What Are Hi-Res Downloads Really?

[Antonmb]

Originally Posted by audio bill

rbbert - that's why in my previous post I referenced some companies that produce and record their own music which is mastered in high resolution. The authenticity of these cannot be questioned since they are mastered directly in high res and not up-converted. I agree with you, that so much of a recording's sound quality is determined by the mastering. It cannot be recreated once it's destroyed by a poor mastering job, regardless of any further processing or up-conversion performed later.

Yes, let's be clear what sources we're referring to. Files from the usual hi res download stores (HDTracks, Acoustic Sounds, Pono) which claim to be hi res remasters of older recordings are sometimes suspect, and information regarding provenance is often lacking, so caveat emptor. Those record companies such as Blue Coast and Channel Classics which produce their own recordings and record and master in hi res, on the other hand, provide clear provenance and are true hi res. Reissue after reissue of the usual old war horses should be viewed with some caution, but there are plenty of quite reputable labels out there producing new recordings in real hi res.

[rbbert]

Again, the overwhelming majority (98+%) of older albums originally recorded on analog tape are true "hi-res" transfers. The albums you should be suspect of are those recorded from roughly 1985 (when digital recording became commonplace, and at only 16/44.1 or 16/48. ) through about 2000, when 24/96 and DSD were more widely available for the original recording. Many mid-90's masters were made at 24/44.1 or 24/48, so if you see something from this era offered at those resolutions it's probably accurate.

[o0OBillO0o]

Originally Posted by Antonmb

Yes, let's be clear what sources we're referring to. Files from the usual hi res download stores (HDTracks, Acoustic Sounds, Pono) which claim to be hi res remasters of older recordings are sometimes suspect, and information regarding provenance is often lacking, so caveat emptor. Those record companies such as Blue Coast and Channel Classics which produce their own recordings and record and master in hi res, on the other hand, provide clear provenance and are true hi res. Reissue after reissue of the usual old war horses should be viewed with some caution, but there are plenty of quite reputable labels out there producing new recordings in real hi res.

Yes, good point. Blue Coast and Channel Classics make high resolution AND well mastered recordings. FACT. Still, at $5 a track, it should. I can find a used record (WHOLE ALBUM) with as much fidelity(mastering quality) for $3 at Newbury Comics in Boston.

Originally Posted by rbbert

Again, the overwhelming majority (98+%) of older albums originally recorded on analog tape are true "hi-res" transfers. The albums you should be suspect of are those recorded from roughly 1985 (when digital recording became commonplace, and at only 16/44.1 or 16/48. ) through about 2000, when 24/96 and DSD were more widely available for the original recording. Many mid-90's masters were made at 24/44.1 or 24/48, so if you see something from this era offered at those resolutions it's probably accurate.

I understand what you are saying, though I caution to statistics. It's nice to throw down numbers, but this is the internet. 98% according to who? References?

Analog tape transfers can be captured at high bite rates. Fact. However it's limited by Analog tape. Music that is captured, such as Bluecoast, directly to a PCM or DSD encoding station is going to have better technical capabilities and quality control.

Now, it's up to the trained engineer to take advantage of these benefits. We could do a whole excursion on loudness wars and what enginners (and the artists) are allowing to be done to their music in order to reach a larger audience. I suggest getting a free subscription to TapeOp magazine (Subscriptions | Tape Op - the Creative Music Recording Magazine) if you are truly interested in developing recording skills or mastering the understanding of each recording medium, be it digital or analog.

Wait- I know we all want this on our music servers and portable devices, in the purest state possible- because audiophile.

Please continue to purchase high resolution audio files if you can guarantee it's provenance. (Does anyone have a list to share?) I am all for getting closer to the music and the suspension of disbelief (the so called you are there experience). I think because we are smart, we must validate what is what before we become overwhelmed with emotion and make irrational decisions based on little to no evidence.


Bill

[rbbert]

That's why I put "high res" in quotes for the analog transfers.

I have several hundred hi-res downloads. Except for the ones that were originally recorded to 16/44.1 digital (usually identifiable beforehand; the Wikipedia will usually tell you the SPARS code of an album) only a handful (i.e about 5 out of several hundred) are not as advertised.

Very few LP's or CD's provide provenance on the package labelling. By the time you get to looking at matrix numbers you can also have easily checked a hi-res file to see what it is.

The Pono store is apparently in the process of trying to provide as much information as possible about their products, so hopefully others will follow suit.

The music reviews at computeraudiophile nearly always tell you the files' resolution, dynamic range, etc

[chessman]

This conversation is really helpful to me. Thanks to all for weighing in.

Now another mud pie: what matters more, sentence length (say 24 over 16) or frequency of sample (say 96 over 44.1)? My two cents is that frequency increase merely seems to be a way to move "pre-ringing" above the hearing spectrum, but sentence length gives a real improvement in fidelity. Thoughts?

[antipop]

Originally Posted by chessman

This conversation is really helpful to me. Thanks to all for weighing in.

Now another mud pie: what matters more, sentence length (say 24 over 16) or frequency of sample (say 96 over 44.1)? My two cents is that frequency increase merely seems to be a way to move "pre-ringing" above the hearing spectrum, but sentence length gives a real improvement in fidelity. Thoughts?

24bit is vastly more important than sample rate. More bits gives you more dynamic range which is always a good thing.

The impact of higher sample rate is always debatable and you have Nyquist formula.

[o0OBillO0o]

Originally Posted by rbbert

Very few LP's or CD's provide provenance on the package labelling. By the time you get to looking at matrix numbers you can also have easily checked a hi-res file to see what it is.

How are you determining this, digitally?

SPARS Codes are useful

Very few? I have not found this to be the case. Discogs or Analogplanet or two great places to seek out provenance.

To the group:
The code allows for six possible combinations:

• AAA – A fully analogue recording, from the original session to mastering. Since at least the mastering recorder must be digital to make a compact disc, this code is not applicable to CDs.[3]
• AAD – Analog tape recorder used during initial recording, analog tape recorder used during mixing, digital mastering.
• ADD – Analog tape recorder used during initial recording, digital tape recorder used during mixing, digital mastering.
• DDD – A fully digital recording, from the original session to mastering.
• DAD – Digital tape recorder used during initial recording, analog tape recorder used during mixing, digital mastering.
• DDA – Digital tape recorder used during initial recording, digital tape recorder used during mixing, analogue (vinyl) mastering.

[o0OBillO0o]

Originally Posted by chessman

This conversation is really helpful to me. Thanks to all for weighing in.

Now another mud pie: what matters more, sentence length (say 24 over 16) or frequency of sample (say 96 over 44.1)? My two cents is that frequency increase merely seems to be a way to move "pre-ringing" above the hearing spectrum, but sentence length gives a real improvement in fidelity. Thoughts?

You should grab a beer or a beverage and watch these.

Main Page:
Xiph.org: Video

Video 1:
Xiph.Org Video Presentations: A Digital Media Primer for Geeks
Video 2:
Xiph.Org Video Presentations: Digital Show & Tell

Draw your own conclusions.

As far as "pre rining" check out linear and minimum phase filters here: https://ccrma.stanford.edu/~jos/fp/M...e_Filters.html

or dig around in the Google: https://www.google.com/search?q=line...m+phase+filter

[chessman]

Bill, thanks (again!) for some great leads. I have been reading about non-linear distortion and what filters do (and don't do), which is what induced me to start this thread. One other motivator for me to learn this is that I do not want to pay for alleged hi-res that is nothing more than what would result from me using the K-01's upsampling and filters on my own stock of Redbook CDs.

[o0OBillO0o]

Randy- There is a difference in upsampling and oversampling algorithms. Some good some not good, some no one can tell.

Here is an Sample Rate Conversion comparison tool. SRC Comparisons

Although it's going in reverse, generally what we want to is this, from 96 to 44Khz- it does provides some insight to help answer the question about sample rate up or down conversion.

I asked this question (see post link below) because I saw a vendor selling software claiming that it's offline upsampling (Both PCM and DSD) improved the sound of the music you own. Right now, I feel that this is snake oil and a load of crap.

http://www.audioaficionado.org/gener...dio-files.html