Audacity truncates audio of 24 bit into 16 bit format.

[Non-$ensoi]

I have just started using Audacity for recording audio. I use ‘Stereo 48 kHz 24 bit PCM’ mode. I want to save the file in uncompressed format of 24 bit audio. To achieve this I save my audio in Wav, Signed 24 bit PCM format. But once processed, I see that the saved file is in 16 bit format. Where do the 8 bits go missing? Can anyone help m saving the audio in perfect 24 bit format?

[Bontu]

To interact with in/out of computers, Audacity uses Portaudio module. In Windows, the version of PortAudio that Audacity supports MME, DirectSound & ASIO. On Windows & Vista, emulated version of MME & DirectSound is used. They access audio hardware indirectly only through WASAPI. Due to this the 24 bit format for audio is not supported on Windows 7 & Vista. Audacity’s newer version might consist of newer version of Portaudio. The current available updates for Portaudio on internet are unstable versions.

[JyotiT]

PortAudio is a cross platform open source audio I/O library. It provides a simple API for recording and for playback of sound and audio files. It is programmed to play sine waves and process input and output of audio devices. It is meant for transfer of audio files from one device to another. It ignores the platform on which it is being used but only checks for availability of Audacity.

[Alfons]

Audacity ignores the quality of hardware available in system and converts the sound in 16 bit format. Even in best audio converters, there is a loss of bit size leading to truncation of 24 bit audio. Even after using high level of equipment and devices, chances of noise level to be converted into 1 bit format is obvious. The only advantage of having a 24 bit format over 16 bit format is that it provides more headroom. Even if you try to convert 16 bit audio into higher version, the closest that will come is 32 bit format.

[Hollyn]

I see what is happening in your case, I am using Windows XP and also I don’t have a 24 bit sound card. As we see that Audacity will automatically dither the quality of Audio when exporting it. To disable this dithering you can go in preferences. In preferences, find Quality tab. Here in the Quality tab, find the ‘Dither’ and disable it. This will work best in high quality conversion settings. Even after disabling Dither, all the mid values won’t be converted into integer. So when exporting the final audio file, the quality will again is downsampled.

[Koman]

You can also try ‘Real Time conversion’ settings. But you will notice a difference in playback. Here too keep the ‘Dither’ settings on off. For testing of exported 24 bit format audio file you can try settings of Dither at Triangle for best quality. Other shapes might even work depending on type of audio.

[GaganjyotTechie]

To compensate for the loss of bits, you can apply various effects like Normalize, Amplify, Fade & Noise Removal. It should not include that edits, cuts or trims the samples. Using this will alter the sample values. Many times a randomized noise will be added to the audio. At this point keeping the dither value on can add this noise of up to -148 dBFS RMS (A-weighted).

[Erakna]

Using ‘real Time conversion’ will lead to consumption of large amount of hard disk space. To avoid this you can save the audio in .flac format. It works similar to .wav format but consumes less amount of hard disk space while maintaining the quality of audio. You can try this option while keeping Dither option on Triangle or some other shape.

[BillieJ]

I have used Audacity on Windows as well on Linux. The Linux version of Audacity uses ASIO SDK instead of Portaudio as in Windows. I have tested the quality of sound generated by using Audacity on Windows and that on Linux. I personally liked the quality available on Linux.