Rupert Neve: The SOS Interview

[Music] some sound took a rare opportunity to talk in depth with legendary audio equipment designer Rupert Neve who at the age of 89 is still designing high-end stereo equipment we opened by asking Rupert about the origin of his world-renowned EQ designs we were talking about how to change the balance in a recording bear in mind that it was all mono it's unthinkable in today's context but you literally had perhaps three or four microphones at the most feeling into a very simple mixer and the only eq's available in those days were things like the baltic an excellent piece of gear but very very gentle goes very low cue and so on and affecting the overall picture if you like but not any kind of surgical importance to what he was doing and you bring a bunch of musicians in and make a recording and they find that the guitar is kind of lost so what do you do you bring all the musicians in again scouring the various nightclubs as on into which they've all disappeared over the last week get them all together again and rerecord very expensive difficult is there any way we can change that mix well EQ occurred to me and I found that the guitar they gave me a tape and had a surprisingly restricted range of frequencies and so I developed a very high Q equaliser circuit which was around somewhere in the region of 2000 cycles and you could swing this a little bit this way in that way and it would actually homed in on the fundamentals of the guitar what it did to the harmonics is different story but and you with with another different equalizer you could play around with the overall picture and actually recover quite a decent signal from the whole thing so we managed to do that and they liked it the interesting thing to me where I started on quality I was always a quality fiend if you like and so the amplifiers into which this EQ is embedded were all well needless to say they were Class A because there were no high C's in those days and so every circuit was single-sided this was transistors by the way not tubes inputs and outputs were transformers I simply didn't know any other way I'd grown up with tubes now tubes have some hundreds of votes on their plates you can't simply and the authorities wouldn't let you put a capacitor from a high the output of a plate circuit into a landline it master thousand you've got deaths and people falling around you dead with the electric shock no it can't be done transformer is the obvious way and I had already in previous jobs that had developed high quality tube amplifiers audio amplifiers it was a few steps to turn them into transistor designs still transformers seemed to me to be the only sensible way of interfacing with the outside world well when we first tried these equalizers at Philips records and first concern was is the basic amplifier good enough all this was on chuckle German connectors which meant you could jack in the input and output circuits simultaneously so they said everything flat and they played tapes that they were familiar with and they spent the afternoon jacking in and out in an hour listen listen listen in circuit our circuit in circuit out of circuit and finally makhmour ERT who was golden eared engineer at the time he said you know is I had to admit it but I think it actually sounds better was that amplifier switched in now there's all kinds of theories as to why it sounded better first of all you are eliminating all kinds of background stuff which is inherent in many systems they were jacking this circuit in and out of a studio complex which went through probably two or three different patch fields power amplifiers the source was a tape machine and by the time you've added in the source impedances the load impedances and so on and so forth and when it came to the transformers that I had introduced I knew although I say it with some arrogance I knew what I was doing when it came to transformers it was my subject and I think it was a matter of interfacing with their system rather than any specific quality that the amplifier itself had although it was a good amplifier when they started to apply the EQ and found that it actually did help them to change the ratio of the guitar in the mix and they could make other adjustments to over the rest of the frequency range um I was you know home and dry on that one that was the origin I suppose of the EQ [Music] so what are the characteristics that make a truly great equalizer circuit an equaliser should have no ideally should have no distortion of its own it's almost impossible to it to do it without some distortion but that's the aim so if you can do that now there are a lot of equalizers about where people have just taken a circuit out of some reference book and hope for the best and they've never really succeeded because I haven't taken account of all these other things specifically the EQ circuit has got to have a certain Q if if you if you make it too if you make the sides of the curve too steep it's going to spoil the sound of the musical instruments that are within the Bell and if it's not steep enough it's not going to be very effective so it's a matter of judgment of listening and listening and listening to try and achieve something that sounds acceptable and to begin to understand what is actually doing to the harmonics for some reason if you were to enhance the harmonic structure such that you've got a lot of fifth and seventh which is an absolute sound no no it's not going to sound sweet whatever the opposite of is you know bitter horrible and so it's a question of listening understanding why it sounds the way it does and then pursuing a particular direction to make it sound that much better avoiding the very high order harmonics the shelving filter which is controllable in terms of its amplitude and its turnover frequency and you've got to have enough of the curve which is approximating or equal to about 3 DB per octave if it's not achieving that degree of steepness it's not an equalizer it's a tone control if you like it sort of changing the overall balance somewhat in order to provide a reasonable range of frequencies over which this 3 DB or 3 and 1/2 DB per octave applies the departure point from flat to where the curve starts takes a dip it's slightly cheating as it were so that you look at these curves you'll find as there's a dip just where the effect starts and then it goes on up to and exceeding 3 DB per octave and if it's a showing curve it'll flatten off so it's really a question of squeezing as much as you can from the the the frequency of where you starting the EQ process to the point where you're finishing it that is if you really want to apply maximum EQ of course you don't always want to apply maximum EQ sometimes it's a very gentlest of slopes that you need nearly every IC is itself push-pull amplifier containing an element of crossover Distortion so if you can avoid that or use the IC in a way that crossover distortion is not oppressive that's the first thing that I listen for and then the easiest way of avoiding crossover distortion is to make it all single sided so by definition it's there's no crossover and then of course you've got various other forms of distortion which come in and you can hear these mechanisms beginning to work and when you're satisfied with the sound that it's a musical sound from a personal point of view I'll sit back and relax and I I do mean relax if I can't relax thing isn't right being known as he is for designing high-end analog circuitry we wanted to know what rivet thought about using ICS or integrated circuits after I sold the old NIMH company there was a huge movement to convince the newer owners that my designs were over the hill and there were much better ways of doing now these integrated circuits had appeared on the scene so much smaller so much more efficient so much cheaper and the consoles from that time that was sort of mid seventies onwards started to lose that precious something because of the insistence of introducing these technologies at the end of the day it's musical it comes to a gut feeling that this isn't the way to go you try it doesn't sound right not going to do it the op-amp is the heart of the amplifier the consoles that we currently manufacture make use of op amps but op amps of our own design they are discreet op amps if you like they had to perform in every way better than the best I see that you can find and avoid the artifacts of crossover distortion of slew rate distortion and and related to that of course is the bandwidth we examined every possible use of existing integrated circuits because the advantages are so obvious if you could use them but the more we delved into it the more we realized that we couldn't so we come into the mic pre and that the first thing the signal encounters is a really nice event and that oppa doesn't have to do much in the way of feeding any power to any way it's got to feed a voltage into circuits sometimes quite a lot of cabling involved the big console has got a lot of wire in it and so you route your signal in the most efficient way that you can to possibly an equaliser circuit or a series of equaliser circuits and now your op amp has got to adopt a format which is going to is going to operate happily in the context of equalizer circuits capacitors sometimes inductors which you're folding around it so it's got to be capable of feeding that kind of circuit again it's not a power thing but it's because an amplifier doesn't know the difference between power and volt amps but it's got to be able to feed the requisite number of volt amps into the equaliser section you then go into your mix bus and again same sort of thing you've got to study what is it expected to do what it's going to feed and how many inputs is that bus going to have to deal with and so on so every stage there are marginal differences to an op amp which is designed for you know for that particular area the output end is a transformer transformers have these amazing advantages of galvanic isolation and the fact that you can choose the impedances that you want them to work between and I'm sort of almost born and bred into transformers so I find it very difficult to be completely objective about transformers so what exactly is it that transformers provide which electronically balanced interfaces don't there is a sound element to it and largely the isolation the fact that you can run a line over many hundreds of feet meters and the transformer will the transformer has leakage inductance your line has capacitance principally it has some inductance as well but if you put a line a long line and an output transformer together you've got a resonance and the resonance will depend more on the lengths of line than anything else so there is a limit now you say well why use a transformer because the fact of a DB and a half or whatever of resonance in a long line plus transformer is a sweet and musical sound if you use an amplifier directly without a transformer the capacitance of that line will cause the amplifier to slew and will give you store ssin of a very undesirable it's not musical so that is really the essence of why a transformer most practical applications is better than a direct output from and from from a transistor or from you and I see of course ribbit hasn't stopped designing new circuitry and one of the latest additions to his products is the silk circuit the equalizer not only was capable of subtly changing the relationship of say a guitar to other instruments which is a fairly heavy use of an equalizer but you can play around for the sake of trying to describe what happens with an equalizer in a very subtle way and produce a sound which you prefer to the totally flat frequency response now this is musical preferences as is down to the actual sound engineer who will play with a sound is to do with the microphone it's to do with the acoustics of the room and quite a number of other factors and I found that in general terms some increase in second harmonic and some increase in third harmonic and we're talking quite subtle amounts actually sounds better than complete elimination of those harmonics so the silk circuit does give you the ability to introduce it it's not that it's a huge amount of equalization is very subtle equalization but it changes the harmonic structure at the mid and low frequencies and that is something which is very very hard to explain in words you just listen to it and say yeah I think I like that better technology has changed a lot in the time that Rupert's been designing circuitry how does he feel about digital modeling for example so everything has become much more compact everything is much more efficient components are smaller better quality for the most part more reliable and you can do things today which would have been very difficult to do 30 years ago we've been working with certain company that has produced some really excellent models and it is very hard indeed to tell one from the other if I were to throw a shadow of a doubt on something that they have done and saying I don't think this is quite you know three months later they they've corrected it and brought back another one to listen to yeah that's been dealt with and it gets better and better and I think that the other thing to bear in mind is that like all digital things it depends upon the resolution standards to which you're working and if you're working to a sufficiently high sampling rate it can be very good indeed certainly the one that we're working with this is astonishingly good so from someone so experienced what are the future what's to come next from analog circuitry design well if everything's been done what would I do next no I don't think everything has been done hmm I think this whole question of pursuing perfection is something which will never finish if you're climbing a mountain you start to climb this mountain and you look up and you see this peak above you it's a boy that'll be great when I get to the top you get the top there's another mountain and you go on climbing and it's like that I can't tell you with any great degree of precision what the next step is going to be I would like to see the buying public becoming sufficiently educated to appreciate the sound and to really understand that they can listen to a first-class recording and relax and enjoy it I would like to play with the I am playing with the harmonic structure of the sounds and we've designed a circuit which will invert the sound structure if you have a sound a musical sound perfectly legitimate musical sound with a certain amount of say third harmonic in it then if you look at the waveform there is a way in which you can as it were suck out that third harmonic invert it and feed it back in again and that will make the sound subtly different you can change the structure of the signal that were examining and you've opened up a whole new possible chapter of a sound manipulation if you call it that it's not all going to be good long chalk but it's interesting it can be done if you liked this video subscribe to the sound on sound youtube channel for all our latest video content also you can find the magazine in your local newsagent download the tablet edition or find us online at Selman sound calm thanks for watching [Music] you

Channel: Sound On Sound magazine

Views: 112,041

Keywords: Rupert Neve, Rupert Neve Designs, RND, Op Amps, Neve, Recording Techniques, recording equipment, recording studio equipment, music recording software, How to record music, microphone techniques, recording at home, Sound On Sound, recording tutorials, mic-pre, Sound On Sound Magazine

Id: jAwwsrnNdzg

Length: 25min 46sec (1546 seconds)

Published: Thu Sep 17 2015