When you want to upload a recording, you can do this two ways:
- Drop the file for the recording, inside the dropzone where it says “Drop an audio file here”. The dropzone is marked by a dotted line, which will turn green when holding a file over it.
- Click the button saying “Choose a file”, and find the file inside the popup window.
You can see a list of supported file types in Info about Mellonn Speak.
When a file has been chosen, the Upload Recording box will expand, with options of giving the recording a title and description. These fields are required to be filled, and the title can be a maximum of 16 characters.
You will also be prompted to give the number of participants and the language spoken in the interview.
- Note: The number of participants, is the amount of people speaking in the interview, not the amount of people in the room. An inaccurate amount of people, will most likely deliver an inaccurate result, the amount cannot be changes after uploading the recording.
- There are 34 different languages/dialects to choose from, it is strongly recommended to choose the one best representing what is spoken in the recording.
When you’ve filled in the information about the recording, you can press “Continue to payment” (this button will say “Upload recording”, if you don’t have to pay for it).
Now you will be presented with the option to pay for the recording, if you already have a payment method associated with your account, you will be able to pay quickly. Otherwise you will have to enter the card details, you can choose to save the payment method for future purchases, this is recommended.
- Note: The payment methods are NOT stored or handled by Mellonn, they’re stored and handled by Stripe.
When you’ve payed for the recording, the upload will start automatically, indicated by a circle showing the upload progress (this can take some time, depending on the internet speed and audio file size). Once the upload is done, you will be taken back to the main page, where the recording should appear with a loading circle. When the AI is done transcribing the recording, the loading circle will disappear (You may need to reload the ).
When the AI is done transcribing
When the transcribing job is done, there will be a checkmark next to the title, on the recording page. When the job is done, you can press the recording to open it, it will take some time downloading the result, depending on the length of the recording.
- If this is the first time opening the transcription, you will be prompted to label the participants. This can be done by listening to short clips from the recording, to make sure your labels are correct. This can also be changed later.
- You will now be presented a chat inspired page, where you can read the transcription.
- Underneath each chat bubble, there will be a description of it, with information about the speaker and what time interval of the recording it is.
- You can press a chat bubble to place your cursor inside it, when this is done the time interval will be chosen in the audio player. To reset this selection, you can press the button “Reset audio selection” in the right window.
- In the orange box with the title, there's a three dot menu, this menu gives you the following options:
- "Edit labels": edit the labels associated with the participants.
- "Download DOCX": save the recording as a DOCX file (Word), on your device.
- “Version history”: look at the 10 previous changes made to the transcription.
- "Info": shows info about your recording, including title, description, file name and number of speakers (participants).
- "Delete this recording": this will delete the recording, this CANNOT be undone.