Handbook for Sound Engineers

1070 Chapter 28

evaluating the distortion, noise, and maximum output

levels for each bias setting.

There are two common methods used for setting the

bias level. One technique is to adjust the bias while

recording a long wavelength such as 1 kHz. The bias is

increased until the recorded signal peaks. The bias level

is then further increased until the recorded signal drops

by 0.5 dB.

A second technique is to use a short wavelength,

typically 1.5 mils, and adjust for a significant amount of

overbias. The bias is increased until the recorded signal

peaks. The bias is then further increased until the

recorded signal decreases from peak by several dB.

How do these two techniques compare? Find the

sensitivity curves S 1 and S 10 near the center of the graph.

These curves show how the 1 kHz and 10 kHz signals

will change in level as the bias is increased. Note that

the S 1 curve is very flat, changing only ¼ dB from peak

over a bias range of 5 dB. In comparison, the S 10 curve

is falling at a rate of approximately 1 dB/dB of bias

increase.

The flat shape of the S 1 curve provides very little

signal drop for a rather large bias change. A 0.1 dB

error in the signal level adjustment, perhaps due to a

sticky meter, may change the 10 kHz sensitivity by 2 dB

or 3 dB. This error would require an additional record

equalization boost of 2–3 dB to correct the overall

response.

In contrast, the rapid signal level change when using

a 10 kHz signal gives a much more precise adjustment

and better uniformity from track to track. It is clear that

both techniques are trying to achieve the same adjust-

ment, but the short wavelength technique offers much

better resolution.

This technique can be a trap for those who don’t

understand what is actually happening. The S 1 and S 10

curves are really curves for the specific wavelengths of

15 mils and 1.5 mils, respectively. If the tape speed is

doubled, these curves will now represent performance at

2 kHz and 20 kHz. The 15 in/s overbias specifications

for 10 kHz must not be used at any other speed! For

example, at 30 in/s the S 10 curve of the example tape

has a downward slope of only 0.5 dB/dB of bias. Why?

Because the wavelength is 3 mils, not the 1.5 mils of the

previous example at 15 in/s. The manufacturer recom-

mends only 1.5 dB of overbias at 10 kHz and 15 in/s. It

is important to use the same wavelength at all speeds by

shifting the test frequency to 20 kHz or 5 kHz at 30 in/s

and 7.5 in/s, respectively.

As mentioned previously, the test data is very depen-

dent on the characteristics of the recorder used during

the testing. In particular, the shape of the S 10 curve

Figure 28-33. Tape performance data in graphical form.

15

10

5

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

Tape speed

Record head gap length

Playback head gap length

Equalization

Reference level

15 ips

7.0 Mm

3.0 Mm

50+3180 Ms

320 nWb/m

nWb

m

514

320

250

160

TDH

% 5 4 3 2 1

0.5

0.4

0.3

0.2

0.1

RLIEC

S 1

10 

 8 

 6 

 4 

2 0 2 4 6

Bias—dB

Record bias

RB

IEC

MTL 16

MTL 10

SOL 10

SOL 16

MOL

S 10

S 16

THDRL+10

THDIEC

BNCCR

S 10 = 4.0 dB