FM recording

[Myrtonos]

Given all the debate about (linear) analog versus digital recording, I thought I'd mention FM recording.
Let's first note that analog audio recording requires a bias, becasue of a magnetic propety called hysteresis. It seems that this bias is higher than the maximum frequency allowed by the head gap.
Also, when you playback a magnetic recording, the playback output increases with frequecy by 6db per octave. Normally in audio recording, an equaliser is used on playback to hold down the higher frequencies. This method (so it seems) works as long as the signal is (entirely) AC and the gain imbalance is no greater than the dynamic range of the medium.
Video signals (especially the luminance part) are fluctuating DC rather than AC, and thus span infinitely many octaves so the only way to record them is to modulate them onto a carrier and record the carrier. Frequency modulation is used becasue it permits recording a signal at constant amplitude, keeping the tape at saturation level, with no bias being necessary, and meaning that a simple limiter is sufficient for correcting gain imbalance and other amplitude variations. There also existed FM tape recorders for recording biological signals.
Yet FM was almost never used to record sound, the only example I know of is the HiFi modes on VHS and Beta videotapes. An audio format using wideband frequency modulation to record sound would have dynamic range and frequency response quite impressive for analog media.
I do wonder whether FM was seen as the future of audio recording as far back as 1956, when ampex introduced quadruplex, which, like all subsequent analog VTR formats used FM to record the luminance portion of the video signal.

 

[jpmorris]

FM does require a much higher frequency response, which generally means either a rotating head and all the complexity and reliability problems that brings, or a higher tape speed (a'la the RACAL 4DS) and lower running time.

AFAIK Laserdisc did this using some form of PWM on the pits and lands of the recording medium. But I don't know whether the audio signal was baseband or whether it was included in the FM envelope. (It wasn't renown for its audio quality either)

 

[witzendoz]

Back before we had DAT tape (showing my age) I bought one of the first VHS HiFi video recorders (panasonic or sony?) So I could record mixes to a format with no loss or added noise, i.e reel to reel or cassette. The recorder I had was a very good unit with inout volume control and level metering. The results were very good sound wise but you had to be careful that the recorder was tracking 100% or you got occasional noise, my machine had manual tracking so you could fix any problems.

This set up worked for me until DAT came along, I still have VHS masters stored, they still play back on my current VHS HiFi machine fine and sound good.

Cheers
Alan.

 

[Myrtonos]

Laserdiscs

Yes and Phillips did propose a dedicated audio verison that did use wideband FM, but this was rejected under Sony's influence in favour of the digital compact disc.

 

[witzendoz]

Yes and Phillips did propose a dedicated audio verison that did use wideband FM, but this was rejected under Sony's influence in favour of the digital compact disc.

Now I think of it my 1st hifi vhs machine was a phillips.

Alan.

 

[Myrtonos]

FM does require a much higher frequency response, which generally means either a rotating head and all the complexity and reliability problems that brings, or a higher tape speed (a'la the RACAL 4DS) and lower running time.

Your comments are valid for tape recording, but, suprisingly not for magnetic recording in general. I don't know about today but back in 1956, the highest practical tape speed was 30 inches per second, and the minimum practical head gap was 1/16000 of an inch. This gave a maximum frequency response of 250khz. Now onto magetic disc recording, while primarly associated with computer disc drives, it did exist in the pre-digital era. Telefunken marketed one once, but as far as I know this recorded sound in the ususal direct way, at a modest writing speed.
And when Valdemar Poulsen invented magnetic recording, he envisioned both cylinder and disc recording as well as tape and wire. In fact, his frist Telegraphone was a magnetic cylinder recorder.
With magnetic disc recording, higher writing speeds can be acheived more easily than with tape, same geos for cylinders, with the additional benefit that ratio of linear to angular velocity is constant.
The largest cyndiner records as far as I know where concert sized and spun at 160 rpm, I have done calculations and the writing speed was well above 30 ips, suc8h much that that a magnetic cylinder of the same diameter spun at the same speed could easily record FM sound.

 

[jpmorris]

Your comments are valid for tape recording, but, suprisingly not for magnetic recording in general. I don't know about today but back in 1956, the highest practical tape speed was 30 inches per second, and the minimum practical head gap was 1/16000 of an inch. This gave a maximum frequency response of 250khz. Now onto magetic disc recording, while primarly associated with computer disc drives, it did exist in the pre-digital era. Telefunken marketed one once, but as far as I know this recorded sound in the ususal direct way, at a modest writing speed.

Ampex made a videodisc recorder like that, the HS-100 and HS-200. It was used for things like instant replay in sports, and in 1970s Dr. Who when they wanted to run a segment of video backwards (e.g. 'Image of the Fendahl' and probably the reverse explosions in 'Pyramids of Mars', but I'm not sure). However, it only had about 36 seconds of video on a disk about 12-14 inches in diameter.
That said, it was running at 3000rpm or something - running at a tenth of that, you'd still only get about 3 minutes per side.
Disks will also have tracking problems unless you're using some kind of servo system which might be a bit much for 1960s technology, which I think is what we're positing, right?

The largest cyndiner records as far as I know where concert sized and spun at 160 rpm, I have done calculations and the writing speed was well above 30 ips, suc8h much that that a magnetic cylinder of the same diameter spun at the same speed could easily record FM sound.

Yeah, but again, they only lasted for about 2 minutes or so. Tape gives you a lot of surface area to record on and fewer tracking problems since the head itself remains fixed relative to the tape.

 

[Myrtonos]

Ampex made a videodisc recorder like that, the HS-100 and HS-200. It was used for things like instant replay in sports, and in 1970s Dr. Who when they wanted to run a segment of video backwards (e.g. 'Image of the Fendahl' and probably the reverse explosions in 'Pyramids of Mars', but I'm not sure). However, it only had about 36 seconds of video on a disk about 12-14 inches in diameter.
That said, it was running at 3000rpm or something - running at a tenth of that, you'd still only get about 3 minutes per side.
Disks will also have tracking problems unless you're using some kind of servo system which might be a bit much for 1960s technology, which I think is what we're positing, right?

What was the track pitch, and what magnetic materials were used?

Yeah, but again, they only lasted for about 2 minutes or so. Tape gives you a lot of surface area to record on and fewer tracking problems since the head itself remains fixed relative to the tape.

What was the groove density? I'm sure with modern magnetic marterials, the track pitch could be greater.

 

[jpmorris]

What was the track pitch, and what magnetic materials were used?

From 'videotape recording' by J F Robinson (1981, 3rd ed):

'the disk is 16 inches in diameter and uses a nickel cobalt recording surface with a rhodium coating.' It has 'a useful recording area on both sides of the disc 4.5" wide from the outer rim.
The number of tracks that can be accommodated on one disc surface depends on the track and guard band width. With a track width of 0.0075" and a guard band of 0.0025" the centre-to-centre track spacing is 0.01" allowing 450 tracks within the 4.5" of recording area. One field is recorded per track which means the total disc capacity, using both surfaces is 900 fields.
On 625/50:
Total time per disc = 900/50 = 18 seconds
On 525/60:
Total time per disc = 900/60 = 15 seconds

Most practical systems use one or two discs.
The tracks, unlike an audio record, are concentric rings and the heads are stepped in when the other head or heads are recording or playing back.'

 

[Myrtonos]

The tracks, unlike an audio record, are concentric rings and the heads are stepped in when the other head or heads are recording or playing back.'

This format only records pictures, not sound, an audio version would have a continuous spiral track. Also because it's video, the highest component frequency (in the baseband signal) is much higher than with audio, which, when combined with the need for a higher centre ferquency, needed a higher writing speed than with audio. That said, with a lower maximum component frequency, one could also have a wider modulation index, and could have a wider carrier deviation with the bandwidth still occupying 10 ocatves.

 

[jpmorris]

This format only records pictures, not sound, an audio version would have a continuous spiral track. Also because it's video, the highest component frequency (in the baseband signal) is much higher than with audio, which, when combined with the need for a higher centre ferquency, needed a higher writing speed than with audio. That said, with a lower maximum component frequency, one could also have a wider modulation index, and could have a wider carrier deviation with the bandwidth still occupying 10 ocatves.

You would probably have a wider usable area of the disk if the bandwidth was cut down. But yes, these things were built to address the fact that Quad couldn't do any kind of freeze-frame or slow motion replay.

 

[Myrtonos]

While many helically scanned formats, especially the Phillips V2000 could do freeze-frame and slow motion replay, tape simply isn't suitable for that, anymore that it's suitable for loops as short as the Sgt Pepper Inner groove.