HD downloads quality

Enrich, yes the concentration distortions are harmonic (periodical refers to time usually) and I suspect caused by digital processing tools and / or the digital audio workstation tools used for mixing and mastering..

The average Crest value is lower than I would have expected for a low compressed hidefinotion studio recorded album.. but perhaps more compression was used than I expected.

However what is also interesting is on how many heavily compressed or loudness optimised CD or hidef masters have TPL levels above 0dB.. (Meet  Me in London shows no issue here) Now this could produce overload in DAC reconstruction filters and cause distortion. interestingly it also means tracks may sound different or distort differently between different DACs.. something I think I had noticed before but didn't know why... As I say an eye opener.. Thanks for the heads up Erich, perhaps we often take our recording masters for granted whilst we sweat and fuss over our replay equipment.

Simon

This strange message is w.r.t. "Meet Me In London"

I bought it from Naim as a 24/192 .wav & in Windows Explorer is shows as .wav & 24/192 correctly. 

However running in MusicScope it's identified as BWF

Per Wikipedia - this is Broadcast Wave Format & is a standard audio format created by European Broadcast Union as a successor to WAV. Among other enhancements, BWF allows more robust metadata to be stored in the file.

The spectrum in the ultra HF has a more or less flat line up to & beyond 96kHz at levels around -90dB.  

This is unlike many other 24/192 files,   (e.g. only) Eagles "Hotel California" shows a strong spectrum up to aprx 48kHz but beyond that up to 96kHz the level is not much more than noise bouncing around -144dB.

Interestingly some exceptions in modern recordings, e.g. Norah Jones "Come Away With Me" shows similar strong content to that shown in the BWF file between 48kHz & 96kHz,  but rather than -90dB its more like -100dB.   My take on that is it was recorded with modern & very high res master & with one of the bitstream formats,  whereas "Hotel California" originally recorded in 1976 started life in analogue & would not contain any HF content.  So the question is is it worth buying 24/192 with old remasters when 24/96 or even 24/48 will probably carry all the data for good SQ.

The problem is you are paying extra money for something which isn't as advertised.

When you pay for a 192 or 96 kHz version and it only contains data of a 44.1 kHz recording, you're being duped.

I recently read the following statement from Mark Waldrep:

“Norah Jones’s only request when it comes to recording is that tracking is done to analogue two-inch, in this instance using a Studer A827 with Dolby SR in conjunction with an SSL 9000J desk.

 One should ask oneself if a two inch analogue tape recording can ever deliver frequency bandwidth that one expects from a high res recording?

I would say analogue tapes cannot be considered high res source material.

Good question Aleg, but I am confident the best quality analogue will capture high fs. 

The studio recording of "Come Away With Me" used a machine rated as the best ever - the Studer A827. This has a record/reproduce at its highest 30ips tape speed of only 50Hz-20kHz +/-2dB, even a cassette is rated better than that,  however that does not mean nothing is recorded at higher & lower fs.  This machine was only used for tracking, once everone was satisfied the recording was then taken to Pro Tools where it was mastered in high bit digital for the original CD (16/44) & SACD (bitstream) releases.   

The later released remasters were intended only for the high-res market & these went in two directions  .............  

Sterling Sounds remastered from the original analogue master tape to 24 bit - available from HDTracks
Analogue Sound did the same but into DSD for the new SACD 

Staying with the HDT release as its what we are talking about - it was remastered by Greg Calbi, the same engineer as worked on the original recording & I understand via a DSD process into the final released & separate 24/192 & 24/96 files.   

Looking at the MusicScope 24/192 spectrum its clear it contains a lot of signal at 48kHz & even at 96kHz, also the high dynamic data does get close to 144dB & as such it does contain all the credentials for high-res, so it seems something in the HF range was recorded on the analogue tape.  

But bugger the technical stuff,  how does it sound?? I have the redbook CD & comparing that to my 24/192 .wav,    this remaster is awesome, its just simply stunning.  Its a fine example of genuine remastering in high res.   

Mike, BWF is simply a WAV file with one or more specific additional chuck identifiers.

 According to the EBU specification a BWF WAV file must contain the following EBU chunk:

• Original Bext chunk (Broadcast Extension - 'bext')

and optionally contain one or more of the following meta data or extension EBU chunk types:

• iXML chunk ('iXML')
• Quality chunk ('qlty')
• MPEG audio extension chunk ('mext')
• Peak Envelope chunk ('levl')
• link chunk ('link')
• axml chunk ('axml')

These meta data types are quite specific and specialised. I will get my file analyser out later to see which of the above chunk types in the Naim WAV file make it a BWF. It might simply be mis labelled of course...

Without the specialised chunks above, the file is a WAV file

Simon

Mike - ok its a BWF as it contains the 'bext' chunk containing encoding history and mastering time. The file also contains the wave file LIST INFO method for metadata.

The bext is largely empty and repeats some of the info in LIST INFO, other than in the bext I can additionally see the Vision track was mastered on the 2011-12-01 at 16:46:21 using a 'Sequoia software' production tool. 

There are two chunks called 'muma' and 'chrp' that contain almost completely blank data - I cant see what these mean - they might be bespoke chunks to the production software.

So yes - the Bext chunk makes this a BWF - other than that its a regular WAV file with list info meta data and two bespoke chunk types.

Simon

PS back to the traces from file sample anaylzer - I would be more concerned at files with sample values set too high for accurate analogue reconstruction without distortion  rather than perhaps the more nuanced hidef sample frequency rate and whether the file contains high frequency timing or ultrasonic frequencies

I'm obliged Simon,  I've been reading up on it myself & found much as you have. I might make a copy to dissect & to see how amenable it is to metadata editing compared to a regular .wav file. 

.......  wonder why Naim chose to use .bwf instead of the regular version??. 

It might be the BWF  enabled WAV file is the standard output from the production suite software?

If you use a quality WAV file editor that understands LIST INFO, such as dBPoweramp, then you should have no issue editing and adding meta data. You might want to disable ID3 extension, but even that should not be an issue - but ID3 is outside the BWF and WAV formal meta data specs

Simon

It is great to have something like MusicScope to take a deeper look into music files. That said, I would be surprised to see much if any real information above 30 kHz. Most microphones, like the popular dpa 4000-series, are simply unable to pick up anything in that range. So its most likely either noise due to dsd noiseshaping, like seen in The Linden Tree above, or artifical.

  dBpoweramp does not read the data in full, stuff is missing & I guess is hidden in those “chunks”, Data lines such as IFRM, IPRT, ISTR, IWRI, Disc (number) Label & Title Sort are not reading & the embedded art cannot be seen.   Needless to say it's missing all this if the file is converted to another format.

Strange, Dbpoweramp sees all those attributes that are present for me. There is no album art embedded in the file. 

All the Wav info attributes in the file that I can see by analysing the file are being displayed by dBPoweramp.

The only exceptions are the mystical muma and chrp chunks which as far as I can see are not defined by WAV or BWF and appear to have no data assigned - just many nulls.

The data attributes you are referring to that start with I are the official WAV list info attributes. An example is  IPRD which is the ListInfo Product Title attribute - and displays as part of the meta data and maps to Album name in ID3 speak. Shows up nicely on the Naim app via my Minimserver media server. Note that the Wav Info attributes you wish to include is purely optional and so do not need to be included if you are not using that value.

The WAV meta data INFO attributes preceded ID3 and so the tag titles within the file itself have different names.

Simon

OK I see what's wrong,  I was reading the file off my back up HD, its the original album as received from Naim,  & yes it does not have embeded art. 

Reading the file on the NAS I can see I have the art picked up from the "folder" .jpg

This is what dBpoweramp sees on HD file,  but the NAS file is exactly the same except it shows the art,  but compared to other .wav files it has a bunch of lines missing.

yeah I was wondering if you were picking up folder.jpg.

Well in my experience different download albums can have different metadata tags.

BTW my dbPoweramp meta data shows the value 'Jazz' for Genre - and indeed is stored in the Wav INFO meta data tag IGNR. Your screen shot shows it blank .... curious

I had edited the genre blank before I uploaded to the NAS share so it does show as "Jazz" in the NAS file.  

Incidentally I have started to use AudioShell for basic editing, simpler & faster than dBpoweramp;  its open source so give it a go. 

http://www.softpointer.com/AudioShell.htm

Simon,

I will look for a track with TPL levels above 0dB (in good quantity), use the tool from  Xivero, see what type of repair it does and hear the results. I will comment on that later.

Regards.

Erich

Erich, I'd be interested to hear how you get on. I was going to spend a tenner myself and have a play.

Simon