Print

Print


FYI, with apologies for the use of bandwidth in quoting the entire original 
post.

Mark Durenberger


----- Original Message ----- 
From: "Art Shifrin"

Hi Mark,

I had a good chortle over (the posting below).

Pete Hammar, Don Ososke (of the standard tape lab) and I more than once
discussed constructing a superior wire playback device in the EARLY
1980s, when we were all working for Ampex.  It was Don's idea that the
ideal machine to morph for this purpose was the ATR 100, because of its
comprehensive servo design and lack of pinch roller.  My on-going
experiments after diving out of VidiPax proved that the less costly and
more expedient use of a capstan / pinch roller combination was
meaningfully inferior to the capstan-only design of the 1946 Brush BK401
and Ampex ATR 100.  That's what my design evolved into: freely borrowing
from those prior technologies.  The machine's shown on my website, which
is www.shifrin.net

The citation of a fishing reel is intriguing, because I emulated the KEY
idea of the wire stroker from the magnificent Magnecord SD-1: that it
should be independent of the head or heads, which should be stationary.
Each time that J.L. brought up that fishy analogy, I reminded him of the
source of that inspiration.  I had to repeatedly remind him about other
comparable issues (...).

They included the fact that I had long before first seen the SD-1 (which
had been on loan from Hal Layer to the Ampex Museum of Magnetic
recording) when, as a result of being introduced by Ampex V.P. Ridley
Rind (sic?) to Pete Hammar, its Founder and Curator.   In my spare and
some Ampex-paid time (while selling Ampex pro audio in the N.E. Region)
I was providing my services to and assisting Pete: some ten years before
I (had joined) VidiPax.  Note that prior to (that), J.L. (or Vidipax...I 
don't
know which entity actually made the purchase) had acquired that SD-1 from 
Hal.

The idea for playing old magnetic media for the advantages of improved
mechanical and electronic methods might have at some time original to
J.L., but it had been conceptualized and done in the late 1920s when
Telegraphone wires were played back not through telephone handsets and
crystal headphones, but electronically amplified loudspeakers.  Another
much more modern example was Ampex's fulfillment of Disney's request
that ATR 100 playback heads be made for the then years-obsolete half
inch 3 track tape format.

BTW the other morning I happened to stumble upon the delightful fact
that there are at least two shots of an SD-1 running (apparently at its
higher speed of 48 ips) in the 1948 feature film "Walk A Crooked Mile".
It was run on TCM. Happily, about three hours later in California, the
same revelation coincidentally occurred to Hal Layer.

You have my permission to post and provide this reply to anyone and or
any forum.

Best Regards,
Shiffy

> ----- Original Message ----- 
> From: "Jim Lindner" <[log in to unmask]>
>> Ah..... summer in North America.... short replies take a while - long 
>> replies take a long time -
>> this is a long reply but I think the time  has come to at least partially 
>> document efforts made
>> by myself and Art Shifrin at my old company VidiPax more then 10 years 
>> ago. The lessons learned
>> will likely be helpful to many on this list now, and perhaps  more 
>> importantly to those who
>> ponder these type of issues in general  in the future - so I am counting 
>> on the Internet to index
>> this somehow  so someone who may have a need in the future can find it.
>>
>> Based on a great deal of experience over time, I developed the theory 
>> that magnetic recorders
>> (and frequently other media types as well)  were in fact capable of far 
>> higher quality recordings
>> then they were  able to play back on the same machine. I slowly developed 
>> this theory  during a
>> period of about 5 years when I first started to do format  migration work 
>> and was dealing with
>> some very difficult issues in the  video area - specifically relating to 
>> signal timing and
>> playback which  is of course critical in video. I noticed that while the 
>> machines  could always
>> record a signal - whether the machine could play back  that signal (or 
>> another signal due to
>> interchange issues) was an  entirely different matter. This theory has 
>> proven to be correct over
>> the years in many instances. Even if a machine was out of alignment it 
>> WOULD record in an out of
>> alignment way - the trick then was to figure  out how to play back that 
>> signal - and it was far
>> more difficult in  many instances to play back that signal with the 
>> electronics of that  era.
>>
>> When one thinks about that  now - it may be no great epiphany - but it 
>> was to me at that time
>> because it made me realize that in fact  virtually millions of recordings 
>> have NEVER been heard
>> or seen with  the same fidelity as had been recorded, and this occurred 
>> to me with a  sort of
>> overwhelming realization at that time. That meant that one  COULD develop 
>> ways to play back
>> recordings with modern electronics  (modern by definition does not 
>> necessarily mean today "state
>> fo the  art" but could mean more modern relative to the time the 
>> technology of  the recording was
>> made - and sometimes in a "golden era" of  electronics when certain gear 
>> that did particularly
>> well with certain  signal types was available).
>>
>> When testing this theory with video, I found that it was sometimes  true 
>> and sometimes not
>> depending on a number of factors - but  primarily one factor specifically 
>> in video that the
>> electronics for a  period got more tolerant (and therefore able to indeed 
>> play back  better then
>> the original recorder could) and then less tolerant - as  state of the 
>> art came to mean that all
>> electronic timings were perfect  from first recording and so the playback 
>> electronics really did
>> not  have to be very tolerant at all (in the analog domain) because all 
>> of  the recorded signals
>> were in good shape (if nothing went wrong). So  when a tape that was in 
>> crapola shape came for
>> playback 30 years  later, playback was exceedingly difficult because the 
>> electronics of  that
>> vintage expected everything to be perfect, there was no tolerance  to 
>> speak of, and so it was a
>> nightmare project to get a decent  playback. This is still the case with 
>> it being far more
>> difficult to  play back a severely damaged digital recording then an 
>> analog one.
>>
>> During this time Art Shifrin came to work for me at my company at that 
>> time - VidiPax. We were at
>> the time business colleagues and friends  (and that unfortunately is no 
>> longer the case). At that
>> time we  discussed this theory I had - which made sense to him. The 
>> business at  that time
>> involved playing back obsolete formats and Wire Recording  playback was a 
>> format we supported. We
>> used period wire recorders, but  you did not need to be too good of an 
>> audio engineer to see that
>> the  playback electronics on those machines (either professional or 
>> consumer) was pretty awful.
>> The transports were largely Rube Goldberg  affairs if you look inside of 
>> one - and we both
>> figured there had to  be a better way. We decided to find out if my 
>> theory was true for Wire
>> Recordings - even considering their age and low coercivity - and  decided 
>> to build ourselves a
>> test bed device to see what might be  possible.  Now remember that this 
>> was for internal purposes
>> only for  our playback business - there was never any intention to 
>> actually sell  one of these.
>>
>> In earlier days Art worked for Ampex as a repair technician and knew  the 
>> insides of the reel to
>> reel decks cold. The Ampex decks were also  in plentiful supply 
>> essentially for the asking, and
>> both of us had  very high regard for the quality of the Ampex playback 
>> electronics  having a
>> clean sound that was analog and appropriate for the playback  of the 
>> wires - at least
>> conceptually. There were many other questions  however - such as 
>> heads.... Would standard
>> playback heads playback the  low level recordings from the wire? We did 
>> not know, but I reasoned
>> that the worse case would be that the levels would be low and that the 
>> S/N would then be in the
>> dumps - but we could deal with that through  preamplification tweaking or 
>> through getting a
>> custom head stack and  preamp made if necessary (we of course had wire 
>> recorder heads and  there
>> are several expert head rebuilders around and could easily  commission a 
>> new head with different
>> characteristics and  electronicsbased on the output level from the wire). 
>> We reasoned that  worst
>> case it would be no lower then the output from a turntable  cartridge and 
>> we could easily deal
>> with that if we needed to. We found  to our delight that we did not have 
>> to go to these extremes,
>> but that  a standard full track head worked just fine with the standard 
>> electronics with just a
>> few tweaks, provided that we could hold the  wire in place (the moving of 
>> the wire across the
>> head made all sorts  of sonic problems as you might imagine). Art devised 
>> a very simple  design
>> whereby he glued tiny wedges directly to the head to hold the  wire in 
>> precisely the same
>> position. Think of tiny triangles placed  against one another with the 
>> bare head inbetween. The
>> wire was  "encouraged" to stay in place by the tension place upon it - 
>> and it  rode in the groove
>> between the two triangle wedges. As you might  imagine the wire had 
>> higher friction then a tape
>> would, so we wore the  heads down a bit faster - but we had plenty of 
>> cheap Ampex full track
>> heads, plenty of super glue, and plenty of plastic wedges so it was  not 
>> much of a problem.
>>
>> We also decided immediately that the transport system should be  capstan 
>> driven. Provided that
>> the wire could be maintained at proper  tension through the entire 
>> playback path we believed that
>> we could get  a much more consistent playback then period machines were 
>> capable of.  This was
>> found to be true. Art did a great job on the transport system  spending a 
>> huge amount of hours
>> and eventually we enlisted the  services of a friend of his that had a 
>> small machine shop, and
>> built  some components to essentially do 2 things - one was to allow the 
>> capstan to move the
>> wire consistently at speed, and the other was to  move the wire at the 
>> take up position so that
>> it was not deposited in  one place. To do that we essentially took the 
>> idea from a deep sea
>> fishing reel and had a bobbin type assembly that moved the wire  forward 
>> and back as the reel
>> turned automatically. This was a tricky  bit but with some 
>> experimentation was shown to be more
>> reliable in  terms of constant speed over the head then moving the reel 
>> up and down  (as is done
>> in some commercial wire recorders).
>>
>> Bottom line - it worked and the results truly amazed both of us. The 
>> theory was more then correct
>> and the results really were amazing - we  had expected better but what we 
>> got was so far better
>> that we truly  were astounded. There was no more sound that sounded like 
>> you were  listening
>> through a cereal box - the sound was almost always clear and  with really 
>> decent frequency
>> response.
>>
>> Art left the company and I believe continued work on his own on the 
>> device. Art deserved Kudos
>> for doing the work that he did, and continuing on his own. I have always 
>> referred this type of
>> work when  it showed up to him in ensuing years whether he realized it or 
>> not.  There are no
>> "plans" and there were no other machines that were built.  I am 
>> absolutely positively convinced
>> that if you want to really hear  what a wire recording has recorded on 
>> it - this approach is the
>> one to  be followed. It will take some time and money to do - but you now 
>> have  the advance
>> knowledge that we did not have - that in fact it DOES work  and is worth 
>> the effort. Try it - you
>> too will be amazed. The recipe -  several used Ampex decks - used but not 
>> abused. Access to a
>> machine  shop to make a few parts. A good understanding of electronics 
>> and the  schematics to
>> make a few tweaks that you will find you need, good  mechanical ability, 
>> lots of coffee, and a
>> great deal of time to fiddle  with it. It will be worth your while.
>>
>> Jim Lindner