Need Spanish teams to record 445 sentences

I need 3 teams from France, each team could organize 100 or more people to record their voice. There are 455 sentences in a APP for each person to record. Only team apply for this job please.

The price is 35 US dollars for each person. Usually it will take 2 hours to finish the job.

Payment for the job finished in this week will be sent by the end the next week, when the quality is confirmed okay.

This job starts now. You need to send me samples first, so please record 50 sentences in the app and upload, we will check and make sure you and I have the same understanding about good quality. Then you could record in batches. I will send you the introduction of APP after you apply for this job.

Guideline

1. Speaker

1) Native Spanish speakers who live in Spain for more than 18 years. No immigrants. No people form another country besides Spain.

2) The speakers need to have a standard English accent.

2. Device for recording

An APP will be specified for recording and uploading. The sentences for recording will be automatically downloaded in app. You just need to read the existing sentence in the app.

3. Requirement of environment

In a very QUIET room, there should be no sound of another people's voice, or noises, No echo, no open sound.

It is better to record in bedroom or car.

Tips related to surrounding noise:

I. The voice of other people is not allowed.

II. The gasping sound of the recorder may be valid, but it is not valid when the sound is too loud.

III. Other invalid situation about noises. For example: the sound of falling things, closing the door, drinking water, laughter, turning over books, music, coughing, cars, etc.

4. Requirement of recording

1) Each person can only record one time. If one person records multiple copies of data, the corresponding remuneration will be cancelled. It must be the same recorder who completes all sentences of a corpus

2) The recorder must pronounce NATURALLY, FLUENTLY AND CLEARLY. It is forbidden to lower the voice, record in whispering, exaggerate and smile, change voice or imitate voice.

3) The recording speed is moderate, the words are clear and the speakers avoid stuttering. It can't be fast to be vague, or slow to stress every word.

4) When recording, keep 20-25cm distance between mouth and mobile phone. It is forbidden to stick the mouth on the microphone for recording, which will cause serious problems of wheat spraying and clipping, and lead to unqualified recording data.

5) When there are reading errors, coughing, stuttering, sudden noise and other phenomena, record again at an interval of at least 2S, and the recording device remains in recording state.

6) When recording, please follow the original sentences in app.

7) When you have a heavy nasal sound or a hoarse voice, please don’t record.

5. Examples of unqualified recording

1) When the accent is not Spanish, the entire output will be defined as invalid

2) Excessive noise, the second person's voice in recording, echo, or open voice are invalid

3) Incomplete recording of a sentence

4) In a sentence, people are far and near from the microphone, but not from the same distance from microphone

5) When registering in App before recording, the information about gender, age, native place, place of growth, dialect, or recording environment is incomplete or false

6) Multi word, missing word, wrong word, semantic confusion and inconsistency of corpus

Tips:

1) Once the text involves violence, politics, religion, pornography and other sensitive content, it can be ignored.

2) When there are unqualified sentences, you may record the second time.

When only a few sentences are unqualified, you only need to record the unqualified sentences again, which will appear in your app.

When the entire recording is invalid, you need to use new Number and record all the sentence again. The number will be assigned before recording.

6. For each recorder, an authorization agreement and information registration form shall be signed.

BPO Introducere date

ID Proiect: #23371303

Detalii despre proiect