From: Patent Tactics, George Brock-Nannestad
Hello,
this is getting exciting; it is a subject that crops up from time to time; I
myself have been active in it since 1983, but I do think (or at least I hope)
that we are adding a little bit every time. At least we can inform new people
as the old ones die away from us.
Mike Biel wrote:
> Most of the time horn resonances from the recording horn result in a
> noticible bump in level at that frequency and at multiples of it to a
> lesser degree.
----- that is a bit of a simplification: it is an alternation of bumps
(resonances) and lack of sensitivity (anti-resonances) that go up the
spectrum. My first presentation at IASA in Washington 1983 was about this. I
had measured an authentic recording horn (borrowed from the EMI Music
Archives) in an anechoic chamber. I also demonstrated that you can EQ your
way out. In this I was aided by the Acoustics Research Institute of the
Austrian Academy of Sciences.
To the extent that you cannot find one curve to fit all, Steven is absolutely
correct--there is no eq curve. But you can do quite a lot, provided you know
where to find your information. I know that Peter Copeland has been maligned
in these pages--with reason--for some of what he wrote, but the acoustic
recording part is not bad at all. I shall not repeat it; you know where to
find it. My paper in Washington was printed in the Phonographic Bulletin, but
you would have to get it from a paper copy, because I have expressly
forbidden IASA to make it available in electronic form.
One of the first devices marketed to EQ these out was
> the Owl 1 and the Owl Multifilter, designed by Tom Owens. The Owl 1 had
> a notch filter which you would sweep up and down in a boost position
> that would greatly exaggerate the resonance, and then you would reduce
> it at that frequency till it sounded more natural. But you can do it
> with a parametric EQ, especially if you can also vary the bandwidth of
> the de-emphesis as well as the frequency.
----- it is very difficult and to a large degree depends on the circuits of
the equaliser, because of its time delay or phase function. In most
equalisers the settings influence each other, except the Orban type. I
personally use an analogue + and - 16 dB third octave equalizer, with a
digital control of the settings; built-in memory of 100 settings, but MIDI
connectable to a computer, which gives you unlimited settings. It is not made
anymore--TC1128--from TC Electronics. It also has a spectrum analyser and a
re-coding unit, in which you can convert a desired transfer function into the
precise setting of the equaliser. It looks very different from the desired
curve! These days you would use digital filtering, although the settling time
of some settings are still larger than for analogue.
Tom Stockham at Soundstream
> took this a great deal further with a digital system of finding the
> multiple resonances and reducing all of them -- sometimes hundreds.
----- the Austrian institute did hundreds for me, only we knew what to aim
for, it was not "blind deconvolution"
He
> also shaped the curve using electrical recordings of the same or similar
> voices, and this is where it becomes controversal, making Caruso sound
> like Bjoerling.
----- as an early demonstrator of what kinds of manipulations you can do with
digital signal processing it was fabulous, but his knowledge of sources was
absolutely insufficient. If you gave his Carusos a slight treble lift, you
obtained Bjoerling's voice with Caruso's enunciation. Fake and lies.
Whatever you do, high quality and low-noise originals are essential, because
if you boost an area where the original response has been low, all you get is
coloured noise at those frequencies. This mixes with the intended and
otherwise well-compensated signal, and it gives a different timbre to the
recording. For this reason I am looking extremely forward to having the time
to work with the Patti, Melba, and Tamagno reissues in vinyl from Heritage
Records.
The NIMBUS Prima Voce approach was misunderstood by the company itself. They
made the best imaginable acoustic reproduction system, i.e. one, in which the
cutoff frequency is between the lower end of the audible range and rumble and
without resonances. It was a huge exponential horn that would have made
Compton McKenzie's (founder and editor of the Gramophone Magazine, London)
mouth water, because he had the gramophone that was the best of its time, the
Balmain straight exponential that was supported without rotating joints on
floats on parallel mercury baths.
The fact that the NIMBUS horn was so linear in its response and had such a
large mouth of the horn had two consequences. First of all, it did nothing to
the recorded resonances of the acoustic recordings, and it was able to trace
almost any groove. This meant that even electric church organ records could
be reproduced acoustically ideally. The needle pressure was obviously much
higher than modern pickups, but not unduly so. NIMBUS insisted on pristine
recordings, and the person repsonsible for the transfer of some Danish
records was actually permitted to take the best copies from the National
Sound Archive to subject them to this. Personally I did not think that it was
worth wearing pristine grooves, however slightly, and a white lie would have
been in place.
But the main problem was that NIMBUS insisted on recording the sound in a
reverberating hall; this gave the treble-deficient (due to the thorn needles)
reproduction an almost artificial boost at around 800 Hz and provided a most
unnatural sound to the voice. But, what the hell, the buyers knew nothing
about vocal production anyway, and the marketing was tremendous. I should
point out that I took 3 years of vocal training at the Department of Music of
the University of Copenhagen when I started my project. I needed to know
about sources (and learnt about breathing and support in the process). I do
not think that you can seriously work with the sound on historical records
without knowing first-hand about sources. Discography is different but
obviously equally essential when you use commercial records.
>
> Mike Biel [log in to unmask]
>
>
> -------- Original Message --------
> Subject: Re: [ARSCLIST] Edison, etc., formerly Polk Miller
> From: Steven Smolian <[log in to unmask]>
> Date: Sun, November 15, 2009 10:31 am
> To: [log in to unmask]
>
> For acoustic recordings, there is no eq curve.
>
> Beyond that, it would be interesting to compare results of eq appiled to
>
> various types of cylinder and disc transfer when adjuted to a flatish
> line
> as guided
> by a spectrum analyzer, keeping in mind the frequency limitations of the
>
> original medium- i.e., don't dispair if nothing but noise comes up at
> 9K.
>
> Steve Smolian
>
>
> ----- Original Message -----
> From: "Tom Fine" <[log in to unmask]>
> To: <[log in to unmask]>
> Sent: Sunday, November 15, 2009 9:50 AM
> Subject: Re: [ARSCLIST] Edison, etc., formerly Polk Miller
>
>
> > This backs up my point. ANY transfer of this material will require some
> > sort of equalization to realize what the original artists and engineers
> > expected for playback to sound like. So how is an electrical transfer,
> > with electrical EQ applied to the transfer engineer or producer's ear any
> > more legitimate or proper than an acoustic transfer made through the horn
> > of the transfer engineer or producer's choice? I'm still trying to figure
> > out the venom directed at acoustic transfers and those who have done
> them.
> >
> > An approach where you know the exact recording condition, the exact
> > parameters of the horn and of the recording device, and you can
> > "un-distort" the sound back to what it was in the room before the horn is
> > a different matter -- although now you are attempting to remove the media
> > which may or may not be a good interpretation of the history. But, in any
> > case, where that sort of electronic/DSP transfer is possible, it's likely
> > to be more pleasing to more listeners, I think.
> >
> > -- Tom Fine
> >
> > ----- Original Message -----
> > From: "Steven Smolian" <[log in to unmask]>
> > To: <[log in to unmask]>
> > Sent: Sunday, November 15, 2009 8:07 AM
> > Subject: Re: [ARSCLIST] Edison, etc., formerly Polk Miller
> >
> >
> >> It was not until 1922 or 3 that the mathematics of horn design emerged
> >> from Maxfield and Harrison's work toward electrical recording.
> >>
> >> Until then, trial and error clashed with the forces of home decoration.
> >>
> >> Steve Smolian
> >>
> >>
> >> ----- Original Message -----
> >> From: "Tom Fine" <[log in to unmask]>
> >> To: <[log in to unmask]>
> >> Sent: Sunday, November 15, 2009 6:19 AM
> >> Subject: Re: [ARSCLIST] Edison, etc., formerly Polk Miller
> >>
> >>
> >>> So Mike, are you and others saying that a horn recording system is a
> >>> one-way system? I thought the point of a playback horn was, partly, to
> >>> undo some of the frequency response characteristics imposed by the
> >>> recording horn, like how an electronic disk recording system uses an EQ
> >>> curve to record and the reverse curve to play back. So, playing back an
> >>> Edison cylinder with no playback EQ -- to use one example -- is not the
> >>> listening experience intended, is it? So then you have to get into EQ
> >>> "by ear" by the electronic-playback engineer.
> >>>
> >>> So how is that any different, at base, from selecting an appropriate
> >>> horn, playing back acoustically in a benign acoustic environment and
> >>> recording the playback with a properly-placed accurate mic (ie a
> >>> small-diaphram instrumentation mic or like -- not an obviously colored
> >>> mic like a U-47)?
> >>>
> >>> I'm not advocating one transfer method or another, I'm just curious why
> >>> there's this hostility toward acoustic transfers -- of material that was
> >>> recorded acoustically to begin with???
> >>>
> >>> -- Tom Fine
> >>>
> >>> ----- Original Message -----
> >>> From: "Michael Biel" <[log in to unmask]>
> >>> To: <[log in to unmask]>
> >>> Sent: Saturday, November 14, 2009 9:09 PM
> >>> Subject: Re: [ARSCLIST] Edison, etc., formerly Polk Miller
> >>>
> >>>
> >>> This discussion sounds very similar to what happened at ARSC back in
> the
> >>> 80s when Bill Storm (as discussed by George) discussed acoustical
> >>> playback in a presentation. The mob (and it WAS a mob!) was about to
> >>> storm the stage (no pun intended) with torches and pitchforks when I
> >>> explained that he was not advocating ARCHIVING the recordings for
> >>> preservation this way, only using it as reference. But it does seem
> >>> from his posting that John Eberle IS advocating archiving and
> >>> distributing recordings played acoustically, much like the Nimbus
> >>> Nitwits and their Crazy Count did with their Prima Voce series back in
> >>> the early days of CDs.
> >>>
> >>> Let me stress once more, playing a record on an acoustical player is
> not
> >>> playing the record it is playing THE MACHINE. As Steve said, each
> >>> acoustical horn introduces resonances on the sound, and thus each
> >>> machine is placing its "stamp" on the sound of every record played on
> >>> it. It is NOT true that "you get to hear the true sound quality of the
> >>> cylinder as the Edison recording staff intended it to be." You get to
> >>> hear it the way they were able to play it back on the primitive
> >>> equipment they had at that time, not the IDEAL equipment that would
> have
> >>> been preferable EVEN THEN. To think they WANTED the recordings to
> sound
> >>> that way is absurd. They did the best they could under the constraints
> >>> of their nearly deaf boss. When electrical recording was available
> they
> >>> WANTED to use it. When electrical reproduction was available they
> >>> WANTED to use it. They (meaning Theodore Edison and to a certain
> >>> extent, Charles Edison) had to practically sneak around The Old Man's
> >>> back to improve their sound. It is an insult to the Edison recording
> >>> staff to say that playing a cylinder acoustically "is how they
> intended
> >>> it to be".
> >>>
> >>> And let me again state that I am not against playing acoustical
> records
> >>> acoustically for the fun of it or to experience what it was like. Just
> >>> like it is fun once in a while to ride around in a Model T Ford or
> some
> >>> other ancient car, but it is not the IDEAL way to travel.
> >>>
> >>>
> >>> Mike Biel [log in to unmask]
> >>>
> >>>
> >>>
> >>> -------- Original Message --------
> >>> Subject: Re: [ARSCLIST] Edison, etc., formerly Polk Miller
> >>> From: Steven Smolian <[log in to unmask]>
> >>> Date: Sat, November 14, 2009 5:50 pm
> >>> To: [log in to unmask]
> >>>
> >>> In recent years I have been privately advocating that each sound
> archive
> >>> and
> >>> other learning venues present a program called "How Our Ancestors
> Heard
> >>> Recordings."
> >>>
> >>> It would sequentially play back the same piece of music, "Stars and
> >>> Stripes"
> >>> or an earlier piece for which recordings exist in all commercial
> media.
> >>>
> >>> These would be reproduced through the actual machines for which they
> >>> were
> >>> intended rather than microphone recordings from them and would cover
> the
> >>>
> >>> time spectrum then to now.
> >>>
> >>> Using microphones to record what is played back on a phonograph,
> >>> graphophone, etc., introduces a further group of variables.
> >>>
> >>> Part of the presentation would include playing electrical recordings
> on
> >>> acoustical machines, mono through stereo speakers, etc. to give an
> >>> inverse
> >>> perspective.
> >>>
> >>> This is a different process than that to which I was objecting in my
> >>> earlier
> >>> message. It brings into play the sound of each instrument in a real
> >>> acoustical setting which is far truer to the ear of the listener in
> the
> >>> room
> >>> than what a microphone can presently create and mimic.
> >>>
> >>> Steve Smolian
> >>>
> >>>
> >>>
> >>> ----- Original Message -----
> >>> From: "George Brock-Nannestad" <[log in to unmask]>
> >>> To: <[log in to unmask]>
> >>> Sent: Saturday, November 14, 2009 1:47 PM
> >>> Subject: Re: [ARSCLIST] Polk Miller
> >>>
> >>>
> >>>> From: Patent Tactics, George Brock-Nannestad
> >>>>
> >>>> Hello,
> >>>>
> >>>> Steven Smolian wrote:
> >>>>
> >>>>
> >>>>> This approach is the shearest nonsense! Horns introduced distortion
> at
> >>>>> the
> >>>>> recording end and also at the playback end. This distortion varied
> >>>>> from
> >>>>> one
> >>>>> horn desingn to another.
> >>>>
> >>>> ----- I am sorry, Steven, but I do think that there is room and even
> a
> >>>> need
> >>>> for this kind of re-recording. It is a part of sound recording and
> >>>> reproduction history. I do not think we can teach our ears to forget,
> >>>> but
> >>>> we
> >>>> can at least be conscious of our ears and try to go back to the times
> >>>> when
> >>>> this type of sound was amazing. Edison is not a good example for the
> >>>> variability of the playback end, because he was the only recording
> >>>> company
> >>>> that sold a complete system, i.e. controlled everything. He may have
> >>>> been
> >>>> idiosyncratic, and obviously we also learn about his preconceived
> views
> >>>> when
> >>>> listening to his products.
> >>>>
> >>>>>
> >>>>> It may sound 'beetter" to the rerecording engineer but is a purely
> >>>>> subjective opinion.
> >>>>
> >>>> ----- I do not think it sounds "better" to anybody, but certainly
> more
> >>>> representative of the sound heard in the parlor than a good
> >>>> interpretative
> >>>> transfer that optimises the access to the sound as it was in the
> >>>> recording
> >>>> studio. A preservation transfer sounds horrible, but it is extremely
> >>>> useful,
> >>>> because it may be used for any purpose.
> >>>>
> >>>>>
> >>>>> It amazes me that so much is made of presumed audio purity based on
> >>>>> the
> >>>>> dictatorial opinions of a deaf listener.
> >>>>
> >>>> ----- I do not hope that it is the purity in an absolute sense that
> >>>> these
> >>>> transfers aim for. But it is rather amazing how good these
> "primitive"
> >>>> machines were. And the secret is, they were not primitive at all, but
> >>>> finely
> >>>> honed to the criteria they set.
> >>>>
> >>>> Bill Storm, formerly of the Belfer Laboratory, which was essentially
> >>>> founded
> >>>> by Walter Welch, was also a proponent of the audio history approach.
> I
> >>>> violently opposed it as a preservation format, you may see my
> >>>> discussion
> >>>> by
> >>>> downloading ARSCJv20n2p156-161 from the ARSC website.
> >>>>
> >>>> Kind regards,
> >>>>
> >>>>
> >>>> George
> >>>>
> >>>>> ----- Original Message -----
> >>>>> From: "John Eberle" <[log in to unmask]>
> >>>>> To: <[log in to unmask]>
> >>>>> Sent: Saturday, November 14, 2009 11:20 AM
> >>>>> Subject: Re: [ARSCLIST] Polk Miller
> >>>>>
> >>>>>
> >>>>> > We are collectors of Edison Phonograph antiquities here at Americana
> >>>>> > CD
> >>>>> > Mastering . In our collection ,
> >>>>> > we have Blue Amberol cylinder records of two Polk Miller songs
> >>>>> > recorded
> >>>>> > in November 1909 for Edison's clientele:
> >>>>> >
> >>>>> > Blue Amberol #2176 "The Laughing Song " and
> >>>>> > #2175 " The Bonnie Blue Flag "
> >>>>> >
> >>>>> > We currently have an mp3 of the Bonnie Blue Flag as played on our
> >>>>> Edison
> >>>>> > Concert Amberola Model A1 and recorded to Ampex GrandMaster 456
> then
> >>>>> > transferred to aif file on a Masterlink at 48khz/24 bit resolution
> .
> >>>>> > This
> >>>>> > way you
> >>>>> > get to hear the true sound quality of the cylinder as the Edison
> >>>>> > recording
> >>>>> > staff intended it to be ; and as it was heard and enjoyed by the
> >>>>> > thousands
> >>>>> > of Edison customers who purchased it 90 years ago !
> >>>>> >
> >>>> ...........
> >>>>
> >>>>
> >>>> P.S. from time to time my mails to the list are not "taken". I wonder
> >>>> if
> >>>> there is some online filtering going on. Just to test the system I
> >>>> recently
> >>>> repeated my message with some variations more than 20 times over a
> >>>> couple
> >>>> of
> >>>> days, but no luck at all. And the loss is definitely with that
> >>>> particular
> >>>> reader who may suddenly read a piece of information he or she did not
> >>>> know
> >>>> about. George
> >>>>
> >>>
> >>
> >
|