Installation of Altered Studio Desktop for Windows
Follow the instructions in the Real-Time window which will prompt you to install Altered Studio desktop exe. Real-Time only runs on a Windows PC. You may need to accept some warning messages from Windows Defender or your Firewall as part of this process.
Once Altered Studio is installed, login with your Altered Studio credentials.
Select the Real-Time tab on from the left menu. You will be prompted to perform a Benchmark test which will notify you if your processor is fast enough to run the voice morphing in real-time
Installation of the Altered Virtual Microphone driver
When Altered Studio is running, you can access the Real-Time feature by clicking the Real-Time headset icon on the left side menu.
Before you can use the feature for the first time, you will need to install the windows driver for the virtual microphone. Follow the instructions at the bottom of the Real-Time screen to install the driver and allow any required access on Windows Defender or your Firewall if prompted.
To access Altered’s Real-Time Voice Changer in your gaming or meeting application of choice, select Altered Virtual Microphone in your Audio Settings.
Once selected, turn on Altered’s Real-Time by selecting a voice model and the On/Off button in the centre of the screen.
Model Selection
There are several models that you can choose from, one for each voice that is available to you plus the pass through model which will use your voice without any morphing applied.
You can use the Voice Enabled slider to speak in your normal voice without any morphing applied. Using this allows you to switch between normal and morphed voices without the need to change any microphone input settings on your gaming or meeting application, etc.
Each voice model has different speaking Styles available to it. The natural speaking style is the voice's normal sound, without any speaking Styles applied.
Each model will be installed by the system on the first time it is used. Simply select a model and click the On/Off button to activate the Real-Time synthesis.
Input / Output Settings
These settings determine how the virtual microphone operates, as follows:
Microphone for Morphing Input - allows you to choose which microphone the system should use for voice input. This is helpful when you have multiple microphones available on your system.
Virtual Microphone - allows you to enable or disable the virtual microphone. This should be set to enabled in order for the morphing to operate.
Headphones - allows you to choose a device for your morphed voice to be played so that you can hear yourself. It is recommended that the output goes to headphones or a speaker that is distant from the microphone to avoid feedback developing in the audio signal. Note there is a slight latency in the playback of the monitor output, dependent on your “Block Duration” settings.
The Microphone and Headphones allows you to adjust the input and output gain. You can use the Refresh Device button to refresh your devices while working in Altered Studio.
Processing Configuration Settings
These settings found in the gear icon allow you to determine how much system resources will be dedicated to the Real-Time morphing. We recommend adjusting these to find the best combination for your machine. In general terms, low block duration with higher thread limit and process priority will give the best synthesis results, however they may impact other areas of system performance in slower machines.
Block duration - defines the length of audio which Altered will morph at a time. Shorter blocks will reduce latency, however they increase demand on system resources and may result in lower quality morphs on slower systems.
CPU Threads - defines the maximum number of CPU threads that Altered can access to process the voice morphing. A higher number will result in better synthesis, however on slower systems with lower thread CPUs a high value here may impact on the available system resources for other applications.
Process priority - determines how the CPU should rank process requests from Altered Studio over other programs.
Voice Detection - determines how sensitive the system is to audio inputs for synthesis. If you are finding that background noise or breath are being morphed then this value should be increased. The best value will depend on your microphone sensitivity and level of background noise.
Auto Gain- Altered automatically adjusts the gain or volume of an audio signal to maintain a consistent and optimal level in Real-Time. You can manually adjust the gain in the Microphone and Headphone settings.
Real-time Morphing Monitors
Once the Real-Time morphing is active the system displays live metrics to indicate performance.
Realtime Factor - shows you how quickly the voice morphing is happening in real-time. This value must be above 1 as anything below 1 indicates that the synthesis is happening in less than real-time. The Processing Configuration Settings can be adjusted to increase the amount of system resources available to Altered Studio to increase the Realtime Factor. Similarly on slower machines it may be necessary to close other applications to free up system resources.
Using the Real-Time Voice Morphing
Once you have adjusted your settings and chosen a voice model, press the blue start button to activate the system.
You will need to choose the Altered Virtual Microphone in the program which you want the voice changer to be heard. For example if you are using the voice changer on a Zoom Virtual Meeting, you should start the RT in Altered Studio, and then change your microphone settings in Zoom so that the active microphone is the Altered Virtual Microphone.
If you want to change back to your normal voice, you can change the voice model to the Passthrough model in Altered Studio, or alternatively change the microphone setting in the output program (e.g. Zoom) back to your physical microphone.
Certain settings will automatically restart the Real-Time effect. When this is required you will see the notification “Changes applied: restarting engine”. For example, you can change between speaking styles on a model without needing to restart, however changing from one voice model to another will restart the Real-Time effect.