DE4426534C2 - Method for controlling the signal recording for a digital answering machine - Google Patents

Description

The invention relates to methods for controlling the Signal recording for a digital answering machine, where a signal recording with different Data rates is possible.

Answering machines were previously usually called Magnetic tape storage devices designed with all known Disadvantages of the magnetic tape storage: noise, large Space requirements, high energy consumption, moving parts and resulting reliability problems, etc.

For some time now, too Answering machine with digital memories (e.g. DRAM, SRAM, EEPROM etc.) offered. When digitally recording Voice signals accumulate large amounts of data, which are caused by the Use of data reduction / data compression can be reduced. There is a compromise between Find data reduction and intelligibility, because with increasing data reduction the Speech intelligibility drops.  

Because of the initially high prices of digital storage with the first devices only the announcement text is saved provided in digital memory. The storage of the incoming reports continued to be made using Magnetic tape.

Such an answering machine is from the patent DE 40 37 514 C2 known. With this answering machine when saying the announcement, the signal at the same time in digital storage and on magnetic tape storage recorded. For digital storage, this is analog voice signal with a given sampling rate digitized and according to this sampling rate in the Memory written. After speaking, the Duration of the announcement text, or that still in digital Memory available storage capacity determined and for optimal use of the storage capacity of the digital memory, a sampling rate is set at their use the digital storage largely filled is. With the sampling rate determined in this way, the signal of the announcement text reproduced by the magnetic memory scanned and written into the digital memory. This way, the available Storage capacity the best possible speech intelligibility achieved.

With newer answering machines, such as. B. in the essay "DESIGN OF A FULL-FEATURED, SOLID-STATE TELEPHONE ANSWERING DEVICE ", by G. Ferris et al, IEEE-CE Vol. 35 No. 3, August 1989, pages 681-686 entire signal recording using digital memory.  

Such a digital answering machine is for example the one manufactured by the applicant and distributed telephone answering machines AM 500 D. Die existing storage capacity allows the recording of Speech signals over a period of eight minutes. In addition, the user has the option of one Select the type of recording in which - with reduced Voice quality - a voice recording of eleven minutes Duration is possible.

The previously known digital answering machines thus point the disadvantage that the total available Recording time is still very short, and that an increase in the recording time by means of Data reduction always leads to a reduction in the Speech intelligibility leads. This is especially so disadvantageous if only a small part .der Storage capacity is needed because only a small one Number of incoming messages from the answering machine must be recorded.

The object of the present invention is therefore Method for controlling signal recording for one specify digital answering machine, which the for Increase available recording time without at the same time in any case to a reduction in the To lead speech intelligibility.

This object is achieved through the features of the claims 1 and 3 solved.  

The method according to claim 1 has the advantage that until a limit is reached incoming calls with the highest available Data rate and thus the highest voice quality to be recorded. Subsequent conversations are with a reduced data rate (or voice quality) recorded, resulting in a longer recording time is achieved. For the first recorded call is thus ensures very good speech intelligibility, while a longer one for subsequent conversations Recording time with lower speech quality Available.

One according to the method described in claim 2 working answering machine has the advantage that the reduction in for signal recording Available data rate in several steps can be made. This will make the Speech intelligibility is slow and gradual reduced.

The method according to claim 3 has the advantage that almost the entire storage capacity first for signal recording with the highest Voice quality is available. Only when the Limit value means that all signals - even those already recorded - with a reduced voice quality to be recorded. So it can be surprisingly simple A particularly good compromise between Speech intelligibility and recording time reached will.  

The method according to claim 4 has the advantage that the reduction in voice quality is slow and done gradually.

The method according to claim 5 has the advantage that all recorded signals have the same speech quality exhibit.

One according to the method described in claim 6 working answering machine has the advantage that the answering machine, by processing the already recorded signals at times without normal operation, can be implemented without circuit expansion.

The invention is described below with reference to drawings described and explained in more detail. Exactly Description of the functions of the answering machine, especially memory control and data reduction, is waived as it is for understanding the invention are not important. Besides, the general one How digital answering machines work and known can for example the fonts mentioned at the beginning as well taken from the documents cited in these documents will.

It shows:

Fig. 1 shows the basic structure of a digital storage device in an answering machine and

Figs. 2 and 3, the breakdown of the capacity of a digital memory.

The digital storage device shown in FIG. 1 comprises an analog / digital converter 1 , a processor P connected to it, a digital memory S connected to the processor P and a digital / analog converter 2 connected to the processor P.

The processor P can be a universal processor programmed for speech processing or a special signal processing processor. The processor P also takes over the monitoring and control of the other components and functions of the answering machine containing it. In particular, the processor P controls the memory S and can therefore continuously provide information about the degree of occupancy of the memory S. It is also possible to combine a signal processing processor for processing the speech signals with a universal processor for the control. As mentioned at the beginning, DRAM, SRAM or EEPROM memories are used for the memory S. If the input signal SE is already available in digital form, for example if the answering machine is operated on an ISDN connection, the converters 1 and 2 can be omitted.

The input signal SE is digitized in the analog / digital converter 1 . The processor P further processes the signal, which can include, for example, suppression of speech pauses and data reduction. The signals generated in this way are stored in memory S, controlled by processor P. When the stored signals are reproduced, they are read out of the memory under the control of the processor P, the processing of the signals carried out during the recording is reversed and the conversion into an analog signal SA takes place in the digital / analog converter 2 in the usual way is processed further.

According to a first procedure voice signals to be recorded at a first data rate recorded which is very good, speech intelligibility guaranteed. When a limit is reached, the for example 60% of the maximum recording time corresponds, the recording takes place with one opposite the first data rate reduced second data rate at the the voice quality compared to signals with the first data rate were recorded - is reduced.

The division of the memory S for this case is shown in Fig. 2a. FIG. 2b shows records stored in the memory S. The recording takes place in the storage area S1 at the first data rate, in the storage area S2 the recording takes place with a reduced data rate. In the case of the recording A6, within which the limit value G separating the memory areas lies, the recording for the part exceeding the limit value G can be continued at the first data rate in order to avoid changes in speech quality within a recording. All recordings following record A6 are made at the second data rate. Of course, it is also possible to switch the data rate directly when the limit value G is reached. It is also possible to switch to the second data rate if the storage capacity remaining in the storage area S1 is only very small.

In a modification of the method, the memory S is divided into more all two memory areas, as shown in FIG. 2c. The signals to be recorded are recorded with different data rates in the memory areas S1, S2 and S3 separated by the limit values GI and G2, as shown in FIG. 2d. The data rate used is greatest in S1, the data rates of the memory areas S2 and S3 are smaller, the data rate in S3 being smaller than the data rate in S2.

According to a second procedure, are to be recorded Voice signals recorded at a first data rate, which ensures very good speech intelligibility. When a limit is reached, they are already saved records read again and from Processor P prepared so that it can be saved again is possible with a lower data rate.

The data is reduced and saved again a time when no incoming reports need to be recorded or none recorded Messages must be reproduced, is enough for this Purpose of the existing processor P out. Additional Circuitry is not necessary. Even on one additional memory can be dispensed with because of the the data reduction reduced data in memory S over the data originally recorded there with higher Data rate can be written without data too to lose. For the data reduction to be carried out, the memory available in the processor or part of the Memory area S2 of the memory S can be used.  

The division of the memory S for this case is shown in Fig. 3a. Memory area S1 comprises the largest part of the memory capacity, memory area S2 only a relatively small area, since it generally only has to have the capacity for a message to be recorded, as will be explained in more detail below.

If the capacity of the storage area S1, as shown in FIG. 3b, is fully utilized and the limit value G is exceeded, the recordings A1 already stored become. . .A6 read out one after the other by processor P, data reduced and stored again ( FIG. 3c). The storage area S1 then has free storage capacity again. Additional recordings A7. . .A9 can be saved. These can either be with the previously used higher data rate or with that for data reduction of the recordings A1. . .A6 used lower data rate can be recorded. If the limit value G is exceeded again, the existing records A1 are further reduced. . .A9 ( Fig. 3d) or at least the records A7. . .A9 etc.

If a message arrives at the answering machine during the data reduction process and is to be recorded, the data reduction is ended when the processor P is sufficiently powerful and the incoming message is simultaneously recorded in the memory area S2. After data reduction has ended, the message recorded in memory area S2 is read out and added to the other recordings at the same data rate. If the processor performance is not sufficient for the simultaneous data reduction and recording of the incoming message, the data reduction is interrupted and after the recording of the message in the memory area S2, the processing is continued as described above. In this case, even when the limit value G is overwritten, as shown in FIGS . 3b, c for the recordings A6 and A9, the data reduction is not started during a recording until the respective recording (A6 or A9) has ended.

In both methods, the step-by-step reduction in the data rate of the recording can be achieved, for example, by a step-by-step reduction in the sampling rate of the analog / digital converter 1 or by using known source coding methods. If the recordings or input signals are already available at a specific sampling rate, processor P can calculate sampling values with other sampling rates by interpolating the existing sampling values. When a predefined speech intelligibility limit is reached, there is no further data reduction, the storage capacity is then fully utilized and a corresponding message is generated. The parameters used for the respective data reduction are stored together with the limit values and / or the individual recordings and taken into account by processor P when preparing the stored signals for reproduction.

The amount of data reduction from storage area too Memory area and the limit values used can either specified or by the user himself can be set. The limit values are thereby advantageous as time limits by the user given.

Claims (6)

1. Method for controlling the signal recording for a digital answering machine, in which a direct signal recording is carried out in a digital memory (S), with different data rates being possible, with the following method steps:

• a) determining the storage capacity available for signal recording during signal recording,
• b) signal recording with the highest possible data rate, as long as the determined storage capacity does not exceed a predetermined or predefinable limit value,
• c) if the limit value signal recording is exceeded with a data rate that is reduced compared to the largest possible data rate.

2. The method according to claim 1, characterized, that multiple storage capacity limits are predetermined or predeterminable when exceeded a respectively reduced data rate for signal recording is used. 3. Method for controlling the signal recording for a digital answering machine, in which a direct signal recording is carried out in a digital memory (S), different data rates being possible, with the following method steps:

• a) determining the storage capacity available for signal recording during signal recording,
• b) signal recording with the highest possible data rate, as long as the determined storage capacity does not exceed a predetermined or predefinable limit value,
• c) If the limit value is exceeded, the signals already stored are read out of the memory (S) and the signals read out are again recorded in the memory (S) at a data rate which is reduced compared to the largest possible data rate.

4. The method according to claim 3, characterized, that each time the limit for the storage capacity a respectively reduced data rate is used for renewed signal recording. 5. The method according to claim 3 or 4, characterized, that the reduced data rates for the subsequent Signal recordings are used. 6. The method according to any one of claims 3 to 5, characterized, that the reading of the signals already stored and the with reduced data rate, new signal recording in There are times when the answering machine does not Receives signals for recording or none reproduces recorded signals.