Microsoft Neural Voices on Speech Central

Speech Central has added the support for Microsoft Neural Voices in the latest version. Those are high quality voices provided online through Microsoft's servers.

There are several apps that provide high quality server side voices like Speechify and Natural Reader, but Speech Central does it differently. Most notably Speech Central doesn't sell those voices to end users or have some fees from their sales which leads to significant benefits:
- There is a free tier provided by Microsoft. As services like Speechify have a cap on how much you can read with high quality voices it turns out that at the moment you get for free in Speech Central what you would pay 139$/year in Speechify.
- If you need more than that cap you can go to a paid tier and pay per consumption without any cap (some guideline is that you can expect around 1$ hour of listening at 1x speed).
- As you are buying directly from Microsoft paid tier is significantly cheaper per hour. You don't have to purchase over the App Store and thus there is no Apple's 15% fee that at the end has to be calculated into the price, and Speech Central takes no profit either.

There are downsides of this approach too:
- it is surely easier to activate the subscription than to register an account on Microsoft Azure and get API key, though I don't think that this is too complicated either and few users have gone through that already.
- if you are using the paid tier you are technically exposed to pay as much as you consume. While I don't expect any unpleasant results it is suggested for everyone to track their cost through the portal at least on the start to avoid such surprises.

Here is the copy of the help article from https://speechcentral.net/ios-help/#azure :

1. Create the Azure account and subscription (note: currently there is a version with free limited tier and paid version) from here.
2. Create the cognitive resource from here
3. You should be able to see the resource in the Azure Portal. When you open it in the left sidebar there is the option ‘Keys and Endpoints’ that you can use to create and get data about API key.

Please note that:

- Speech Central is in no way affiliated with Microsoft. Your relationship with Microsoft is in no way connected to your relationship with Speech Central as Speech Central doesn’t take any fee from your payment to Microsoft.
- Microsoft paid tier is fairly expensive (the price at the moment of writing of this article is 16$/million characters which means that listening to larger books would cost 10$ or more). We have picked to support Microsoft as all other providers (Google, Amazon, IBM) charge the same amount but don’t have a free tier at the moment.
- If you use their paid tier it is strongly suggested for you to track your costs in their portal and avoid unpleasant surprises (there is a page ‘Cost Analysis’ that can be activated from the option right below ‘Keys and Endpoints’ that you have used in the previous step).

To activate those voices go to Extras tab > Settings > Voice > Microsoft Neural Voices (activated from the menu at the top left of the screen).