Compaction method for waveform storage

• US 4,617,645 A
• Filed: 09/10/1984
• Issued: 10/14/1986
• Est. Priority Date: 09/10/1984
• Status: Expired due to Fees

• Current Assignee: Sierra Entertainment, Inc. (Vivendi SE)
• Original Assignee: First Byte
• Inventors: Sprague, Richard P.
• Primary Examiner(s): Fears, Terrell W.
• Application Number:
  US06/649,227
• Time in Patent Office: 764 Days
• Field of Search: 365/45, 365/189, 307/229
• US Class Current: 365/45
• CPC Class Codes: G11C 27/02 Sample-and-hold arrangements G11C27/04 takes precedence
• CPC Class Codes: H03M 7/3044 Conversion to or from differential modulation with several bits only, i.e. the difference between successive samples being coded by more than one bit, e.g. differential pulse code modulation [DPCM] H03M7/3004 takes precedence; voice coding G10L19/00; image coding H04N19/00

Abstract

The memory needed for the digital storage of a waveform having an amplitude range of ±128 units is compacted by nearly 50% through the use of a 16-position table containing selected increments. Each waveform sample value is first predicted from at least two prior samples by linear prediction, and the predicted value is then incremented by the most nearly accurate increment in the table to obtain an approximation of the actual waveform value at the sample point. The appropriate increment for each sample can be defined by a 4-bit table address. The best-fit increment values for any given waveform can be calculated by a reiterative trial-and-error computation process. By the use of a best-fit table, inaccuracies in the waveform reproduction can be held below audible levels.

Claims

1. A method of compactly storing and retrieving a waveform, comprising the steps of:

(a) establishing a table containing a relatively small number of predetermined increments;

(b) storing a series of addresses, each address identifying one of said increments in said table;

(c) successively producing a series of samples representing generally the value of said waveform at time-spaced intervals, each of said samples corresponding to one of said addresses;

(d) predicting the value of each new sample to be produced, said predicted value being a fixed function of the values of at least a current and a previous sample;

(e) incrementing said predicted value by the increment identified by said address for said new sample, to produce a sample value at least approximating the actual value of said waveform at the sampling point; and

(f) reconstituting said waveform from said approximating samples.

2. The method of claim 1, further comprising the step of minimizing the approximation errors between said sample values and said actual waveform values.

3. The method of claim 2, in which said minimizing step includes the steps of:

(i) assigning generally arbitrary values to said increments;

(ii) reiteratively computing a predetermined function of said approximation errors using various values for one of said increments;

(iii) fixing said one increment at the value for which said error function is a minimum; and

(iv) repeating said computing and fixing steps for each of the remaining increments in said table.

4. The method of claim 3, further comprising the step of re-performing steps (ii) through (iv) until said minimum error function for the last of said increments is identical in two consecutive performances.

5. The method of claim 3, in which said increments are varied between the values of their next adjacent increments for the purposes of said reiterative computations.

6. The method of claim 3, in which said reiterative computations are made for the various increments in descending order of their absolute magnitude.

7. The method of claim 3, in which said error function is the sum of the squares of the approximation errors of all of said samples.

8. The method of claim 1, in which said relatively small number is the square root of the number of possible values said samples can have.

9. The method of claim 1, in which said new sample value is predicted by a linear combination.

10. The method of claim 9, in which said fixed function is 2S_c -S_p, where S_c is the current sample and S_p the next previous sample.