Please see article below, which supports this. I am not the author and the comments following the article are not mine.
Lot's of varying opinions on this, even in instruction manuals.
Myself, I clean the heads and pinch roller on my RT-707 and sit back and enjoy the music.
I have more important things to worry about....
- Operating a tape recorder does not cause anything in the tape path to become magnetized. Not even a tiny bit.
2. All parts including heads have made from materials with a natural neutral magnetic state, to which they tend to return even if deliberately magnetized.
3. The magnetic flux required to affect a recorded tape is much higher than anything anywhere around the tape recorder, except in the erase and record head gaps. The effect of a steady-state magnetic field (a DC field) on recorded tape is first noticed as partial erasure of high frequencies, and that occurred with a field somewhere above 200 Gauss. Tape has very high coercivity by design, meaning you need a high field strength to cross the threshold above which the tape will become magnetized. That's what the bias oscillator does. By being many times hotter than the actual audio signal on the head, the bias forces the signal over the coercivity threshold into the more "linear" range of the tape. But that means it's harder to erase tape than you might think. Typical guide and head residuals were in the zero to 20 gauss range, far below tape coercivity levels.
4. Erase and record heads are self-demagnetized by the bias oscillator, which provides an AC field in the gap that is thousands of times higher than what a de-magnetizer can induce.
5. The field found on recorded tape is incapable of magnetizing anything. It is minuscule, and when the tape moves, it becomes an AC field, which would demagnetize if it had any effect.
6. The effect of a DC field near or in a record head produces two measurable results: First, a significant elevation of even-order harmonic distortion, and second, an elevation in low frequency noise (a sort of gravelly sounding noise). But again, the field has to be fairly high for this to happen. You need a spectrum analyzer to differentiate even harmonics from the normal odd harmonics created in the recording process.
7. If tape machine parts become magnetized, there is a defect in manufacture (the material doesn't have a natural, neutral magnetic state), or a defect in the tape machine, either due to a failure or design flaw. Asymmetrical bias waveforms will present a DC component, any DC on a head due to a circuit defect (bad blocking cap, for example). Demagnetizing may temporarily remove the residual, but if it creeps back, something is "broken", and needs repair or replacement.
8. Lastly, casual use of a demagnetizing device, even if the classic proper techniques are observed (slow removal, switching of when several feet away), sometimes resulted in an increased magnetic state rather than a decrease. There's no way to tell if you are demagnetizing, or magnetizing, without further testing for the effects of magnetism. Once again, these are HF erasure of a recorded tape, or increase in noise and even-order harmonic distortion during record. Yes, you need instrumentation to do this.
I don't know who wrote that long quote above, but most of it is pretty accurate.
I hardly EVER demag....maybe once a year(?)...and I only do that just for the heck of it, just so's I can say I've done it once in awhile ...just in case.
See...I have a Gauss meter, and no matter whenever I'd check my deck with it, the magnetism was absolutely nominal. IOW, I could demag it a thousand times, and the teeny-tiny bit of needle deflection I would see on my Gauss meter would always be there. It's partly just the metal mass of the whole deck and partly the background magnetism of good ol' mother earth.
So...I never give it much thought any more, Once in a blue moon I'll demag, but it's probably not needed.
When I had my first 4-track, I use to demag ALL THE TIME....but not now with my current 16 and 2 track decks.