Open Ou-Jie opened 3 years ago
the result i have tested is right, you can You can compare the result with librosa.feature.melspectrogram in python
Hello xu, the features I extracted with this code at that time were aligned with the features extracted by python, and the difference between the data was very small. I have forgotten the specific python version,But I don't think this has much to do with the python version.
发件人: Yu @.> 发送时间: 2022年3月31日 5:55 收件人: ewan-xu/LibrosaCpp @.> 抄送: Ou-Jie @.>; Author @.> 主题: Re: [ewan-xu/LibrosaCpp] Can the extracted by python be aligned with the c code? (#4)
@ewan-xuhttps://github.com/ewan-xu may i know which version could i use to compare the result? it seems like some of my result may not exactly same as the current version 0.9.1
― Reply to this email directly, view it on GitHubhttps://github.com/ewan-xu/LibrosaCpp/issues/4#issuecomment-1083664217, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ASEBARHHDERLFNAJ2UEYQJDVCTETXANCNFSM45BTUZGA. You are receiving this because you authored the thread.Message ID: @.***>
Hello author, I recently tried to extract log-mel features with your code, but found that the actual value is not the same as the value of the c code you provided. Did you get the same result?