Digital sound

This, the thirteenth post of this blog deviates from previous post by discussing a subject not encompassed by my book.   Although a few sources of popular analogue sound media still exist, most modern day electronic contraptions such as PCs, CDs, DVDs, portable music players, cell phones, satellite radios and TVs (now) deliver sound to us in a digitally reproduced form.  There are many differing methods employed to convert, store and play back this digitized sound.  Digital video usually contains sound and is therefore germane to the overall topic of digital sound.  When considering the Internet one should be aware that inserting images into HTML is easy and straight forward.  There is no corresponding specification however for inserting audio or video into an HTML page or even an expected interoperability between browsers to play same back correctly.  There has been and continues to be a behind the scenes warfare between large corporations to control the lucrative acceptance and distribution of certain digital media formats.  Many digital sound formats have fallen by the wayside; some because of obsolescence but others because of warfare and special interest.  This post proposes to sort out some of these digital sound formats and finish by providing interested readers with some examples of largely abandoned but hopefully public domain synthesized music.

In the image above a digital signal with limited discrete value samples overlays a smoother continuous analogue signal in the background.   From the picture one might deduce that “digital” does not implicitly infer accuracy.   Here the number of samples collected in a second (t) and the number of values possible for each sample f(t) describe the accuracy of digital reproduction.  On a personal computer, digital sound comes in one of about four different forms (uncompressed or raw, lossless, lossy & synthesized sound).  As far back as 1937 a British engineer working on telephones figured out a way to convert analog sound to a digital format.  Pulse-code modulation (PCM) was the original raw data / uncompressed digital format and it is still being used today in CDs, DVDs and digital phones.  Music contained on a common CD has the file extension “.cda” (Continuous Disk Audio) and is not compressed nor does it contain information in an internal “header” tag.   Related uncompressed music or sound formats include:  AIFF, AU, WAV, and BWF.  These ‘raw’ formats provide the highest sound quality but are often inefficient because they require copious amounts of memory or disk space.  Often the bit rate and bit dept of these formats can be modified however when converting an analogue waveform to digital.  The bit rate (sampling rate) is the number of times in a second that a sample is taken, and bit depth is the number if digital values possible for each sample.   In pulse–code modulation an analog-to-digital converter encodes sound into a digital signal and conversely a digital-to-analog converter allows the restoration of a digital signal back into an audible sound.

AIFF (.aif) meaning Audio Interchange File Format is a pulse-code modulation developed by Apple Computer but modeled on an earlier Amiga (IFF) format.  AU (.au) was developed by Sun Microsystems and used in Sun, Unix and Java platforms; the early formats 2 through 7 are uncompressed PCM.   WAV (.wav or WAVE or Waveform Audio File Format) was developed by Microsoft and like AIFF files, performs on Windows, Apple and Linux machines.   BWF (Broadcast Wave Format) is an extension of Microsoft’s WAVE format; it is more popular in Europe than in the U.S. and differs only in having metadata or extended information in an appended file header.

The second type or category of digital sound is “lossless”.  This contrived name refers to data compression without the loss of any information.   Audio lossless compression is achieved by reducing data during periods of quite or during durations of continuous tones, which can later be reconstructed without the loss of fidelity.   Formats in the lossless category include: WMA or WMAL (Windows Media Audio) lossless, FLAC (Free Lossless Audio Codec), APE (Monkey Audio), WV (WavPack – open source), ATRAC (Adaptive Transform Acoustic Coding -from Sony – lossless) and MPEG-4 (SLS, ALS & DST versions and Apple lossless – .m4a).

* MPEG was formed in 1988 by the ISO (International Organization for Standardization) and IEC (International Electrotechnical Commission).  MPEG (Moving Picture Experts Group) specification number four defines lossless compression standards for streaming both audio and visual digital data on the web.   MPEG-4 is very important for today’s Internet multimedia, but it is a complex can of worms that can feature user interaction, support DRM (Digital Rights Management) or require licensing for the patented technologies it contains.  Abandoned here to continue with the subject of digital sound formats, it could be mentioned in passing that MPEG’s adoption of Apple QuickTime’s container in 1998, only briefly frustrated Microsoft’s bullying behavior in its quest for multimedia dominance.  The two links that follow go into more detail.   Quoting the second link:  “Microsoft’s sabotage included creating misleading error messages and introducing technical bypasses that deprived QuickTime of the opportunity to process certain types of multimedia files.  In some instances users were left with the false impression that QuickTime was not functioning properly.”

http://www.appleinsider.com/articles/09/07/06/ogg_theora_h_264_and_the_html_5_browser_squabble.html

http://www.roughlydrafted.com/RD/RDM.Tech.Q1.07/5F0C866C-6DDF-4A9A-9515-531B0CA0C29C.html

* QuickTime itself has been displaced as the web’s most ubiquitous multimedia player, by Adobe Flash Player.  Freely given away to developers and users alike, Flash Player has a high market penetration on the majority of computing devices which by coincidence happen to share a Microsoft operating system.   Flash Player as incorporated by Google’s YouTube videos is a limited, restricted format where the viewer usually has no option to copy or save a video.   

iPod DRM logo

To the strict audiophile lossless compression may be a bit less desirable than pulse-code modulation but it is certainly more efficient.   More efficient and space saving yet are the “lossy” formats.  Lossy compression differs from “lossless” compression in that lossy only approximates the original sound.   While some data is lost (discarded) the tradeoff is that lossy offers much better compression rates while providing a sound quality that is usually acceptable.   Psychoacoustics (or the study of the perception of sound) allows for lossy digital signal compression without disrupting the apparent sound quality.   Formats with lossy compression include MP3, AAC, ATRAC-lossy, MPC– Musepack, Vorbis and WMA-lossy.  The three most honorable mentions in this group are MP3, AAC and Vorbis.   MP3 (MPEG-1 Audio Layer 3) is showing its age and is somewhat proprietary but it is still extremely popular.   MP3’s don’t waste file space by storing excess metadata or DRM information.  AAC (Advanced Audio Coding– 1997) is an improvement upon and was intended to replace the MP3 format.   AAC’s sound quality is superior to MP3’s at low bit-rates but about the same at higher ones.  M4p’s (which exploit DRM) and .m4a’s both incorporate the ACC codec.   No licensing or payments are required for AAC and it’s used in Sony Walkmans, PlayStations, Nintendos, Apple’s iPhones, iTunes, iPods and iPads.  Ogg Vorbis is a free / open source format that is similar to MP3.  The “Ogg” portion of the name originally referred to a shell or container that was designated to hold just data but now the “.ogg” extension is commonly associated with sound data.  Because of its high fidelity and open source nature (no patent restrictions) lossy Ogg Vorbis is growing quickly in popularity and is used in many of the most popular video game titles.

Two different list of all (?) common digital sound formats:

http://www.fileinfo.com/filetypes/audio

http://www.digitalpreservation.gov/formats/fdd/browse_list.shtml

1st commercial Moog synthesizer (circa 1964)
https://en.wikipedia.org/wiki/Moog_synthesizer
attrib – http://twitter.com/AngryChad
Moogs were analogue – not digital

Synthesized sound

Before DOS, Windows, Macintosh or Linux the Atari 400/800 PC’s of 30 years ago and then later the Commodore 64 PC, were the favored machines for homespun musicians.  The “MOD” (Module) format was later defined in 1987, concurrent with the Commodore Amiga PC.  Similarly the MIDI (Musical Instrument Digital Interface) was defined as an industry standard protocol in 1982.   MOD and MIDI files contain no actual sound data at all, compressed or otherwise.

MIDI sound instructions are comparable to a roll of perforated sheet music used on an old self playing “player piano”.  The MIDI format was also created so that MIDI enabled musical instruments could communicate with or control one another.  Sounds are created by modifying the pitch, intensity, volume, panning, vibrato and tempo of a note.  MIDI playback relies upon a dedicated sound chip mounted on the motherboard or sound card, to interpret instructions and then synthesize an imitation of the intended sound.  The problem with MIDI music, one that still persist today is that the quality of the PC hardware and therefore the sound reproduced can vary considerably between machines.  Less common today, quality (expensive) ‘Wavetable’ sound cards actually store digital samples of individual musical instruments.  The cheaper sound cards or chips used by the majority of PCs employ a stripped-down version of Yamaha’s frequency modulation (FM) synthesis to imitate a note.   Most people perceive MIDI as being low quality audio because they have only experienced its playback on inadequate hardware.

Whereas MIDIs rely upon calls to FM synthesis or a set of instrument samples, MODs carry their own samples with them– in the file.  This means that MOD and similar module formats should offer more uniform and consistent playback between individual machines.  Module formats might offer a wider dynamic range and the ability to imitate human voice – a feature improbable with MIDI sound.  Additional popular module based/ chip-synthesized sound formats offering various features are the XM (extended module) audio file format, IT (Impulse Tracker) and the very impressive S3M (Scream Tracker 3.0) format.  There are multiple tracks (channels) to a module song.  “Tracking” is composing such a song and a “tracker” is the software that helps you do it.

Synthesized sound and music is much less common in mainstream media than it once was.  This is mainly due to the fact that improvements in communications technology allow more bandwidth for the streaming of higher fidelity lossless and lossy digital sound.  The quality can vary widely depending upon the example but synthesized music offers an elegance of efficiency that is unmatched by its high bandwidth counterparts.  A good 51 kb MIDI music file can perform just as loudly, be just as impressive in its presentation and play for the same 5 minutes as its 6,829 kb MP3 alternative.

Synthesizer rack
https://en.wikipedia.org/wiki/MIDI
attribution – Blurred Ren

Audio Players

There is not one single audio player or multimedia player that will play all sound formats.  Windows Media Player and Apple QuickTime have very limited capabilities and usually don’t cooperate with each other’s offerings.  Neither do these two media players show any interest in supporting alternative or open source music formats.  Acquiring both WINAMP and VLC media player will allow playback of many other music formats, but certainly not all.  Curiously, WINAMP won’t play music stored in Apple’s ‘.aif’ format and VLC won’t play ‘.mid’ files.   It may be necessary to download either jetAudio or MediaMonkey in order to play ‘.flac’, ‘.ape’ & ‘.mpc’ formats.   UMPlayer is an all-purpose multimedia player that will play many alternative sound formats as well as streaming media or movies from DVD disk.   XMPlay is a tiny but capable player that will play some of the rarer synthesized module formats.  * Reminiscent of what Winamp used to be before it was purchased by AOL and became bloated with questionable features, XMPlay.exe represents rare and efficient programming skill which accomplishes much with its mere 287 KB file size.   

Browsers

It has already been stated that Adobe Flash Player is the most common video player on the Internet by virtue of the fact that so many videos are embedded in that format.  Sound files opened on the web used to be handled by the file associations intrinsic or present on the client computer.  Nowadays a web browser is likely to direct sound files to open and perform with a player like QuickTime (irregardless of established file associations); whether your computer is Windows or Apple based.  One wonders if alternative browsers like Firefox, Chrome and Opera on Windows machines are as unbiased and as open-minded as they should be.   Within IE and these browsers, the contents of many open-source and uncommon music formats are mysteriously altered or scuttled as they pass through the MIME encoding and translation barrier during their transfer.   In many cases the only way to faithfully acquire music stored in an alternative music format is to download it in a ‘zipped’ form.

Linux   

   Apple and Microsoft either already own or are wealthy enough to purchase certain licensed technologies which they can then bundle with their operating systems.  While Linux OS distributions might embrace open-source media formats like FLAC or Ogg Vorbus they carefully avoid association with more common and probably patented or copyrighted formats.  Most vendors working on open source software wisely avoid packing and shipping any ‘codecs’ that might allow more mainstream media to play, simply to avoid potentially costly legal entanglements.  Licensing a particular codec (usually an algorithm capable of encoding or decoding a digital data stream or signal) might be possible but potentially costly for a small enterprise that often gives its software away freely.  This situation hampers Linux acceptance.  There are partial workarounds to this dilemma in some instances.

Examples

MIDI music files have generally been assumed to be public domain either because they weren’t important enough to worry about or because the mechanisms to protect intellectual property were not present within the format.  Improvements were made to the MIDI format with the “General MIDI specification” of the early 1990’s.   It became possible to brand MIDI files with a DRM type of individual key code.

Industry standard MIDI files seem not to require the archival (zip like) protection that other MOD type files do to be transferred across the Internet.  Where once Windows machines copied and saved a MIDI file to disk memory before the music was played, OuickTime will now likely intercept the MIDI and play it on the fly.  This change in the way of doing business began after Win XP SP2 (about 2004) and was likely a way to counter vulnerabilities discovered in MS’s “Active X”.   The ‘free’ version of QuickTime offers Windows users no option to save a MIDI file but perhaps after purchase the ‘Pro’ version does.

The MIDI format may not be more desirable to the creative musician than are the MOD type formats, but they are more widespread.   MIDI’s are found embedded in MS Power Point Presentations and in some web pages (where IE may play them but alternatives like Mozilla Firefox may not).  Unlike MOD type files a given MIDI file may sound surprisingly different when played on another computer.   If there is a choice, even individual players on the same computer can reproduce MIDI music in very different fashions because the players may employ different synthesis.   Also a user of a Windows machine may have a choice of different drivers (codecs) for MIDI playback (selectable from – Control Panel  / Sounds and Audio Devices /Audio tab).

The few examples that follow are simply for demonstration purposes and are assumed to be in the public domain.  Attempts to place a few  sample songs and zipped files on this page have proven futile however – because cactusbush@wordpress.com is presently a free account and an upgrade must be purchased to do so.   Therefore the following 17 small zipped together music files now reside on ‘Google docs’ where they can be downloaded by clicking on the following links.   Perhaps the links will need to be copied and then pasted into a new browser tab to accomplish the desired result  (at least until cactusbush upgrades his account)…

6_MIDIs.zip https://docs.google.com/file/d/0B-ogNSTYDAGQNlEtVTFsYjZJVGs/edit?usp=sharing

4_S3M.zip https://docs.google.com/file/d/0B-ogNSTYDAGQQUxjTDFneGlObXc/edit?usp=sharing

3_XM.zip https://docs.google.com/file/d/0B-ogNSTYDAGQenpEbVFJdXdoY2c/edit?usp=sharing

2_MOD.zip https://docs.google.com/file/d/0B-ogNSTYDAGQVDNXcElFczFzanM/edit?usp=sharing

2_IT.zip https://docs.google.com/file/d/0B-ogNSTYDAGQSTNjZmxzeUVUaTg/edit?usp=sharing

Other Synthetic Music Links

All these old synthesized music formats are largely irrelevant in today’s high-bandwidth, multimedia rich cyberspace environment.  The details of digital sound itself are largely ignored by the bulk of the online public, and are of interest perhaps to only a few geeks among us.   Some authorities and technically inclined musicians may take exception to observations made here.  Several years ago there were numerous web pages dedicated to synthesized music but that number is in decline.  Synthesizers frequently accompany live music but purely synthetic music itself may stage a comeback in popularity some day because it still possesses great potential.

http://www.ajsmidi.com/ (executable MIDIs)

http://w3.uwyo.edu/~dwwilson/music.htm (executable MIDIs)

http://www.classicalarchives.com/midi.html (classical music MIDIs – subscription required)

http://modarchive.org/

http://mody_collection.tripod.com/#MOD   (mostly MODs – some others)

http://www.dcee.net/Files/Music/S3m/  (S3m  music wrapped in old form – .arj – archival compression)

——————————

Update added 9/26/2014

Although the title of this post is “Digital Sound” it talks a bit about synthesized sound.  The first electrically synthesized sound was analogue and created using vacuum tubes (invented around 1907), followed by synthesis from transistors (invented 1954).

A Russian inventor named Léon Theremin created the first electronic instrument in the 1920’s.   This link  provides a video of a Theremin in action and then goes on to discuss Moog analogue synthesis.   * An interesting aside about Lev Theremin is that he designed a couple of slick eavesdropping systems for Russian espionage during the Cold War.   His so called “Buran” could intercept sound vibrations from glass windows using an infrared beam.   His creation known as “The Thing” was a passive “bug” which needed to be stimulated by an external microwave source before it could transmit its intercepted conversations.   

Pre dating the transistorized Moog of course were the vacuum tube electric organs.   Quote: “The Novachord is a gargantuan, all tube, 72 note polyphonic synthesizer with oscillators, filters, VCAs, envelope generators and even frequency dividers.”

The Hammond B3:

A Hammond Novachord (precursor to the B3) organ dissected:

—— More links ————–

A Moog in action:

Another Moog demonstration:

A more recent, multitouch controller instrument known as the LinnStrument in action:

A short demonstration of a ROLI Seaboard GRAND: