Beside the rather obvious answer that EVERYTHING matters, which is true, nevertheless not all factors determine or contribute to good sound quality. One may do so with an impact of say 50%, while another will do barely 1%.

In your view, what gives audio electronics the most in terms of saound character? Power? Feedback? Power supplies? Circuit topology? Its parts? Parts selection? Specific tolerances and/or matching?

Well as u say everything! But a decent to very good power supply can improve a somewhat mediocre unit to something quite exceptional, whereas a well thought out short circuit with quality parts and a shight psu isv a complete waste of time IMO...

There are a lot of constraints on what is possible in this day and age when peoplevwantbsmall elegant one box solutions, I mean what can u possibly cram inside them? A good external psu will help but even these are confined to the usual wall wart size.

Most manufacturers are chucking out so many new revisions and new lines the lifevexpectancy and quality of internals are just not fit too last, quantity over quality and looks over substance!

The designer/engineer.

In order of importance and at a risk of generalising, this is my opinion.

Speakers: Well matched drivers installed in non resonant cabinets and crossovers that don't create phase anomalies.

Amplifiers: Massive, clean power, combined with capable power supplies and creative topology and low tolerance components.

Preamplifiers: Low noise components and circuits assisted by (preferably) outboard power supplies. Circuits that offer above average channel separation and low levels of distortion.

CD players: The separation of power supplies ( digital / analogue circuits) seem to have a very positive effect, as well as - again - low tolerance components in both digital and analogue circuits. Also, keeping jitter under control seems to make a difference in the overall sound.

Finally, I've always been a great supporter of synergy between components, and I would put this at the top of the list of priorities when setting up a good sounding system.

A bit of a trick question that dvv you rascal

Too many factors to give the right answer or is there really a right or correct answer, but I think there are things that can improve the final sound, things do effect the sound because we know many manufactures may use the same chips or DAC chips but end up sound different.

So if any thing I would go for meaty power supplies as something very important to get right.

Good answer arthur, agree with that too.

All is nowt without a designer...

MONEY!!!

Arthur, I said "electronics", so forget speakers for the moment. They are a separate story altogether.

Frank says designer; I would agree with it for most part, but there's another thing before the designer himself, who will put the finishing touces.

In my view, a topology is about 50% of the answer. Power supplies account for the next 40%, and the rest all goes for no more than 10%.

Topology, or circuit type and implementation, are the key to everything. I got really interested in this subject some years ago, so I did a little experimenting. I took three different topologies and played around with them for around three months, just to see what would happen if I .... something.

The first is the text book topology, as was used by Naim for example, in their early 80ies models, but is still much used today. It turned out to be hard to get it just right and was very sensitive to component choice. It was inherently rather slow, and produce pleasing, warm sound, but a sound which was dynamically compromised. It just couldn't keep up, and the subjective feeling on dynamic material was that was a lazy amp, nice to casually listen to, but not up to dynamic material.

The second was Marantz's P700 platform, as used in a number of their arguably best models 1978-1980. It is a deceptively simple layout, but in fact, an uncommonly capable one. Whoever derived it from whatever did so by using an impressive amount of brainpower. It defied many of my attempts to improve on it in terms of exchanging a simple resistor for a transistor based ciruit; the sound changed very little, if anything. However, on the other hand, it reacted very positively to the exchange of original semiconductors with newer and better ones, still in the same mass produced category (i.e. not some exotic component which costs an arm and a leg), improving performance quite a bit. In terms of dynamics, it was hard one to catch at any fault, or in terms of response under open loop (no overall feedback) response.

Lastly, I used a classic Bongiorno fully complementary topology, as used by 90% of US high end companies in one version or another. This topology yielded the lowest natural distortion factors, and was consequently ideal for low feedback designs. But where it excelled was just where you'd expect it to excell, which is under dynamic conditions. Since music is almost never ideally symmterical in the plus and minus halves of the compound sine wave, this topology, which also divides the amplification of the sine wave in the same two halves, reacted with least transient distortion factors, and could be made to have an extraordinary bandwidth.

While this may appear at first glance to be a technical game play, it really isn't so. The P700 topology (Marantz) is an excellent example of how well a circuit can be designed to be very stable no matter what we put in it. However I changed components, used standard El Cheapo 5% carbon resistors, crappy capacitors, etc, it never failed to deliver the basic clean and clear sound it was designed for. Going to the other extreme, using hand matched high quality transistors unheard of in its design time, 0.1% metal film resistors, expensive capacitors, etc, obviously improved performance and sound, but left the original sound type and quality more or less intact, it just made the amp sound more clean and clear. Ditto for the complementary version.

The power supply also plays a tremendous role. It cannot make a poor design sound great, but it will make even such a design sound better than it used to. Nothing can match a fully electronically stabilized power supply, but unfortunately, these quickly become very expensive both in terms of real esate and sheer price; they are practically another set of power amps serving the first set. However, a good compromise is to use separate power supply lines for the voltage amp, typically say 5 Volts higher than the current gain amp line, and since the voltage amplification lines are low power even in pure class A (typically 20...30 mA), it's a good idea to have them fully regulated. That way, if the current gain stage is struggling with a wild speaker, the voltage gain stages have no idea any of that is going on.

In my experience, the reports claiming revelations and heavely lights coming on by changing a few resistors are simply some bad ganja being used during evaluation and/or uncoomonly large powers of self convincing.

That is a interesting set of comments.

From my listening experiance I would have said that the early naim amps (Nait 1 and 2, never did like the 3) where faster sounding than Marantz, with a good dynamic response. The down side to the speed and response was that with the wrong speakers they could be headache inducing.

If I had to choose which amp would be slow and warm sounding from Naim and Marantz, I would have said Marantz. I have owned Marantz, late 70's or early 80's, but can not remember the model number.

Colin, I limited myself to just one Naim model, the small 30 Watter from 1983, quite simply because I have its service manuals. I have no idea how other Naim models of the day were made, perhaps the same way, but perhaps not.

The Marantz P700 platform is, as far as I am aware, limited to the aupper half models of Marantz in their penultimate "Designed in the U.S.A., manufactured in Japan" series 2, 1978-1980. Starts from model DC 1152 and goes on upwards, including power amps DC 170 and DC 300.

As I write this, I am listening to my own headphone/preamp driving a refreshed Marantz DC 170 power amp, and trust me on this, there is no shortage of dynamics. I have not modified the power amp in any way, I just exchanged all of the electrolytic caps, however, because I couldn't find the dual concentric 2x12,000 uF capacitor as originally used, I installed two BC Components 22,000 uF caps. These are 70% larger, however, they are not exclusive parts, they are very solid middle of the road quality components.

From past experience, I believe Marantz' preamps are the bottleneck, somehow they tend to choke the power amp. In later models, the preamps typically consist of op amps, so changing the ubiquious NJRs for some quality AD op amps usually does the trick and makes it come alive.

given that in the late 70's, I was newly wed, more interested in beer, kart racing and rugby than music reproduction, the chances of anything in my system being from the upper reaches of any manufactures range are at best remote. I will therefore assume it was ''jap crap''. To be fair it can not have been that bad, other wise I would possibly have bought a Nait to go with another 'old' british brand of speakers, Royd Conistons.

I always found the early Naims to sound a bit brash in the mids and rolled off at the frequency extremes and two dimensional. All the modern Marantz amps have sounded flat and weedy, I hope the 70s version were better.

I think you make a good point about topology dvv, I think a simple but good good circuit design using modest components can sound just as good as a design that used top grade audiofilo parts, sometimes just changing a component to an higher spec part won't make the sound always better, it has to fit in with the rest of the circuit design and implementation.