It would be nice to have a meter that tells you if something is right or if something is wrong. We have "state-of-the-art" broadcast quality test gear and even that cannot give us the full picture. However, the musician can! Provided you know what you're listening for.

Songs tell a story that words in speech alone simply cannot. Simple songs are easily understood. Songs accompanied by a big band, orchestra or rock ensemble, often have that particular lyric that makes sense of the song masked. But if we think in the time domain we find that the vocal differs in many ways to the sound of the instrument we are led to believe is playing over it. If we were there at the recording and able to step closer to the singer, we would hear that word. What we need to do is separate each performer in his or her own space. Because of stereo, aren't we supposed to be doing exactly that?

Bandwidth limiting was brought in as a concept to pre-condition the Hi-Fi buyer into believing CD and all things digital are better. Look at the arguments any way you want, but in analysis you will find glaring errors in them that you'd missed. But they served their purpose of making the chosen ones rich.

You bandwidth limit to make sure you don't waste amplifier power amplifying things you're not going to hear. At least that's the marketing spin chorused by all the Hi-Fi magazines all those years ago.

I have never known an amplifier waste power amplifying something that isn't there! After all you have a mains socket in your house, but it doesn't use power until you plug a load into it! So your amp could go from DC to light and won't waste any power at all amplifying DC or light if you are using it to play music.

So why didn't all those Japanese amplifiers boasting DC to light bandwidth sound better than the bandwidth limited western designs? Stability! I repeat, stability!

You can make an amplifier as wideband as you want but if there's a chance of it going unstable, it will. I could tell you about each and every travesty (means misrepresentation) but that would require a book which one day I'll write, but stability is one subject few amplifier "designers" understand.

Instability doesn't mean it'll blow, and it can easily pass-by EMC monitoring by use of "sticking plaster" techniques. There have been lots of unstable amplifiers that still work to this day, but don't work all that well. However, by bandwidth limiting you can make an unstable design stable. And it'll sound better than when it was unstable because there's less RF at the high end, and sub-sub-sonic oscillation at the low end to modulate and thus alter the audible signal.

There's one exception to the above rule. It seems that unnatural signal modulation is turning some people on. A variation of the class D principle uses a "floating clock" so there is no defined modulation "signature" in the audible band. Instead, the modulation phases up and down in a way the ear cannot latch-on to, which would otherwise be fatiguing. Phasing is actually a quite nice listening experience. That's why musicians use it, but our job is not to change their work, but to reproduce it the way THEY intended it to sound.

It has been through a deeper exploration of stability in the context of our wide bandwidth designs that the REFLEX has come into being. The way it has told us it is right, is through the musicians and singers - the way they have become "unstuck" from one another. The REFLEX has the ability to separate the performers further than before, but still keeps the togetherness or cohesiveness of the performance. It can enable you to hear that missing word while still hearing the instrumental accompaniment - that's what stability with wide bandwidth does - and nothing else can.

Picture it this way: You would never dream of buying a phono stage with zero headroom (overload margin) in the amplitude domain, because it would distort any transient signal above the nominal output. However, they can boast +26dB headroom, but that transient still distorts. It distorts because there is absolutely no headroom in a bandwidth limited stage in the time domain - It's that simple!

In the REFLEX the emphasis has been in making the extremes, both of the high and the usually forgotten low frequencies, absolutely stable, and that means a gain of unity (1) at those extremes, while still providing the right gain with the necessary time domain headroom for the music signal and all its harmonics.

As for the low frequencies, that's something that's impossible to achieve in a truly DC coupled design - this is why such stages need tremendously big power supplies to maintain a just-adequate degree of ground stiffness - the foundation from which the music is launched.

The REFLEX is therefore unashamedly ac (read capacitor here) coupled. It is amazing what happens once pride and an age of brainwashing is thrown to the winds. But we do understand that Electrolytics can distort. They distort when there is no polarising voltage across them - what do you think ours have? Cheap sexy looking black versions distort too. Ours are the very best, bought in bulk at considerable expense from the likes of Nichicon, Elna and Panasonic. They are special capacitors with many years of development behind them. Using capacitors we can ensure the REFLEX is absolutely stable at a gain of twelve hundred plus that's required due to the recording characteristic of your records down at 4 Hertz (thus ensuring faithfulness of phase from 40Hz upwards), ensuring the circuit is stable right down to one twelve hundredth of that frequency where its gain has fallen to just 1 - that "magic" gain number which says there can be no oscillation here!

At the high frequency end it is just as important otherwise you have an RF generator modulating not only your signal, but transmitting its own little radio programs too. Our use of bypass capacitors is nothing new, but maybe using the right types is?

Pride and brainwashing also says active EQ is wrong, but we have read Orwell's Animal Farm. The REFLEX is unashamedly active equalised! This usually leads to propagation delays which distort not only the transients at the start of a note, but "comb" the harmonics leading to a richly detailed mash of gross distortion (what some reviewers think is "warm and detailed"). It also leads to emphasis of record noise, producing massive spikes from all but the most virginal vinyl (unless further bandwidth limited). However, if there are no propagation delays then there is no such distortion or emphasized surface noise.

The REFLEX has taken the Era Gold V's audio-propagation-delay-free active equalisation stage (250 volts per micro second) a step further, using two of the V's noise matched op-amps in parallel, which increases current density, reducing circuit noise further than its already silent operation, and increases the already more-than-adequate "drive", but also takes the output into class A bias for virtually all output conditions that will ever be encountered.

With all this "processing power" it would be foolish not to try and get every single part of the signal through. In the REFLEX we have examined the effect of the micro-currents that traditional amplifier star-grounding often neglects. What a waste not to hear what is in-fact there on the record by forgoing this opportunity. Usually grounding is used to simply get the noise performance down within acceptable levels. It should come as no surprise then to find that all we did amounted to not one single difference in S/N ratio, but it made a big difference to the amount of signal retrieved from the record. We have not re-tracked the existing Era Gold V board because PCB tracks are too inductive to make this sort of difference. As such, the grounding scheme within the REFLEX is hard wired on - not just to the earthy side, but to the supply side that is essentially part of the signal ground too.

In the event we also found how alarmingly well the IEC's standard 61938, recommendations for cartridge loading, work. There must be hundreds of phono stages with loading adjustments to twiddle that leave the listener frustrated because "no difference could be heard". The REFLEX, in keeping with all our RIAA stages has no such adjustment. It is quite simply optimised to the standard as all references are. In other words, we calibrate it so you don't have to do our work for us.

I suppose we could have taken a gamble on the above explanation making us rich, but in keeping with our philosophy of "being the common man for the common man" we have priced the REFLEX at an hopefully affordable £620.

Hereby hangs a warning: If other components in your system are simply not capable of this level of performance, then you'd probably be better-off choosing something else from our range of phono preamplifier stages. Until next year that is. Because the next thing we do will be to make a preamp to allow the REFLEX the room to work.