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。

Listen to what gets lost when an MP3 is made (2015)

142 点作者 teleforce6 个月前

26 条评论

gwbas1c6 个月前
In 1999 when MP3 was getting attention, I tried to do this. I encoded a file, then inverted it, and mixed it back into the original.<p>It didn&#x27;t cancel anything out.<p>The reason: Mp3 <i>dramatically alters phase.</i> Because all the phases are different, it&#x27;s hard to naively determine how the signal is altered.<p>Years later, I took the time to write a series of tools to investigate lossy audio: <a href="https:&#x2F;&#x2F;andrewrondeau.com&#x2F;blog&#x2F;2016&#x2F;07&#x2F;deconstructing-lossy-audio-the-case-for-lossless" rel="nofollow">https:&#x2F;&#x2F;andrewrondeau.com&#x2F;blog&#x2F;2016&#x2F;07&#x2F;deconstructing-lossy-...</a>
评论 #42183458 未加载
评论 #42221947 未加载
a-french-anon6 个月前
This article could at least have a paragraph explaining (in a dumbed-down way) why various kinds of psychoacoustic masking (temporal, frequency) make what&#x27;s removed almost inaudible anyway. Reading the linked source (<a href="https:&#x2F;&#x2F;www.theghostinthemp3.com&#x2F;theghostinthemp3.html" rel="nofollow">https:&#x2F;&#x2F;www.theghostinthemp3.com&#x2F;theghostinthemp3.html</a>), he at least used LAME, but at a fixed 128 kbps bitrate, not in VBR mode =(<p>EDIT: nerds should read about sfb21 (<a href="https:&#x2F;&#x2F;wiki.hydrogenaud.io&#x2F;index.php?title=LAME_Y_switch" rel="nofollow">https:&#x2F;&#x2F;wiki.hydrogenaud.io&#x2F;index.php?title=LAME_Y_switch</a>), AAC, Vorbis and Opus (CELT) aren&#x27;t just theoretical improvements
评论 #42182268 未加载
cladopa6 个月前
There is a trick there. A sound can mask another sound. You will not be able to tell the difference with both sounds playing at the same time, but if you subtract them you can hear it because there is no masking.<p>I always loved to test the ears of my &quot;Audiophile&quot; friends. They will tell you how different MP3s are. You make a bet they can not differentiate them in 20 trials better than chance. I won with most people but some professional musicians that can identify little differences.
评论 #42183011 未加载
评论 #42183955 未加载
评论 #42182950 未加载
评论 #42183169 未加载
评论 #42183008 未加载
评论 #42183167 未加载
评论 #42184278 未加载
BurpyDave6 个月前
Ironically, the &#x27;diff&#x27; is compressed anyway, because it&#x27;s on Vimeo, so that&#x27;s not the actual diff either!
评论 #42183271 未加载
jonathanstrange6 个月前
So what. People listened to music on mechanical gramophones and enjoyed it. Too many audio engineers think it&#x27;s all about the sound, when in the end it&#x27;s about the music and the feelings it expresses.
评论 #42182542 未加载
评论 #42181865 未加载
评论 #42184247 未加载
评论 #42183766 未加载
Quarondeau6 个月前
Interesting approach. So are we only able to hear those sounds now because the rest of the music was removed, which would ordinarily mask the missing sounds?<p>To say that the mp3-encoded version is not &quot;what the artist recorded and wanted for us to hear&quot; would imply that we can hear all sounds in the uncompressed recording.
评论 #42181991 未加载
reliablereason6 个月前
You can try it yourself:<p>ffmpeg -i original.wav -codec:a libmp3lame -b:a 192k output.mp3 &amp;&amp; \<p>ffmpeg -i output.mp3 decoded.wav &amp;&amp; \<p>ffmpeg -i original.wav -i decoded.wav -filter_complex &quot;[1:a]aresample=async=1,volume=-1.0[inverted];[0:a][inverted]amix=inputs=2:weights=1 1&quot; difference.wav
评论 #42182406 未加载
评论 #42183211 未加载
jonnycomputer6 个月前
&quot;What MaGuire has proved here is that the songs we listen to every single day are not the exact master copy that the artist recorded and wanted for us to hear. Instead, they are slightly stripped versions of their art run through a set of standards created by a bunch of engineers in 1993. For many people, that won’t matter. The songs sound almost the same, but the compression of music into an MP3 format is an important question to weigh when considering artistic intent and analyzing songs that aren’t exactly the original.&quot;<p>I feel like this analysis isn&#x27;t well grounded in what artists and sound engineers actually do, or how they think.
评论 #42187776 未加载
评论 #42183936 未加载
NoPicklez6 个月前
Fairly lackluster article.<p>Not all .mp3&#x27;s are created equally and can vary in how lossy they are based on the bitrate.<p>If you care enough to want to hear exactly what the artist wants you to hear, you just listen to the lossless version.
评论 #42183823 未加载
评论 #42183320 未加载
kazinator6 个月前
&gt; You can hear so many unnecessarily rejected sounds.<p>That accusation requires evidence based in psychoacoustics. Just because you can hear it in isolation doesn&#x27;t mean you can hear it if it is added back to the host audio.<p>For instance when some quiet sound that is masked by immediately preceding loud sound is removed, of course you can hear that quiet sound in isolation! Your hearing has something like 120 decibel dynamic range, or better.<p>You can hear differences in the compressed audio. Nobody can claims that there&#x27;s no degradation in quality. Artifacts are obvious. Much more so at lower bit rates, though. MP3 starts to sound quite good around 192 kbps.<p>The removal of those components is necessary. It is necessary to the algorithm so that it can achieve compression.<p>Also there&#x27;s this issue. If we take a signal and apply some modest EQ to it. Say we boost the bass and treble and cut me a little bit. Or any other EQ profile. If we then level match the two signals and subtract them from the other, there will be a difference: some aspects of the original material will be recognizably heard. For instance the difference between a slightly treble cut signal and the original will be the treble. But the trouble was not completely cut from the original. What you&#x27;re hearing in the difference is not something that was <i>entirely</i> removed.
CGamesPlay6 个月前
Found the original author&#x27;s page about the project (no longer on the internet): <a href="https:&#x2F;&#x2F;web.archive.org&#x2F;web&#x2F;20211011015410&#x2F;http:&#x2F;&#x2F;ryanmaguiremusic.com&#x2F;theghostinthemp3.html" rel="nofollow">https:&#x2F;&#x2F;web.archive.org&#x2F;web&#x2F;20211011015410&#x2F;http:&#x2F;&#x2F;ryanmaguir...</a><p>One interesting thing to note: this is a composition, not an analysis. It&#x27;s not fully documented exactly what modifications to the &quot;raw data&quot; were made.
评论 #42181851 未加载
Agraillo6 个月前
Why Tom&#x27;s Dinner? Because it is a cappella. There&#x27;s a book &quot;How Music Got Free&quot; by Stephen Witt [1] detailing the history of mp3 format and related events. It is a very good read and there&#x27;s an explanation<p><i>Increases in processing power spurred progress. Within a year Brandenburg’s algorithm was handling a wide variety of recorded music... But one audio source was proving intractable: what Grill, with his imperfect command of English, called “the lonely voice.” (He meant “lone.”) Human speech could not, in isolation, be psychoacoustically masked. Nor could you use Huffman’s pattern recognition approach—the essence of speech was its dynamic nature, its plosives and sibilants and glottal stops. Brandenburg’s shrinking algorithm could handle symphonies, guitar solos, cannons, even “Oye Mi Canto,” but it still couldn’t handle a newscast. Stuck, Brandenburg isolated samples of “lonely” voices. The first was a recording of a difficult German dialect that had plagued audio engineers for years. The second was a snippet of Suzanne Vega singing the opening bars of “Tom’s Diner,” her 1987 radio hit.</i><p>[1] <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;How_Music_Got_Free" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;How_Music_Got_Free</a>
sdk776 个月前
Very interesting! The audio of Tom&#x27;s Dinner rejected by the encoding sounds mesmerizing to me. I still find it to be musical - it reminds me of a record I bought a really long time ago, it was called modulation &amp; transformation on mille plateaux, it&#x27;s a collection of songs in the abstract and experimental genre.
pvillano6 个月前
Two instances where lossy compression failed for me are the movie Koyaanisqatsi and songs by the artist TOBACCO. Koyaanisqatsi has a lot of film grain and TOBACCO uses a lot of distortion. There is noise in there, but it&#x27;s very deeply mixed into the signal.
0points6 个月前
This is why we dont encode mp3 in 96kbps or whatever.
moomin6 个月前
I think the thing that&#x27;s really sticks out is that the breath noise are gone, which is one of the things that gives the track its character. Willing to bet the same kind of thing happens to fret noise as well.
评论 #42183712 未加载
no-such-address6 个月前
Funny article.<p>&quot;The exact master copy that the artist recorded and wanted for us to hear&quot; In the digital era, does that even, uniquely, exist?<p>&quot;a set of standards created by a bunch of engineers in 1993&quot; Nice!<p>Was hoping the article would mention double blind studies about the ability to perceive differences and the quality between various audio file format, available elsewhere. Interesting, though not as overwrought as the reporting in this article.
评论 #42184009 未加载
HPsquared6 个月前
This could also be done on visual compression with JPEGs.<p>Or on video compression, for that matter.<p>It just shows though that these diffs are invisible to a human - by design.
评论 #42182671 未加载
chrsgrrtt6 个月前
I developed a streaming service many years ago; Dolby wanted us to use their codec for the audio, and they used a track just like this as the primary basis of their sales pitch. Was quite impressive at the time.
ezconnect6 个月前
When I first experience CD audio it was too high pitch compared to tape versions. MP3 came along and each song sound different depending on the MP3 compression settings.
评论 #42182388 未加载
Timwi6 个月前
I would have liked a comparison with Ogg and perhaps other formats. I hear a lot about MP3 throwing away a lot more than Ogg but I&#x27;d love to see real data on it.
评论 #42183566 未加载
评论 #42182658 未加载
Traubenfuchs6 个月前
...I think this person just created a new genre of music. Something like: &quot;What&#x27;s lost noise.&quot;<p>I immensely enjoyed listening to the &quot;lost material&quot; of Tom&#x27;s Diner and would like to hear more of this!<p>Maybe one could diff with a lower quality version, one where more has been cut away, more is lost&#x2F;left over? There are so many possibilities!
评论 #42183006 未加载
ipunchghosts6 个月前
There&#x27;s still audio motifs in there that can be further optized out.<p>If the remaining audio was noise like, I would say we reached the compression limit.
Klaster_16 个月前
The article doesn&#x27;t mention at what bit-rate the difference track was made, anyone knows? Seems disingenuous and pro-&quot;authentic&quot; otherwise.
评论 #42181761 未加载
评论 #42181775 未加载
评论 #42181758 未加载
kazinator6 个月前
The funny capitalization of moDernisT instantly gives away that it is an anagram of Tom&#x27;s Diner.
grishka6 个月前
Now I want this comparison for Opus. It doesn&#x27;t do that whole psychoacoustics thing, does it? But it also somehow manages to ~double the compression ratio compared to MP3 without any noticeable difference in the sound quality.
评论 #42183638 未加载