BNC

It's a funny old world where a range of speakers pretty well equals the costs of a range of cars from nice saloon to supercar. We are just talking about a box, a few magnets, coils of wire, cones etc. My view is that supposedly high tech engineering doesn't do any harm, but is not actually crucial - it really doesn't matter whether strut 3A is 4.2mm or 4.5mm across or, probably, even there in the first place. These chassis look like demo files for a CAD program, created in about 10 minutes to look complicated and impress the punters, but which hardly cost any more to make once the CNC machine is purchased - cyncial, moi? Overt over-engineering isn't all that expensive to do once a decision is made to market one's company as 'high end' in contrast with typical mass production and its often absurd cost trimming.

That's not what I understand the term "bit perfect" to mean. I assume it to be a purely 'mathematical' term that says that the numbers that go to the DAC should be the same numbers stored in the file. It makes no claims for quality beyond that (jitter, noise etc.). But given a particular setup with certain levels of noise and jitter, it is preferable to feed the DAC with the numbers in the file at the intended sample rate, rather than to re-sample to a different sample rate, or change the resolution, or add dither. A truly "bit perfect" system could still have worse sonic specs (noise, jitter) than a non-bit perfect system with great noise performance and a near-transparent re-sampling algorithm.

I think this is where the 'old ways' meet the new. A lot of speaker 'folklore' still persists even though DSP completely changes what is possible. Done with DSP and measurements, three or even four ways will integrate perfectly, with no more effort needed than designing a two way. And of course it then works much, much better than a two way (drivers not working outside their comfortable frequency zones, less excursion on the drivers etc.).

My system is homebrew: My own software running on Windows XP acts as a real time convolution engine giving N way crossover with 'live' calculation of the filters, so I can tweak crossover frequency and slope, driver delay etc. and hear the effect. The filters are overlaid with inverse impulse response measured for each driver in the near field (using REW), plus adjustable baffle step compensation (essential I suggest). I'm not trying to to do room correction, merely to make the speakers 'correct' in themselves. A single PCI Creative X-Fi sound card running in an old but silent Dell PC, acts as the destination for any music app, or can take inputs from SPDIF or line in. My software can then access that stream, process it and send out multiple processed channels to the analogue outputs over ASIO. No re-sampling is necessary as the input and output are locked to the same clock. By default the Creative drivers respond to multimedia keyboard volume. My calculations are floating point and the card is 24 bit, so I don't think I'm losing much quality using software volume control. I've tried a few cards, and the Creative is the only one so far that behaves perfectly in terms of what it does when the software isn't running, and total lack of unpleasantness at power up and down. The crossover software uses about 20% of the CPU, but this is a 9 year old PC. Six analogue outputs feed three stereo amplifiers, and then my homebrew speakers: 12" polypropylene woofers in 90l cabinets, 4" polypropylene mid and 1" silk dome tweeter in separate smaller cabinets, all sealed. As far as I can tell, it's a 100% success (I would say that wouldn't I!). In terms of convenience, I press the PC power button and all the apps load at startup (or left running at the last hibernate). I turn on the three amps. Select a track and press play, and that's it. My software is 'transparent' and runs all the time, so any and all multimedia apps will be heard through it. I favour a more complex crossover filter, so latency would be a bit high for video, I fear - I'm happy with just using the system for music. Obviously I could have as snazzy a media player as you can get, but I have a soft spot for Spotify (pay extra and you can have 320 kbps Ogg Vorbis which is pretty acceptable). Spotify responds to standard multimedia Stop, Play, Pause etc. so using a wireless keyboard it's all very convenient. You could probably do it all with your mobile phone if you knew about that sort of thing. I don't know if you're familiar with the sound of active crossovers yet, but it is certainly worth it (an understatement). If I had tried to build my own passive speakers it would have been a disaster, but doing it actively there is little mystery or black arts involved: make a swept sine measurement of each driver to use for inverting the impulse response, select the correct crossover frequencies based on rule of thumb or where the responses aren't yet wobbly, set the delay for each driver based on its distance from the listener and set the amp volume levels by ear by playing a sinusoidal tone at each crossover frequency and alternating between the two relevant drivers in each speaker, listening and bobbing about a bit to get an average level without room effects. Adjust baffle step compensation to taste. No need to touch the settings again - but you will again and again until you think you've got them just right! Forget all the awful so-called 'high end' speakers you've heard: DSP active speakers are phenomenally clean, rich and powerful. Amaze your friends! I would happily give you the software, but there's a lot of hard-coded stuff in there at the moment. I'm gradually working around to a nice GUI and all that. There must be off-the-shelf alternatives out there, but maybe not quite as tweakable at run time. Once you've got your ideal settings, it doesn't matter anyway.

I would have said the opposite. The PC with sound card gives you instant access to all sources: digital files, CD, DVD, SPDIF, analogue. No re-sampling necessary, and only a single A/D stage for the analogue input.

I once tried to set up a comparison test where I could switch between two sources using my amplifier source selector, listening on headphones. Using a variety of music, I could hear the difference between the two quite clearly. As expected, one of them had a lack of definition at the top end and a lack of 'bite'. It was only then, that I noticed that my headphones were not plugged into the amplifier, but directly into the headphone output of one of the sources. Since then, I've found it quite hard to trust my ears when it comes to subtle differences!

A worthy experiment, and I would be fascinated by the results between different cables. If I had to bet, it would be that you see no difference between cables, except for, depending on the output impedance of your source, extremely minor phase and amplitude effects at high frequencies strictly related to cable capacitance and not directly related to price. I suppose it is also conceivable that some cables might suffer more from susceptibility from hum from nearby mains-powered equipment. What you are really looking for, I suppose, is distortion in the form of spurious harmonics related to what? Micro-diode effects? Impurities in the conductor? etc. I look forward to seeing the results. Here's my view on how to resolve the timing issue: As has been pointed out above, you are effectively timing your recording randomly to within, perhaps, one sample period. What is stored in the file, however, has not been passed through a reconstruction filter such as a DAC would provide on playback. This turns the discrete samples into a continuous analogue waveform, and it 'resonates' at the sample rate such that it will actually overshoot and exceed the amplitude of the stored samples, producing smooth peaks to waveforms in between discrete samples. What you need to do, I think, is to software-upsample your recorded waveform to some arbitrarily high sample rate (using a software sinc filter), then line up the two waveforms (and optionally downsample again) and subtract them.