This essay is in a continual state of revision thanks to postings on this board. It is designed to both act as a guide and a stimulant for further investigation of the intricacies of obtaining high quality audio. We are all learning from one another.
Some Audio Basics
More detail can be found ion these topics at Wikipedia or by a Google search
AES/EBU: This is the digital audio standard frequently , officially known as AES3, and is used for carrying digital audio signals between various devices. It was developed by the Audio Engineering Society (AES) and the European Broadcasting Union (EBU) with final standards established in 2003. Several different physical connectors are also defined as part of the overall group of standards.
Analog: A continuous signal, necessary for ultimate recognition by our ears.
Amplifier: An analog device which accepts in a small voltage of 2 mv (usually referred to a “line” voltage) and increases it to an output of volts suitable for a loudspeaker system.
AV Receiver: A complex component which includes audio and video switching circuitry, a DAC, audio amplifiers and in some cases video upscaling.
CD: Compact disc, introduced in the 1980, these have digital data encoded at a sampling frequency of 44.1 kHz and a word length of 16 bits in a format known as PCM – pulse code management. This limits the frequency of each channel to 22.05 khz. This format is often referred to as “Redbook” because of the standards Sony & Philips laid out when it was first introduced. PCM can be copied.
Crossover: A device which splits up a full frequency input into a number of different frequencies suitable for different speaker drivers, most commonly into treble, mid and bass in a 3 way speaker system. Most speaker systems use passive crossovers meaning no external power is required. The other type of crossover is active and uses sophisticated electronics to split up the frequencies. Active crossovers are generally of higher quality than passive but are more expensive than passive.
Biamplification: This usually involves replacing a passive crossover with an active one, usually the unit between the mid range speakers and the subs. The line level signal from the preamplifier is then fed to this active crossover and the split signal out -> two amplifiers, one feeding the mid and high speaker drivers, the other amp feeding the sub. This arrangement is generally felt to improve the bass reproduction.
DAC: Digital to analog converter. This “unsamples” the digital data and reconstructs it to an analog output.
Digital: The common form now being used as an intermediate inbetween the original analog audio in and the final analog audio out of a system. That digital signal samples the analog and converts it into 0’s and 1’s. Different sampling rates are used in different formats, the most common for CDs being 44.1 kHz. DVD, SACD etc use different sampling rates.
Driver: Describes a cone speaker in a system of 2 or more units. The cone has a coil at its focus and that is centred in a strong permanent magnet. Varying currents from an amplifier through the coil cause the magnetic field around the coil to alter and this reacts with the permanent magnet to cause the cone to vibrate and give sound. High frequency drivers are termed tweeters while low frequency units are woofers.
DSP: Digital Signal Processing. This is performed by the DAC in the player but some components use DSP to change the amplifier input to correct room acoustic problems.
DVD: Digital video disc is the most popular video format and its audio uses a sampling rate of 48 kHz.
DVD-A: A high resolution audio format using higher sampling rates than CD (up to 192 kHz) and a longer word bit length. Initially in competition to SACD but as well favoured so there are few recordings being released using it now. A special player required.
Electrostatic Speaker: This is of fundamentally different design to the commonly used driver. Basically it is made of a thin coated Mylar membrane suspended between two stators or highly charged perforated metal plates. Different voltages applied to the membrane from the amplifier cause it to vibrate to give sound. The membrane is low mass compared to the cone of driver so responds more quickly to input so such speakers are admired for their accuracy and quick response to input. However most ES speakers do not reproduce well below 200 hz so are usually supplemented by a bass driver. Matching this to the ES is difficult because of the basic differences between these two transducers.
GIGO: garbage in -> garbage out. The quality of the signal going in to a component cannot be improved, only suppressed or filtered.
IC: Integrated circuit. This replaces a number of discrete items in a circuit and simplifies circuit design so lowers cost meaning ICs are in common production use.
Integrated amplifier: A combination of a preamplifier and amplifier.
Jitter: This is basically a distortion caused by timing errors in DSPs when handling digital data. Better timers and circuitry in DACs and transports will minimise but not eliminate all jitter. Jitter can destroy the stereo soundstage and, in its worst form, causes a harshness and edginess to the sound of the music. Jitter is measured in ps (picoseconds - one millionth of a millionth of a second or 10 to the power -12 ) and it is only recently such small timing errors have been capable of being measured. The subject of jitter is a complex and poorly understood topic so the following references are suggested for reading-
5. http://www.stereophi...m/features/368/ This link describes nicely the different affects of jitter.
6. For the technically minded - http://www.lavryengi...pers/jitter.pdf
Jitter is discussed on this site in the thread at http://www.dtvforum....rt=#entry734864
PCM: Pulse code modulation – the digital format used in CDs, DVDs, DVD-As and now specified for the audio in high definition DVD. There are a number of different forms of PCM as they vary in both frequency sampling and the bit length of the words.
Preamplifier: This accepts inputs from a number of devices and switches between them to give output at line level for an amplifier. Some preamplifiers accept the very low level voltage from a cartridge and amplify this to line level after suitable compensation.
Redbook: See CD
Resonance: Every item has a frequency at which it vibrates naturally and this can cause problems in a system, particularly in speaker drivers where the design of the cabinet is usually as rigid and well damped to minimise this. Rooms are particularly susceptible to bass resonances.
SACD: Super Audio CD. A high resolution audio format. Recordings use a different encoding method than PCM known as DSD, direct stream digital, a one bit process with a very high sampling frequency. This digital data is encrypted so it cannot be easily copied in digital form. Many claim SACD to be far superior to CD but much of this can be traced back to the extra care taken in the production of SACD recordings. A special p0layer required.
Soundstage: The illusion of an even spread of sound between the two stereo speakers.
S/PDIF: Stands for Sony/Philips Digital Interface Format and this is part of the AES/EBU digital standards for carrying digital audio signals between components. S/PDIF is basically a minor modification of the
AES/EBU standard for consumer use, providing small differences in the protocol and requiring less expensive hardware.
Transducer: A component that converts between a physical vibration and an electrical current. The needle in a cartridge for playing LPs is vibrated by the grooves in the record to generate variable current. Conversely when varying current is fed into a speaker driver it vibrates to give sound.
Universal Player: designed to handle all popular formats – CD, SACD, DVD-A and DVD.
The Audio Chain
1. Software source (CD, SACD, DVD-A or LP)
2. Player (CDP or separate transport/DAC combination)
3. Amplifier (Integrated or preamp/amp combination)
4. Loudspeaker system
5. Listening room.
Weaknesses in the Audio Chain
1. Software. Unfortunately these often contain ingrained distortion so the disc to be chosen when making a judgment of an audio component should be chosen with care. It is best to use a recording with a wide variety of different types of music.
2. Player. The cost of these varies from under $100 and up to or over $20,000. The biggest weakness in players is generally recognised to be distortion caused by jitter. This can occur in either or both of the transport and the DAC whether they be separates or together in the one unit. The observable distortion caused by jitter will vary according to components used further down the audio chain as it might be masked by amplifiers or speaker systems.
3. Amplifier. There are a huge number of variations in circuit design and quality of items used in construction. Different amplifiers do embed a different sonic character to the resultant sound and this is often very dependent on the interaction between the amplifier and the speaker system. The general aim should be to have a neutral sounding amplifier, particularly in the mid range of frequencies. Note that in higher end amplifiers, the quality of cables linking from the player and to the speaker can alter the sonic signature. Amplifiers need to have sufficient power to drive the speaker system so they do not overload and start clipping as this can destroy speakers, particularly tweeters. The damping factor of the amplifier can be important for some speaker systems. Generally the synergy between the amp and speakers is particularly important.
4. Loudspeaker System. This is generally accepted as the weakest link in the audio chain. However, apart from the quality of design and the quality of crossovers and drivers used, its effectiveness depends both on the interaction with the amplifier and the room acoustics. Speaker sensitivity needs to be matched carefully with the amplifier and the room. Resonances in the speaker cabinet give a woolly, boomy and muffled character to the audio and destroy music fidelity. Placement of the speakers in the room and the seating position relative to those speakers can make a huge difference to the perceived musical result. Experiment in placement is essential
5. Listening Room. Bare floors and walls cause excessive reflection and cuboid shaped rooms can be responsible for unpleasant room resonances which can detract from appreciation of the music played. There are a number of ways of offsetting problems. These can involve judicious placement of carpet, wall hangings, bookshelves (very effective behind speakers) or even commercial “tube traps” designed to minimise room resonance problems. If the acoustic problems are not severe, a DSP electronic correction could be the answer but might introduce more (jitter etc) problems again if not carefully implemented.
So, What to Upgrade First?
Introductory Note: The "weakest" link in a particular setup will vary from one system to another and might be the player, the amps or the speakers. Generally audiophiles correctly regard the speaker system as the most compromised part of the chain but it does not follow that it is automatically the weakest. Changing speaker systems is an expensive process and the money could be better spent elsewhere in the chain. This might be the amplifier or better cabling. However the author urges audiophiles to critically examine the player as a possible best first place for improvement, not necessarily by purchase of a newer CDP, but to have the present player modified with an improved timing control. Such modification on a 2 CH player could cost up to $1500 if done by experienced technicians such as those at Soundlabs Group http://www.soundlabsgroup.com.au/. Such work can yield dramatic and clearly audible improvements in resolution, soundstaging and retrieval of detail in recordings. Further upgrades of amps or speakers etc can then further reveal the quality of the signal coming from the player. This philosophy is expanded in furtherdiscussion below.
Obviously there is no simple answer for everyone to the question of what to upgrade first. We are dealing with a system and that includes the room acoustics. The total system can never be better than the weakest link and exactly what is the weakest link in any system can be difficult factor to determine. Upgrading can be an expensive and frustrating merry go round if not approached carefully.
The author has been playing around with audio (and now video) for over 60 years since first being hooked by cylindrical records found in an attic in the 1940’s. Progression followed from the Edison cylinders to 78’s, to mono, then stereo LPs and ultimately to modern day CDs, SACDs, DVD-As. These were played on wind up gramophones with outboard rigger pickups to Linn Sondek etc tables and now to Esoteric universal players.
So what has been learned in this long voyage? I put forward the following thoughts about audio to consider:
1. Items change in character after a period of warm up which can extend to several weeks;
2. The most consistent audio results are obtained if gear is left permanently on (yes, I know, most environmentally unfriendly, but your choice);
3. GARBAGE IN -> GARBAGE OUT so it is pointless owning the best possible amps and/or speakers if the player is not inputting the best possible signal;
4. Most, if not all, commercial players of digital data (audio & video) can benefit from upgrading with better components, particularly timer clocks. The player is often the weakest link in the audio chain;
5. Most modern amplifiers, and that includes receivers, are fine. No, not all amplifiers sound the same, but in most (not all) cases this would be the last item to upgrade;
6. Positioning of speakers is a bit of a black art and can make a huge difference to the sound, particularly the bass. Experimenting with even small positioning changes is a well worthwhile tweak.
7. Loudspeakers are often the first target to replace because they are the obvious final source of the sound but this needs to be approached very carefully and can be a waste of money (reasons given later in 32);
8. Loudspeakers sound as they do because of input AND reaction to room acoustics. A reflective room with bare floor or cuboid shape poses as very hostile environment for the reproduction of quality sound;
9. Items which have good specifications on paper might not mate together smoothly. Unfortunately assembling a good system is not a simple Lego exercise;
10. Insertion of a different component can alter the sound to the joy of the owner. However this alteration might not actually have improved the sound, only made it different, something accepted by the owner who had become bored with what he/she heard previously. The net effect, at expense, was merely a move sideways, not forward;
11. The law of diminishing returns can set in quite rapidly after a certain point has been reached;
12. The audio quality of the final result in any system is not necessarily proportional to the amount spent on purchasing it. The most expensive component is not necessarily the best;
13. No system can replicate the live experience but many can reproduce a very satisfying replica;
14. Auditioning components in a showroom or another audiophile’s home might not be useful as the resultant sound is the result of the chain of items used (including the acoustics of the demo room) and that same sound might not be replicated when the item is inserted into the owner’s home system. Similarly insertion into that different home system could yield disappointing results but that might be a reflection of weaknesses elsewhere, not of the new component;
15. Auditory memory is highly unreliable so this makes opinions of “A” compared to “B” suspect even if there is only a short time interval between the audition and of even less value if the time interval is longer;
16. The numbers game continues in advertising hype with the angle now on digital figures (e.g. bits, upsampling rates) which mean little to most people and, may or may not, relate to sound quality;
17. Very good sound from either analog or digital sources is neither easy nor cheap to attain;
18. Software in all formats varies greatly in quality but the best analog and digital are remarkably good. However the software, particularly of rock, pop etc material is often very flawed and the quality of classical recordings also varies a lot so any audition of components must be performed carefully with recordings known to be of high standard;
19. CD digital can sound very close to analog BUT it takes expensive players and DACs to achieve that. And Hi-rez SACD is not necessarily better than CD although it generally is because more care is taken in its production;
20. DAC’s have significantly improved in the last few years making older, used units doubtful propositions. Upsampling does appear to improve accuracy and sound quality in some units but is no guarantee of a better sounding DAC;
21. Prejudices abound in audio with many defending valves vs solid state or vinyl vs digital, 2 ch vs surround sound, SACD vs DVD-A etc with religious fervour. There is good and bad in all systems and formats;
22. Top audio for Home Theatre is just as important as for audio only entertainment as there is a lot of well recorded music in video documentaries and movies apart from opera, ballet etc. (I suspect HT Luddites have not ever heard a really good audio system with HT);
23. Smooth bass is difficult to achieve, particularly with bass units attached to the mid and top as is usual. Speaker placement for best sound staging is not necessarily the best for room loading of bass.
24. Biamplification might be an answer to bass problems and the reasons are well argued by Rod Elliott at http://sound.westhost.com/bi-amp.htm. However it seems many audiophiles have naïve ideas about what is involved in biamping and think they can rush in when they have a spare amp. Done properly with active crossovers biamplification is great but it should follow after most other upgrades, particularly those to the front end;
25. Many systems can appear “too bright” and harsh– the result of poor bass reproduction and/or jitter in the digital player;
26. Commercial press reviews in reliable publications like TAS and Stereophile are to be believed BUT it is important to read between the lines;
27. Internet postings may or may not be valid and should be treated with caution. Most audiophiles are entranced by new toys and often seek to justify their audio idiocy;
28. Audio-video dealers are knowledgeable and are often enthusiasts themselves BUT they are unlikely to recommend brands they do not carry so their advice is based upon what they have in stock. It must therefore be treated with caution;
29. Ultimately choice of type of music and quality of sound is so subjective that it must be personal so that in the absence of opportunity to audition for oneself, it is important to canvass as many different opinions as possible to be able to form a sensible consensus. The internet is making that much easier these days and is an invaluable resource;
30. All high end systems with low distortion and high resolution etc have their own sonic signature. Judgement of other systems is highly coloured by love of the sound from one’s own. After having worked upon and refined a system to personal taste, a differing sonic signature might not appeal, regardless of whether or not that differing system is closer to the original sonic signature. This makes acceptance of other audiophile opinion a minefield;
31. Musical enjoyment should be at the heart of it all and some brilliant artistry can overcome poor technical software e.g. the 1958 recordings of Callas singing in Verdi’s “La Traviata”. However bad technical reproduction from software or hardware can make enjoyment of the music very difficult. Regrettably surface imperfections on LPs fall into this category;
32. Attempts by audiophiles to counteract the “yin” in one component with a “yang” in another is absolutely futile and bound to end up as a continual dog chasing its tail situation. Great for dealers but a very (literally) poor approach. The aim should always be to start with a neutral, clean amp, match it with appropriate (sensitive wise) smooth speakers and then feed in the best front end the budget allows. The adage “garbage in -> garbage out” applies so amp and speaker should not be changed to minimize the garbage effect unless either are proven responsible for it. The audio chain is only as strong as its weakest link and unless the front end signal is as good as the budget can stretch to, many following upgrades can be a frustrating waste of time and money. Loudspeakers are popularly regarded as the best starting point but there is a trap in that. No component is faultless. All have their weak points, particularly digital players which have a tendency to harshness, sometimes referred to as digitalis. Much of this harshness is due to jitter caused by timing errors in reading the software. These timing errors are of no consequence in reading data to a computer but are fundamental to the quality of reproduced audio. Digitalis is unpleasant so the first response to ameliorate this by using valve amplifiers or speaker systems that “soften” this problem. So what is being done is to counteract a “ying” of “X” with a “yang” of “Y” in an attempt to give a listenable result. So, choosing the speakers with the best “yang” is actually choosing speakers with a fault, not obvious until the “ying” is eliminated from the other component.
33. The bottom line in all the audio-video mania should be for someone to be satisfied with their system because they are achieving musical enjoyment from it.
So what are the traps with upgrading and where should one start? If someone is happy with the present sound and do not plan to upgrade then no problem. Problems only arise if a new component really does improve an element of the reproducing chain as the system can then be thrown out of balance. The process of upgrading can then become an expensive merry go round with unsatisfying results.
If anyone is serious about hearing better sound from digital media then they will not do so with a $100 player. Using such an input and then spending hundreds of dollars on other amps or speakers is a waste of money in my experience. Maybe some $100 players could be massively improved by spending $1000 or more on modifications or maybe not. Many of the latest players use cheap internal parts (they have to to remain competitive) and might not be good candidates for upgrading. However some might have good potential – the opinion of a technician with experience in a number of players e.g. at Soundlabs Group Melbourne and Sydney, should be sought about that. I had an OPPO 971 player modified and the result was really good although this player costs well over the $100 mark if purchased locally or imported as I did.
So folks, my leaning is towards getting the best input possible from the player as the first best move. This does not have to mean spending excessive amounts on a player. But if the input is faulty there is no way you can use neutral sounding component elsewhere in the chain because that faulty input will sail right through with unpleasant results. This essay makes no attempt to define what is the "best buy" in players or any other component because the answer to that is very subjective as it depends not only on budgetary constraints but also depends on how well the item integrates into the rest of the system.
The argument in favour of speaker upgrade first has merit IF care is taken to purchase units that are neutral, particularly in the midrange. The advantage of this is that improvements made earlier in the chain can be more easily recognised. The trap is to reject such good speakers because the sound is not good, when the fault lies not in the speakers but in the signal going in to them. Yes, that fault might lie in the amplifier (or an impedance mismatch) but is more likely to be in the player. And do not forget the software in all this. Few CDs are perfect.
To finish I guess I should give a reminder of the system here so you can judge where I’m coming from. First up, around 18 years ago I designed the main entertainment room in this house with a pentagonal shape and cathedral ceiling to avoid opposing parallel surfaces. It has worked exceedingly well. The rest of this house was fitted in around that room. You can see pics of it and the general layout at http://gallery.audio...p...58&session=
Audio Input: Esoteric UX1 LE, mated to an Esoteric G-0s Master-Clock regenerator
DAC: Highly modded (by Steve Nugent at Empirical Audio USA) Benchmark DAC 1 (used as the DAC for laserdisc and other digital outputs)
Preamp: McIntosh C200
Stereo amps: Halcro DM68’s
Stereo speakers: Sound Labs full range electrostatic “Majestic” units
Cabling: Furutech to amps, Siltec, and various other interconnects
Power: Regenerated for the player, preamp and speakers by a PP600
This essay is not made as a flag flying exercise but has been put up to possibly help others avoid the many traps (and associated waste of dollars) that I’ve incurred over the last 6 decades. Any corrections or additions - please post as this is intended to help, not hinder. In all upgrading please approach prospective new toys with care rather than love, or worst lust Easierto say than do. I know, been there, still doing that!
2010 POSTSCRIPT: A less expensive way of getting great 2 CH audio is via headphones. I hate wearing them but cannot deny the great quality - better than anything else here. See my post below at http://www.dtvforum....showtopic=90992 for details.
Edited by Tassie Devil, 09 October 2010 - 03:36 PM.