Products: Voice Synthesis ICs [ Speech & Audio ]
Communication between people and machines: “Speech Interface”
-- Opening new worlds through spoken communication --
What is sought after by Epson is total speech solution that is simple and handy. It does not require recordings at a studio. It is capable of generating high-quality natural sounding voice instructions with melodies with the built-in melody synthesizer. The TTS function that converts text data into voice data, and other audio LSIs supporting accPlus, AAC-LC and MP3 are also available. These are leading players of secure, safe, and convenient life produced by Epson.
Speech & audio product lineup
Epson offers total support for speech and audio, from voice and music creation through to automatic speech recognition.
So let us take care of all your sound-related product needs.
Product Lineup
- Voice guidance LSIs
Spoken operating instructions
(White goods, home electronics, car telematics, OA and FA equipment) - Melody & Voice guidance LSIs
Melody playback system,
Fire alarm system - Text-To-Speech LSI
Text reading (E-mail, e-books, manuals, news, traffic information, etc.) - Audio LSI
Detail music players (Silicon audio and mobile phones)
Rated Information
- Automatic speech recognition LSIs (Under development)
Voice operated controls (White goods, home electronics, car telematics, OA equipment and educational aids) - Introduction to speech & audio users' site
Voice guidance LSIs
Features
Feature 1. No need for studio recording
Using Epson speech & audio LSIs for your products featuring voice guidance functions means there's no need to arrange recording studios or hire professional voice performers. This allows you to significantly reduce the product development costs and time to market. The TTS tool available from Epson makes it easy to create or modify your speech phrases. Japanese, English, and Korean language versions are now available, with Chinese and Spanish language versions currently under development.
Feature 2. Excellent sound quality for messages up to 4 times longer
| EPSON (16kHz) |
ADPCM (16kHz) |
ADPCM (8kHz) |
|
|---|---|---|---|
| Bit rate | 16[kbps] | 64[kbps] | 32[kbps] |
| Memory size (10 sec.) | 160[kbit] | 640[kbit] | 320[kbit] |
| Memory size (1 min.) | 960[kbit] | 3,840[kbit] | 1,920[kbit] |
| Memory size (2 min.) | 1,920[kbit] | 7,680[kbit] | 3,840[kbit] |
| Memory size (3 min.) | 2,880[kbit] | 11,520[kbit] | 5,760[kbit] |
Epson's highly efficient encoding format allows for longer phrases with excellent sound quality. The integrated high-quality 16-bit DAC produces a "silky voice" that can be coupled with high quality sound effects.
Feature 3. Easy to integrate
Epson's voice guidance LSIs use a message protocol based on asynchronous serial interface. This means they are easily integrated into a wide range of host devices and microcontrollers, reducing development time and costs, and helping achieve faster time to market.
Feature 4. Terminal Compatible Flash Memory version support
The S1V3S344 is an IC chip for the S1V3034x series with built-in flash memory that is terminal compatible. This makes it possible to conduct voice evaluation in product development boards and product packages, and allows additions and alterations arising from last-minute changes and upgrades to be performed easily. When used in combination with Epson's natural-sounding voice data creation tool, this chip provides an efficient and convenient route to implementing voice guidance in consumer products.
Feature 5. Easy to modify/add voice data by external flash IF, streaming function support
S1V3G340 makes it possible to modify/add sounds by following specification.
- Playing the sounds onto external flash memory(serial interface) by just specifying the command to S1V3G340.
- Playing the sounds onto Host Memory by streaming the sound data from Host to S1V3G340.
(Notes: All S1V3x3xx series support voice data streaming function.)
"Silky Voice" Creation tool Voice Sample
Here are some sample voices for Epson's audio and speech LSIs created by our PC-based TTS tool. Click the following links to experience studio-quality voices with high quality sound effects.
Voice guidance LSI lineup
| Series name | S1V30300 | S1V30331/2/3 | S1V30341/3/5 | S1V3G340 | S1V3S344 |
|---|---|---|---|---|---|
| Status | MP | MP | MP | MP | MP |
| Download | - | - |  96kb |
195kb |
196kb |
| Power supply | 3.3v/1.8v | 2.2v - 5.5v(single power supply) | |||
| Clock | 32.768kHz | 12.288MHz | 32.768kHz or 12.288MHz | ||
| Host interface | SPI (Command control) |
SPI/UART/I2C (Command control) |
|||
| Sampling frequency | 8/16kHz | 16kHz | |||
| Decode format | AAC-LC (mono) *1 ADPCM |
Epson highly efficient format | |||
| Bit rate | - | 16/24/32/40kbps | |||
| Voice phrase combination function | - | No limitaion on combination | |||
| Delay setup between phrases | - | 0 ms, 20-2047 ms (1 ms step) | |||
| Integrated ROM size (record duration) | Supported by the command control through the host interface | ||||
| Integrated ROM size (record duration) | - | S1V30331： 1 min.*2 S1V30332： 2 min.*2 S1V30333： 3 min.*2 |
S1V30341： 1 min.*2 S1V30343： 3 min.*2 S1V30345： 5 min.*2 |
- | 4 min.*2 |
| External serial flash memory interface | - | - | - | Support | - |
| DAC | Highly accurate 16-bit DAC | ||||
| Package | QFP52 (10mm x 10mm) (0.65mm pitch) |
QFP52 (10 mm x 10 mm, 0.65 mm pitch) TQFP48 (7 mm x 7 mm, 0.5 mm pitch) |
QFP52pin (10mm x 10mm, 0.65mm pitch) | ||
| Languages supported by TTS tool | Japanese, English, Korean （Chinese/Spanish: under development) | ||||
| - | - | Supported by S1V3S344 | - | - | |
*1: License fee incurred for use of AAL-LC.
*2: Estimate when used at 16 kbps
Voice guidance LSI specification document download
Voice guidance LSI evaluation environment
Under construction
Melody & Voice guidance LSIs
Features
Feature 1. Melody Synthesizer function included
Epson’s Melody & Voice guidance LSI integrates a simultaneous 5-channel Melody Synthesizer and supports 5octaves, Enabling sophisticated melody sounds to be generated from a very small amount of note data.
Feature 2. No need for studio recording
Using Epson speech & audio LSIs for your products featuring voice guidance functions means there's no need to arrange recording studios or hire professional voice performers. This allows you to significantly reduce the product development costs and time to market. The TTS tool available from Epson makes it easy to create or modify your speech phrases. Japanese, English, and Korean language versions are now available, with Chinese and Spanish language versions currently under development.
Feature 3. Mixing
Epson’s Melody & Voice guidance LSI can mix generated Melody sound & Voice Data from on chip ROM, with each volume level set individually.
Feature 4. Simple control using Standalone Mode
In addition to SPI and I2C support, Epson’s Melody & Voice guidance LSI supports a Standalone Mode in which dedicated pins are used to specify the sound data ID number that is to be played. In this way, the Melody & Voice guidance LSI can be controlled without the need for any code on the host processor, or, if desired, without the need for a host at all.
Melody & Voice guidance LSI lineup
| Series name | S1V30080 |
|---|---|
| Status | MP |
| Download | 245kb |
| Power supply | 2.2v - 5.5v |
| Clock | 16.384MHz(Sampling frequency:16kHz) 8.192MHz(Sampling frequency:8kHz) |
| Host interface | SPI、I2C(command control) Standalone Mode |
| Sampling frequency | 4/8/12/16kHz |
| Voice phrase combination function | No limitaion on combination |
| Delay setup between phrases | 0-1000ms(10ms step) |
| Integrated ROM size (record duration) | Approx 30sec (Sampling frequency:8kHz) Approx 15sec (Sampling frequency:16kHz) |
| DAC | 10bit DAC |
| Package | SSOP-16pin(4.4mm x 6.6mm, 0.8mm pitch) QFP-48pin(7mm x 7mm, 0.5mm pitch). ... External serial Flash Interface integrated |
| Natural-Sounding Voice Data Creation Tool Support | Japanese, English, Korean (Chinese/Spanish: under development) |
Melody & Voice guidance LSI specification document download
Under construction
Melody & Voice guidance LSI evaluation environment
Under construction
TTS (text-to-speech) LSIs
Features
Feature 1. All you have to do is input the text
TTS LSIs just need text to output speech.
And text is easy to create and modify.
Feature 2. Minimal data size
TTS LSIs just need text to output speech.
Therefore the speech data size is very small.
e.g. 20 minutes of speech spoken at 180 words/min. only requires around 30 KBs.
Feature 3. Easy to integrate
Epson's voice guidance LSIs use a message protocol based on asynchronous serial interface. This means they are easily integrated into a wide range of host devices and microcontrollers, reducing development time and costs, and helping achieve faster time to market.
Feature 4. The S1V30120 is multi-lingual
The S1V30120 supports the following languages with one chip:
- US English
- Castilian Spanish
- Latin American Spanish
In addition, 9 different voices are supported.
TTS (text-to-speech) LSI lineup
| Series name | S1V30120 |
|---|---|
| Status | MP |
| Power supply | 3.3v/1.8v |
| Clock | 32.768kHz |
| Host interface | SPI (command control) |
| TTSSampling frequency | 11.025kHz |
| Voice stream data replay | Supported by command control through the host interface |
| Decode format | ADPCM |
| Bit rate | - |
| DAC | High quality 16- bit DAC |
| Speaker amp. | - |
| Package | TQFP 13-64 (10 mm x 10 mm, 0.5 mm pitch) |
| Supported languages | US English Castilian Spanish Latin American Spanish TTS Engine: Fonix DECtalk(R) *1 |
*1: Fonix DECtalk(R) 
is a registered trademark of Fonix Corporation.
TTS (text-to-speech) LSI specification document download
TTS (text-to-speech) LSI evaluation environment
Under construction
Audio LSIs
Features
Feature 1. One-chip solution for aacPlus decoder
The S1V30200 can decode aacPlus, AAC-LC, and MP3 with no external memory, supporting streaming from a host, as well as reading from a flash memory (e.g., SD card).
Feature 2. Easy to integrate
Epson's voice guidance LSIs use a message protocol based on
asynchronous serial interface. This means they are easily integrated
into a wide range of host devices and microcontrollers, reducing
development time and costs, and helping achieve faster time to
market.
Audio LSI lineup
| Series name | S1V30200 |
|---|---|
| Status | MP |
| Power supply | 2.9v/1.8v |
| Clock | 32.768kHz |
| Host interface | SPI (command control) |
| Decode format | aacPlus*1 AAC-LC*1 MP3*1 |
| Sampling frequency | MP3: 48, 44.1, 32, 24, 22.05, 16kHz AAC-LC/aacPlus: 48, 44.1, 32, 24, 22.05, 16, 12, 11.025, 8kHz |
| Bit rate | All bit rates supported |
| Voice data stream input | Supported by the command control through the host interface |
| DAC IF | I2S |
| Package | 100-pin PFBGA (7 mm x 7mm, 0.65 mm pitch) |
*1: License fee incurred for use of aacPlus, AAL-LC and MP3.
Audio LSI specification document download
Under construction
Audio LSI evaluation environment
Under construction
Automatic speech recognition LSIs (under development)
Features
Feature 1. One-chip ASR solution
Epson's ASR (Automatic Speech Recognition) chips integrate
the analog circuit, noise canceller, and ASR engine.
Therefore, customers do not need to verify the compatibility between
the analog circuit and the voice recognition engine, nor do they have to pay
royalties for the speech recognition middleware. With these benefits,
customers can significantly reduce product development time and costs.
Feature 2. Easy to integrate
Epson's voice guidance LSIs use a message protocol based on asynchronous serial interface. This means they are easily integrated into a wide range of host devices and microcontrollers, reducing development time and costs, and helping achieve faster time to market.
Introduction to speech & audio users' site
Epson user's site for speech & audio products is available to customers who have already purchased or are planning to purchase Epson Speech & Audio products. Registering for the site allows you to download detailed technical information and specification documents free of charge.
Click here to register for the user site.
For inquiries about user registration, please contact an Epson sales representative or a distributor in your region.
Notes:
-We recommend Internet Explorer 5.01 or higher, or Netscape Navigator 7.02 when using this service. Other browsers may not show the web contents properly.
-Enable JavaScript in your browser to log in for user registration.
Copyright © SEIKO EPSON CORP 2009