THREE-BODY TECH / TUBE BELL download delivery

This EQ plug-in precisely processes the midrange to enhance the clarity and presence of instruments. Available for download.

■Plugins
■EQ
====================================
■Download Delivery
- Approximately 10 minutes after completing your purchase, the activation code will be delivered to your registered email address.
- Activate the code you received on the D2R page below and you will receive product installation information via email. (You can also view the information by clicking the email icon on the D2R page.)
D2R Page
*If you do not have a D2R Page account, please register by clicking ""Register here"" on the D2R Page.

- You can also check your D2R registration activation code by clicking ""Order Details"" in your Sound House My Page or Purchase History.
*It may take a few minutes for the information to be reflected in your purchase history.
- Click ""Support Information"" under ""Order Details"" to display activation information.
- For detailed installation/usage instructions, please contact the manufacturer.
====================================

""TUBE BELL"" is an EQ plug-in that primarily processes the mid-range, complementing the functionality of ""TUBE SHELF."" It features two peak sections and one dip section, all of which use bell-shaped filters.

AI Hardware Simulation Plug-in Series: ""Deep Vintage""
The Deep Vintage series is a group of plug-ins that harness the power of AI to simulate authentic hardware. These machines are capable of withstanding the Turing test of the human ear, faithfully capturing the true ""soul"" of vintage equipment.

Three-Body Tech's proprietary APNN (Audio Processing Neural Network) machine learning technology, specialized for simulating analog effect processors, ensures a listening experience that most closely resembles legendary hardware models while maintaining fully real-time processing.

APNN (Audio Processing Neural Network) 2.0
Three-Body Tech has experimented with numerous emulation techniques, including physical modeling and convolution. Regardless of the technology used, THREE-BODY TECH's ultimate goal remains consistent: to reproduce the sound of real-world devices as faithfully as possible.

Trained by THREE-BODY TECH's proprietary APNN (Audio Processing Neural Network) 2.0, Deep Vintage delivers a listening experience indistinguishable from the original hardware.

APNN is a neural network optimized for audio processing, consisting of a large number of audio processing modules (EQ, compressor, overdrive, etc.). During the training process, the same audio is input to both the APNN 2.0 and the hardware, and APNN 2.0 learns how the hardware alters the input signal in both waveform and spectral dimensions. To capture the essence of the modeled hardware, it automatically adjusts its structure and parameters until the difference between the hardware and APNN 2.0 outputs is small. In other words, a fully trained APNN 2.0 instance is able to capture both the dynamic and tonal characteristics of the source hardware.

In the end, the APNN achieves phase cancellation with an error of approximately -40dB to -75dB (depending on the hardware model). This exceptional level of error control, even exceeding the variability between different manufacturing lots of the same hardware model, allows us to confidently state that the APNN can ""fool"" the human ear.

After the APNN 2.0 instance completed training, THREE-BODY TECH conducted rigorous human testing and continued tuning until all APNN 2.0 instances passed the ABX test. Nothing in your DAW is as close to real hardware as a plug-in equipped with APNN 2.0.

■Features
Drive
Adjusts the saturation caused by the drive effect. If you only want to use the equalizer function, disabling Drive will also disable Output Gain.

Low Mid Peak
This section adjusts the frequency range and gain level of the low-mid peak.

Mid Dip
This section adjusts the frequency range and gain level of the mid-dip.

High Mid Peak
This section adjusts the frequency range and gain level of the high-mid peak.

Dual Switch (Link Mode)
Unlike other Deep Vintage plug-ins, the Tube model's link mode switch is located as a ""Dual"" button. When disabled, it operates like other plug-ins with L/R link mode enabled. When enabled, the GUI expands to two panels, allowing you to process the left and right channels independently. In M/S mode, adjustments for the Mid and Side channels are linked/unlinked.

Other Features
- Switchable Noise Simulation
Real hardware is inherently noisy. In fact, some vintage gear produces very high levels of noise. When an APNN trains from a hardware model, the network cannot be instructed to distinguish between sound and noise during the training process, so it captures all sound, including noise. Fortunately, APNNs are not completely black boxes, and you can control whether or not to generate noise after training. You can also disable/enable noise simulation in each DEEP VINTAGE product according to your preference. * Green AD AI does not have a 'Noise' button; this is because the noise level of the hardware models is low, making it unnecessary to disable it.
- Harmonics
Generates a certain amount of harmonics based on the modeled hardware. The Drive knob controls saturation by adjusting the input gain, while Harmonics allows you to directly edit the network model, forcing a reduction in the amount of generated harmonics. At 100% (default), the harmonic characteristics are closer to the hardware model, and lowering the value reduces the amount of harmonics proportionally (though due to the nature of the network architecture, harmonics cannot be completely bypassed). This allows the plug-in to operate in a cleaner state, preserving mostly only the frequency response and dynamics.
- Oversampling
While Deep Vintage leverages the power of AI to approximate the hardware models that inspired it, digital audio cannot escape the ""curse"" of aliasing. Enabling 2x, 4x, 8x, or 16x oversampling can significantly reduce aliasing.
- Undo/Redo
- A/B Switching
- L/R/M/S Channel Monitoring
You can solo and monitor any of the left, right, middle, or side channels.
- Input/Output Level Meters
- Mono Mode
- LR/MS Processing
- Phase Inversion
- GUI Rescaling

Latency
Approximately 0.6 ms at 26 sample points and 44100 Hz

■ System Requirements
For the latest system requirements, please refer to the manufacturer's product page.