I'm having issues with my project at 96k sounding slightly different from my LV formatted submission. A few of my words that have tiny differences like precisely ending up sounding more like precise. So far I've only tried Audacity to convert.
Has anyone found a particularly good program/tool for downsampling and getting LV format?
Thank you,
Nodoka
Downsampling from 96k
I presume you're recording using a sampling rate of 96k.
What process of 'downsampling' are you using? In Audacity, are you using Tracks >> Resample... and entering 44100? That uses the same method as any other program, and it should be a transparent conversion. It shouldn't be cutting out parts of words.
Librivox also requires you to convert your audio to MP3 format (at 128k bitrate and 44.1k sampling rate). That compresses the audio to a smaller file size, and it will slightly lower the audio quality - though not so as to truncate words.
By the way, there's no advantage to recording in 96k, despite what some people claim - you just end up with uncompressed audio files that are more than twice the size. Most people here use Audacity. The built-in LAME MP3 converter works very well, so there aren't other encoders that are 'better'.
What process of 'downsampling' are you using? In Audacity, are you using Tracks >> Resample... and entering 44100? That uses the same method as any other program, and it should be a transparent conversion. It shouldn't be cutting out parts of words.
Librivox also requires you to convert your audio to MP3 format (at 128k bitrate and 44.1k sampling rate). That compresses the audio to a smaller file size, and it will slightly lower the audio quality - though not so as to truncate words.
By the way, there's no advantage to recording in 96k, despite what some people claim - you just end up with uncompressed audio files that are more than twice the size. Most people here use Audacity. The built-in LAME MP3 converter works very well, so there aren't other encoders that are 'better'.
Yes on recordingGrayHouse wrote: ↑June 25th, 2023, 3:56 am I presume you're recording using a sampling rate of 96k.
What process of 'downsampling' are you using? In Audacity, are you using Tracks >> Resample... and entering 44100? That uses the same method as any other program, and it should be a transparent conversion. It shouldn't be cutting out parts of words.
That's definitely not what I was doing. I just picked the export function at 44100, constant 128 bit rate, mono. I'll try that and report back.
Thank you,
Nodoka