# Release notes (/en/realtime-media/rtc-server-sdk/reference/release-notes/linux-cpp)

> For AI agents: see the complete documentation index at [llms.txt](/llms.txt).

This document tracks important changes and improvements to the Server Gateway.

    ## v4.4.32 [#v4432]

    Released on April 28, 2025.

    ### Improvements [#improvements]

    * Improved the concurrent performance of the SDK when sending and receiving PCM audio streams.

    ### Fixed issues [#fixed-issues]

    * Fixed a color cast issue that occurred in specific scenarios.
    * Fixed occasional app crashes and memory leaks.

    ## v4.4.30 [#v4430]

    Released on December 11, 2024.

    ### Compatibility changes [#compatibility-changes]

    #### Log encryption behavior changes [#log-encryption-behavior-changes]

    For security and performance reasons, as of this release, the SDK encrypts logs and no longer supports printing plaintext logs via the console.

    Refer to the following solutions for different needs:

    * If you need to know the API call status, please check the API logs and print the SDK callback logs yourself.
    * For any other special requirements, please contact technical support and provide the corresponding encrypted logs.

    ### New features [#new-features]

    #### Audio send buffer settings [#audio-send-buffer-settings]

    In use-cases such as AI interaction, the frequency of sending audio frames may be uneven. If the SDK sends to the Agora server according to the actual sending frequency, the receiver may experience issues such as stuttering or frame loss. To optimize audio transmission performance, this release adds methods to set the audio send buffer in the `ILocalAudioTrack` class:

    * `setMaxBufferedAudioFrameNumber`: Sets the maximum capacity of the audio send buffer.
    * `ClearSenderBuffer`: Clears the current audio send buffer, applicable to interrupt audio in human-machine dialogue use-cases.
    * `setAudioFrameSendDelayMs`: Sets the minimum data duration allowed to be sent by the audio send buffer, applicable to precise control of audio transmission delay.

    #### Audio labels [#audio-labels]

    This release introduces the audio labeling capability (`AudioLabel`). By adding audio labels, you can intuitively obtain audio information such as user audio energy, fundamental frequency, and voice type, making it convenient for post-processing. Contact technical support to enable and use this feature.

    #### 22050 Hz audio send sample rate [#22050-hz-audio-send-sample-rate]

    This release supports a 22050 Hz audio send sample rate, and you can set the sample rate when calling the `sendAudioPcmData` method to send raw audio data.

    #### SDK global custom parameters [#sdk-global-custom-parameters]

    This release adds the `getAgoraParameter` method in the `IAgoraService` class. The difference between this method and the `getAgoraParameter` method in the `IRtcConnection` class is as follows:

    * `IAgoraService::getAgoraParameter` is effective globally for the SDK.
    * `IRtcConnection::getAgoraParameter` is only effective for the current connection.

    ### Improvements [#improvements-1]

    This release reduces CPU consumption in multi-channel use-cases.

    ### Issues fixed [#issues-fixed]

    This release fixed issues caused by a null `channelId` pointer in the `onPlaybackAudioFrameBeforeMixing` callback.

    ## v4.2.32 [#v4232]

    Released on August 30, 2024.

    ### Improvements [#improvements-2]

    This release enhances the efficiency and performance of the FFmpeg library when decoding H.264 and H.265 video data, improving the user experience.

    ### Fixed issues [#fixed-issues-1]

    This release fixed the issue of socket file descriptor leakage, ensuring proper management of network connection resources.

    ## v4.2.30 [#v4230]

    Released on February 6, 2024.

    ### New features [#new-features-1]

    #### Multipath network transmission [#multipath-network-transmission]

    This release adds the multipath network transmission feature, which allows simultaneous transmission of audio and video streams over multiple network paths. For example, when you insert two IoT SIM cards in the device, the SDK can transmit audio and video streams through the network of both SIM cards, thereby improving the stability and reliability of audio and video transmission. Contact [support@agora.io](mailto\:support@agora.io) to enable and use this feature.

    #### Hardware encoding acceleration [#hardware-encoding-acceleration]

    This release supports utilizing NVIDIA GPU hardware acceleration for video encoding, which reduces the power consumption of CPU and improves hardware encoding efficiency and performance. Contact [support@agora.io](mailto\:support@agora.io) to enable and use this feature.

    ### Improvements [#improvements-3]

    #### Audio sending timestamp [#audio-sending-timestamp]

    To set the timestamp at which you send the audio data, this release adds the `presentation_ms` parameter in the `sendAudioPcmData` method. In use-cases such as video conferences and remote collaboration, this parameter allows you to achieve audio and video synchronization.

    ### Bug fixes [#bug-fixes]

    This release fixed the following issues:

    * App memory growth caused by receiving H.264 video raw data.
    * Garbled data stream when using both string-type user IDs and data stream functions simultaneously.