Transferring Your Audio

Audio Restoration: Maryland, Virginia, West Virginia, Pennsylvania & Washington, DC

RECORDS

Audio transfer of records and their subsequent restoration is a rather inexact science.  Transferring vinyl records (33.3 and 45 r.p.m. discs) ought to be a straightforward procedure, and usually is.  Nevertheless, problems such as skips/hung grooves, off-center discs, warps and very dirty or otherwise damaged records, can show up from time to time.  These can slow down the transfer process considerably.

78 rpm records present additional challenges:

Groove size— the 78 r.p.m. format went through a lot of changes between the 1890s and the 1950s. The “78 needle” that can be purchased for use with many cartridges today is really only optimal for the very last (post-1955) 78s made. Earlier records typically have a wider groove that sounds best when played with a custom stylus. Often, records that seem almost unplayable with one size of stylus will perform marvelously with another. Any good transfer engineer should have many sizes and shapes of styli.

Playback speed— 78’s is what we call them, but in actual practice, the rotation speed can vary from 60 to 90 r.p.m.! Many early Decca recordings for example, were recorded at 80 r.p.m.; and if they’re played back at the 78 speed found on most modern turntables, the music is too slow and in the wrong key. Various Okeh recordings from 1926-29 were recorded between 76 and 84 r.p.m.; and even Victor, the “gold standard” record company of the day, used varying speeds.

Turnover frequency— When recordings became electric around 1925, record companies reduced the bass response when recording and boosted it upon playback, to avoid certain problems  that could be caused by bass-heavy grooves. The problem facing engineers today is that not all record companies adjusted the bass the same way. Typically, ‘turnover frequencies’ between 200 Hz and 800 Hz were used: with very different results. Using the wrong turnover (or even  worse, using the standard RIAA LP curve found on modern phono preamps,) can result in too  much or too little bass, and therefore, inaccurate reproduction of the recording.

Excess noise— Unlike the pristine vinyl that was used for modern records, 78s are made of a variety of different materials. Between ca. 1900 and ca. 1945, ‘shellac’ was a common medium, though the quality of the shellac varied widely. Modern reproduction quickly exposes the quality of the material used in a pressing: bad shellac creates loud steady-state noise and crackle. Another common product is the ‘laminated’ record, which uses a smooth, clean sheet of high-quality material for the record surface; but is prone to  rumble when played back on modern equipment. In addition to the original composition of the disc, we have to look at how it has been treated over the years. Early recordings played on Victrola-style phonographs with heavy steel needles tended to wear out very quickly the more they were played; but any well-played (well-loved) record will have audible signs of use.

Odd Formats— Many ‘home recordings’ exist which are made from either aluminum or glass discs coated with acetate. Others are made from laminated cardboard or even uncoated, solid aluminum. These recordings present unique problems of their own, and must be examined case-by-case.

MAGNETIC TAPE

Reel-to-reel (also known as open-reel) tapes and audio cassettes often contain one-of-a-kind recordings that degrade over time, even just sitting on a shelf. It is important to transfer these to a safer, more durable digital format for future preservation. Tapes can break and require splicing. They can also ‘shed’ oxide upon playback; which is why certain tapes should be heat-treated (baked) before they are played. Head azimuth usually varies if the recording and playback decks are different; a problem that is particularly noticeable on audio cassettes (it’s that ‘wow’ or ‘swoosh’ sound that gets in the way of the music.) Cassette shells sometimes become warped, requiring replacement before the tape can be played at all.