Unfortunately Frank, few of us can be that purist to eliminate a local source "switchboard", in form of a preamp. For sources, I have my TT, my CD player, my tuner and my cassette deck. I tend to use them all, some more, some less, but having to change the wiring every time I want to change a source is not really practical.
I agree that bypassing an entire device like the preamp should yield better audio results, but I also find that the deterioration of the sound through it is tryuly minimal, while the practical benefits are signidficant. Until you come to the TT, which will not work without an eg/amp.
Ooooh - you can make indoor shish kebab. You lucky man! 😀
You could pack my son's entire apartment into your living room 15' x 30' and have a bit to spare. So, room sizes indeed do vary around the world.
Fortunately, so do the speakers. A friend neck deep into loudspeakers recently completed a small box using Morel drivers. As expected, it was very clean and detailed, but what floored me was the bass extension - I have never heard a small speaker manage that so well. I estimate it goes down to just under 50 Hz, a first for such a small box. He explained that he used a usual trade-off, he sacrificed efficiency for extension, trading off the drivers' reasonably high efficiency for low roll off. Still, he managed to get effciency to some 90 dB/2.83V/1m, which is still reasonable. He reckons it worked out to about €600, but this includes shipping, ciustoms duties at 9% and VAT at 18%.
Barleywater, I haven't got round to doing measurements with a test signal going through the convolver. Your instruction were clear, but I already got lost taking another route: VSTHost with virtual cables. Will come back to it later.
In the meanwhile, I found the following on the audibility of different filter types set off against different filter slopes. http://lib.tkk.fi/Dipl/2008/urn011933.pdf
My summary brain has condensed it to the following bits of knowledge I may be able to retain :
- LR filters of 8th order are audible @ 3 KHz, but at lower xover frequencies, 8th and even larger filter orders may be used.
- FIR filters can lead to audible ringing off axis at filter slopes > 600th order.
- the use of the concept of 'group delay' obfuscates and the conclusions would have been even more useful if another metric had been used. But it is tradition.
Richard,
I haven't really thought about it in these terms, not have I come across it in measurements (but I haven't been looking for it either). If two drivers have the same radiation pattern at the crossover region, it appears that there cannot be a problem of this nature. Another good argument for controlled directivity.
What you describe is rather a boundary effect than room gain. Your listening room is large enough for room gain to only play a role in the case of minor nuclear explosions. Boundary effects are present in small and large rooms alike.
Dejan, just note what inspired that post - Bonsai wanted a "real set up" that gave me good results, and the first one is probably the best because it used better components, and required less internal fiddling. I have no hangups about any particular combination being able to get over the line - a well enough engineered preamp should be able to do it. But I will state a key requirement: I have yet to hear a system that uses a standard volume pot being able to sustain the quality needed - every system I've fiddled with, or others I've heard good things from, has bypassed using such a part.
My design philosophy has resulted in opposite approach - preamp covers all the gain, it has high output voltage swing. Power amplifier has +1 gain, it is a power buffer. Link cable between preamp and power amp carries high signal level. Noise is greatly reduced.
I had an SD375. A friend at MIT built one of the first hardware FFT engines ever as his masters thesis circa 1970.
Yes, everything in the one box is a pretty good solution - but that means you really have to engineer things well - can it be done? 🙂
Can't help with specifics, but I think swept heterodyne analysers arose from the war in the 40s ?
I am glad we share the same opinion.
I am testing your advice.
I’ll be back soon with screenshots and more questions
Very good approach Pavel. And you won’t have many competitors
More inside stories
http://www.dataphysics.com/30_Years_of_FFT_Analyzers_by_Sri_Welaratna.pdf
and a treasure found during my latest search for Rodney Potter’s work
https://ia800301.us.archive.org/29/items/ElectronicMeasurementsAndInstrumentation/OliverCage-ElectronicMeasurementsAndInstrumentation.pdf
George
1965 or so Altec Lansing was doing "Acousta Voicing" http://www.preservationsound.com/wp-content/uploads/2011/08/Altec_Worship_1966.pdf
Boner was hand building filters before that.
RTA's certainly preceded FFTs.
Now the first hobbyist to publish a circuit I don't know. But the circuitry was certainly in National's handbooks by the early 70's although not explicitly shown. Clay Barclay built his own which was later sold by Crown Audio.
I certainly had one in the early 70's and by around 76 had one based on the Reticon Fourier chip built into a Heathkit oscilloscope. Cost just under $2,000.00.
Shure did one that used window comparators to avoid a costly LED display. You kept decreasing the window size as you tuned.
The biggest issue was not the circuitry but the measurement microphone. Still is!
Boner was hand building filters before that.
RTA's certainly preceded FFTs.
Now the first hobbyist to publish a circuit I don't know. But the circuitry was certainly in National's handbooks by the early 70's although not explicitly shown. Clay Barclay built his own which was later sold by Crown Audio.
I certainly had one in the early 70's and by around 76 had one based on the Reticon Fourier chip built into a Heathkit oscilloscope. Cost just under $2,000.00.
Shure did one that used window comparators to avoid a costly LED display. You kept decreasing the window size as you tuned.
The biggest issue was not the circuitry but the measurement microphone. Still is!
Last edited:
Thanks for link.
From above in 5.2.3:
This is typical example of bad driver layout. Driver center to center appears to be 14cm; For 3kHz crossover this is roughly one full wavelength; a 1/4 wavelength is much better target. Vertical lobes are horrendous with any crossover. None the less the paper shows that some really steep crossovers by FIR work quite well with their setup.
Low frequency audibility of phase in this study with high order IIR filters is worthless.
Here is 10Hz square wave response of Sennheiser HD-600 headphones:
It sounds nothing like 10Hz square wave. High pass filtering of system acts effectively as a differentiator circuit. The test signal functions as positive and negative Dirac pulses at 50ms intervals. I've set up cardioid sub-woofer that does 10Hz square wave. The entire room throbs even at low levels. Window glass flexes in and out. Once the phase is this messed up, a few thousand degrees of phase shift don't do much more damage. This accounts for paper's results for high order IIR crossover all pass filtering at LF sounding no different than with FIR, or just the speaker's native response.
As a test signal for 3kHz crossover it returns positive and negative going waveform of system's IR. A great test signal for 3kHz testing polar pattern of a speaker in terms of crossover behavior.
So happens I've tested Sennheiser HD-600 and results are comparable to results from Build A Graph | headphone.com. Their 50Hz and 500Hz square wave results are good indicator of how difficult it is to know when a real square wave in being produced acoustically.
Perhaps it was Marcel Wallace's 1932 French patent for 'Panadapter' ? First consumer commercialisation 1946, featured on front cover of Radio News March 1948, product PCA-2 T-200 ?
These are heterodyne sweep anaylsers by any other name.
The difference is I do not need a huge power transformer in the high-output swing preamp (big transformer is a source of magnetic field interference). It has to drive only input impedance of the power buffer which is placed in a separate case. The improved S/N is the fact.
Ijust leafed through my copy of the November 1925 (yes) Popular Radio and there is no sign of anything looking like FFT.
Although they did praise a new use of photocells to sort cigars by color, and an ad for an 'A-C Dayton Radio - for the Man who believes his Own Ears'.
Not kidding. 😉
Jan
Although they did praise a new use of photocells to sort cigars by color, and an ad for an 'A-C Dayton Radio - for the Man who believes his Own Ears'.
Not kidding. 😉
Jan
- Status
- Not open for further replies.