audio-talk

Indeed - I sure one can make the argument both ways round as well so its meaningless without context setting out what unmusical means in this statement... or its a debating trick to always enable one to always be right when making the statement by shifting its meaning... I'm tempted to ask what non ...

Its interesting that DEL Shorter & Norman Crowhurst back in the late 50s and early 60s described how the harmonic progression of distortion affected our perception of sound quality and how feedback and especially global negative feedback in commercial amplifiers tended to make this worse than it...

Not really alluding to anything in specific - the 989 use a LR network for the delays and, I believe, that they are more accurately trimmed than the 63s were. One of the reasons for the LR network on the 989s is that it worked worked with the capacitive panels to produce an easier speaker to drive t...

Its all about balance and nuance. And one big difference is between an amplifier for the general public that has to work and be stable in all possible systems and a diy amplifier that can be optimised for one system... Global feedback is hard to get right for all possible systems deployments and muc...

The 989 have a different arrangement of electrostatic panels that mimic a point source more accurately than the ESL-63 or other speakers and have more control of the panels for better bass extension and dynamics compared to the ESL-63s. A better speaker all round than the 63s but still not to my tas...

I worked with a college who had a pair - actually was on the design team at Quad that developed them so they were well setup and driven by his own custom amps. They were superb on most small piece music but didn't go very deep in the bass and the sound tended to breakdown on large music pieces such ...

All visible to me. Using Vivaldi browser on Ubuntu 22.04 LTS.

Splendid!

The frequency plot shape is pretty typical of an OPT. Here is an indepth measurements of a Willianson OPT with explainations of what the plots are showing - its long and semi-technical not a lot of equations for fully technical :D https://dalmura.com.au/static/Williamson%20output%20transformer%20mea...

good point - I've certainly seen that before - usually when chasing stability issues...

When doing the frequency sweep it is sometimes instructive to compare the input on one channel with the output on the other - sometimes an amplifier loads the input device such that the input sweep isn't flat. If that is the case then subtracting one trace from the other will show the amplifier resp...

The 10KHz square wave edges are showing signs of slew rate limiting when asked to exceed 1Vper10usec that is 1 V slew at 100KHz - probably OK for audio at our age :-) Agree with the 100Hz or so roll off visible on the 100Hz waveform. It could do with some tweaking to optimise the hf compensation but...

I designed for 20mm and the only reason I used 15mm in the end was that this was the thickest perspex that the baffle maker could bend and guarantee no stress marks... 20mm is audibly better than 15mm but not by a lot - it sounds a little more open in the upper bass and midrange to my ears on a flat...

The perspex is pricey in the thickness required to be acoustically transparent and it is very difficult to bend in that thickness without producing stress marks in the material- all too visible. I ruined the best part of €4000 worth of material before I saw the light and got the baffles made profess...

Interesting to see the cost ramp up... At, say, £6000 for parts and the usual ballpark commercial multiple of six times for consumer cost compared to cost of production they should be close to £36.000 - which is getting a little expensive so say that a commercial production cost was £4000 compared t...