Thanks for looking.
My main theme is this:
There is more detail and information trapped within recordings than most audio systems can reproduce. A system can have a flat frequency response and low harmonic distortion and it can play cleanly at high volumes, yet it can still fail to properly reproduce transients and the ‘acoustic space’; there is more to an audio system than ‘good tone’.
Accessing that beautiful sound could be a very simple problem to solve – there is no mystery. The magic happens when – stated simplistically – the pressure variations that emerge from the speaker are a duplicate of those that reached the microphone when the recording was made. This is only possible through the use of digital audio, DSP crossovers and driver correction. There may be debate on whether the speaker should be more, or less, omnidirectional and what size and shape it should be, but these aspects are relatively fine tuning.
For many reasons, the audio world chooses to avoid doing this. Most systems, regardless of price, are still built from simple analogue electronics and uncorrected speaker drivers that can reproduce steady state sine waves accurately, but which modify dynamically changing waveforms by ‘smearing’ the timing of their various components. Conventional measurements of frequency response show only the steady state sine wave performance.
Even if designers go high tech, they usually merely duplicate the building blocks of the traditional fifty year old audio system using more efficient technology. They also often have a fascination with expensive materials and massive components that have no effect on the sound. Digital audio and DSP are often seen, even by their own practitioners in audio, as just a low cost substitute for expensive hardware rather than an outright superior way of doing things. It is very limiting.
In a new twist, the designers who do embrace DSP often take it too far, assuming that the room is a problem that can be ‘corrected’. The opposite is true: because the recording captures only a static viewpoint of the audio ‘scene’, we need to listen in a real room in order to restore some impression of natural acoustics that respond coherently to the sound and to our head movements. This is what makes listening to speakers superior to headphones. Attempting to ‘correct’, or in some cases eliminate, the room is a mistake.
A single speaker cone cannot properly cover the full frequency range. Blending several cones of differing sizes without the use of DSP and active amplification can only be approximate. But accurately coordinating several cones using DSP is easy – the problems that plague traditional audio systems just fall away.
My argument is backed up by two audio systems I built myself. One of these systems has been heard ‘in public’, with favourable comments.
I cannot claim to be an authority on anything, but I have played with, and built, hi-fi and recording equipment since a young age.
I was very young when the hi-fi industry took off in the 1970s, and I was obsessed with it, reading all the magazines, and understanding the basics pretty well, I think. After that, I always had audio equipment – some of it DIY – but didn’t necessarily follow the trends and gossip within ‘the industry’. I made the move to CD in the eighties, and went through a headphone phase in the nineties and noughties, using a self-designed headphone amplifier for quite a few years.
Then, about five years ago I began taking more of an interest in the technicalities of speakers – something I had previously left to the experts. Very interesting! Is it possible that the vast majority of audio systems are built around a quaint configuration that made sense in the 1950s but can now be surpassed using more complex, but lower cost, technology?
I thought it would be fun to ignore much of the conventional speaker building wisdom, and have a bash at building a system using DSP, active linear phase crossovers, driver correction, unfashionably large sealed woofers and the ‘high end’ configuration of three drivers per speaker. As a result, I seem to have built something that, to my ears, works tremendously well, and which attracted some very positive comments when I showed it ‘in public’.
I then had a go at a different pair of speakers which are ‘concordant’ with my love of 1970s hi-fi styling. I am confident that these speakers, too, sound excellent despite their smaller size. This time I am indulging in DSP EQ to extend (not “boost”) the bass, which seems to give remarkable results – I know it is not ‘a free lunch’ but the upsides may outweigh the downsides – and it can be turned off of course.
Apologies if my online persona is dogmatic, dismissive, abrasive or sharp; I’m quite harmless if you meet me and would be only too happy to share your enjoyment of turntables, R-R tape recorders and valve amplifiers – and I love vintage equipment. But my particular theme is that if we are serious about hi-fi, then digital audio and DSP are mandatory, and the lovely turntables and valve amplifiers must be relegated to nostalgia duties. You may disagree!
My qualifications for all this pontificating are simply that I am an ideas-oriented R&D engineer with a B.Sc.. I can wield a soldering iron, use basic woodworking tools such as a router, write software, and I have some slight knowledge of how to process signals in real time using DSP. I am also a bit musical, and can doodle on the piano.
I also think that, at heart, I really am quite rational about this stuff – a philosophy that doesn’t quite mean what many people think it does.
In addition to the theory and practice of audio, I am also interested in vintage equipment, technology in general, the aesthetics of technology, and music.
Please post a comment. It would be great to make contact with anyone who would like to discuss the ideas behind music, hi-fi and audio.
You can also contact me directly on firstname.lastname@example.org
I do frequently amend and edit my posts retrospectively when I realise I have written something very stupid or badly worded, so if you think a new post is particularly bad, check back in a couple of days when it may have been improved!
[Last edited 20/05/17]