does TranslationRecognizer use "Speech to Text" or "Speech Translation" as per the pricing on the website.

Neil Payne 0

Hello,

Im using both TranslationRecognizer and SpeechRecognizer classes from the azure Microsoft.CognitiveServices.Speech sdk but unsure what costing bracket they fall under.

there are prices for "Speech to Text" or "Speech Translation" but cant quite determine which one exactly it will use.

For reference, and example,

A) im converting an audio file of spoken french to french text using SpeechRecognizer and RecognizeOnceAsync()

B) im converting an audio file of spoken french to english text using TranslationRecognizer and RecognizeOnceAsync()

Im not using continuous recognition, and im not automatically detecting the language, im specifiying up front.

Please could you confirm which price brackets they fall under.

Thank you!

Saideep Anchuri 325 Reputation points Microsoft Vendor

2024-09-27T17:36:51.8533333+00:00

Hi Neil Payne,

Welcome to the Microsoft Q&A Platform! Thank you for asking your question here.

The TranslationRecognizer class falls under Speech Translation, while the SpeechRecognizer class falls under Speech to Text. So, use the pricing for Speech Translation for TranslationRecognizer and Speech to Text for SpeechRecognizer.

For your examples:

A) SpeechRecognizer (converting spoken French to French text) falls under the "Speech to Text" pricing bracket.

B) TranslationRecognizer (converting spoken French to English text) falls under the "Speech Translation" pricing bracket.

So, you'll be charged according to the respective pricing models for each class used.

For the most accurate and up-to-date pricing details, refer to the Pricing Calculator | Microsoft Azure

For reference, please review the documentation: - Speech to text overview - Speech service - Azure AI services | Microsoft Learn

Speech translation overview - Speech service - Azure AI services | Microsoft Learn

Hope this helps. Do let us know if you any further queries.

If this answers your query, do click Accept Answer and Yes for was this answer helpful. And, if you have any further query do let us know.

Thank You.
Neil Payne 0 Reputation points

2024-09-30T08:59:54+00:00

accidental repost, please see my above comment to your answer
Neil Payne 0 Reputation points

2024-09-30T09:25:34.59+00:00

Thank you very much

What is confusing is there is not a pricing option for batch transcription for speech translation, there is only real-time speech translation

I assumed as i was using RecognizeOnceAsync() for both classes with a prerecorded audio file this would be using batch processing?

Edit: i see now batch transcription requires us to use the rest API, are these not available for the azure Microsoft.CognitiveServices.Speech sdk?

And please confirm im currently using "real-time transcription" for both classes.

Thank you!
Saideep Anchuri 325 Reputation points Microsoft Vendor

2024-10-01T15:25:01.73+00:00

Hi Neil Payne,

Following up to see if the given response was helpful.

Thank you.
Neil Payne 0 Reputation points

2024-10-02T12:07:26.96+00:00

Thank you the response was indeed helpful!
Saideep Anchuri 325 Reputation points Microsoft Vendor

2024-10-03T14:52:10.9566667+00:00

Hi Neil Payne,

I'm glad to hear that my response was helpful to you. And thanks for sharing the information, which might be beneficial to other community members reading this thread as solution. Since the Microsoft Q&A community has a policy that "The question author cannot accept their own answer. They can only accept answers by others ", so I'll convert the previous response to an answer in case you'd like to accept the answer. This will help other users who may have a similar query find the solution more easily. If you have any further questions or concerns, please don't hesitate to ask. We're always here to help.

Thank You.

1 answer

Saideep Anchuri 325 Reputation points Microsoft Vendor

2024-09-30T18:41:36.9733333+00:00

Hi Neil Payne,

For batch transcription and other, please refer to: Azure AI Speech pricing

No, using 'RecognizeOnceAsync()' with prerecorded audio files is not batch processing. It's designed for real-time, single-shot recognition of short audio clips. For longer files or large datasets, you should use Azure's batch transcription services.

Azure Speech batch transcription typically requires using the REST API, as the SDK primarily focuses on real-time speech recognition and single-shot methods. While the SDK provides functionality for real-time processing, batch transcription features are more efficiently handled through the REST API. If you need to perform batch transcription, you'll need to make API calls directly rather than relying solely on the SDK. Please refer to: What is batch transcription?

If you're using RecognizeOnceAsync() with both Azure SpeechRecognizer and TranslationRecognizer, you're currently implementing real-time transcription.

Please see: How to recognize speech.

I hope you understand!

Thank you.
Please sign in to rate this answer.

0 comments No comments
Sign in to comment

Use comments to ask for clarification, additional information, or improvements to the question.

Share via

does TranslationRecognizer use "Speech to Text" or "Speech Translation" as per the pricing on the website.

1 answer

Your answer