The theme of the next issue of “High Fidelity” (No. 83) are products related to high resolution files, CD quality files and sound from a computer. In the audiophile world all those are usually regarded as rubbish and only rarely distinguished from mp3. I will write more in the editorial to the “file” issue, but I want to let you know, that we are not interested in mp3. I am not pretending, that there is nothing like that, but lossy compression excludes this phenomenon from my area of interest. But… There is always a “but” coming in the least expected moment. During the experiment described below we compared also a mp3 recording with its master, the 24/96 master tape. Based on the listening session, it turns out that this “but” is very uncomfortable. Details below.
Anyway, the thing I want to take a closer look at, are the files themselves. And in such context the mp3 would be a part of this phenomenon – the fading away from physical carriers with digital recorded music (because the vinyl is a different case) and the change of the paradigm related to distribution of music – from physical stores to virtual stores, artists web pages, etc, where we can buy the music, and then – this is important! – download it immediately to the hard disk of our computer.

And computer is probably the key word here, which scares away the music lovers and audiophiles from new technologies. I think, that companies in the file playing business, mostly Linn, Naim Audio and Sooloos know that well, and do not use that word nowhere in the company materials in any kind of relationship to their products. And if they do, then they accent its supporting function, helping buying files, cataloging them, etc.
And yet – all file players of the companies I mentioned are in fact computers – specialized computers with a classic processor, software (usually Linux), operating system, etc. I’ll repeat: those are computers specialized in one function – playing audio files. This of course frees them from having the processor deal with other tasks, but it does not change the main thing. Anyway – all players working with hard disks are computers. And that’s it.
Hard disk players are the main group of file players. The second one are the Flash card players (or with built in flash memory). Usually SD cards are used. Do you remember my editorial about the SDMusA player from Ancient Audio? Many things were said about it since – good and bad things. But it is worth mentioning that it is nothing new – new is only the way, the platform, where on one carrier audio and video files, photographs, etc, were united, with a very simple, intuitive navigation. And for me SDMusA is a platform, and not a new “format”. SD card players are known from some time, with examples like HIFIDIY Mini and HiFiMan HM-602 players. According to Mr. Waszczyszyn, players of that type are based on “dedicated hardware based on FPGA. There is no operating system, or a central processor (there are a few small ones, controlling the display, keyboard, etc).” So there are no components of a computer, from a technical point of view those players resemble much a CD player. This is a big asset – the simpler something is, the better.
Both groups of devices have one goal: play files, starting from mp3 and ending with high resolution files. Currently the upper limit is 24 bits and 192kHz, but soon there will be 32 bit files. The upper sampling frequency will probably also double (there are Zodiac USB converters of that type) although that is not so sure.

Looking at all that we should ask ourselves a question – is it more about a way of selling and distributing music, freeing us from a corset of a physical carrier, or about the quality of sound. The pessimistic version of the response is simple – it is about sales. The quality of the sound does not interest anybody – companies are ensuring us about that, compressing the sound to extremes, having a “loudness war” on, where someone louder (more compressed) is winning.
For me the quality aspect is more important. The question if the 16/44.1 files sound better from the disc or file is important, but in context of the capabilities of this new technology, it is negligible. More interesting is that, if the high resolution files will become a part of the audio market, even in the way now occupied by SACD discs (but not vinyl), in a niche. How I love niches!!! It seems I am a niche animal.

This was the basis for the Krakow Sonic Society meeting I am writing about. I wanted to learn from its members, how the sound changes after adding the bits and increasing the sampling frequency. I mentioned that before, but the direct impulse for this comparison was the issue of the first edition of a Japanese magazine “NetAudio” (vol. 01, 2010 Winter, PhileWeb), to which a disc was attached - Audio Accesory - T-TOC Records High Quality Data Master Comparison (TDVD-0002, DVD), made to allow such comparisons. It contained files with different word lengths and sampling frequencies, prepared directly from master tapes.
I made also some comparisons between files ripped from CDs (this is something else than a file of that kind downloaded from the internet) and hi-res files.

The system on which we made the comparisons is owned by Tomasz, one of the members of the Krakow Sonic Society. Almost everything there is new – the house, the player, the cables and the loudspeakers.
The system is composed of:

• CD player – Ayon Audio CD-2s, test HERE
• integrated amplifier McIntosh MC7000, test HERE
• loudspeakers Dynaudio Special Twenty-Five Signature Edition [2010 version], test HERE (earlier edition, test HERE)
• interconnects – Acrolink 7N-D5000
• loudspeaker cable – Acrolink 7N-S10000
• power cable (CD player) – Acrolink Mexcel 7N-PC9300, description HERE)
• power strip – Gigawatt PF-2, test HERE

We used the player only as a DAC with an USB input. The source of the files was my laptop Hewlett Packard Pavilion dv7 (Windows Vista, 320 HDD, 2 GB RAM), with the Foobar2000 player. The files from CDs were ripped using dBPoweramp in Ultra Secure mode.
I decided to use a computer for utilitarian reasons – I wanted instantaneous access to the files, switching between them without silence, etc. From the computer the signal went by a 5m long Acoustic Revive USB cable (a brilliant cable) USB-5.0 PL, where the power runs and the signal flow through separate runs. The USB cable was plugged directly into the USB socket on the player.
Initially I planned it differently – the company KingRex sent me a super novelty, a USB to S/PDIF converter called UC192, which allows to convert a signal up to 32 bits and 192 kHz! But because I got it only two days before the meeting I did not try it out and took it to the session as it was. And that was a mistake – the unit required installing a driver from a CD-R and a bit of juggling around in the computer. In the meeting there was no time to do it, so I resigned from using it. This shows also, that our meetings do not always run as we plan them.
Resigning from the converter limited us to the sampling frequency of maximally 96kHz, as this is the limit imposed by the Ayon. We will deal with the 192kHz frequency at a later meeting.

Like I said, our comparison was about recordings made in different ways. Those were comparisons of recordings made in a studio, without transferring them on a physical carrier and rips from DVD-A and CD discs. Each time it was about comparing a certain file, with a certain way of creating the file.
This is the reason, that I divided the description in the same way we made our comparisons. Each step will start from a description of the files, I will say what we compared, and why, and then I will mention a few representative opinions of the meeting participants. One more thing – the comparisons were made in AB test, blinded. Only after the opinions the listeners learned, which file was which.

While reading please remember that:
16 bits = 2 to the power of 16 = 65 534 sound levels
24 bits = 2 to the power of 24 = 16 777 216 sound levels SOUND

I. Audio Accesory - T-TOC Records High Quality Data Master Comparison, TDVD-0002, 2010

1. Kankawa, Dear Myself; ripped from DVD (availability: CD – HQCD, HQ – Master CD-RII).

Assumptions
Comparison of two files – 16/44.1 and 24/96, digital recording. We compare at the same time the change of word length and sampling frequency. The files are from the same source – they were prepared in a mastering studio and had WAV format, ripped from the same medium – DVD.

Opinions
Ryszard B. – we can hear immediately which file is better. For me the 24/96 has better breath, space, better saturation and was more pleasant to listen to. The 16/44.1 version was just weak.
Janusz – I will absolutely not agree, that the 16/44.1 version was bad. Dynamics was shown really well. Only that I then listened to the 24/96 version – that one was the second one, wasn’t it?
Wojtek – yes, we listened first to the 16/44.1 and then 24/96.
Janusz – well, that can be heard immediately. I had no doubts, that we listened to the “dense” file second. I just do not want the readers to have the impression, that the CD quality version was bad – it was not! There was everything that should have been, better drawing of the instruments, dynamics. Maybe it was a bit overblown, clearer, because there was less information, and those that remained was shown stronger – I do not know. But well, the 24/96 sounded better.
Tomek – frankly speaking I liked the file nr. 1 – as it turns out the CD. Everything was clearer and closer to me.

2. Grand Trio, Schubert, Piano Trio No. 2, Allegro; files ripped from DVD (availability: CD – HQCD, HQ – Master CD-RII)

Assumptions
Comparison of two files – 16/44.1 and 24/96, digital recording. We compare at the same time the change of word length and sampling frequency. The files are from the same source – they were prepared in a mastering studio and had WAV format, ripped from the same medium – DVD.

Opinions
Tomek + Janusz – this time, the file no. 1 was better [it was the 16/44.1 file – but the listeners did not know about that – note WP]. Much more peace, fullness, a kind of leisure. It was noticeable, that the piano sounded stronger, more distinct.
Rysiek S. – form me the file no. 2 sounded better [meaning 24/96 – WP]. I cannot agree with you. For me the first file was too shrill, sometimes even unpleasant. Probably when I would listen to it without making comparisons then it would be OK – this was a good recording. But the next file was splendid, deep and vivid. Yes, the no. 1was on a leisure, but for me it lacked coherence.

Tomek – but the difference was not big.
Andrzej – no, in this case the difference was negligible for me, I would not be able to show which file is which.
Wojtek – the first one was 16/44.1 and the second one 24/96.
Janusz – together with the piano a viola enters. On the 16/44.1 file it is stronger, it is an equal partner, and it is a bit to the back on the 24/96. This is a surprise – but maybe this is the way it should sound, and the compression necessary for the CD brings the levels closer together?
Wojtek – in general we did not agree in this case: Janusz and Tomek preferred the version 16/44.1 and the rest the 24/96 although they did not know it is a hi-res file and chose blindly.

3. Yarikoyazina, Percussion&Bass; files ripped from DVD (availablity: CD – Master CD-RII, HQ – Ultimate Master Data Disc).

Assumptions
This is one of the more interesting sessions that night. It was about comparing two 24/96 files, digitally recorded, where one was uncompressed (WAV) and the other one was compressed in a studio using the mp3 algorithm to a file 24/96 192kbs mp3. The files came from the same source – they were prepared in the same mastering studio as WAV and were ripped from the same medium (DVD).

Opinions
I will say it in one text, distilling that, what was most important, because we had a lengthy discussion about this comparison. The mp3 file was a surprise to us. Everybody pointed to it, as to the compressed file , but we were shocked, how well it sounded. Although our audiophile hearts wanted to escape our chests, but everyone said, that if mp3 would be prepared like this one, then it would be absolutely satisfactory on less expensive equipment. In the real world, where we have the compression made from 16/44.1 files it is not so good, but this experiment showed, that the Fraunhoffer Institute knew what it was doing, and the bad opinion about mp3 is mostly due to bad preparation of such files. Yes, on a better system the difference is evident, but – I’ll repeat this – mp3 was shockingly satisfying in this form.

II. Zbigniew Namysłowski, Open, Polish Jazz, vol. 74, SX2539, 1987

• 24/96 – source: CD-R copy of the master tape (before processing) – louder by 3dB
• 24/44.1– source: CD-R copy of the master tape (before processing) – louder by 2dB
• 16/44.1– source: CD-R copy of the master tape (before processing)

Assumptions
Comparison of three files – 16/44.1, 24/44.1 and 24/96, analog recording. First we compare the change of the word length, and in the second part the change of the sampling frequency. The files come from the same source – they were prepared especially for us by Ms. Karolina Gleinert. Ms. Karolina is responsible for the re-masters of a part of the recordings from the Polish Jazz series – an interview prepared by me on this topic appeared in “Audio”. The material was recorded directly from the analog master tape to a computer in the 24/96 format, and then converted to 24/44.1 and 16/44.1. Besides that, the files were not processed in any way, those were direct copies of the analog tape.

Opinions
Rysiek R. – without a doubt the 24/96 file was best. I have no doubt about that. The recording itself is not a good one, the piano was not recorded well, but the CD version is too flat, to appeal to me, the 24/96 version is much easier to absorb.
Janusz – the difference is audible, the 24 bit recording has more breath, is calmer. Going to 96kHz is less audible. I am not sure, if such changes will be visible on less expensive equipment.
Rysiek S. – for me the 24/96 was clearly the best. It sounded nice.
Andrzej – without a doubt the last file, as it turned out, 24/96 was best. We cannot pretend, that there is no difference. And I agree with Janusz, the change from 16 to 24 bits was more audible than going from 44.1kHz to 96kHz.

III. Charlie Haden & Antonio Forcione, Heartplay, 2006
Recordings: Anna, La Pasionara

file 16/44.1 – source: CD, Naim CD098

Assumptions
Comparison of two files – 16/44.1 and 24/96, analog recording. The files come from two different sources – the file 16/44.1 was ripped from CD, and the 24/96 was bought in the internet shop HD Tracks. Both files were prepared in the same mastering studio and it is the same master.

Opinions
Janusz – lets’ leave this – I am sure that file no. 2 is the 24 bit file. What an embarrassment with the CD. An I heard it before and it sounded well. Only after comparison it is shattering – in this case the difference is colossal!
Rysiek R. – Oh my – I did not know, that CD can sound so bad! Or in other words – 24/96 is brilliant.
Rysiek S. – indeed – the hi-res file shows the timbre of the guitar fantastically and brilliantly places the contrabass in its spot. Beautiful reverbs – really incredible.
Andrzej – this does not leave a trace of doubt, this is not a subtle change, but a complete change of reality. If all hi-res files would sound like this, then it would mean the death of the CD.

IV. Stan Getz & Joao Gilberto, Getz/Gilberto, 1963
Recording Corcovado (Quiet Night of Quiet Stars)

• 24/96 – source: HDTracks – louder by 7-9dB
• 16/44.1 – source: K2HD, Verve/Lasting Impression Music, LIM K2HD 036
  

Assumptions
Comparison of two files – 16/44.1 and 24/96, analog recording. The files come from different sources – the 16/44.1 was ripped from the reference disc K2HD CD, prepared by Winston Ma, issued on silver by First Impression Music. The 24/96 file comes from Verve and its master was prepared by their engineers.

Opinions
Janusz – ach, now something different – the CD file does not seem worse than the 24/96. FIM is brilliant!
Rysiek S. – indeed, the intensity of the sound, its softness, etc, show FIM from the best side. But the voice of the vocalist is better from the 24/96, more fluent and less mechanical.
Andrzej – yes, the vocal sounds better from the 24/96. But the rest seems better from CD.
Tomek – for me both are good but different.
Rysiek B. – I was trapped and said, that the FIM file is 24/96. This because I probably did not know what to expect from the hi-res file. In direct comparison the CD seems more intensive, stronger. But indeed – although FIM seemed better to me, the vocals were better on the other file, deeper, free from any clipping and plastic tarnish.

SUMMARY

I think, that the ambiguousness of the opinions of the listeners is clear. The difference between the files was always clear, most listeners said that they seem better (when not knowing what file it was, and choosing the “better” file). But not always.
A certain tendency was clear – the 16/44.1 files (except for Forcione) sounded in a more substancial way, some elements of the sound were extracted to the surface. On the other hand, the 24/96 files sounded in a more relaxed, calmer way. But the listeners missed that, what they had with the CD, meaning palpability.
If I would point to something from myself, then it would be something, we could call a “lack of history” of files – the files are mute. I explain – having a CD we know exactly, who issued it, who prepared the master or who re-mastered it. The files are “mute” – we do not know where they are coming from – we can edit the ID of the file, change it. And as we can see, a well prepared CD master can sound not worse than a standard 24/96.
This is why I am sure, that soon there will be an industry of Japanese files – similar to the CDs from Japan are almost always better than their European counterparts, so I expect the Japanese files being that, what we are searching for. It seems, that the art of making 24/96 files is still developing – increasing the word length and sampling frequency gives big potential, but how it is utilized depends on people, as usual.
The hi-res files only begin their career, and there is a lot to do to have this switch make sense. Nevertheless it can be done splendidly, as shown by the recordings of the Japanese sampler and Forcione. With the latter the advantage from 24/96 was shocking.
And 24/192 files? They will have their time, we will check that. I even prepared a SD player HIFIDIY Mini, with a 24/192 comparison, but we did not have time for that. The listening session took 3.5 hours, and we were very tired at the end.

Previous meetings of the Krakow Sonic Society:
#74: Acoustic Revive RST-38 and RAF-48 – anti-vibration platform
#73: Kazuo Kiuchi (Combak Corporation) in Krakow
#72: Acrolink Mexcel 7N-PC9100 vs. Acrolink Mexcel 7N-PC9300
#71: Capacitors in loudspeaker cross-over
#70: Blue Note on XRCD24 Audio Wave
#69: Power in the walls
#68: McIntosh – MCD7000+MA250 vs. MCD301+MA275
#67: Tara Labs The Zero
#64: HQCD + Blu-spec CD
#63: Acrolink Mexcel 7N-D6300
#62: System RCM
#61: KT88 tubes
#60: Power cable Acrolink Mexcel 7N-PC9100
#59: Power amplifiers - Ancient Audio Silver Grand Mono Vs. Reimyo PAT-777
#58: Cables over the top
#55: Racks, part 1
#54: Power, part 2 (Furutech power cables)
#53: K2HD discs
#52: SACD player Accuphase DP-700
#51: Isolators (anti-vibration platforms)
#50: Power, part 1 (power plugs)
#49: Accuphase – digital (r)evolution
#40: Digital vs analog or 25 years of the silver disc
#35: CD players – Accuphase vs. Ancient Audio
#34: Jazz in The Pawnshop. Vol. 1
#33: Power amplifier Ancient Audio Silver Grand pre-premiere
#32: Power amplifiers Nagra MPA vs Ancient Audio Silver Grand (prototype)
#31: SONUS FABER Minima Amator vs Electa Amator vs Guarnieri Homage
#30: Loudspeakers Eryk S. Concept Nuvo in Krakow