickwall filters close to 20kHz have detrimental effects on sound quality, and those that are pushed higher up the band can have a slower sloping filter avoiding this issue.

 

I fully agree with this, which is why I suggested downsampling this file to 24bit 96kHz and not 24bit 48kHz.

[One and a half]

I generally enjoy DSD, but in the one instance when I could compare DSD and PCM head-to-head (2L's La Voie Triomphale in 24bit 192kHz PCM and DSD on my Benchmark DAC2), PCM won by a slender margin. The 24bit 192kHz PCM offered a bit more detail and presence than DSD, which sounded a bit laid back (though still extremely good).

 

Could it be the emphasis on the Benchmark DAC2 leans towards PCM rather than DSD to give you that impression. My DAC works well on DSD, but misses the mark with some PCM material, the opposite. I often wonder if there is a need to buy a PCM DAC and a DSD DAC and switch (or mix) between the two at the analog outs.

[Kees de Visser]

It is often noted on pages at CA that brickwall filters close to 20kHz have detrimental effects on sound quality, and those that are pushed higher up the band can have a slower sloping filter avoiding this issue.

Would the be the effect (distortion) of the filter, or the absence of the high-frequency content that was filtered out ?

[Jud]

It is often noted on pages at CA that brickwall filters close to 20kHz have detrimental effects on sound quality, and those that are pushed higher up the band can have a slower sloping filter avoiding this issue.

Would the be the effect (distortion) of the filter, or the absence of the high-frequency content that was filtered out ?

 

The former.

[Teresa]

I generally enjoy DSD, but in the one instance when I could compare DSD and PCM head-to-head (2L's La Voie Triomphale in 24bit 192kHz PCM and DSD on my Benchmark DAC2), PCM won by a slender margin. The 24bit 192kHz PCM offered a bit more detail and presence than DSD, which sounded a bit laid back (though still extremely good).

 

I also downloaded 2L samples from their free test bench high resolution comparison files. I preferred the DXD 24/352.8kHz over the DSD on my Teac DAC. I believe it is because the original masters are DXD 24/352.8kHz. If the original masters were DSD I expect it would be the other way around.

 

Could it be the emphasis on the Benchmark DAC2 leans towards PCM rather than DSD to give you that impression. My DAC works well on DSD, but misses the mark with some PCM material, the opposite. I often wonder if there is a need to buy a PCM DAC and a DSD DAC and switch (or mix) between the two at the analog outs.

 

That is possible. However, since the original was 24/352.8kHz I would expect the PCM to sound better to the DSD if all things are equal on the DAC. My Teac sounds great with both high resolution PCM and DSD, although I think DSD sounds the best especially at 5.6MHz but in the case of 2L music files I liked the 24/352.8kHz PCM better. However, I have read that many DACs are better at DSD or PCM.

[Hailey]

Nice, but with HDTracks record- who knows? It's interesting though, isn't 352K the equivalent of DSD128?

 

-Paul

 

 

Curious, what's the rap on HDTracks?

[wgscott]

Sometimes they take a small amount of wine, put it in a large bottle, and charge more.

[Hailey]

Sometimes they take a small amount of wine, put it in a large bottle, and charge more.

 

Gotcha, thanks. I figured they're all like that, though: Ripping CDs at high resolution, which results in...CD-level resolution.

 

i downloaded a Munch/BSO album from them and that's what I got, anyway.

 

Is any other audiophile download source any different?

[firedog]

No, mostly they aren't like that either at HDT or the other hi-res sites. The sites for the most part are just vendors-they are supplied files by labels and resell them. And with relatively few exceptions, the material is hi-res.

 

So the main culprit in this area are labels/distributors. The download sites are secondary culprits.

 

The sites could be more vigilant about preventing any fake hi-res from appearing at all. Till now they haven't done this well enough.

 

 

 

Gotcha, thanks. I figured they're all like that, though: Ripping CDs at high resolution, which results in...CD-level resolution.

 

i downloaded a Munch/BSO album from them and that's what I got, anyway.

 

Is any other audiophile download source any different?

[Boris75]

It's interesting though, isn't 352K the equivalent of DSD128?

 

Mark Waldrep has indicated in this post on his blog that DSD128 provides the same resolution as 24-96PCM:

[...] DSD up to 128 (equivalent to about 96/24...if you forget about the noise)

[tailspn]

No, mostly they aren't like that either at HDT or the other hi-res sites. The sites for the most part are just vendors-they are supplied files by labels and resell them. And with relatively few exceptions, the material is hi-res.

 

There are some exceptions.

 

First understand something of the recording process. By in large, audio recordings can be separated into studio and acoustic. Studio recordings as the name implies, are recorded in a space where the space adds little or nothing to the ambiance you hear from the recording. That's all added synthetically in the production process. Bottom line; there's little or no reference to those musicians performing in an acoustic space, for which we have a life's accumulated experience of judging the reality. These also account for at least 80% of all available recordings. They can sound very nice, just not realistic.

 

Acoustic recordings are those recorded in an acoustic environment, where the environment makes up the spaciousness information, and can therefore be judged from our natural hearing experience.

 

Since ambience/spaciousness information is very low level, the choice of recording formats has little bearing in recording and delivering of studio music. The natural spaciousness information is not there to begin with. However, acoustic music recorded in a performance space, where the space is an integral ingredient of the recording, is rich with very low level sound cues, which we easily recognize as reality.

 

With that very overly simplified definition, the choice of delivery format and sampling rates from download sites for studio recorded music is largely marketing BS. That's not true of acoustic performance recordings however. And to your point, it goes a step further.

 

With very few exceptions, the content sold on download sites, high-res claimed or not, are from rips from CD's and SACD's. That's how the labels distributed their projects to those download stores. They're simply selling an alternate delivery modality, one step further removed from the "studio master" they all tout.

 

The processing chain of a high-resolution recording (recorded and delivered in higher than 44.1/16 CD quality) consists of the original session recording(s), edited, and optionally post processed sweetened, ending in a form named the Edit Master. That Edit Master, to be made deliverable, must then be authored to the deliverable process (SACD, DVD-A, Blu-Ray), which involves further processing. For SACD, that involves "lossless" compression and water marking, yielding a .dst format (Direct Stream Transfer) file, and is known as the ISO Cutting Master. From this the DSD layer of a SACD is manufactured. DVD-A and Blu-Ray have similar, but different authoring processes for their optical disk manufacturing.

 

The download site I work with, nativedsd.com does not offer ripped media downloads. Only original Edit Masters from the labels are offered for sale. Does this make a difference? For acoustic recorded music it does, especially in multi-channel. This is the factor nativedsd is using to differentiate itself from the other bulk quantity download sites. The downside is the agonizingly long amount of time to get labels to retrieve their Edit Masters to be uploaded.

 

There are other download sites employing the same care. As mentioned, Opus 3, Blue Coast, Acoustic Sounds (for THEIR tape > DSD remastered recordings) come to mind. But the point is, it's the Edit Master that's the crown jewels in this quality download business.

[Hailey]

There are some exceptions.

 

First understand something of the recording process. By in large, audio recordings can be separated into studio and acoustic. Studio recordings as the name implies, are recorded in a space where the space adds little or nothing to the ambiance you hear from the recording. That's all added synthetically in the production process. Bottom line; there's little or no reference to those musicians performing in an acoustic space, for which we have a life's accumulated experience of judging the reality. These also account for at least 80% of all available recordings. They can sound very nice, just not realistic.

 

Acoustic recordings are those recorded in an acoustic environment, where the environment makes up the spaciousness information, and can therefore be judged from our natural hearing experience.

 

Since ambience/spaciousness information is very low level, the choice of recording formats has little bearing in recording and delivering of studio music. The natural spaciousness information is not there to begin with. However, acoustic music recorded in a performance space, where the space is an integral ingredient of the recording, is rich with very low level sound cues, which we easily recognize as reality.

 

With that very overly simplified definition, the choice of delivery format and sampling rates from download sites for studio recorded music is largely marketing BS. That's not true of acoustic performance recordings however. And to your point, it goes a step further.

 

With very few exceptions, the content sold on download sites, high-res claimed or not, are from rips from CD's and SACD's. That's how the labels distributed their projects to those download stores. They're simply selling an alternate delivery modality, one step further removed from the "studio master" they all tout.

 

The processing chain of a high-resolution recording (recorded and delivered in higher than 44.1/16 CD quality) consists of the original session recording(s), edited, and optionally post processed sweetened, ending in a form named the Edit Master. That Edit Master, to be made deliverable, must then be authored to the deliverable process (SACD, DVD-A, Blu-Ray), which involves further processing. For SACD, that involves "lossless" compression and water marking, yielding a .dst format (Direct Stream Transfer) file, and is known as the ISO Cutting Master. From this the DSD layer of a SACD is manufactured. DVD-A and Blu-Ray have similar, but different authoring processes for their optical disk manufacturing.

 

The download site I work with, nativedsd.com does not offer ripped media downloads. Only original Edit Masters from the labels are offered for sale. Does this make a difference? For acoustic recorded music it does, especially in multi-channel. This is the factor nativedsd is using to differentiate itself from the other bulk quantity download sites. The downside is the agonizingly long amount of time to get labels to retrieve their Edit Masters to be uploaded.

 

There are other download sites employing the same care. As mentioned, Opus 3, Blue Coast, Acoustic Sounds (for THEIR tape > DSD remastered recordings) come to mind. But the point is, it's the Edit Master that's the crown jewels in this quality download business.

Thank you so much for this tour d'horizon. Most illuminating. And nativedsd.com looks really interesting. Signing up now...

[bmoura]

Thank you so much for this tour d'horizon. Most illuminating. And nativedsd.com looks really interesting. Signing up now...

 

Once you've signed up, be sure to check out the free Just Listen music downloads on NativeDSD. Some very nice music there.

[firedog]

With very few exceptions, the content sold on download sites, high-res claimed or not, are from rips from CD's and SACD's. That's how the labels distributed their projects to those download stores. They're simply selling an alternate delivery modality, one step further removed from the "studio master" they all tout.

 

I know you are in the industry, but I think you are overstating the point. The discussion was about whether the material sold as hi-res downloads is actually hi-res, not what the ultimate delivery method for hi-res is.

 

Much of the hi-res being sold today is either derived directly from DSD or SACD, or is a remaster made from an analog master transcribed to hi-res digital. Lots of other material is recorded in 24/44.1 or higher bitrates. All of those are legitimate "hi-res" as I see it. The exception would be when it is derived from an SACD or other disc which itself was derived from a 16 /44.1 source.

 

Saying it is mostly derived from CDs I think is quite an exaggeration.

 

Maybe offering only edit masters for download is a superior method, but if a recording is sourced from hi-res (even a legit DSD/SACD disc) I don't see any reason to not call it hi-res. Maybe it isn't the best or perfect delivery method; but lots of converted SACD albums turned into downloads that I have are the best sounding version available of that album. Since no other hi-res version is likely to become available in many cases, I don't have any complaint about that, as long as they don't claim that it is the actual "studio (edit) master". Most of this music is not recorded in anything like an acoustic environment anyway...

[REShaman]

There are some exceptions.

 

First understand something of the recording process. By in large, audio recordings can be separated into studio and acoustic. Studio recordings as the name implies, are recorded in a space where the space adds little or nothing to the ambiance you hear from the recording. That's all added synthetically in the production process. Bottom line; there's little or no reference to those musicians performing in an acoustic space, for which we have a life's accumulated experience of judging the reality. These also account for at least 80% of all available recordings. They can sound very nice, just not realistic.

 

Acoustic recordings are those recorded in an acoustic environment, where the environment makes up the spaciousness information, and can therefore be judged from our natural hearing experience.

 

Since ambience/spaciousness information is very low level, the choice of recording formats has little bearing in recording and delivering of studio music. The natural spaciousness information is not there to begin with. However, acoustic music recorded in a performance space, where the space is an integral ingredient of the recording, is rich with very low level sound cues, which we easily recognize as reality.

 

With that very overly simplified definition, the choice of delivery format and sampling rates from download sites for studio recorded music is largely marketing BS. That's not true of acoustic performance recordings however. And to your point, it goes a step further.

 

With very few exceptions, the content sold on download sites, high-res claimed or not, are from rips from CD's and SACD's. That's how the labels distributed their projects to those download stores. They're simply selling an alternate delivery modality, one step further removed from the "studio master" they all tout.

 

The processing chain of a high-resolution recording (recorded and delivered in higher than 44.1/16 CD quality) consists of the original session recording(s), edited, and optionally post processed sweetened, ending in a form named the Edit Master. That Edit Master, to be made deliverable, must then be authored to the deliverable process (SACD, DVD-A, Blu-Ray), which involves further processing. For SACD, that involves "lossless" compression and water marking, yielding a .dst format (Direct Stream Transfer) file, and is known as the ISO Cutting Master. From this the DSD layer of a SACD is manufactured. DVD-A and Blu-Ray have similar, but different authoring processes for their optical disk manufacturing.

 

The download site I work with, nativedsd.com does not offer ripped media downloads. Only original Edit Masters from the labels are offered for sale. Does this make a difference? For acoustic recorded music it does, especially in multi-channel. This is the factor nativedsd is using to differentiate itself from the other bulk quantity download sites. The downside is the agonizingly long amount of time to get labels to retrieve their Edit Masters to be uploaded.

 

There are other download sites employing the same care. As mentioned, Opus 3, Blue Coast, Acoustic Sounds (for THEIR tape > DSD remastered recordings) come to mind. But the point is, it's the Edit Master that's the crown jewels in this quality download business.

 

Excellent exposition explained easily comprehended.

 

Thank you,

Richard

[Hailey]

I know you are in the industry, but I think you are overstating the point. The discussion was about whether the material sold as hi-res downloads is actually hi-res, not what the ultimate delivery method for hi-res is.

 

Much of the hi-res being sold today is either derived directly from DSD or SACD, or is a remaster made from an analog master transcribed to hi-res digital. Lots of other material is recorded in 24/44.1 or higher bitrates. All of those are legitimate "hi-res" as I see it. The exception would be when it is derived from an SACD or other disc which itself was derived from a 16 /44.1 source.

 

Saying it is mostly derived from CDs I think is quite an exaggeration.

 

Maybe offering only edit masters for download is a superior method, but if a recording is sourced from hi-res (even a legit DSD/SACD disc) I don't see any reason to not call it hi-res. Maybe it isn't the best or perfect delivery method; but lots of converted SACD albums turned into downloads that I have are the best sounding version available of that album. Since no other hi-res version is likely to become available in many cases, I don't have any complaint about that, as long as they don't claim that it is the actual "studio (edit) master". Most of this music is not recorded in anything like an acoustic environment anyway...

 

 

Question: Why assume a recording is sourced from hi-res if it is not explicitly stated to be so?

[firedog]

Question: Why assume a recording is sourced from hi-res if it is not explicitly stated to be so?

 

Because that is the inherent claim of all the high res sites. They all claim they don't upsample material, and have been known to remove files like that when they've been shown to be fake hi-res. I'll reverse the question: on what basis do you assume the default is that the files aren't hi-res? Just an assumption on your part, isn't it?

[tailspn]

Maybe offering only edit masters for download is a superior method, but if a recording is sourced from hi-res (even a legit DSD/SACD disc) I don't see any reason to not call it hi-res. Maybe it isn't the best or perfect delivery method; but lots of converted SACD albums turned into downloads that I have are the best sounding version available of that album. Since no other hi-res version is likely to become available in many cases, I don't have any complaint about that, as long as they don't claim that it is the actual "studio (edit) master". Most of this music is not recorded in anything like an acoustic environment anyway...

 

I couldn't agree with you more! For me, two issues:

 

The term High Resolution Audio has become a meaningless circus. Even the CEA, with their fellow agencies/groups have recently gotten in on the act:

 

DEG, CEA, The Recording Academy® and Major Labels Reach Agreement on Definition for High Resolution - CEA

 

Even good old 44.1/16 CD now has a High-Res classification; Master Quality MQ-C.

 

And I agree with your point about Hi-Res Audio sourced content, including SACD rips being called Hi-Res. But it's the very over used/abused "Studio Master" slogan many download sites use to promote their products, when the source of those products is actually consumer available SACD/DVD-A/optical media rips, which I believe is misleading.

 

Like many here, including yourself, my view of audio delivery quality is from the top down, not the bottom up. I admit we're the tiniest imaginable market, but one I think deserves servicing, now that the specification limitations of optical media distribution have been overcome with download sites. I believe it's important to understand the stages of music content production available for distribution, and if those levels are desired, insist on their availability. And support them!