al det e la
bx II more apprendent of the or of the office
The ultimate aim in recording is to capture. dynamic range but without adding audible
learn the reasons, it is necessary to consider the physical dimensions of musical sound. In concert, the sound level fortissimos may be as much as 105 dB
100 dB (plus a safety margin to handle audible
is necessary to subtract at least 10 dB 60 dB to provide an adequate safetu reduces the usable dynamic range to own capability.
range is recorded on a 50 dB dynamic dB of the music will be distorted and
tape recorder technology have improved experts agree that future developments in nothing near the 50 dB improvement that
limits of the tape medium and the
dynamic range of the music to fit within
The traditional solution used by the during recording. This restricts the dynamic range of the music to fall within permitting most of the quiet sounds to be recorded above the tape noise level while tape which are only slightly (although audibly) distorted. The dynamic range of quietly. This produces a limited dynam range for the microphones to pick un often results in artistically poor peroverload of the tape and audible distorquieter signals. A limiter acts more
drastically to restrict any loud signal that
the limits of the recording medium, some rtion of the music is irretrievably lost in ne process, and this is why recordings fail hisswhich has always been audible since the invention of the tane recorder and can be heard in the guiet portions of your tapes (and even your discs, since they originate on tape). Clearly, some means of "noise reduction" and retrievable the live listening event
ise and letting the music flow through two classes of noise reduction systems ended or "after the fact" systems, which material to suppress its noise content.
Noise filters can reduce noise by Most of the "after the fact" systems use
electronic gate in each band opens to let stops to prevent noise from getting
filters and auto-correlators is that at devices are unable to electronically
dbx offers a line of expanders expand uniformly over the entire audio frequency spectrum, and use no filters or which would rob the music of its high
uses a two-part system consisting of a
recording studios throughout the world, optimized this system for discriminating superiorities of dbx professional studio
What do we want from the recording process?
dbx 120 noise reduction is a sophisticated
The dbx 120 system makes possible for audible noise in the recording process
original dbx professional system except for the specific requirements of the audiophile and home recordist.
The dbx 120 series of audiophile noise with any reel-to-reel, cassette or good quality. While the dbx noise reduc-
dbx 120 audiophile noise recording of live programs, for dubbing noise present in the source, but allows
For tape dubbing and duplication and
but not both at the same time. When without changing cords or connectors.
the 122. It also has additional capability three-head machines while recording is
impossible to manufacture a conventional
totally eliminate disc surface noise. Only
Hersteller does not require. Inasmuch as dbx reduces noise to inaudibility, we discourage Nitchael Otto the use of Dobly B in addition to dby Copying Doble P
Copying Dobly B tapes or broad the casts on the dbx system is another matter.
medium to another, as long as you do
As stated above, you can copy dbx
Yes, definitely, although it probably would
a dbx dynamic range enhancer such noise removal from conventional discs, a
and an expander (Figure 1). The signal or mixer and prior to tape recording to decrease the dynamic range of the original signal to pass through the usable dynamic range "window" of the above the noise floor, and the loudest signal below the level of distortion or signal is expanded to precisely recreate program material. For a compander to expander must operate as exact mirror signal detector (see Figure 2). There peak, average, and RMS.
All tane recorders introduce substantial amounts of phase shift. As the Dolby
and average detection that is sensitive recorder, it is bound to introduce some inadequate for demanding tape recording applications RMS detection as developed by dbx for its companders. energies present in the signal and is introduced by the tape recorder. In this true mirror image compression/expan-The Dolby "B" system operates
only in the high frequency portion of the audio spectrum and is based on the is subjectively called tape hiss is most annoving at higher frequencies. This approach ignores the fact that noise is a broadband phenomenon present at system covers the entire audio
The Dolby "B" system reduces high frequency tape noise by approx-
imately 7 to 10 dB which is a noticeable reduction, but which still leaves the hiss plainly audible during the quieter passages of music. The dbx system achieves over 30 dB of noise reduction that absolutely no hiss is audible at any heard through the most revealing reproducing system.
The Dolby "B" system do not compress and expand the program result is that, while hiss is reduced in of the total program is not compressed room." The ever present tendency to is as great with the Dolby "B" as with the dbx system compresses the entire that headroom is increased by some 10 dB and the risk of tape overload distortion is virtually eliminated. The dynamic range of the original signal.
The Dolby "B" system is inherently non-linear and operates above a certain threshold which must be identical during recording and playback if the compression and properly. Although most Dolby "B" step is difficult and bothersome at best and is impossible on some Dolby "B" equipped units due to manufacturers' failure to include level matching controls Bu contrast dby companders operate linearly over the complete dynamic range making complex level match and alignment procedures
The dbx 120 system uses a linear decibel form of compression/expansion at a factor of 2:1 (see Figure 3). Unlike decibel approach has no threshold at does Dolby B. With dby all levels and range in excess of 100 dB. For convenience in operation, a specific input level has been selected as the unity gain point shown as 0 dB on Figure 3. Notice that the entire signal is compressed to half its original value or by a factor of 0.5. It can be observed that with this approach high level signals are least a 10 dB increase in tape recorder meter reading before distortion occurs). Notice also that the low level signals are raised to a level above the noise
Upon playback, the signal a precisely complementary factor of 2. exact reconstruction of the original setting for this system is not critical for accurate record/playback tracking. The dbx system has level adjustments for compression/expansion, but these accurate reproduction.
Historically, other compander systems exhibited audible variations in back-To minimize this effect a high
during compression and a reciprocal
emphasis reduces audible variations in high frequency "breathing" or
This large amount of preemphasis could cause high frequency added to the RMS level detector
The dbx approach effectively lowers the noise floor by 30 dB and without introducing audible breathing
Expanders must accurately track the compressor's operation, dbx uses RMS due to inherent tape recorder phase shifts. It is also possible for an expander to mistrack due to subsonic filters are placed in the signal path The handpass filters used in
the compressor and expander signal portant to note that the signal detector range where most of the musical signal filter in the detector functions only from 30 Hz to 10 kHz. The frequency response of the system is the full audio spectrum of 20 Hz to 20 kHz: however. This allows live recording of the full due to passing trucks, subways, ventil-
The dbx 120 compression/expansion vise reduction system elegantly complishes the elimination tible tape his the system elegantly and prevents the build-up of noise taping off-the-air. The system accomplishes these benefits without annoving side effects and with no level setting or adjustments necessary by the user.
2 | 0 50 | 0 100 | 5 | DO IK | 2.6 | SK IOK 2 |
)
OK |
20 5 | 0 100 |
500 |
IK | 2.K | 5 | K 101 | ĸ | ||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
- 5 | - | _ | |||||||||||||||
0 | _ | - | 0 | ||||||||||||||
+5 | / | 1 | • | ||||||||||||||
10 | / | 0 | - | - | - | ||||||||||||
15 | + | 5 | - | ||||||||||||||
] | 1 |
Tracking accuracy
± 1 dB/20 dB for complete code
Frequency response -3 dB at 27 Hz for disc decode IM distortion (60 Hz and 7 kHz 4:1 per SMPTE) 0.15% typical; 0.3 maximum Note: Above measurements at 1 volt RMS input and output
Overshoot = 12 dB for 1 kHz tone Release rate = 240 dB/second
audio stages: level match setting
The characteristic of usable dynamic range is one of the most important qualitative aspects of a tape recorder. It is generally not specified on manufacturers' data sheets, but it is roughly equivalent to the machine's signal-to-noise ratio, which usually is stated. Unfortunately, there is presently no accepted industry standard for defining or measuring signal-to-noise ratio in tape recorders and it may be years before there is such a standard.
In practical terms, the usable dynamic range of a tape recorder is a measure of the difference between the loudest and quietest passages of music which may be recorded on the tape and played back without audible noise or distortion. Usable dynamic range may be illustrated as a "window" of limited height through which the musical signal passes on its way through the tape recorder.
The height of this window is determined by the location of the distortion "ceiling" and the noise "floor." This is shown graphically in Figure 8. Any signal which is below the distortion ceiling and above the noise floor will pass through the tape recorder without being audibly changed or degraded. Some information on the distortion ceiling and the noise floor will help to define the signal window and relate it to everyday reality in tape recording.
As the signal level (loudness) presented to a tape record head is increased, a point is reached where any additional increase in level overloads the tape and audible distortion results. The point at which 3% harmonic distortion occurs is generally accepted as the maximum recording level by most tape and recorder manufacturers. The problem for the recordist is that he has no good way of determining when this ceiling has been reached, short of hearing distortion played back off the tape.
Unfortunately, there is no industry standard to relate the 3% distortion point to the recorder's VU meters or "overload" indicators. Tape recorders may
go into audible distortion at anywhere from 2 to 12 dB* above 0 VU, depending on the quality of the machine and the conservatism of its manufacturer in attaching performance specifications to it. The difference between the maximum VU meter reading and the 3% distortion point is a measure of the recorder's "headroom", or safety margin. One of the recordist's major problems is not knowing, and not being able to determine in advance with any reasonable degree of certainty, the headroom of the recorder.
An additional complexity facing the recordist is that most VU meters give only average readings and do not measure peaks. These peaks are loud instantaneous sounds like drumbeats and cymbal crashes which are able to cause distortion but are of such short time duration they do not show up in the meter readings. The extent to which the peaks exceed the VU meter reading has been observed to be as high as 20 dB. Thus, it is possible to use all the recorder's headroom and drive it into severe distortion on program peaks, even though the VU meter readings never exceed +3 VU.
Headroom is a critical factor determining the practical ceiling for distortion-free recording. Overestimating the recorder's headroom will result in 1. tape overload distortion, 2. the loss of the sharp edge of percussive attacks, 3. a smearing of peak transient sounds, and 4. a general muddying of the loud passages when many instruments are playing together. Underestimating headroom and
dbx, Incorporated
296 Newton Street Waltham, Massachusetts 02154 (617) 899-8090
recording at too low a level results in using less of the recorder's dynamic range and places the recorded program where the quieter portions of the music may be below the noise floor of the recorder.
The noise floor is the restricting factor on the guiet end of a tape recording and is defined as the level at which the recorded program signal is equal to the total hiss and background noise of the recorder. If a tape with nothing recorded noticeable sound coming off the tape. This sound is subjectively characterized as tape hiss. The reasons for this hiss are beyond the scope of this brochure. but they are explained in any book on tape recording theory. This hiss plus any electronic noise contributed by the tape recorder circuitry defines the noise floor. which is the lowest level at which a signal may be recorded on the tape and played back without being covered or masked
The usable dynamic range, or "window", then, of any given tape recorder is the number of dB difference between the audible distortion level on the noise level plus tape hiss on the "quiet" end. For the very best professional studio recorders the theoretically attainable dynamic range is about 68 dB at 15 inches per second tape speed, but this performance is seldom, if ever, met in practice. A more realistic dynamic range for professional recorders is approximately 60 dB. For good quality audiophile reel-toreel recorders of contemporary manusecond this number is typically 50 dB. For cartridge and cassette recorders, even fairly good ones, this number is typically 45 dB, because of limitations imposed by slower tape speed, narrower track width and other considerations. The addition of dbx noise reduction to any tape recorder will nearly double the usable dynamic range capability.
Four channel record or playback noise reduction system (with channels switched in pairs) and dbx disc decoder. Can also be used as a two channel simultaneous record and playback noise reduction system.
The dbx 120 series of audiophile noise reduction systems is designed for use with any reel-to-reel, cassette or cartridge tape recorder of reasonably good the better the quality of the associated components, the better will be the audible result. The units in this family interface with any device operating at line level and having single ended lines terminated by RCA type phono plugs.
dbx 120 audiophile noise reduction units may be used for original recording of live programs, for dubbing or duplication of tapes from one machine to
another or one medium to another, or for taping broadcasts off the air, for transferring material from discs to cassettes,
For tape dubbing and duplication and recording off the air, the dbx system is normally connected to the tape monitor inputs and outputs of the audio system preamp or receiver. The dbx unit may recorder and another, or between tuner preamp or receiver. For live recording,
a 30 dB reduction of noise contributed by the tape recording process. (We do not, however, attempt to remove noise present in the original input signal.)
uses 2:1 linear decibel compression and expansion over the entire dynamic range. High frequency pre-emphasis mize mid-frequency distortion and tape frequency tape saturation. The 120 headroom and in excess of 30 dB of
For companders to work effectively. the expander must be able to exactly restore on playback the dynamics which phase shifts which make it difficult for conventional compander systems to properly re-create the input signal. The dbx system precisely reproduces the original sound regardless of recorder phase shifts over a 100 dB range even on music containing percussive peak system eliminates the need for level match tones.
An additional feature of the 120 series is its ability to reproduce dbx encoded discs. The limitations in the disc and tape recording process make it impossible to manufacture a record capable of reproducing the full dynamic range (commercially available) provide the listener with the full excitement of the original performance and totally eliminate disc surface noise. Only when tape hiss and disc surface noise are completely eliminated via the dbx noise reduction system does one realize how pervasive background noise has been
The dbx noise reduction system uses dynamic range compression to encode the signal before recording and a matched expander to restore the original dynamics on playback. The compression and expansion factor is 2:1 and is linear in decibels over a 110 dB range. A 20 dB change in the program is recorded as a 10 dB change on the tape. Thus, a recording medium capable of only 55 dB signal to noise ratio can record and reproduce the full dynamic range of live music which often exceeds 100 dB. Level matching on plavback is unnecessary for proper decoding as this factor is the same at all levels. The adjustments provided for playback level are only for convenience
Frequency weighting networks are used on both the signal and level sensing circuits in the encoder and decoder. This optimizes performance in terms of mid-frequency distortion, tape modulation noise, the audibility
of phonograph mistracking, and high frequency overloading of recording media.
The dbx compression-expansion noise reduction system consists of two voltage controlled amplifiers and two level sensors, there being one of each in the encoder and decoder. If the two level sensors track, the gain change instructions given to the encode and decode voltage controlled amplifiers will be equal and opposite.
All recording systems have some frequency dependent phase shift which changes waveforms considerably even though the sound is not audibly degraded. dbx gets around this problem by using wide range RMS (root-mean-square) level sensing circuits which calculate the sum of the signal energies, regardless of their phase relationships. This results in excellent code-decode gain tracking even with extremely fast transient rise times and a wide range of input levels.
Dynamic range (weighted background noise to peak signal ratio) Designed to feed tape recorder inputs 5 k ohms or greater Maximum output level 7 volts RMS at 1 kHz into 5 KΩ load Effective noise reduction 30 dB for tape recorders with S/N 40 dB for dbx encoded discs ± 1 dB/20 dB for complete code Frequency response ± 0.5 dB 50 Hz to 15 kHz. ± 1 dB 30 Hz to 20 kHz single sinewave code-decode cycle plex music program -3 dB at 20 Hz for tape noise -3 dB at 27 Hz for disc decode 2nd harmonic ≤ 0.1% code-decode 30 Hz to 15 kHz 3rd harmonic ≤ 0.1% code-decode 100 Hz to 15 kHz 3rd harmonic ≤ 0.5% code-decode 30 Hz to 100 Hz IM distortion (60 Hz and 7 kHz 4:1 per SMPTE) 0.15% typical; 0.3% maximum Note: Above measurements at 1 volt RMS input and output Impulse response Risetime = < 20 µ s Overshoot = 12 dB for 1 kHz tone Release rate = 240 dB/second Level match range
100 mV to 3 V for unity gain; individually adjustable each channel, for convenience in level setting and avoidance of overload in succeeding audio stages; level match setting unimportant for code-decode tracking
Manufactured under one or more of the following U.S. patents: 3,681,618; 3,714,462; 3,789,143.
dbx, Incorporated 296 Newton Street, Waltham, Mass. 02154