Loading ...
Sorry, an error occurred while loading the content.

Re: Converting piano rolls to notation

Expand Messages
  • Ron ODell
    ... transcribing notes, not time values....to get the time values down, you d have to determine the length of each perforation and assign that a time value...
    Message 1 of 7 , Jun 30, 2008
    • 0 Attachment
      --- In EliteSyncopations@yahoogroups.com, Bryan Cather <catt967@...>
      wrote:
      >
      > The problem with doing that, however, would be that you'd only be
      transcribing notes, not time values....to get the time values down,
      you'd have to determine the length of each perforation and assign
      that a time value...

      That isn't so difficult. You can see the rhythm and thus know
      how long each beat is, and easily work out what duration to give
      to a shorter hole.

      I hand-transcribed the first few bars of a piano roll to MIDI once.
      Figuring out the durations was no problem, as I recall.
      (The MIDI file I made is:
      http://www.armory.com/~keeper/1/work/hmonkeyt.mid .)

      The only advantage to doing it by hand is that you can adjust
      volume levels based on the lines printed on many 88-note rolls.
      But it would take a heck of a long time to do an entire roll,
      and messy without a spool.
    • Rob
      Yes, it IS very tedious. As for the timing problem, since I only did the rolls I really like enough to invest the time, I knew the music well enough and had
      Message 2 of 7 , Jul 1, 2008
      • 0 Attachment
        Yes, it IS very tedious. As for the timing problem, since I only
        did the rolls I really like enough to invest the time, I knew the
        music well enough and had the timing in my head. So that part was
        easy.
        Bb




        --- In EliteSyncopations@yahoogroups.com, Bryan Cather <catt967@...>
        wrote:
        >
        > The problem with doing that, however, would be that you'd only be
        transcribing notes, not time values....to get the time values down,
        you'd have to determine the length of each perforation and assign
        that a time value...
        >
        > Its still impressive, though....and sounds profoundly tedious!
        >
        > BryanC
        >
        > Rob <yosemitewriter27@...> wrote: Once
        upon a time, way back in the 60's, and before I even had a
        > player piano, I was collecting rolls and wanting to do
        > transcriptions on a few of them. So I got an 88-note tracker
        bar,
        > put a piece of masking tape just above the holes including all 88
        > notes, but so you could still see the holes below the masking
        tape.
        > Then I wrote above each hole, what note it played on the
        keybaord.
        > Then when the paper roll went over the tracker bar, I could read
        > thru the holes what note I had written on the masking tape and
        wrote
        > it down on ledger paper. (Not Heath Ledger, but the music-
        writing
        > kind).
        > Then I built a box in which I could mount the tracker bar and
        > have an uptake spool for the paper as I rolled it out. So,
        > essentially it looked like the unit in a regular player piano and
        > allowed me to trascribe rolls. I still have this contraption in
        my
        > attic.
        > See....Adam isn't the only child genius in the group :-))
        > Bb
        >
        > --- In EliteSyncopations@yahoogroups.com, philharpn@ wrote:
        > >
        > > I've got quite a few QRS piano rolls for my 1978 Aeolian player
        > piano.
        > >
        > > I have in my archives somewhere a kit (which I haven't
        recovered
        > yet) that
        > > had a clear plastic piano keyboard template that fit over the
        > piano roll to tell
        > > you what notes the holes represented. Doable but would take
        > forever.
        > >
        > > I also could slow the rolls w---a--y down and write down each
        key
        > > depressed... Likewise forever,
        > >
        > > But is there an easy way to convert the rolls to piano notation
        > for my own
        > > personal use?
        > >
        > > On a related subject, has anybody come up with a portable
        device
        > that "plays"
        > > piano rolls?
        > > (other than record the piano while playing to CD)
        > >
        > > Phil Lloyd
        > >
        > >
        > >
        > >
        > >
        > > Phil Lloyd
        > >
        > >
        > >
        > >
        > > **************
        > > Gas prices getting you down? Search AOL Autos for
        > > fuel-efficient used cars.
        > > (http://autos.aol.com/used?ncid=aolaut00050000000007)
        > >
        >
        >
        >
        >
        >
        >
        >
        >
        > test'; ">
        >
      • rag1916
        Yes, it s tedious but important. To transcribe, get a MIDI file of the roll made from a high-quality scan of it. First, check websites such as the IAMMP
        Message 3 of 7 , Jul 10, 2008
        • 0 Attachment
          Yes, it's tedious but important.

          To transcribe, get a MIDI file of the roll made from a high-quality
          scan of it.

          First, check websites such as the IAMMP archives, Terry Smythe's
          "Player Piano Rebirth", Warren Trachtman's piano roll scans, and
          Robert Perry's Pianola site to see if the roll you want to transcribe
          has already been scanned.

          http://www.iammp.org/rolldatabase.php

          http://members.shaw.ca/smythe/rebirth.htm

          http://members.shaw.ca/smythe/archive.htm

          http://www.trachtman.org/rollscans/index.htm

          http://www.pianola.co.nz/index.asp

          (The stuff that is greyed out on IAMMP and Warren Trachtman's site
          means it is copyright and not easily downloadable... just send them an
          email requesting that particular MIDI file and they will email it to
          you personally).

          Check the rollscanners group to see who has a scanner in your area,
          you might need to join or at least email someone who can put in a
          query to the group.

          http://launch.groups.yahoo.com/group/Rollscanners/

          Not all roll scanners are perfectly accurate! The
          twin-array scanner developed by Spencer Chase is one of the best, as a
          re a few others, but some roll scanners are not perfectly accurate and
          the resulting scans are not suitable for either reconstructing the
          orignal master roll to make recuts, or making a decent note-accurate
          transcription (the rhythms will lurch all over the place).

          Anyway, once you have a MIDI file of the roll you want to transcribe,
          get Cakewalk (preferably an earlier version like 5 or 6 which I think
          will still run on XP), and open it in their "piano roll editor". You
          will then want to listen to it several times through and click to
          define where beat one of each measure is (so don't start with Dave
          Brubeck!)... and then use Cakewalk's "Fit Improvisation" feature to
          snap the notes to where you have individually clicked is the beginning
          of each measure (yes, you have to do a separate click for the
          beginning of EVERY SINGLE BAR! It's tedious but totally WORTH IT!).

          http://www.mmdigest.com/Archives/Digests/200605/2006.05.04.08.html

          http://www.mmdigest.com/Archives/Digests/200705/2007.05.14.03.html

          Then you can quantize the notes within each of the defined bars to
          8ths, 16ths, triplets, whatever, etc. to make it easier to edit in the
          notation view. This is suitable for rolls that are already quantized
          into some kind of mathematical rhythm, such as all straight-8ths, all
          swung-eighths, where the swing is often a division of 5 (3:2), 3
          (2:1), or 11 (6:5).

          For the really neat performances, such as early Charley Straight and
          Max Kortlander rolls, and many real hand-played rolls such as QRS
          Autograph (100,000 series), Aeolian Uni-Record, Vocalstyle, etc. you
          might want to try a different approach and just define the beginnings
          of the bars, and then DON'T quantize within them but rather edit the
          notation itself in the sheet music view to something more readable,
          since it's inevitable that such performances will have various types
          of swing and rhythmic variety that makes them sparkle in a true
          hand-played manner.

          You'll notice that most people who do piano roll transcriptions do not
          bother to write down these variety of rhythms, but simplify it all
          into one kind of notation, such as dotted eighth - sixteenth. Thus,
          when a person sits down and plays one of these transcriptions, it
          sounds kind of flat, while the actual roll sparkles. So, try to do
          better than they have!

          ---

          If you take a look in Bob Pinsker's book "Piano Roll Sampler" you will
          see that these particular rolls had relatively few extra notes added
          by the editors (which, by the way, he has printed in grey instead of
          black ink to differentiate them from the presumed "playable" notes of
          the original performance). By "these particular rolls" I mean hot
          blues, ragtime, and jazz performances by well-known pianists.

          Of course, if you transcribe an 8-hand march or rag roll, such as one
          of the arranged ones on U.S., Aeolian ("Eighty-Eight Note"),
          Connorized, etc. you certainly get more notes than you could ever play.

          The same with "hand-played" duet rolls, such as the excellent Arden
          and Kortlander stuff on QRS. For those, you make four staves and
          transcribe it as you would a piano duo, 4 hands, one piano (which is
          who will be playing the transcription, anyway). I have seen very few
          "hand-played" rolls of any description that have more notes than can
          be handled by a duo of two very good pianists sitting at one piano.
          (and I mean VERY GOOD pianists)

          Be forewarned... there are some "hand-played" rolls listed as being
          played by one pianist that have a LOT of notes... and if transcribed
          would best be tackled by 2 or 3 pianists. A good example are the Zez
          Confrey rolls for QRS word roll series, and also Ampico. These are
          generally phenomenal drafting-board arrangements (such as all the
          faux-oriental songs he did in circa 1919-1921), except for the true
          hand-played Confrey performances, which are his instrumental rolls on
          the QRS Autograph series (such as "My Pet" and "Kitten on the Keys").

          Another example are the "Alvin Gardner" rolls on Arto (or other labels
          made by the Standard Music Roll Co. of Orange, NJ). I don't know who
          Mr. Gardner was, but there is no way he could have played all those
          notes! Probably drafting-board arrangements as well, though in my
          opinion his stuff is no-where near as good as Confrey's.

          Check out how Artis Wodehouse writes out Confrey's Ampico roll of "By
          the Waters of Minnetonka":

          http://depthome.brooklyn.cuny.edu/isam/artis.html

          Though I disagree with her about the ending of this roll being
          "impossible" for two performers to execute (I can already play some
          things that are supposedly only possible for two people), and also
          with her notation (she uses dotted rhythms as a short-hand for
          different kinds of swing; something I don't like about most modern-day
          transcriptions or scores from the era), she does give you a basis for
          getting started.

          Many people have done piano roll transcriptions. I recommend talking
          to Bob Pinsker, Vincent Johnson, and Benjamin Intartaglia (all of whom
          are on this group, I think) for tips.

          One trap that's easy to fall into: don't simplify the left-hand
          oom-pah rhythms into simple quarter notes. I don't even write my new
          rags that way. Instead, in order to get a decent-sounding performance,
          note the differences between bass notes and chords played as quarter
          (long), eighth (medium) and sixteenth (short) notes, the usage of
          which can give the performance a different feel, and also any
          held-over tenor-line notes which should be written as a separate voice
          on the bass staff (they are often playable; you sustain them with the
          sostenuto pedal on a grand piano).

          -Andrew


          --- In EliteSyncopations@yahoogroups.com, "Rob" <yosemitewriter27@...>
          wrote:
          >
          > Yes, it IS very tedious. As for the timing problem, since I only
          > did the rolls I really like enough to invest the time, I knew the
          > music well enough and had the timing in my head. So that part was
          > easy.
          > Bb
          >
          >
          >
          >
          > --- In EliteSyncopations@yahoogroups.com, Bryan Cather <catt967@>
          > wrote:
          > >
          > > The problem with doing that, however, would be that you'd only be
          > transcribing notes, not time values....to get the time values down,
          > you'd have to determine the length of each perforation and assign
          > that a time value...
          > >
          > > Its still impressive, though....and sounds profoundly tedious!
          > >
          > > BryanC
          > >
          > > Rob <yosemitewriter27@> wrote: Once
          > upon a time, way back in the 60's, and before I even had a
          > > player piano, I was collecting rolls and wanting to do
          > > transcriptions on a few of them. So I got an 88-note tracker
          > bar,
          > > put a piece of masking tape just above the holes including all 88
          > > notes, but so you could still see the holes below the masking
          > tape.
          > > Then I wrote above each hole, what note it played on the
          > keybaord.
          > > Then when the paper roll went over the tracker bar, I could read
          > > thru the holes what note I had written on the masking tape and
          > wrote
          > > it down on ledger paper. (Not Heath Ledger, but the music-
          > writing
          > > kind).
          > > Then I built a box in which I could mount the tracker bar and
          > > have an uptake spool for the paper as I rolled it out. So,
          > > essentially it looked like the unit in a regular player piano and
          > > allowed me to trascribe rolls. I still have this contraption in
          > my
          > > attic.
          > > See....Adam isn't the only child genius in the group :-))
          > > Bb
          > >
          > > --- In EliteSyncopations@yahoogroups.com, philharpn@ wrote:
          > > >
          > > > I've got quite a few QRS piano rolls for my 1978 Aeolian player
          > > piano.
          > > >
          > > > I have in my archives somewhere a kit (which I haven't
          > recovered
          > > yet) that
          > > > had a clear plastic piano keyboard template that fit over the
          > > piano roll to tell
          > > > you what notes the holes represented. Doable but would take
          > > forever.
          > > >
          > > > I also could slow the rolls w---a--y down and write down each
          > key
          > > > depressed... Likewise forever,
          > > >
          > > > But is there an easy way to convert the rolls to piano notation
          > > for my own
          > > > personal use?
          > > >
          > > > On a related subject, has anybody come up with a portable
          > device
          > > that "plays"
          > > > piano rolls?
          > > > (other than record the piano while playing to CD)
          > > >
          > > > Phil Lloyd
          > > >
          > > >
          > > >
          > > >
          > > >
          > > > Phil Lloyd
          > > >
          > > >
          > > >
          > > >
          > > > **************
          > > > Gas prices getting you down? Search AOL Autos for
          > > > fuel-efficient used cars.
          > > > (http://autos.aol.com/used?ncid=aolaut00050000000007)
          > > >
          > >
          > >
          > >
          > >
          > >
          > >
          > >
          > >
          > > test'; ">
          > >
          >
        Your message has been successfully submitted and would be delivered to recipients shortly.