Provides good speech quality and robustness to the channel errors at very low consumption of the computing resources and memory (~2.5 times less than MELPe).

For low cost Digital HF Radio, Digital Mobile Radio (DMR), Voice Storage devices and other markets.

SPR Technology Features. The SPR 1200 bps Vocoder is based on Sinusoidal-Pulsed Representation (SPR) model developed by DSP Innovations.

SPRTM model is based simultaneously on two well-known models: Sinusoidal (Harmonic) Coding (SHC) and Linear Predictive Coding (LPC), where system function is presented by LPC-filter and excitation function is formed by Sinusoidal Harmonic model. 

However, SPR model uses more complex excitation of the synthesize LPC-filter that contains three components:

  • voiced (sinusoidal harmonics of the “fundamental frequency”),
  • unvoiced (noise),
  • pulsed (aperiodic pulses)

As rule, the first two components are used in Sinusoidal Coding, and third component is used in Multi-Pulse Excitation coding on high bit rates. 

We combined advantages of these models to achieve high speech quality at low bit rate and low resources costs. 

So, for example, SHC model can’t represent complex speech intervals with aperiodic fluctuations of vocal chords. However, such intervals can be represented by pulse excitation very well. And vise versa, MPE- model can’t provide high speech quality if number of pulses is not enough, therefore this model is used on relatively high bit rates (more 4500 bps). 

The combining of these models to form an excitation signal for LPC- model allows achieving unique characteristics of vocoders based on the SPR model.

Therefore, bit stream on output of the Voice Encoder contains the following quantized parameters:

  • LPC-parameters (as LSF),
  • “Fundamental frequency” (Pitch),
  • Frequency-depended Voiced/Unvoiced decisions,
  • Pulse parameters,
  • Gains

Speech Quality. Here is the comparison with MELPe 1200 bps vocoder in noiseless channel. SPR 1200 bps vocoder, MELPe 1200 bps vocoder and TWELP 1200 bps vocoder were tested, using ITU-T P.50 speech base for 20 different languages. ITU-T P.862 utility was used for estimation of the speech quality in PESQ terms:

 
A diagram demonstrates superiority of the TWELP 1200 bps over MELPe 1200 bps and SPR 1200 bps  in speech quality in clear channel. TWELP 1200 bps vocoder provides the best speech quality, but consumes much more (~3 times more) computing resources and memory in comparison with SPR 1200 bps vocoder. Exact numbers of the speech quality are presented in the table below.
LanguageTWELP 1200MELPe 1200SPR 1200
American 2.997 2.876 2.784
Arabic 2.917 2.809 2.690
British 2.9 2.826 2.710
Chinese 2.898 2.71 2.651
Danish 2.98 2.797 2.650
Dutch 2.751 2.646 2.518
Finnich 2.74 2.641 2.489
French 2.934 2.897 2.784
German 2.906 2.803 2.671
Greek 2.933 2.753 2.651
Hindi 2.927 2.875 2.736
Hungarian 2.991 2.831 2.727
Italian 3.102 2.989 2.888
Japanese 3.113 2.983 2.877
Norwegian 2.941 2.800 2.678
Polish 2.922 2.792 2.704
Portuguese 2.986 2.900 2.785
Russian 2.837 2.735 2.658
Spanish 2.976 2.851 2.756
Swedish 3.009 2.958 2.792
Average2.9382.8242.710

Superiority of the MELPe 1200 bps vocoder over SPR 1200 bps vocoder in noiseless channel is on average 0.114 PESQ


Speech Samples (WAV-files). A few independent experts listened SPR 1200 bps vocoder in comparison with MELPe 1200 bps vocoder and TWELP 1200 bps vocoder, using method of preferences. Majority of experts preferred TWELP to SPR and MELPe, having noted much more natural human-sounding of voice in the TWELP vocoder. Experts haven't found significant difference in sounding between SPR 1200 and MELPe 1200 vocoders.
You can play and listen short samples of the source speech as well as the speech processed by these vocoders for any of 20 languages, using links in the table below.
Also, you can download full set of the P.50 samples as zip-files for all languages simultaneously, using the links in the "Downloads" para in a bottom of the page.

LanguageSource speechMELPe 1200 bpsSPR 1200 bpsTWELP 1200 bps
American
Arabic
British
Chinese
Danish
Dutch
Finnich
French
German
Greek
Hindi
Hungarian
Italian
Japanese
Norwegian
Polish
Portuguese
Russian
Spanish
Swedish

Superiority In Quality Of The Non-speech Signals. In contrast to other LBR vocoders (MELPe, AMBE+2, etc.), TWELP vocoders provide high quality of non-speech signals, including police, ambulance, fire sirens, etc. This feature in conjunction with high quality natural human-sounding of voice makes TWELP vocoders well suitable for replacement of analog radio by digital radio and also for other applications where high quality transmitting of non-speech signals is relevant along with high quality transmitting of speech signals.

Source typeSource signalMELPe 1200 bpsSPR 1200 bpsTWELP 1200 bps
Siren only
With voice

High Robustness To Acoustic Noise. In contrast to other LBR vocoders, TWELP vocoders are well robust to acoustic noise thanks to robust reliable method of pitch estimation and other features of TWELP technology.

Moreover, vocoder includes in-built Noise Cancellation—Speech Enhancement (NCSE) functionality that improves speech quality in noisy acoustic environment.

NCSE ModeSource signalMELPe 1200 bpsSPR 1200 bpsTWELP 1200 bps
Disabled
Enabled

High Robustness To The Channel Errors. The diagram and table below show a dependence of the averaged speech quality for AWGN-noisy channel for different BER in comparison with other vocoders.

SPR vocoder is robust already to the channel errors. Also, special "robust" versions of the SPR and TWELP vocoders include FEC that are integrated with vocoder on base of "joint source-channel coding" approach that provides high speech quality simultaneously in noisy channel as well as in noiseless channel. FEC can operate with "soft decisions" as well as with "hard decisions" from a modem. "Soft decisions" mode provides much better robustness in comparison with the "hard decisions" mode.

 
BER %MELPe 1200TWELP 1600 RobustSPR 1200
0.00 2.824 2.937 2.710
0.10 2.72 2.913 2.651
0.20 2.649 2.885 2.600
0.30 2.588 2.866 2.559
0.40 2.522 2.837 2.513
0.50 2.463 2.814 2.469
0.60 2.3867 2.781 2.425
0.70 2.338 2.743 2.393
0.80 2.272 2.717 2.350
0.90 2.223 2.676 2.317
1.00 2.184 2.651 2.290

Additional Functionalities. The following additional functionalities are developed by DSPINI and can be integrated (on a request) into vocoders:

  • Automatic Gain Control (AGC),
  • Noise Cancellation for Speech Enhancement (NCSE)
  • Voice Activity Detector (VAD),
  • Tone Detection/Generation (Single tones and Dual tones). The tones are transmitted by the vocoder facilities.

Each functionality has unique features, performance and characteristics, providing significant superiority over any well-known implementations on the market.

Technical Characteristics And Resource Requirements:

Technical characteristics
Bit Rate
(bps)
AlgorithmFrame size
(ms)
Algorithmic delay
(including frame size)
(ms)
Sampling rate
(kHz)
Signal formatBit stream format
1200 SPRTM 40 60 8 Linear
16-bit
PCM
48
Additional functionalities
NameFunctionalityTechnical characteristics
NameValue
AGC Automatic Gain 
Control
Control range: 0 ... +20 dB
NCSE Noise Canceller -
Speech Enhancer
SNR increasing > 6 dB
Speech quality
improvement
> 0.1 PESQ
Tone
Detector
Single/Dual tones 
detection
In accordance with international standards
Tone
Generator
Single/Dual tones 
generation
Special generator, kept continuity of signal 
(phase and amplitude of signal of previous frame)
VAD Voice Activity 
Detection
Reliable detection speech 
in background noise
CNG Comfort Noise 
Generation
Type of noise "white"
Level - 60 dB

Resources for ARM Cortex-M4 platform
ModuleMIPS*
peak
Memory (KBytes)
ProgramData
ConstantsChannelHeapStack
Voice Encoder 22.9 29.8 10.3 3.6 2.4 0.5
Voice Decoder 11.0
Total 33.9


Resources for TI's C64 DSP platform
ModuleMIPS*
peak
Memory (KBytes)
ProgramData
ConstantsChannelHeapStack
Voice Encoder 6.7 39.5 10.3 3.6 2.4 0.5
Voice Decoder 2.9
Total 9.6

Resources for TI's C55 DSP platform
ModuleMIPS*
peak
Memory (KBytes)
ProgramData
ConstantsChannelHeapStack
Voice Encoder 10.5 14.4 10.3 3.6 2.4 0.5
Voice Decoder 4.9
Total 15.4

* DSPINI continues optimization of the SPR algorithm and code in order to minimize computational complexity of the vocoder.

Guarantee And Support.  DSPINI guarantees a quality and accordance of all technical characteristics of the product to requirement of current specifications. Testing and other method of quality control are used for guarantee support.

Any Platforms.  DSPINI can port this vocoder software into any other DSP, RISC or general- purposes platform inshort time: 1-2 months.

Licensing Terms.  To use the vocoder, customer should obtain a license from DSPINI only.

Customization.  The vocoder can be customized under any specific requirements- other bit rate, frame size, any other robustness to channel errors, etc. Please contact with us for details.

Prospects.  DSPINI is impoving and developing continuously a set of new vocoders with range from 300 bps up to 9600 bps, based on SPR and TWELP technologies.

Related Software.  This vocoder may be effectively used in a bundle with other DSPINI's products:

  • Linear and acoustic echo cancellers,
  • Multichannel noise cancellers (including two-microphone adaptive array),
  • Wired or radiomodems for any types of channels and bitrates,
  • Other products.

Downloads: