You've recently published an article about music signals above 20 kHz: may you briefly explain to our readers what have you discovered and how do your studies relate to home music reproduction?
I've discovered that each family of musical instruments (strings, woodwinds, brass and percussion) has at least one member which can produce energy to above 40 kHz, that is, at least an octave above what is normally taken as the limit of human hearing (20 kHz).
Even if it were known for sure that sound this high did not matter to human perception, the fact that the energy is present means that recording equipment must deal with it.
That is, at some point in recording, it will be filtered out, intentionally or unintentionally; and the equipment up to that point had better be able to handle the ultrasonics without creating spurious sounds or distortion. (In some percussion instruments, the amount of ultrasonic energy is huge!).
The cited paper is found at my site
The real content is in the graphs, which should be intelligible even to those who do not speak English; the purpose of the text is mostly just to convince the reader that the graphs are legitimate.
The paper also refers to papers by other authors that indicate that humans can perceive ultrasonic sound by bone conduction (one paper) and air conduction (the other paper).
Many audiophiles are always searching for a reproduction that is "absolute" and many of them believe to have found this "absolute".
I think this "absolute" is just merely "relative" and I mean "relative" to our ears, our experience, the way our brain elaborates the data from the ears and many other variables that make the term "absolute" sound silly, at least.
Agreed. One fundamental reason that there is no absolute is that no record/play system can match live music in dynamic range (see the AES papers by Fielder). Nor can any system match live sound in beauty.
I think everyone should search his OWN "best way" to reproduce the Musical emotions at home. Your views on this topic?
I agree with the thrust of this; but as I know from 20 years' experience, it is very difficult to find out what one needs to know about the sounds of various components and about their interactions.
Virtually none of the relevant information is available from manufacturer or from reviews, so one must discover it for oneself.
If the goal is satisfaction in music listening, and if there is a real expert available, one is better off using the expert for advice.
If on the other hand the goal is to learn for oneself about sound reproduction, one should become involved in *recording*.
An hour spent in recording will teach more about reproduction than ten hours spent with the reproduction equipment.
However, the most important thing to do in learning about audio is to start making music oneself: take lessons on an instrument, join a choir, or whatever else is possible.
The quality of attention that a musician pays to musical sound is different in kind, it seems to me, from that of others.
Recording live Music is a very difficult task. Especially because live Music dynamic content is hard to transfer on a reproduction system.
The minimalist approach seems to be the preferred one by most audiophile labels. May you explain us the pros and cons (if any) of a live-to-track few-mikes recording?
Recording live music is indeed difficult, and you are right that the dynamic range is what makes it so. Attempting to capture dynamics puts at risk every other aspect of the musical sound.
And if you read the papers by Fielder of Dolby Labs, you find that there is no medium in existence that can capture the dynamic range of live music.
There is a lot of nonsense talked about dynamic ranges of various media: digital vs. analog, for instance.
While many labels claim to take a "minimalist" approach, few understand what is involved, say, in spatially-accurate stereo recording; and few people in the industry actually spend time evaluating equipment directly against the live music.
That is, you can listen all day in a control room to the sound direct from your microphones. This is useless.
What you must do is to go into the room with the musicians and listen for all you're worth there, soaking up the beauties of the live sound, and the way *the sound itself is part of the meaning of the music* ---- and *then* you go back into the control room and see how much of that you've caught.
If you get 30%, you're doing well. (I don't know what "30%" means, but I'm trying to convey a certain impression.)
Correct coincident-mike recording by a single pair gives: accurate stereo (within a limited area at the front), correct transients and smooth frequency response. The moment the mikes are spaced, all of these disappear.
Yesterday I was listening to a percussion ensemble, live at few meters of distance. And I was wondering how much of that energy and dynamics is lost with recording and home reproduction.
Do you think that in the near future the HUGE gap that tears live vs recorded Music apart will be filled in some way? What do you think of these new digital technologies (24/96 kHz, multichannel reproduction etc.)?
*Most* of *that energy & dynamics* is lost, isn't it?
Do I think that the gap will be filled? No!
As for new technologies, the only one I've heard in any meaningful way---we auditioned it against live music in my lab---is a 24/96 recorder.
This was pretty impressive, but still not equal to the direct feed (the sound fed directly from microphones to speakers); and in the opinion of musicians listening to it, not equal to analog.
According to you, which should be the role and the scope of the HiFi magazines?
What they should do --- but none of them do this --- is to give trustworthy guidance to those who love music and want to hear it reproduced accurately in their homes while keeping expenditure within their budgets.
This already is impossible (!), because the choice for any one individual will depend on so many factors about which the magazine knows nothing. (See discussion of consulting, above.)
But at least a magazine could give information based on Valid Listening Tests as to how a unit sounds.
None of them do this, either, though Hi Fi Choice used to do so. (Maybe it still does. I can't find it any more.)
The magazines should also offer enthusiasts a look at the latest developments, but should report on them not as sycophants of the manufacturers' publicity departments, but as independent thinkers and workers in the field.
Some of these things are literally impossible, while others are never going to happen for economic reasons.
I did think that the British magazine HiFi News & Record Review, when it was edited by John Crabbe, was excellent, as was HiFi Choice in that same era (when Paul Messenger was publisher).
I have only English, so cannot comment on magazines from other countries.
As for the Usenet, I have on occasion looked up the discussions among audiophiles and also audio professionals on the one or two subjects about which I know something of my own experience; and I find such nonsense that it's beyond belief.
A friend of mine says that the Usenet has "a negative signal-to-noise ratio" and he's clearly right.
The piano is a very difficult musical instrument to reproduce at home. Which are the reasons for this?
Apart from dynamics, imaging (virtual image of the instrument) is HARD to reproduce realistically.
Some recordings have the right part of the keyboard placed in the left and vice-versa.
May you unveil some secrets for a good piano recording?
One reason is that few piano recordings are any good. Listen to a good recording and see how much better your system sounds!
Forgive me for contradiction, but the imaging is the one thing that *is* easy! Just use a Blumlein pair of ribbon microphones. (Listen to the Performance Recordings(R) Demonstration of Stereo Microphone Technique, catalog pr6cd; information at http://www.cco.caltech.edu/~boyk/records.html ).
As for secrets: record with audience, use coincident miking, and use ribbon mikes and tube electronics.
Of course any such summary is subject to misunderstanding, and I have no doubt that a very bad recording could be made this way.
*Any* approach can produce a bad recording if followed blindly.
I wrote about my approach to recording in "On Both Sides of the Microphone" ; the full text is at http://www.cco.caltech.edu/~boyk/bothsides.html
My other articles and papers are listed at http://www.cco.caltech.edu/~boyk/ARTICLES.HTM and more than a dozen are given in full there (English only; sorry).
Would you please explain us what does the Magnesaurus exactly consist of?
It is a modified Ampex 351 tape recorder.
Modifications consist of rebuilding the transport completely, carefully balancing all rotating parts, and replacing all bearings with ones hand-selected using FFT analysis of the machine's wow & flutter spectrum; replacing the heads and head assembly with custom ones for half-inch tape; modifying the record electronics; constructing a huge outboard power supply (energy storage over 100 joules); etc. etc.
There was a feature interview with me about Magnesaurus (TM) in R*E*P magazine some years ago. I'm trying to find a copy to include on my web site.
Magnesaurus (TM) specs include the following: wow & flutter unweighted <0.06%; DIN peak weighted <0.03%. Frequency response -6 dB @ 13 Hz and 30 kHz, +/- 0.5 dB from 50 Hz to 15 kHz; etc.
I'm not giving a figure for dynamic range because understanding what such figures really mean takes a lot of study and technical understanding; and I object in principle to the simple-minded dynamic range comparisons that have been floating around the audio world ever since digital recording first appeared!
Which are your future plans, both as a pianist and a recording engineer?
Hope to record a new album this spring, with Magnesaurus (TM), my custom analog recorder and simultaneously with a top digital machine, maybe the Nagra/dCS 24/96 combination. I'll play Prokofiev 7th Sonata & Beethoven Opus 126 Bagatelles.
Playing chamber concerts in May with my clarinetist friend Dr. Margaret Thornhill.
Giving master classes to introduce music students and teachers to my book, _To Hear Ourselves As Others Hear Us_
Would love to record major symphony orchestra in Sibelius 2 with my mike system and Magnesaurus as well as digital state-of-the-art. (My microphone system was used by Sheffield Lab for the _Passions_ album of music for two classical guitars.)
Courtesy James Boyk for TNT.
Copyright © 1998 Lucio Cadeddu