[Generative AI] DeepVoice - Text To Voice

[Announcement]

Automatic Quota Reset to 60,000 characters. [Quota changes]

60,000 characters allotted for the period 1-03-2024 to 31-03-2024.

 

Is there any way I can get that voice back? I have been making English pronunciation program and I need them desperately.

 

Hello,
Any updaes on v2.2 mentioned in reviews?

 

Please bring back the old Lily. We love old Lily. Create new one for New Lily. How can I trust on the consistency of the voice if the voice has be changed without any notice! Give Us the old Lily back please.

 

Hello Micho_Dev,

We have added the old Lily. Please find it in the next update.

 

Hi,

what should i do if i get this error anytime i try to import Deepvoice?

Assets\DeepVoice\Editor\Scripts\DeepVoiceEditor.cs(7,13): error CS0234: The type or namespace name 'EditorCoroutines' does not exist in the namespace 'Unity' (are you missing an assembly reference?)

 

Good day, I verified my invoice number but am still unable to generate any audio. The Generate button is missing, actually. Am I missing a step?

 

Hi PlasmaFox,

Please import the EditorCoroitines package from the package manager (Window > Package Manager > Unity Registry > Editor Coroutines). You should be able to get it working then.

 

Hi theunigned,

This is quite strange. If the generate button is missing, there can be a variety of reasons.
If you are in the demo scene, then zoom out and fit the screen to 16:9 ratio. The generate button should be present at the bottom. If you are instead in the editor window, which we recommend, please scroll down on it.
Please also ensure there are no errors in your project from any other assets that hinder the compilation of the editor window. Please let us know what you see at info@aikodex.com

 

[Announcement]

Automatic Quota Reset to 60,000 characters. [Quota changes]

60,000 characters allotted for the period 1-04-2024 to 30-04-2024.

 

did you change lucas voice? if yes im screwed thanks, its impossible to works with this the voice change everytime

 

Hello KenzoGames38,

We have not made any significant changes to the voice data, if you feel there is a discrepancy in the voice of Lucas, please let us know at info@aikodex.com so we can examine the voice clips and add the voice in the next update if that is possible. Please rest assured you will still be able to use the voice you’ve generated as we can reconstruct and fine tune voices.

 

Thank you!

 

I can't seem to get DeepVoice to verify my invoice number any more, in any project that I try in in both Unity 2023.2 and 2022.3 across 2 seperate computers all I get is an HTTP/1.1 429 Too Many Requests error and it's been nearly an entire day. What can I do?

 

Hello BDBlaffy,

We’ve sent you an email on how you can fix this. Please let us know if it works.

 

Waiting a full 24 hours did indeed work, all is good now, thank you!

 

[Announcement]

Automatic Quota Reset to 60,000 characters. [Quota changes]

60,000 characters allotted for the period 1-05-2024 to 31-05-2024.

 

Hi @AiKodex I have a question about the 60,000 limit.
Will this limit always be in place?
Is DeepVoice AI more suitable for 'voicing' a text project in parts over several months?
Is DeepVoice AI not suitable for use in live streaming text-to-speech applications (because the 60,000 limit would be quickly exhausted)?

 

And another question. If I purchase VoiceGPT together with DeepVoice AI, will the character count per month be 560,000? As stated in the line 'Also can be used to extend the character count of DeepVoice up to 560,000 char/mon'?
What is the key difference between DeepVoice AI and VoiceGPT?

 

Hi @AiKodex ,
Could you change the way Deep Voice_Multi voices are handled, so that we could specify the language we want to use before the sound is generated?

I have small text strings like "Jupiter" or "V1400 Centauri" which don't contain any hint of the language I want it pronounced in. I'm generating voices in Englsih and French : for french to work correctly, I need to prepend my text with "En français :" (meaning "In French") so that it gets the pronounciation right.
And then I have to trim each generated clip one by one, which is kind of annoying

Apart from that, great asset overall, love using it!

 

Hello Orexx,

DeepVoice was built keeping in mind the need for developers to voice their characters while working on their game rather than itself being a feature that developers could add to their game.

For a voice model of this quality, it would require several gigabytes of space and memory alone locally to be bundled with the additional game files.

We have built a smaller English-Only asset called VoiceGPT for this purpose. If you are interested, please feel free to check it out.

 

The limit for DeepVoice will not increase, only that VoiceGPT + DeepVoice will give you 560,000 characters (500k (VoiceGPT with DeepVoice voice clone) + 60k (DeepVoice voices)). Apologies for any confusion.

 

Hello PolarEclipse,

We understand your issue, and this is a unique use case that we admit that we did not think of. Currently, this is not possible due to the way the model is trained. It makes sense to have a language field - hence we created another model for this exact purpose (VoiceGPT).

We can however offer some advice in regards to your use case. You could try and add all the text you want in one go, for example: "En français : Jupiter - V1400 Centauri - Saturn - Sirius - Neptune - Kepler 186f...". You will not only get all of the voices in French but will also have the same tone and tempo.

 

[Announcement]
It seems that the server is experiencing downtime at the moment. We are working on fixing it. We will keep you updated here.

Edit: UTC: 07:25: The service should be functional now.

 

Hello,
I can't verify the invoice number anymore today.

I got this message:

Code (CSharp):

1. Request timeout
2. UnityEngine.Debug:Log (object)
3. AiKodexDeepVoice.DeepVoiceEditor/<Verify>d__81:MoveNext () (at Assets/DeepVoice/Editor/Scripts/DeepVoiceEditor.cs:1291)
4. Unity.EditorCoroutines.Editor.EditorCoroutine/YieldProcessor:MoveNext (System.Collections.IEnumerator) (at ./Library/PackageCache/com.unity.editorcoroutines@1.0.0/Editor/EditorCoroutine.cs:80)
5. Unity.EditorCoroutines.Editor.EditorCoroutine:ProcessIEnumeratorRecursive (System.Collections.IEnumerator) (at ./Library/PackageCache/com.unity.editorcoroutines@1.0.0/Editor/EditorCoroutine.cs:134)
6. Unity.EditorCoroutines.Editor.EditorCoroutine:MoveNext () (at ./Library/PackageCache/com.unity.editorcoroutines@1.0.0/Editor/EditorCoroutine.cs:115)
7. UnityEditor.EditorApplication:Internal_CallUpdateFunctions ()

 

Today I started to get this error when creating audio clips.

Code (CSharp):

1. InvalidCastException: Specified cast is not valid.
2. AiKodexDeepVoice.DeepVoiceEditor+<Post>d__79.MoveNext () (at Assets/DeepVoice/Editor/Scripts/DeepVoiceEditor.cs:1241)

I am getting the .mp3 file returned from the server, but it appears like maybe its corrupted because the cast to the AudioClip is failing.

When I check the .mps file, I see this error:

Code (CSharp):

1. Errors during import of AudioClip Assets/DeepVoice/Voices/Lotte_Voice_3.mp3:
2. FSBTool ERROR: Internal error from FMOD sub-system.
3. FSBTool ERROR: Internal error from FMOD sub-system.

I updated to latest version, which enabled me to get a couple of generations completed before it started erroring again.

Any ideas?

 

[Announcement]

Automatic Quota Reset to 60,000 characters. [Quota changes]

60,000 characters allotted for the period 01-06-2024 to 30-06-2024.

 

Hello McBauer,

Some of the users have had issues with the FSBTool. Please try resetting audio settings in Unity / open a fresh project without any audio assets that modify the FSBTool. This is a bug in some 2022 and 2023+ versions. In this case, we just email them with a custom compiled FSBTool of another Unity Version which works for them.

 

Hello BenoitFreslon,

This is quite strange. Please could you let us know if you are still experiencing the error? Please could you send us an email specifying what date and time you experienced this issue?

 

Hi, I see Voice GPT is released. Do I have to pay another $50 to get new fantasy voice? There will be no fantasy voices for DeepVoice?

 

Hello lofwyre,

VoiceGPT will soon be a local TTS model inside Unity with a single-shot voice cloning ability. If you think you would prefer to have an unlimited offline model in Unity, then you should pick VoiceGPT, else, you can go with DeepVoice.

 

I purchased DeepVoice some time ago. My question was, will it be getting updated fantasy voices or are the fantasy voices only going to be in VoiceGPT?

 

Hello Iofwyre,

DeepVoice was built with highly realistic speech output in mind. So fantasy voices are quite difficult to get as an output. VoiceGPT, on the other hand, has a wider range of control which can create some fantasy sounding voices.

 

Hello. I am planning to buy this asset and would like to know when version 2.2 is scheduled to be released?

 

Hello! I ordered your product yesterday and it's great - however, I need help with its runtime functionality. Do you have any example projects or scripts that show how to set DeepVoice up such that it reads text on a Canvas? I can't figure out the Example scene.

 

Hello Xinnys,

For now, we are working on a new update release of VoiceGPT because of the high demand. DeepVoice still remains our primary area of focus and we are planning to add new voices and improve the GUI for trimming as these are the two most requested updates. Is there anything else you'd like to add to the update roster?

 

Hello Blinxer,

We have sent you an email with detailed instructions on how you can use to generate audio in runtime and build.

Basically, you can take a closer look at the "Generate()" function which is the most important when it comes to generating the audio file. You then need to perform CRUD on the platform you want to build on. You can modify it to suit your needs whether you are building a game or automated voice over clips.

 

Hi. Thank you for your reply. Actually I'm mainly interested in non-verbal sounds functionality to make dialogues more various, simillar to what we have in VoiceGPT. I noticed that you mentioned before that this will be available in 2.2, so I thought I'd clarify

 

This plugin isnt working for me... there's no sound at all. None of the preview voices make any sound, and nothing I generate can be previewed either... its just silent. I really need to hear the sample voices so I know which one to use!

All other sound works fine in the unity project.

 

...and when I try to generate anything I get - There was an error in generating the voice. Please check your invoice/order number and try again or check the documentation for more information.

But my invoice is fine as that gets me - Your invoice is verified. Thank you for choosing DeepVoice!

 

Hello Mulbin,

It may be possible that the DeepVoice folder is not in the root directory. This can cause some issues like the ones you’re mentioning. If the folder is in the root directory, there may be an issue with the FSBTool / Unity Internal Audio settings. Please could you email us with a screenshot of the voice Preview section at info@aikodex.com so we can suggest fixes.

 

I like this software/asset, I've been testing it for a while now and got really good results, however I'd like a true Spanish voice accent, because when I try to generate a Spanish voice i always get this kind of latin spanish accent, and that's not what i was going for.

Also, usually AIs have a "seed" option, this could help if you run out of characters in a single prompt for the voice to be consistent with eachother, or if you need to update it in the future by adding extra lines.

I'll link this as a reference:

At 1:00 it has the difference written which basically are the "Ustedes" vs "Vosotros" and the pronounciation differences of the letters C, S, Z, J and G. Also the expressions used and the accent are quite different. There are a few teachers joining the video from Spain, Argentina, Venezuela and México, so that could help noticing the differences too. The examples start at around 3:15 in that video.
Just trying to provide constructive criticism with examples. I hope this helps!