The Talking Machine Forum — For All Antique Phonographs & Recordings

My friend restored this latest disc of mine, a high fidelity disc recorded by Victor in Hollywood. It sounds like it could have been done yesterday and yet it was 1937!

The explanation is in the video - does anyone else own these hi-fi discs?

[youtube]http://www.youtube.com/watch?v=UW3qJo65kJg[/youtube]

Victor was doing some impressive truer hi-fi things closer to the early thirties. This record is not hi-fi. Looking at the audio on a spectrum analyzer shows the audio to drop off exactly at 10,000 cycles, somewhat short of the range of 15,000 to 20,000 cycles needed for true hi-fi in the treble end. The video caption claims 18,000 cycles.

If you want that, I'd say the Decca FFRR 78's from the later 1940's are the most obvious examples, but there are plenty more.

If we call hi-fi the range of 30 cycles to 15,000 cycles and beyond, the bass range had been achieved by the late 1920's, but the treble range took until the late 1940's to reach on regular commercial records unless someone has some earlier examples otherwise.

By the later 30's Victor had reverted to the deader closeted kind of sound you hear on that record, though it still sounds pretty good, relatively speaking, I'm not complaining. Good music too.

The problem, of course, is not the frequency response limitation in and of itself, as far as reproduction of instrumental and vocal music is concerned, because no instrument or voice produces fundamental pitches in the 10k Hz range. Where the limitation is "heard" is in the timbre of the instruments and voice, which depends on the presence (or absence!) of the upper partials of the harmonic series in the tones, and their relative strengths. The particular and unique mix of these upper partials for each instrument/voice is what enables the listener to distinguish one instrument from another (say, flute and trumpet), or one singer from another (say, Pavarotti and Domingo). The frequency response limitations of acoustic recordings account, in large measure, for the unsatisfying reproduction of the female soprano voice (with rare, qualified exceptions, e.g., Amelita Galli-Curci), because the full richness of tone cannot be captured due to the cut-off of the upper partials. Only the advent of electrical recording and playback, with their greatly extended frequency range ("high fidelity"), allowed artists and ensembles to be heard in something approaching what the ear hears in the concert hall.

I was, a few months ago, making a transfer of Fritz Kreisler's Liebesfreud which he had recorded on a late 1920's Victor 78.

For reference I pulled a CD of the same tune that he recorded in 1938, to help ascertain the pitch. It was really striking how much the earlier Victor recording lost on the violin overtones, and yet I didn't feel like I had been missing too much, musically. I preferred the earlier recording, most of what was needed to prefer the musical performance in my opinion.

It was just the inital juxtaposition of the recordings that was jarring, but the ear adjusts.

I always wonder about those people who heard Caruso and said that his records didn't give you a fraction of his sound. He sounds great to me.

Wolfe wrote:Victor was doing some impressive truer hi-fi things closer to the early thirties. This record is not hi-fi. Looking at the audio on a spectrum analyzer shows the audio to drop off exactly at 10,000 cycles, somewhat short of the range of 15,000 to 20,000 cycles needed for true hi-fi in the treble end. The video caption claims 18,000 cycles.

Hmm, how odd. All I went on for that is what my friend said - I haven't the technology to do that myself. Perhaps youtube does something to it?

I use to have a London FFRR 78 from mid-50s featuring some Operatic excerpts by Mario del Monaco, Renata Tebaldi and many others - and they sounded fantastic. Also, I was lucky enough to hear some bits from the legendary 1931-1932 Bell Labs Stereo High Fidelity recordings in their original form; and they sound absolutely fantastic, too!

Of course, any 78s from Major labels of the 1950s sounds absolutely equal to Hi-Fi LPs and even much superior on their brightness of sound compared to later "reissues/remastered editions". When playing these records on acoustic machines, one can even hear some sort of "reverb" effect since the acoustic reproducers couldn't cope with all those high frequencies...even some of muddy-sounding Sun Records 78s sounds like that.

I think that YouTube does compress audio a bit, doesn't it?

Post the original transfer to box.net and let's see!

Oh -- NICE RECORD!

Beaumonde, the link from box.net can be found here. I had to reduce the file to mono to save space...but we're not missing much there anyway with stereo

http://www.box.net/s/xkligcy3b3mmodxkfj1n

YouTube user Prozoot has some impressive things on his channel. There are some early 1930's Victor sides in particular that approach true hi-fi.

http://www.youtube.com/playlist?list=PL49D23FA4225F3057

If one has 78's with a lot of extended treble content, I think it's important to avoid playing them on old machines with heavy soundboxes / electric pickups as the delicate modulations of the groove wall will be worn away. I've heard theories that it's why Victor started recording with less range after their initial attempts in the early 1930's. Heavy pickups on home phonographs and jukeboxes wore the grooves out too quickly.