18

I'm currently using espeak, an open-source software. Not so bad, even if the voices sound artificial. Details here. E.g. : (Seneca, Epistulae ad Lucilium, 1.1.1) espeak -v eo --ipa -s120 -p60 -a20 "[['ita'fak'mi:lu:'ki:li:]]" i.e. speak with an Esperanto voice (-v eo), print the IPA on the console (--ipa), set the speed at 120 words per minute (-s120; ...


17

IPA Reader is a nice front-end to Amazon's Polly service, configured specifically for IPA text. Click on the “Read” button for it to read the word “ad hoc” out loud.


14

Amazon Web Services' Polly text-to-speech service supports Speech Synthesis Markup Language (SSML) and specifically its <phoneme> element. You will need to create an AWS account, but you can then use the 'get started' demo to hear the speech of any (supported) SSML. The demo is here. For example, I put in the IPA from the Wikipedia article for the ...


9

You might try using mbrola. http://tcts.fpms.ac.be/synthesis/mbrola.html It's a downloadable speech synthesizer, where you type in a sound and it uses diphone synthesis to play it back. To play a given sound, you would type in something like this: _ 200 10 120 a 300 10 120 _ 200 10 120 The above text means, line by line "play silence for 200ms, and ...


7

There is a phoneme synthesis tool that runs entirely in your browser and converts IPA into sound that you can play in your browser or download as a WAV file. It requires no software to install, no notation to learn (beyond IPA itself), and works quickly and well. The program is Javascript and uses several of the tools described in other answers. In ...


6

You might like Eckher IPA to Speech. It supports multiple text-to-speech engines and comes with the IPA keyboard.


5

A colleague had such a device some years ago, but it depended on a touch interface that was a decade before there were touch interfaces. Theoretically it would not be hard to create such an interface, where a touch corresponds to F1/F2 values; then you simply synthesize the vowel with those formant values. Praat allows you to play around like that: see this ...


5

IBM Text to Speech also supports IPA per their docs here: https://console.bluemix.net/docs/services/text-to-speech/SPRs.html#ipa https://console.bluemix.net/docs/services/text-to-speech/SSML-elements.html#phoneme_element Here is an example of SSML in ipa to generate audio for the word tomato: <speak version="1.0"> <phoneme alphabet="ipa" ph="...


5

A lot of systems make little to no use of phonologically informed prosodic models. Some just can't, others are designed to and ostensibly do, and others have the potential to but don't. One widely used type of synthesis system involves the storage of hours of digitized human speech. In this type of system, which uses a method called unit selection, pieces ...


5

It seems to be a custom system developed by Google. Wikipedia lists it as the "Google pronunciation dictionary" scheme in its table of pronunciation respellings. In particular, Google spells the SQUARE vowel as ehr, which is not used in any other system in Wikipedia's table.


4

Here are some resources, in no particular order: The Wikipedia article on Speech Synthesis gives a pretty good overview on the topic, although it's a bit thin in its discussion of statistical parametric (i.e., HMM-based) synthesis. Dennis Klatt's History of Speech Synthesis gives example audio clips that exemplify the evolution of rule-based formant speech ...


4

Try ekho if you want to synthesize Mandarin and other Chinese dialects. It's open-source and hosted on github.


4

The rhythmic grouping of numbers is usually called "phrasing", e.g. "4-3 2-1-7 9-1-5-6". Within a "phrase", at least in English, there are still options regarding reading the numbers as individual digits vs. number-expressions (one-nine-eight versus one-ninety eight). I've never heard that sub-pattern be given a separate name.


3

I am not aware of any linguistic terminology for this particular kind of conventions. However, there is some applicable terminology from software engineering, particularly from the field of localisation: It some kind of locale. While locales typically apply to the visual formatting of dates, numbers etc. in terminal output, stretching it to spoken formatting ...


3

Unfortunately, any algorithm for determining the pronunciation from its spelling would be rife with special cases and exceptions. That's simply the nature of our spelling system. In the distant past (Old English) it was fully regular, but many, many things have changed since then. There are some general rules, certainly, which can be found in resources for ...


3

I'm pretty sure the short answer for this is: "there's no algorithm that will always agree with your judgements about what is/isn't a vowel/consonant (cf. glides) and will always agree with your judgements about where said phones start/end in the signal" BUT, a bit of quick Googling turned up the following, which may be useful if you're familiar with ...


3

York University (Ontario, Canada) has a site where you can hear all the IPA sounds: http://www.yorku.ca/earmstro/ipa/


3

No, lack of vowels or diacritic vowel marks is not a particular problem for text-to-speech systems, which usually make use of a dictionary of phonetic transcriptions anyway. Very few languages have a so tight mapping between writing and speech that correct pronounciation can be deduced by its written representation alone. Without a phonetic transcription ...


3

I found this: http://ipa-reader.xyz Multiple accents are offered. At least on my phone I note that mini IPA symbols are not displayed at all. But they work: they are pronounced even though they are not visible on the screen after being pasted in.


3

The AT&T text to speech wizard moved to www.wizzardsoftware.com/text-to-speech-sdk.php


2

Without question, it is an order of magnitude more difficult to implement speech recognition. You can do passable speech synthesis knowing nothing about context, grammar, etc. You will get some issues with homographs but other than that you will get recognizable speech. Now, it will be greatly lacking in prosody and stress, so you need to do some language ...


2

In a word, no. There are more organs involved in speech production than the ones you list. Bear in mind that "the throat" is not a single organ, but a set of organs and cavities. The ones involved in speech production include the true vocal folds, which are part of the larynx, which lies below the pharynx. The true folds phonate when they are brought ...


2

Nasals You may be able to approximate a nasal sound with closed nostrils, but if you pinch your nose shut and try to pronounce a nasal sound you will find you cannot do it. You can feel the air building up behind where you have pinched it shut. Also, the nasal cavity acts as a secondary resonator. It is involved in producing nasal vowels from e.g. French ...


2

Some additional resources to musicallinguist's answer: eSpeak -- This is a formant synthesis-based speech synthesizer along the lines of the Klatt synthesizer that supports a large number of languages. I also have a version at https://github.com/rhdunn/espeak that makes it easier to build the program and associated voice data on POSIX systems (Linux, Mac, ...


2

HMM-based systems and concatenative waveform-based unit-selection systems are similar in that they both involve the use of a large database of recorded human speech, and they both involve concatenating stored units to construct something that results in a whole output utterance. However, whereas waveform-based unit-selection systems store actual waveforms ...


2

If you know beforehand the content of the speech signal but you just don't know which parts of the waveform correspond to which parts of the utterance, then your problem is reduced to an alignment problem--much easier to deal with than a full-fledged speech recognition (plus alignment) problem. You already know what language it is, and you know what ...


2

It depends on which word, and what person. At the most basic level, you have learned the various rules of English (subjects before verbs, adjectives before nouns, and so on), and also the morphemes of the language, like "cat". When you learn "cat" you learn the pronunciation ([kæt]) and the real-world referent of the word. If you grew up speaking English, ...


2

I'm not familiar with this toolkit, but given your description my suggestion would be to check formant bandwidths. You are creating rounded vowels, and formant bandwidths are correlated with the degree of lip rounding (see Stevens 2000 p 292). Inaccurate formant bandwidths may lead to perceptual differences even if the formant positions are correct since ...


1

You can only synthesize actual sounds, not phonemes, which are abstract constructs in phonological analysis. IPA letters stand for a range of actual sounds which may be particular phonemes in a language, for example [a] or [ʕ]. So there isn't one correct pronunciation of these letters, there are very many. The reason why there is no program that tells you if ...


Only top voted, non community-wiki answers of a minimum length are eligible