EP1125274A1

EP1125274A1 - Digital audio signal processing apparatus comprising a delay line

Info

Publication number: EP1125274A1
Application number: EP00960453A
Authority: EP
Inventors: Richard D. Gallery
Original assignee: Koninklijke Philips Electronics NV
Current assignee: Koninklijke Philips NV
Priority date: 1999-08-28
Filing date: 2000-08-14
Publication date: 2001-08-22
Anticipated expiration: 2020-08-14
Also published as: DE60026719D1; EP1125274B1; JP2003508803A; KR20010089369A; DE60026719T2; KR100717519B1; WO2001016934A1; GB9920326D0

Abstract

In a digital audio signal processing apparatus for karaoke systems, echo cancellation and like applications, a digital delay line (10) is connected to receive an input audio signal (DRY) and has a plurality of outputs, each at a respective delay with respect to the input signal and each with a respective signal filter (12, 14) before the delayed and filtered signals are combined by a mixer stage (20). The signal filters (12, 14) are provided by a programmable digital data processor which receives the delay line outputs and processes each according to a respective current filter setting (SE, SC). On receipt of instructions to change a filter setting (SC) for a particular delay line output, the processor generates a second filter (114) for that output in parallel with the first (14) and at the new setting (SF), filters the output at both the old and new settings, and outputs a signal fading from the first to the second filter output. In order to avoid processing overloads, the processor creates only so many respective parallel filters (114) as can be accommodated at one time within predetermined restrictions on the data handling capability of the processor, effectively changing the filter settings one at a time or in small groups.

Description

DESCRIPTION

DIGITAL AUDIO SIGNAL PROCESSING APPARATUS COMPRISING A

DELAY LINE

The present invention relates to the processing of digitised audio signals and particularly, although not exclusively, to the handling of processor resources during the addition of echo or reverberation effects by such processing.

One example of an audio signal processor is given in US Patent

5,774,560 and involves a reverberation network for processing including a plurality of wave-ladder filters, each of which simulates a corresponding reflected audio signal from wave propagation in a selected direction. Each of these filters receives as its input a signal derived from the original audio signal and the outputs of the filters are combined in an adder to create the reverberation effect.

Further examples may be found in the field of karaoke systems where echo/reverberation effects are often applied to the signal from a singers microphone. The effect is achieved by forming an output signal which consists of a generally unmodified signal (a so-called "dry" signal) summed with an additional signal component formed from a weighted sum of filtered and delayed versions of the dry signal, with the filtering providing the reverberation and the delay producing the echo. Such a system will be described hereinafter with reference to Figure 1 of the attached drawings. As will be readily understood, for digitised data, the filtering may be applied by one or more suitably programmed processors which can vary the treatments applied in response to user or system commands. Filtering can, however, prove relatively computationally expensive and, for relatively low powered systems the processor will be working at or near peak capacity for the majority of time. Unless particular measures are provided to avoid it, changes in the setting of the filters can either lead to unacceptable delays as the processor runs out of capacity to handle the changing instructions in real time, leading to audible "clicks" appearing in the output signal, or a more powerful processor has to be used with the capability to handle these short- term requirements for additional processing capability.

It is accordingly an object of the present invention to provide an audio signal processing apparatus having a plurality of programmable filters which avoids audio discontinuities as filter settings are changed.

It is a further object to provide such an apparatus which may achieve this without requiring excessive processor capacity for the filters. In accordance with the present invention there is provided digital audio signal processing apparatus comprising a digital delay line connected to receive an input audio signal and having a plurality of outputs, each at a respective delay with respect to the input signal; a respective signal filter coupled to each delay line output, each filter having at least one variable setting; and a mixer stage arranged to receive and combine the outputs of the signal filters and to output a processed audio signal; wherein the signal filters are provided by a programmable digital data processor arranged to receive the delay line outputs and process each according to a respective current filter setting and wherein, on receipt of instructions to change a filter setting for a particular delay line output, the processor generates a second filter for that output in parallel with the first and at the new setting, filters the output at both the old and new settings, and outputs a signal fading from the first to the second filter output. By the creation of a second filter, a smooth transition between the settings may be achieved.

In order that this creation of additional filters for the duration of a setting change does not create excessive processing overheads, on receipt of instructions to change a plurality of filter settings for respective delay line outputs, the processor is preferably programmed to create only so many respective parallel filters as can be accommodated at one time within predetermined restrictions on the data handling capability of the processor. In a particularly processor-restricted embodiment, the processor may change the settings of the filters, including the creation of respective parallel filters and fading between outputs thereof, one after the other.

The apparatus may further comprise two or more audio signal inputs each with a respective delay line, with the output filters for all delay lines being provided by the same processor, which processor is arranged to allow changes to filter settings for only one of the delay lines at a time - again preventing overloading of the processor. With not all filter settings being changed simultaneously, the apparatus may further contain storage means holding an instruction queue in which are placed those instructions to change one or more filter settings that are not immediately acted upon.

The apparatus may further comprise user-operable input means coupled with the processor and by means of which a user can specify, directly or indirectly, changes to one or more filter settings. Additionally, the mixer stage may be further coupled to receive the or each input signal and arranged to combine this with the outputs of the signal filters to output said processed audio signal and/or coupled to receive one or more further digital audio signals and arranged to combine the or each such signal with the outputs of the signal filters to output said processed audio signal.

Further features and advantages of the present invention will become apparent from reading of the following description of preferred embodiments of the invention, given by way of example only, and with reference to the accompanying drawings in which:

Figure 1 schematically represents a multiple output digital delay line with output filters on each output for the creation of reverberation and echo effects;

Figure 2 is a block schematic diagram of a data processing apparatus suitable to embody the present invention;

Figures 3 to 6 sequentially represent the changing of filter settings in an apparatus embodying the present invention; and

Figure 7 illustrates the sequential handling of filter change operations with multiple delay lines. In the following description, a dual microphone karaoke system embodying the present invention will be described with reference to the attached Figures 1 to 7: it will be understood, however, that the present invention is not limited to such applications and may be used in other digital audio signal processing systems such as echo cancellation circuits in telecommunications applications, as will be readily understood by the skilled reader.

As previously mentioned, in karaoke systems echo and/or reverberation effects are often applied to the microphone input. As shown in Figure 1 , this effect is achieved by forming an output (OUT) signal which consists of a so- called DRY signal (the microphone output following such analogue to digital conversion and other localised signal processing as necessary) summed with an additional signal component formed from a weighted sum of filtered and delayed versions of the dry signal. The echo results from passing the dry signal through a multiple-tapped delay line 10 where the amount of available delay ranges from T=0 (no delay) to T=End (maximum delay) with tapping outputs being shown for four increasing delay values, D(1), D(2), D(n-1 ), and D(n): other values between D(2) and D(n-1) may of course be inserted. On each of these outputs, providing reverberation effects, are respective filter stages 12 (F1), 14 (F2), 16 (Fn-1), 18 (Fn), with each filter having a respective initial setting 18 (SA), 16 (SB), 14 (SC) and 12 (SD). In order to produce the output signal OUT, the outputs from the filter stages 12, 14, 16, 18 are summed by mixer 20 together with the dry signal. Also shown coupled to each of the filter stages 12, 14, 16, 18 is a memory 22 which acts as a queue Q or buffer for user or system commands UIP to change one or more of the filter settings SA, SB, SC, SD.

Figure 2 represents the functional components of an apparatus embodying the present invention in a twin-microphone karaoke apparatus. Each of the two microphones 24, 26 outputs its signal to a respective analogue to digital (A/D) converter 28, 30 and from there the respective digital bit streams are read into respective multiple-tapped random access memories MEM 32, 34. Each of these memories MEM is coupled with a programmed digital data processor PROC 36 and forms a respective digital delay line through reading out of the contents thereof at predetermined times D(1), D(2)...D(n) after they were read in. As shown, the processor 36 may be a slave unit operating under control of a central control and processing unit (CPU) 38 for the apparatus as a whole, or the functions of processor 36 may be carried out by CPU 38 as a dedicated subset of its functions.

The processor stage 36 applies the treatments specified for each of the filter stages 12, 14, 16, 18 (Fig.1) to the digital data read from the memories 32, 34 and outputs this as a digital audio stream to mixer stage 40: as in the case of the mixer 20 in the Figure 1 embodiment, the output signal may include an unprocessed (or dry) input signal component routed directly through the memories 32 or 34 without delay and via the processor 36 to the mixer 40.

Also connected as input to the mixer 40 for the illustrated karaoke embodiment is a source of background music read from optical disc 42, which music the holders of microphones 24, 26 are singing along with: the visual display of lyrics for the singers on a display screen is a well-known feature of such systems and, since it bears no relation to operation in accordance with the requirements of the present invention, it is neither illustrated nor discussed further.

The desired track of background digital audio that the user wishes to sing along with is indicated to the apparatus by means of user input controls 44, which controls may comprise a menu selection from a touch screen display, the input of an alphanumeric code to indicate the selection as one from a group or in other ways as will be readily understood. Following selection, the digital audio is read from the disc 42 by disc reader 46 under direction of the system CPU 38. As the audio on disc will generally be in encoded form (e.g. MPEG), the digital audio read by reader 46 is suitably passed via a decoder stage 48 before being read, in decoded form, into a buffer 50. From the buffer 50, any necessary signal treatments such as pitch shifting 52 are applied under direction of the CPU 38 before the background audio is added as a further component to the mixer 40. From the mixer 40, the output digital audio signal is subjected to any such further signal processing as required (at stage 54) before being output to amplifier 56 and from thence to left and right speakers 58, 60 respectively: note that amplifier 56 is presumed to be a digital signal amplifier and hence no separate digital to analogue (D/A) stage is shown.

The arrangement whereby the settings of one or more of the filters (12, 14, 16, 18; Fig.1) changes setting will now be described with reference to the sequence of drawings in Figures 3 to 6. In Figure 3, a command to change settings UIP has been received and queued in buffer 22. The change command is communicated to the filters affected (F1 , F2 in this example) and then either due to restrictions in the available processor capability in the processor (38; Fig.2) hosting the filters, or as a matter of routine practise for the system, it is decided to change the settings sequentially.

Figure 4 shows the situation as the setting SD of filter F1 is changed to setting SE. The change involves the creation by the hosting processor of a second F1 filter 112 in parallel with the first 12, the second filter being at the new setting SE. The tapped signal from delay line 10 is passed to both copies of the filter and, as indicated by arrow 62, a cross fade is applied on the outputs to mixer 20 such as to give a smooth transition from the first setting SD to the second SE.

In Figure 5, the change from setting SD to SE for the F1 filter has been concluded and the copy 12 of the filter at the SD setting has been deleted. The next (F2) filter 14 is being changed from setting SC to SF and again a parallel filter 114 is created at the new setting. Also as before, there is a cross- fade (indicated by arrow 64) to give a smooth transition between the settings.

In Figure 6, the settings change has been concluded and only those copies of the filters F1 , F2 at the new settings (SE, SF respectively) remain. As will be clear from the progress of Figures 3 to 6, overloads in required processing capability are avoided by scheduling the filter changes to occur sequentially. Where there are a larger number of filters to change settings for, and sufficient processing capability to manage without slowing operations unacceptably, the filters may be changed in small groups (perhaps two or three filters) simultaneously, with the groups being handled sequentially.

In general, changing the individual filter settings sequentially is preferred to give an even loading to the processor. This may also apply in situations as shown in Figure 7 where there are two delay lines 10, 110 (a separate one for each microphone MIC1 , MIC2) and it is desired to change the settings for the filters 12, 14 on the first two line taps of the delay line 10 and the filters 72, 74 on the first two taps of the second delay line 110. Where the filters are all provided by the same processor, the rule that only one may change at a time is applied, with the filters being treated in the order 12, 14, 72, 74 and all settings for the delay line 10 output being changed before the settings for the delay line 110 output.

From reading the present disclosure, other variations (including conformance or otherwise with other optical or magnetic disc recording standards) will be apparent to persons skilled in the art. Such variations may involve other features which are already known in the field of apparatuses for replaying of audio and/or video signals and component parts thereof and which may be used instead of or in addition to features already described herein.

Claims

1. Digital audio signal processing apparatus comprising a digital delay line connected to receive an input audio signal and having a plurality of outputs, each at a respective delay with respect to the input signal; a respective signal filter coupled to each delay line output, each filter having at least one variable setting; and a mixer stage arranged to receive and combine the outputs of the signal filters and to output a processed audio signal; wherein the signal filters are provided by a programmable digital data processor arranged to receive the delay line outputs and process each according to a respective current filter setting and wherein, on receipt of instructions to change a filter setting for a particular delay line output, the processor generates a second filter for that output in parallel with the first and at the new setting, filters the output at both the old and new settings, and outputs a signal fading from the first to the second filter output.

2. Apparatus as claimed in Claim 1 , wherein on receipt of instructions to change a plurality of filter settings for respective delay line outputs, the processor creates only so many respective parallel filters as can be accommodated at one time within predetermined restrictions on the data handling capability of the processor.

3. Apparatus as claimed in Claim 1 or Claim 2, wherein on receipt of instructions to change a plurality of filter settings for respective delay line outputs, the processor changes the settings of the filters, including the creation of respective parallel filters and fading between outputs thereof, one after the other.

4. Apparatus as claimed in any of Claims 1 to 3, further comprising two or more audio signal inputs each with a respective delay line, wherein the output filters for all delay lines are provided by the same processor, which processor is arranged to allow changes to filter settings for only one of the delay lines at a time.

5. Apparatus as claimed in Claim 4, further comprising storage means holding an instruction queue in which are placed those instructions to change one or more filter settings that are not immediately acted upon.

6. Apparatus as claimed in any of Claims 1 to 5, further comprising user-operable input means coupled with the processor and by means of which a user can specify, directly or indirectly, changes to one or more filter settings.

7. Apparatus as claimed in any of Claims 1 to 6, wherein the mixer stage is further coupled to receive the or each input signal and arranged to combine this with the outputs of the signal filters to output said processed audio signal.

8. Apparatus as claimed in any of Claims 1 to 7, wherein the mixer stage is further coupled to receive one or more further digital audio signals and arranged to combine the or each such signal with the outputs of the signal filters to output said processed audio signal.