IBM Watson Text to Speech is an API cloud service that enables users to convert written text into natural-sounding audio in a variety of languages and voices within an existing application or within Watson Assistant. It can be used to give a brand a voice and interact with users in their native language. Increase accessibility for users with different abilities, provide audio options to avoid distracted driving, or automate customer service interactions to eliminate hold times.
N/A
Phrase
Score8.6 out of 10
Small Businesses (1-50 employees)
Phrase is a Language Intelligence provider. Its enterprise platform automates, manages, and delivers multilingual content. Global brands use Phrase across hundreds of languages to reduce time to market and deliver consistent brand experiences worldwide. The Phrase Platform brings together translation management, software localization, multimedia localization, machine translation, workflow automation, and language AI in a single integrated environment. From marketing…
I advise Watson for all scenarios where you need AI to speak. I don't use this for this goal, but I suppose that could be a good solution for tourism and travel: companies in the hospitality industry can make it easier for people to get around and offer tours in numerous languages, all at the same time. In telecommunications, could be used to create customized messaging that the caller can use with customers, and it can generate words from a customer’s records that are read to them in a professional and friendly voice. It's very good for English, for Italian is a little "robotic" but the pronunciation is right.
A de minimis incentive was given to thank the reviewer for their time. The incentive was not used to bias or drive a particular response, nor was the incentive contingent on a positive endorsement. More Info
The system is well suited if you also run Trados or any second system. As the system is very expensive and the use of MTUs is rather complicated to understand the best thing is to run most larger projects through Trados or any other CAT tool and then to import it. Large projects are expensive in Phrase Localization Platform so I would recommend it if you have small projects and have enough capacity to carefully check the terminology as the data pulled from the term base is inconsistent. The same applies to the find replace feature which can work great.
A de minimis incentive was given to thank the reviewer for their time. The incentive was not used to bias or drive a particular response, nor was the incentive contingent on a positive endorsement. More Info
A de minimis incentive was given to thank the reviewer for their time. The incentive was not used to bias or drive a particular response, nor was the incentive contingent on a positive endorsement. More Info
In the case of EN to JA translation, we enter text and then convert it to get the correct final text (word or phrase) which uses correct Chinese character(s), as there are often multiple Chinese characters with the same reading but different meanings. When those conversion options are displayed, it is usually possible to change our selection among them by hitting the Tab key, but in Phrase, hitting the Tab key makes us leave the text conversion and move to the source segment, so we must make sure to use arrow keys to choose the right conversion option when using Phrase.
When there are tag elements used in the source text, the tags must exist in the target text of course, but the tag order also must be the same. The tags cannot be moved around in the segment, which causes problems in the case of the Japanese language, because the word order differs between EN and JA.
In the Japanese language, Italics are basically not used, so there must be no texts between the tags which specify Italic font. But then the segment cannot be confirmed and the job status cannot be changed to "Complete".
I use and manage an Academic edition (approx. 15 students) and the corporate license (10 users). In both cases, the features are easy to find, set and use. My students and my translators understand the dynamics of the platform easily and get used to them quickly.
It is designed to be a lean system, but that means that things are clustered in weird groups and a lot of trial and error is involved getting the settings right. If something goes wrong first level support is very quick and has very fast answers which are quite typical of first level support and essentially equate to asking you if you have switched it off and on yet. Anything more involved is challenging for support to address.
We rarely use support, but most questions were answered in a timely fashion, although we didn't exactly find them satisfactory. That's mostly the fault of the software and not the Support team because we asked for things that Phrase couldn't do.
A de minimis incentive was given to thank the reviewer for their time. The incentive was not used to bias or drive a particular response, nor was the incentive contingent on a positive endorsement. More Info
I've never been a fan of the robotic voice of text-to-speech tools and I've always thought they sounded flat and bland. Until I tried IBM Watson Text to Speech, that is. This service uses advanced deep learning techniques to synthesize speech output in natural-sounding voices. I was blown away by how good the voices were that I've been using it for personal stuff like journaling and brainstorming sessions. And I'm definitely going to be using this for podcasting and my YouTube channel.
A de minimis incentive was given to thank the reviewer for their time. The incentive was not used to bias or drive a particular response, nor was the incentive contingent on a positive endorsement. More Info
Phrase has a very good user interface that people can quickly learn and start using. Also, its analysis features are top class which helps me provide a detailed estimate to my client based on the file particulars. It has a QA feature that helps me do a very high-quality check before I send out my translation to the client.
A de minimis incentive was given to thank the reviewer for their time. The incentive was not used to bias or drive a particular response, nor was the incentive contingent on a positive endorsement. More Info
