Effective acoustic center redefined

[Joachim Gerhard]

Yes i have but it is too late now in Germany to answer that. Let me only say that it is not easy to build a digital chain that is audibly transparent enough. For example inside a computer are many problems with the PSU. The algorithms may be correctly implemented but the signals that come out of the DA converter can be highly polluted. There can be problems with jitter and I/V conversion too. As i said we build everything to audiophile standards, including the choices of the passive components.

 

[john k...]

John, i was there in 1993. We build a digital speaker in Essex that could be switched from linear phase to 4th order allpass. With some artificial signal like short pulses, one version sounded "click-click" and the other "clock-clock". With music the effect was lost. Very linear frequency response and off axis radiation pattern emerged as the most important parameters, of cause with good dynamic range and low distortion.
Nevertheless i listen to phase linear wideband drivers. Not the stuff you might know like Lowther and such. Mine is plus-minus 1dB on an octave to octave basis, 1/6 oct. smoothed, extended to over 20kHz, analog equalization, high sensitivity. I prefer it to multiway, but maybe out of other reasons. Except the Essex system where we build everything ourself, even the DSP and the DA converters not to speak about the cables, i never heard a digital system that had even remotely the fine dynamics. But when have you a team of one Professor, 3 PHDs, two engineers and one designer totally obsessed with sound quality ? This back and forth switching on a low key computer based DSP does not cut the mustard. Anyway, our research and others did not change the situation a jota. Commercial speakers got worse the last 20 years. Everybody is searching for that magic transfer function that makes the damn junk sell. That is one of the reasons i spend so much time here with "amateurs". Amare is latin and means love.

What can be done on a PC today with a quality sound card is light years ahead of what you could do in 93 with SOTA dsp. PCs today have more computing power and speed that most main frames form the 95 era. Not to mention the progress in implementing dsp algorithms. And let's not forget 20 years of develoment and understanding.

 

[Joachim Gerhard]

OK, play it for me and i am convinced. Where do you life ? If you have nothing against it i would like to listen to your system.

 

[Joachim Gerhard]

One last thing. The electronics we used where designed by Prof. Hawksford, Dr.Greenfield and Dr.Bews. The DA converters where PCM63 selected for 20Bit, interleaved to cancel jitter. The I/U converter was discrete, super high speed with error correction. I have not seen something like that on a soundcard.

 

[Joachim Gerhard]

We did not need much computing power ether. The FRD was constructed on IIR filters and the phase correction with FIR filters. Dr.Greenfield was very good with compactification. That is also the reason he was later a Chief Scientist at RISC computers.
Also we committed sin. We only equalized over 200Hz. Anyway, i trust you that progress has being made. I have simply not heard it.

 

[GDO]

What i find most desappointing with all those DSP based systems is to see how little the hardware manufacturers care about the analog part of the signal treatment chain, especially in the case of PC soundcards, always done on the cheap with plenty of low cost (...and of course blameless) opamps and e-litics...Not to mention their blameless low cost psus too...

Ultimate EQ offers stunning specs on paper, but when i see the sort of sound card they recommend, ( not will mention any brand but...) i practically lose any faith that i will be pleased with the results. I own one of these sound card i use for measurements, nothing more because that stuff does not play music at all... ( although claims being made for musicians...)

 

[john k...]

We did not need much computing power ether. The FRD was constructed on IIR filters and the phase correction with FIR filters. Dr.Greenfield was very good with compactification. That is also the reason he was later a Chief Scientist at RISC computers.
Also we committed sin. We only equalized over 200Hz. Anyway, i trust you that progress has being made. I have simply not heard it.

You make my point. To do the FIR filters correctly to eq done to and below 20 Hz you need long FIR filters. I serioucsly doubt that there were any DSP chips available in '93 that had sufficient power to do that as there are very few available today. This has been one reason the UE remains PC based. Power supply and other electronic noise are problems with a PC based system, but those can be dealky with.

Anyway, let's not turn the thread in to a head banging on DSP as that is not the topic. I'll take that was was done in 93 was as good as it could be back then.