TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Looking for the Slow Build in the Million Song Dataset

38 点作者 utcursch超过 13 年前

4 条评论

omn1超过 13 年前
Interesting approach but a lot of songs I would consider falling into this category score very low:<p>Here are a few examples from the original reddit post[1]:<p><pre><code> Song (Correlation) The Velvet Underground - Heroin (2.95) The Beatles - A Day In The Life (0.00) </code></pre> Other songs with low correlation:<p><pre><code> Lynyrd Skynyrd - Free Bird (0.38) Metallica - Hero Of The Day (3.72), One(3.47) </code></pre> I think the reason is that tempo and compression (which I associate with tension) are not taken into account but play a big part in our hearing experience. This data may be harder to extract, though. Anyway interesting work!<p>[1]: <a href="http://www.reddit.com/r/Music/comments/hnc3b/what_is_your_favorite_song_that_builds/" rel="nofollow">http://www.reddit.com/r/Music/comments/hnc3b/what_is_your_fa...</a>
StrawberryFrog超过 13 年前
Others come to mind:<p>Ravel's Bolero has a "huge build" at least in some versions: <a href="http://labs.echonest.com/loud/?songId=SOKNXIQ13134384E1B" rel="nofollow">http://labs.echonest.com/loud/?songId=SOKNXIQ13134384E1B</a><p>Metallica's To Live Is to Die actually has quite a quick build <a href="http://labs.echonest.com/loud/?artist=&#38;title=to+live+is+to+die" rel="nofollow">http://labs.echonest.com/loud/?artist=&#38;title=to+live+is+...</a><p>Godspeed you! Black Emperor's works are more ebb and flow than sustained build throughout the track <a href="http://labs.echonest.com/loud/?songId=SOJBUTX1311AFDAC27" rel="nofollow">http://labs.echonest.com/loud/?songId=SOJBUTX1311AFDAC27</a>
iampims超过 13 年前
I find this truly amazing, not because of the technologies involved, but mostly because I find it fascinating that we can now easily process a library of over a million song and find which have a “slow build”.
dodo53超过 13 年前
This is very cool - one quibble though, there doesn't seem to be any test of 'slow' in the slow build. You could add a linear regression slope and select for songs where Pearson coefficient is high, the slope is small and the last decile volume/first decile volume is high is large.