Replacing MP3: Is A New Format Likely For DJs + Listeners?

Any DJ with a sizable collection of music on their hard drive knows that the MP3 is far from dead, despite a rash of clickbait articles in recent weeks. In today’s article, contributor Steven Maude assesses what it would take take for a new file format to replace MP3. What would have to change? Is it even a vaguely realistic possibility considering what tools people are using to play music? Keep reading for an in-depth analysis.

The MP3 Is Dead? Not Really.

Why not declare MP3 dead? We’ll get more clicks!

Over the past week, you might have read articles announcing MP3 as one of the casualties of 2017.

Especially if you’re listening to an MP3 while reading, it’s probably surprising to discover what you’re actually hearing in your ears is not actually music, but the lamentable cries of a now-departed format.

Fraunhofer, the German research organization, where Karl-Heinz Brandenburg and colleagues developed the MP3 format, declared MP3 dead to them. This NPR story appears to have been the patient zero of this story, although they were quite deliberate about adding a caveat:

But now, 22 years later, the MP3 truly is dead, according to the people who invented it.

Unfortunately, other writers (going for clickbait gold) were happy to inform their readers that — sorry folks — the MP3 is dead.

“Dead” = “We Can’t Make Money Off It”?

The original audio team at Fraunhofer IS (image via MP3-history.com)

MP3 is dead to Fraunhofer because the patents recently expired on the algorithm itself. From a business perspective, Fraunhofer cares about the formats for which they can still charge money for patent licensing (for example, the AAC format). In that vein, it makes sense for them to promote newer formats they’ve developed that can make them money, and not the one that doesn’t any longer.

But this is a very different story, one of soulless balance sheets, and more mundane than the death of MP3. Motherboard offered one of the more sober responses around, arguing that MP3 isn’t really going anywhere just because the creator of it says so: what happens now is entirely out of Fraunhofer’s hands.

That said, the same NPR article does conclude that:

“when the people who spent the better part of a decade creating it say the jig is up, we should probably start paying attention.”

What’s A Potential MP3 Successor?

Since DJs playing digital formats are likely to still rely on audio files, it’s timely to consider what the future of MP3 really is.

AAC is a successor to MP3 that Fraunhofer and other organisations created, and yet hasn’t supplanted MP3 as the format of choice. From memory, my instinct is that most of the audio I’ve downloaded in the last year (tracks, podcasts, and albums) have been MP3s, with a small sprinkling of WAVs and FLACs.

Notable also is Bandcamp’s view; they default to variable bit rate MP3 for downloads, “The default download format is MP3, and this is probably what you want.”

If, then, MP3’s health hasn’t then taken a sudden turn for the worse, then what would a format need to get rid of MP3 for good?

#1. A much more efficient format

AAC, despite being more efficient at lower bit rates hasn’t dethroned MP3, so what would? Maybe if there was a format that’s more efficient by, say 10-100 times, i.e. giving comparable listener quality to high quality MP3, that might provoke people into finally switching. Being a little more efficient and a little smaller isn’t really going to cut through the inertia of users switching to it.

Opus can perform even better than AAC at low bit rates, and, in contrast to AAC, is freely licensed, but, again, neither has its development prompted a shift away from MP3.

#2. Bandwidth contention

MP3 was designed for a different era of internet.

There’s another hindrance to switching from MP3. Since it was made available in the 1990s, bandwidth and storage capacities have increased. Distributing 24-bit lossless WAV or FLAC is now standard in some digital music stores. The issue of MP3 being small enough isn’t really important unless you have relatively poor internet connectivity. They’re usually quick to download, and even moderate phones can carry thousands of MP3s. This supports the idea that a real revolution in compression technology would be needed to convince users to switch to something new.

Maybe something like a lossless compression that was the same size or smaller than MP3 might be a selling point. But generally, high quality MP3 (e.g. 320 kbps constant bit rate or LAME high quality variable bit rate modes) largely sounds good enough for most people, so such a change is unlikely to convince the public.

Like developer and blogger Marco Arment observed, JPEG has been surpassed, yet it’s not “dead”. It’s likely most of the images you’ve seen today online are still JPEGs.

#3. Very fast encoding and decoding

Who wants a smaller or better audio format that killed MP3, but also their phone's battery life?Click To Tweet

Another barrier: any new format requires sufficiently fast encoders and decoders to be available. No-one but the most dedicated are going to spend, say, several minutes or an hour encoding an audio file, when they can the same in a few seconds now.

The decoding aspect is perhaps even more important. Much listening is done on mobile devices with still relatively poor battery life (charging a phone daily isn’t out of the question for some devices with substantial use). MP3 on the other hand has fast decoders available for many hardware platforms: MP3s were even playable back on PCs with Intel’s original Pentium processors 20 years ago, and today’s devices don’t even break a sweat to play MP3.

#4. Playable everywhere

The defunct Zune won’t be getting an update anytime soon….but they play MP3s fine already, and will for generations

Such a format needs to be supported so that creators know that their content is playable everywhere. Computers and mobile phones are extendable through new apps, but there’s lots of legacy hardware that won’t receive an upgrade to play this new imaginary MP3-killing format. If that hardware plays MP3 now, it will still continue to do so quite happily.

This is a vicious circle that a new audio formats would struggle to break:

  • It could take a long time before potential listeners have a convenient way to play the new format. This depends on how eager device manufacturers and software developers are to use it.
  • In the meantime, anyone providing the new format only would inevitably get complaints from listeners who can’t playback those files.
  • Audio producers would just continue to put out audio in multiple formats, including MP3, which increases the inertia for switching in the first place, or for anyone to implement it.

#5. Feature enhancements?

Maybe adding features that offer more options for listening might help a new format dethrone MP3. One thing that comes to mind is functionality like Native Instruments’ Stems, which I suggested here on DJTT might be sold more generally as a way of providing different ways of listening (e.g. bundling instrumentals with the usual track). That might even extend outside of dance music.

But this hasn’t happened, which either means no-one’s thought to push this, or that it’s just not viable.

Still, it’s not completely impossible that someone comes up with an ingenious idea that disrupts how we think about listening to music. There could be an innovation analogous to how FaceApp has changed recently changed how users think about and interact with images of people.

But, this type of addition may not even need a new format at all, and thus doesn’t offer a nice route for would-be assassins of MP3.

#6. Streaming

Photo credit: William Iven on Unsplash

MP3 isn’t being killed by another format, but rather a delivery mechanism: streaming.

With streaming, it’s the software makers of standalone apps, or browsers supporting new formats that web app developers can take advantage of, that ultimately control the delivery format. Solving the “playable everywhere” problem is then done by the streaming app developer, and the format decision passes the user by entirely (unless, perhaps, there’s some explicit option for setting streaming quality).

Even if there’s no format revolution, it can benefit streaming providers to follow format evolution. Small savings in file size can add up to a huge saving of bandwidth. Saving a few hundred kilobytes if you stream a track isn’t a major concern, unless you’re on a slow connection or a metered usage plan. However, scale that up to the billion-plus users that a site like YouTube has, and that can provide a big motivation.

Spotify, Pandora and Apple Music are among those that have already paid their respects and thrown soil on MP3’s coffin. Notably though, SoundCloud are still providing their streams as 128 kbps MP3 today: “For all uploaded files, we provide a 128 kbps MP3 stream version alongside with the original audio file.”

DJs and MP3s

In a Vice Motherboard article, Stephen Witt, the author of How Music Got Free (a well-worth reading history of MP3 and digital music piracy) also adds a comment:

“I don’t think the locally stored file is completely dead and it won’t be for some time,” Witt said.

This is especially true for DJs.

Locally stored music, whether MP3 or not, isn’t going to suddenly be washed away by streaming any time soon. Even if streaming music in DJ software became more common, there are still big hurdles to overcome. Many (if not, all) of the streaming services have a no performing rule, as Spotify’s user guidelines.

YouTube aside, streaming catalogues are far from comprehensive, especially outside of the realm of popular dance music. Lots of past dance music is still buried underground, maybe because the creators and/or labels have disappeared, because master tapes or source files are lost or no longer exist, or just because no-one’s got round to getting them on there yet. (This chaos also explains why YouTube’s the exception: defunct labels aren’t around to arrange licensing. YouTube’s model is opt-out, creators have to request material is taken down, whereas the likes of Spotify carry music they’ve actively licensed.) For plenty of music, the only way of getting hold of a digital copy is ripping from your own CD or vinyl copy, a problem that affects download stores too.

And relying on internet connectivity for music is another possible point of failure that DJs playing out may be reluctant to add to their workflow.

The future, for now, is the past

Photo credit: Simone Acquaroli on Unsplash

It’s also worth taking a look at what’s happened to other successful audio formats:

  • Records are resilient despite being more than a century old
  • Cassette tapes have seen a surging demand in recent years for indie releases
  • CDs are in decline, but some new releases are still sold in that format even after nearly 40 years.

MP3s might not see that dedicated long-term loyalty once it finally is overthrown. A digital file is a digital file, and there’s no corresponding attachment to the form or design of the format itself or its packaging. Part of the love for vinyl and tape too perhaps also comes as a deliberate holding on to analog in the face of a digital future that increasingly permeates our lives.

Nonetheless, MP3s are now easy to create and license free to create, there’s a huge back catalogue of music in this format, and they are playable on just about every media playing or computing device made in the last 20 years.

It’ll be a while before MP3 is finally laid to rest.

  • Pingback: Unit 8: Final Major Project | Music Production LV3()

  • Per Jakobsen

    half the world, is still struggling with connection speeds below average hence mp3 will stay on for loong time

    • Very good point – but one of the things discussed here is that a better compression format than MP3 that delivers lossless-level quality at MP3-size files would be something that might make DJs and listeners switch. Imagine the quality of a 50MB WAV file but at a 5MB file size…

      Then again, most people won’t have the systems to tell the difference 😉

      • Per Jakobsen

        “Imagine the quality of a 50MB WAV file but at a 5MB file size”
        That would definitely open many eyes, and if systems implements support for backward’s file formats then Im in

  • Reticuli

    AAC>Vorbis>MP3 in pure technical terms based on preserved transients and frequency response, though sometimes I find Vorbis sounds slightly more natural and pleasant at equivalent rates. I have no idea why.
    Lossless is the way to go, though. Bandwidth and storage is cheap. The issue with that is the amount of processing to unpack/decode compressed lossless formats like FLAC, so that leaves WAV and AIFF. WAV has no tagging built in, so that leaves AIFF if you need tagging, but WAV is nearly universally supported if you can get by with just the filenames. There are so many brandings of the Hanpin stuff out there, and they’re all universally WAV and MP3.

  • TinGA

    MP3’s are like an airplane wing, mature in its design and function and very new innovations are required except to save ‘fuel’. We can look for smaller more efficient options, but the cost of portable storage (USB Sticks) has steadily come down. In addition any shortfall in MP3 can be mitigated by creating an accompanying database which provides any additional data information we might need. MP3’s in my opinion are sufficiently mature and should remain in the free public domain.

    • Reticuli

      MP3s sound crunchy and artificial (from slightly to very depending on the track) to me compared to lossless files, especially when processed by digital equipment.

  • killmedj

    As much as I love high quality audio, I would love the whole tagging system to be revamped. I mean seriously, “Track Number?” not to mention the hidden layers of old info that sometimes reappears when you need it least! I reckon (for DJs at least) a comprehensive tagging alternative would have me ditching MP3’s instantly!

  • DjChachi

    Dj TechTools this is a hell of an article. Five star in my opinion.

  • this is interesting

  • James Z.

    Realistically there are only a handful of viable paths here. The vast majority of music is going to end up being streamed and at some point will be of the same quality of Wav and lossless files. Since drive size is no longer a concern, the real need for a format to replace mp3 is nil when talking about file size. The only real step up from MP3 that some people may receive a benefit from would be the move to FLAC or a similar lossless codec. Apart from the minor improvement in sound quality, it give a much higher quality starting point for sampling and remixing. Its a point that seems to be overlooked by the DJ and producer community quite often and it is audible.

  • aphocus

    Opus is the format to dethrone mp3, it’s more efficient, from 20kbps through 512kbps than AAC-HE (low bitrate quality), AAC (high bitrate quality), it’s low decode/encode latency (5-20ms) makes it Ideal for live internet stuff (voip, online jamming), and is superior to ACC-LD.

    It’s a patent free IETF standard (2012), mandatory for WebRTC, and webm, that is already supported by YouTube, Edge, Chrome, Firefox, and Android.

    It has the power and momentum to take over, with Google and Microsoft behind it.

    Ironically Skype contributed to the spec, but has not added it to their app.

    • Partially. Opus is a format optimized for high compression, not for high quality or post processing. Mostly streaming content to mobile devices. It might do mp3s job for that but not for archiving or post-processing. FLAC does a much better job there.

    • Be

      Have you tried converting a lossless file to Opus? It doesn’t sound nearly as good as AAC, Vorbis, or even MP3. Opus isn’t really meant for offline storage; it’s meant for streaming.

      • aphocus

        I don’t care for personal anecdotes, there are a multitude of reasons why your experience might not align with the science (bad tools, sighted tests / placebo, exceptional sample), I have samples here with 64kbps opus sounding better than 128kbps mp3, but that wouldn’t be representative.

        What matters is the science, and the science is clear: http://listening-test.coresv.net/results.htm

        • Be

          I compared higher quality settings of each codec, not low quality. Opus is better for what it is designed for, which is lower bitrate streaming.

          • aphocus

            Do you want me to concede that your anecdote is somehow superior to science?

            Come back to me when you’ve done real research, you didn’t even claim your testing was done blind, which is probably the biggest reason why you’re getting biased results.

          • Be
  • jm2c

    AIFF ftw. Lossless, metadata, corrupted file can still be partially readable/recoverable unlike flac/mp3, where the blocks become garbled upon file corruptuon.

    As soon as internet speeds allow it realistically, that is.

    • Or ultimately just some form of high speed lossless compression algorithm that makes our current internet speeds a bit faster. We could call it…. middle-out compression? 😉

  • MP3 will survive for the same reason the CD and DVD survived. It still has a market for its widespread familiarity, ease of use, and the fact most people listen on poor quality speakers/HPs, which mask the dynamic range and spectrum loss in 256/320 kbps MP3.

    Once people start buying higher grade speakers, they may start to hear the differences in lossless and lossy audio.

    From a technological standpoint, FLAC should have replaced the MP3 entirely, given the cheap cost of storage space. I have 20k songs in 500 GB… that was a lot of space 10 yrs ago, but its nothing on a 2-3 TB HD. I am actively replacing all mp3 with flac (thanks to cheap new/used cds, dbaudio amp accurate ripper, and 7digital.com)

    STEMs are good too (tho not lossless, are less likely to have loss due to the split tracks), but I they havent taken off enough to convert to fully.

    Streaming is of course the modern consumer standard, and it will stick to MP3s as long as the market will allow it. Its simply not cost efficient for them to offer it all in lossless.

    • Jyve

      Even on a consumer-marketed high-end audio system, you wouldn’t be able to hear the difference of 256 vs 320 kbps MP3

  • Jman Yiisshh

    this is interesting