With cassette decks and recorders
you can read things like "Dolby B" or "C" or even "S". Sometimes, "HX PRO"
is also mentioned. And then: dB, equalization,
bias. . .
What does this all mean???
1. What's Dolby
B,C and S?
What they are good for
A bit of history
The beginning: Dolby B
What happened next: Dolby C
The newest thing: Dolby S
What they gain
at -30 dB
at -40 dB
2. What is tape saturation?
3. Overview of the different types of cassettes
4. What are dB?
5. What does Equalization mean here?
Overview of the eqalization time constants
How to calculate with time constants
6. What is Dolby HX PRO?
7. What is the head gap and what is azimuth?
8. What is bias? - this was too complicated to translate!
Sorry, you'll have to refer to my page in German
9. My sources
Dolby B, C und S:
Dolby B, C and S are so-called noise reductions systems. Although this is not quite correct, this is what they seem to do. The right term is compander.
How they work (in principle):
During recording, quiet tones are boosted (with B and C only in a part of the frequrncy spectrum) and cut back during playback.
It's not that easy because a level-dependant gain must be >realisiert<. Fiortunately, I needn't explain this here. If you want to know exactly how it works, have a look at the website of Dolby, where an exact explanation can be found.
Why they are there:
Everyone who has already listened to recordings without Dolby on a "normal" cassette deck with headphones: cassettes are quite noisy. The reason for this is they relatively small dynamic range, the "distance" from the highest volume without terrible distortion and the noise floor (which is caused by the particles on the tape itself): mostly a bit above 50 dB (corresponding to ~316:1) (for comparison: the dynamic range of a CD is about 96 dB [although this is a "digital" dynamic range, and for a direct comparison you have to subtract 12dB to 22dB] ). In the bass and treble range, this is even worse.
|Type of cassette ("normal"name-Type-IEC-chemical)||Einführung||dynamic range||equalization|
|Normal - Type I - IEC I - ferric oxide (Fe2O3)||1963 with compact-cassette||~50 dB||3180 + 120 µs|
|Chrome - Type II - IEC II - chromium dioxide (CrO2)||some 1970||~53 dB||3180 + 70 µs|
|Ferrichrome - Type III - IEC III - ? (FeCr) (died out)||only available in early 1970s||~52 dB?||3180 + 70 µs|
|(Pure) Metal - Type IV - IEC IV- iron (Fe)||some 1979||~56 dB||3180 + 70 µs|
As you can see, the dynamic range
of cassettes is sufficient for the (quite old) DIN definition of HiFi(40-16000Hz,
50dB[?], wow & flutter 2% [!!!!!!!]). But today, this is not sufficient
any more, considering that wow & flutter cassette decks should not
exceed 0.2 %, and the minimum signal/noise ratio should be 60dB.
In 1968, Dolby Corp. developed their first noise reduction system for "slow-speed consumer tape formats". The result was called Dolby B and offered a maximum of 10 dB noise reduction. Their first NR system, Dolby A, was used by professionals. Recordings encoded with Dolby B can easily be exchanged, as it's little sensitive to tape deck parameters.
In 1980, Dolby C was presented, which was specially designed for cassette decks. It offers 20dB of noise reduction and anti-saturation and spectral skewing, which are used to avoid tape saturation. How anti-saturation works: High tones are cut back during recording and boosted during playback, as tapes saturate much more easily in the highs.
Dolby C is not suited for the exchange of recordings, as it is VERY sensitive to tape deck parameters (azimuth, frequency response etc.).
In 1990, Dolby S was presented. All decks supporting Dolby S must meet higher performance standards, for example
"- extended high frequency response
- tighter overall response tolerances
- reduced wow and flutter
- a new head azimuth alignment standard"
Dolby S provides up to 24 dB of noise reduction, and up to 10 dB in the bass range. With Dolby S, anti-saturation in the bass range has also been included
Recordings encoded with Dolby S can be exchanged more easily, as it's less sensitive.
Tapes cannot saves tones or any other sounds at any volume. If the magnetic fields of all particles on the tape point into the same direction, you can record as loud as you want - the volume of the output signal will not rise any more; the tape is saturated - otherwise there were magnets weighing 1g being able to lift 10 kg of iron. So if the recorded signal is too loud, the tape will not reproduce the input signal correctly; it will be very distorted, and this is why tape saturation should be avoided.
What do NR systems gain in practice?
The following chart shows the maximum NR values at -30dB, with different frequencies, based on Dolby's diagrams.
|NR||40||60||120||250||500||1 k||2 k||5 k||10 k||12 k||15 k||20 k||Hz|
|NR||40||60||120||250||500||1 k||2 k||5 k||10 k||12 k||15 k||20 k||Hz|
And what does that mean?
|In the critical 5 - 10 kHz frequency range a maximum of NR of 10 dB using Dolby B, 13 dB using Dolby C (I expected a similar value) and 15 dB using Dolby S are to be gained, and normally 6 - 8 dB using Dolby B, 8 dB using Dolby C, and 9 - 10 dB bei Dolby S.|
What do these dB's mean?
dB, Dezibel, 1/10 Bel, after A. G. Bell (=>telephone), is a logarithmic unit of measurement for differences in voltage, volume, performance etc.
With voltages and volumes, the dB's
are calculated this way: (attention: they're calculated differently with
this means 20 times the logarithm (key "log" on the pocket-calculator) of the difference, or if you want to calculate the difference:
(10 raised Dezibel divided by 20).
20 dB of signal to noise ratio: How much difference (d) is there between noise and signal?
=101 (10 raised 1)
=10 , that means signal / noise=10:1 .
With differences in performance, the voltage appears twice (P=U*I; U~I), and that's why the following formula must be used:
[dB]=10*10log[difference in performance] or 10*log10[difference in perf.]
(depends on schoolbook)
[difference in perf.]=10^([dB]/10)
And what is Dolby HX PRO?
For the explanation of HX Pro, the
control should be mentioned first. With it, you can adjust the bias according
to the tape material. Higher bias means less distortion, but causes a drop
in frequency response toward the highs.
Lower bias causes more distortion, but the drop in frequency response is reduced or even reversed (which causes the recording to sound shrill).
Dolby HX PRO adapts the bias level to the recorded signal for an optimum balance between frequency response and distortion.
1.) Deep tones are boosted during
playback (where the time constant 3180 µs defines the
frequency where the boost is more than 3 dB) and cut back during playback
to reduce hum mainly caused by the engines and the transformer.
The disadvantage is that loud deep tones can saturate the tape. This is why this type of equalization is not used with recordings with Dolby S. Once there was a time constant of 1590 µs (until 1974), which was even more problematic.
2.) The maximum recording level of a cassette tape drops towards the highs. This is why high tones are cut back with a time constant of 120 µs (ferric oxide tape) or 70µs (chromium dioxide tapes, metal tapes, ferrichrome tapes [died out]) during recording (to avoid tape saturation) (hope that's right) and boosted during playback. The difference to the bass equalization is that the boost gets larger towards the highs.
Overview of the equalization
time constants and theit introduction
|bass EQ||EQ of the highs||tape type||introduced in|
|1590 µs||120 µs||Typ I||1966|
|3180 µs||120 µs||Typ I||1974|
|3180 µs||70 µs||Typ II||1970|
this means: transitional frequency is equal to 1 divided by (2 * * time constant).
time constant =3180µs
ft=1 / (2*3180*10-6s)
= (½ *106) / (*3180s)
= 50,04872424273 s-1
= 50,04872424273 Hz (quite exactly 50 Hz)
What's the head gap and what is azimuth?
These terms are related to each
other and are presented together.
Every record/playback head has a gap, the so-called head gap, the angle of which in relation to the tape movement direction is or should be exactly 90 degrees. The magnetic field coming out of the head gap is used for the magnetization of the tape. The head gap should be smaller than half of the wavelength (symbol: lambda) to be recorded; if its size is equal to ½ lambda, the attenuation is 4 dB. As you can see, it is frequency dependant, as the wavelength is decreased with an increase of frequency, so the attenuation rises with higher frequencies.
The angle between tape movement direction and the head gap is called azimuth and is exactly 90° in an ideal case. A difference of some angle minutes (!) can cause great loss. Luckily, this is less critical when the tape is more narrow.
Any questions? (I know, this is really a bit complicated and my English is not the best) Or any remarks, errors or...?
If this was too fast for you, you
can save this page (and don't forget the pictures)!
2. Schallspeicherung auf Magnetband, AGFA-GEVAERT AG (publisher), 1975
3. B. Krieg, Praxis der digitalen Audiotechnik, Franzis', 1989
Thanks to Mr. Richard H. Kuschel, Montana / USA for his information about equalization and the photocopies of Nakamichi's publications.
4. Nakamichi USA, manual of the
cassette deck 680ZX
5. Nakamichi USA, Nakamichi Technical Bulletin 2
Maybe there are also contents from the Phish.Net FAQs about Taping Live Shows.
Dolby's web site
Phish.Net FAQs about Taping Live Shows
This page was created when listening
to music in HiFi quality - from cassette, of course! (no music during translation)
Thanks to Mozart, Chopin, Roxette, S. Brightman & A.Bocelli, Beethoven, Mike Oldfield...
This is finally the EOF.