TAPE GUIDE (Mar. 1993)

Dolby, dbx, and HX Pro Circuits

Q. I am still somewhat in the dark about noise-reduction systems. Please explain the differences among Dolby B, Dolby C, dbx, and Dolby HX Pro. Does HX Pro work in conjunction with noise reduction, or can only one or the other be used?

-Jeff Fehring; Milwaukee, Wisc.

A. Dolby B and C and dbx are noise-reduction systems. HX Pro is not; it seeks to increase treble headroom during recording.

The Dolby and dbx NR systems work on the principle of compression during recording and expansion during playback.

The downward expansion in playback also takes noise down with it. Dolby B and C NR operate only on the upper frequencies, where hiss and tape noise are most audible, and their effects vary with the signal's frequency content. In recording, there is treble boost which varies inversely with treble level; the lower the treble level (and hence the less the masking effect of strong treble), the greater the treble boost. In playback, there is complementary treble cut to restore flat response, again varying inversely with treble level; the lower the level, the greater the cut-and this cut reduces noise.

For Dolby NR to work properly, the level in playback must be matched to the level in recording. Without such "tracking," treble frequencies can be muffled or suppressed. The Dolby B circuit reduces noise above approximately 500 Hz and achieves about 8 to 10 dB of noise reduction; Dolby C NR is effective above approximately 150 Hz and achieves about 18 to 20 dB of improvement in S/N. Although Dolby C NR employs more treble boost than Dolby B NR, the former includes a special recording characteristic to avoid saturating the tape at extremely high frequencies. Therefore, Dolby C NR tends to have more headroom than Dolby B NR-that is, it is freer from tape saturation when recording high frequencies at high levels. Saturation not only causes distortion but also reduces the recorded treble level--a phenomenon called fold back.

The dbx NR system operates throughout the audio range and does not require matching record and playback levels. Thus, there is no tracking problem that can affect treble response. However, it has been noted that response at the frequency extremes, particularly at the treble end, tends to be not as good with dbx as with Dolby NR. The dbx system achieves about a 30-dB reduction in noise.

At this point, mention should he made of the newest development in Dolby noise reduction, Dolby S, which has made its appearance in high-quality cassette decks. It achieves a bit higher S/N ratio than Dolby C NR, covers the entire audio range, provides an anti-saturation characteristic for the bass as well as for the treble, is less subject to the effects of mistracking, is more resistant to such side effects as breathing (varying noise level), and tends to provide somewhat cleaner reproduction.

Dolby HX Pro is a variable-bias scheme to avoid tape saturation in the treble range. It takes into account the fact that treble frequencies add their own biasing effect to that of the high-frequency bias supplied by the deck's oscillator during recording. The HX Pro circuit senses the treble level and correspondingly adjusts the output of the deck's bias oscillator so as to keep total bias more or less constant. Excessive bias causes erasure of high frequencies; by preventing this treble loss, HX Pro allows improved treble response with less treble boost required in recording. This reduced treble boost lowers the chance of tape saturation, which produces distortion and treble loss.

Going Half Speed

O. I plan to modify a cassette deck for half-speed operation. I am an electronics engineer and have the equipment to do so.

What are the equalization, bias, and Dolby NR issues involved in half-speed operation?

-Raymond J. Berry, Bellevue, Wash.

A. The usual requirement for bias is that its frequency be at least five times the highest audio frequency to be reproduced; this is to minimize audible interaction between bias and harmonics of the highest audio frequencies. Thus, if you seek response to, say, 10 kHz, bias should be at least 50 kHz.

To avoid an increase in distortion, it would be desirable to keep the amount of bias the same at half-speed as at full speed. However, if all else remains the same, treble response drops with a reduction in tape speed, so you may want to decrease bias somewhat to avoid excessive treble loss. This entails an increase in distortion; a moderate increase in distortion may be tolerable.

At half speed, you will probably want to use something other than 70-uS ("Type II" or "CrO2") playback equalization, because the more bass boost employed in playback (which may conversely be visualized as treble droop), the more treble boost is required in recording to maintain good treble response. But excessive treble boost may overload the tape, causing distortion and treble loss. Therefore, 120-µS ("Type I" or " Normal") playback equalization would be more satisfactory, requiring less treble boost in recording.

Given a specific playback curve, the standard requirement for record equalization is that it should be such as to achieve substantially flat response over the intended audio range. At half speed, I think you should not aim at overall response beyond 10 kHz; perhaps 8 or 9 kHz would be a more realistic objective.

There might be a minor interaction with Dolby noise reduction because a reduction in tape speed entails a reduction in recorded level. This would affect tracking, namely the matching of recorded and playback levels necessary for proper Dolby NR performance.

(adapted from Audio magazine, Aug. 1992, HERMAN BURSTEIN)

= = = =