Hacker News Reader

Comment by sarreph | Hacker News Reader

For all those looking to make a SoundCloud clone: What is HN's opinion on its failings (from a user perspective) — i.e. how would you make SoundCloud better right now?

Also interested to hear those who can balance with commercial viability.

View on HN

Go back to basics.

Keep it focused on community facing artists, remixes, dj sets, etc.

From what I've seen they suffered most from lack of direction, weak leadership, and no real business plan from the beginning. It seems (correct me if I'm wrong) it started as a hobby project and got twisted into a SaaS model that didn't fit it's purpose or audience.

As far as monetization - why not have a Patreon-esque system of support specific content producers/artists? I don't think a blanket monthly subscription is really a good fit but personally I would pay 1-5$ / month for specific artists/content providers and follow only them.

> why not have a Patreon-esque system

You've lost me there.

You can't, because people don't pay and don't want to pay, unless they are forced to, as in there's actual scarcity (i.e. no way to easily get that content from anywhere else).

Scarcity is part of Patreon. I'm guess/assume that much (most?) of the support comes from backers wanting exclusive content that is sometimes offered as a reward.

Fair point.

Maybe a limit on how many streams/month that could be lifted for a certain fee?

Regardless - point is I think there are ways to monetize this enough to keep it running. I doubt any streaming service will ever be a cash cow but I'm sure there are ways to keep them running.

Why do you think "you can't", despite Patreon (for now) looking like it can work? (Although they of course don't have to finance an entire music sharing service with this)

Music fans seem comparatively loyal and willing-to-spend to me, so I'm curious about your reasoning.

(EDIT: reworded for clarity/less confrontational)

Clearly it can work, because it works for Patreon. It also works for twitch, who allow people to "subscribe" to streamers they like. This entails paying $5 a month (half to the streamer, half to twitch) to receive exclusive chatting rights, special "emotes", and other non-essential features. People do it not because they want access to the special features, but because they want to support the streamer. It seems that people are adverse to paying only when it is on an-item-by-item basis, but are not adverse to it in general.

People will pay for indie music, it's been happening for a long time. It still happens today. Particularly if, as we've been talking about here, it's focused on electronica/hip-hop people will pay so they can perform the track in public. There's a real stigma against playing stolen music like that. Plus you can pay for higher quality versions. Lower quality versions don't sound as good when timestretched so they dont work as well for mixing.

People listen to a lot more than a couple artists. No one is going to Patreon fifty bands.

See Bandcamp. The main thing that it doesn't have that SC does is a feed, but otherwise, they do have a Patreon-esque system - you can listen to the music on the website for free, but you can either buy the albums you like (and download them) or subscribe to users you like (and get subscriber-exclusive releases, which may include earlier versions of songs, stems etc)

Bandcamp does have a feed.

It is also not particularly Patreon-esque. You don't pay the artist a recurring fee. You can listen to the music for free only a handful of times, then you need to buy it, in a one-time transaction.

Bandcamp is really a place to discover music, buy it and download it. It's not similar to Soundcloud, or Spotify, which focus on streaming. A more apt comparison would be Beatport.

The target audience of Bandcamp are DJs and music connoisseurs, who want to own the music they like, in high-quality formats.

Bandcamp does a really good job of staying focused on these core user groups, and apparently is also doing rather well, financially:

https://daily.bandcamp.com/2017/01/24/everything-is-terrific...

I guess they're something of a reverse Soundcloud.

I think that's a brilliant idea.

There's no real evidence that there is a profitable business model for streaming music except insofar as subsidies from other parts of the business make sense overall. Just because a lot of people like something doesn't mean that there is a path to making money with it.

I don't have strong feelings about what's good and bad about their UI, etc. Their main failing is that they're probably not going to exist much longer unless Apple, Google, or someone along those lines snaps them up at fire sale prices and is OK with losing money on them.

Focus on DJ sets, remixes, hip-hop/electronica producers like others have said. Don't try to be Spotify.

I second this, the mixes are my favorite part as well as DIY producers.

I would say focus on sets over tracks, and develop partnerships to stream exclusive live concert sets- would be really easy to scale if they focused on electronic first as those sets are easy to record/stream.

Mixcloud has that area covered and I hope they become a much more popular with the recent Soundcloud drama.

Thanks for the vote of confidence! (Mixcloud founder)

I’ve commented extensively on this before if you check my comment history but basically they need to go back to their roots. No commercial music. No copyrighted music. It should be a great place and community for indie and unsigned artists to post content and they can monetise it using paid artist accounts as they originally did (pay for more storage, better listener stats, etc). Commercial music means licensing and lawyers and paying money you can’t afford. Stay away from it. You can’t compete with Spotify et al. Maybe try to implement the tech Moxcloud used to allow DJ mixes but I wouldn’t go further than that commercially.

They need to find a much better way to let their listeners discover new music. SoundCloud groups used to be an alright way to do that, and when SC ditched them they really crippled themselves, and now I as a small, independent music producer don't even bother to upload my music to SC because it gets zero listens since there's virtually no way for anyone who might appreciate it to discover it.

Back when SC had groups, I didn't have a lot of listens, but it wasn't zero, and some people actually liked my music. Now SC is a ghost town for me, so why even bother?

Very much this. Some simple search tools would go a long way - multiple tags, sort by upload date, "not including" negative search, filter by like/download count, downloadable only etc

Many are admittedly hard challenges, but;

  > Give me a non-browser client.
  > Make the browser client not leak memory.
  > Don't interleave Soundcloud Go previews with free music.
  > Give me reasonable discovery of similar music.

Until Go happened, Soundcloud was my everyday work listening. Now I only use it for long mixes that I can't find (or am too lazy to find) elsewhere.

The challenge for any clone isn't going to be features, it's going to be content.

Don't make a business out of it, at least not directly. Make an open source, federated clone, ideally with something like GNUSocial driving it - something that doesn't go down under the pressure of endless bullshit DMCA claims or investor pressure. Sell your services as a consultant maintaining it for artists who can't get the technical bits down.

We have got to quit trusting commercial entities with our culture. They are not good stewards of it.

Honestly, an open source solution backed by an automated hosting company (like how Ghost) is probably the best solution.

Charge your customers for DMCA claims and just skim enough off the underlying hosting company to cover your expenses.

+ soundcloud redesign

+ merge the same shared track > bob shared "sounds01" > randy shared "sounds01" {at a later time} >>merge names if the same track is shared > Bob & Randy shared "sounds01" >>ui animations if in real time change of merge, otherwise don't animate

+ make an artist collaboration area > see Reestablishing Groups/Clubs

+ different subscription models > free artist account > paid artist account [removes ads + more upload space] ~ > free listener account > paid listener account [removes ads + more playlists] ~ > advertiser account [design fill-form ads that link directly to soundcloud track/album (all ads are in the same style, text and links.)]

+ less hassle create playlists - ui > see old soundcloud, pre-redesign

+ no hidden ui buttons [download, buy etc]

+ having another subcategory called 'reposted/shared' along side tracks, etc. > toggle reposts on paid accounts

+ groups/clubs readded > Reestablishing Groups/Clubs to make soundcloud more open to collaboration and like minded playlists. > Group playlists

I'm probably not the typical user, since I like to listen to music off-line. One thing that annoyed me about SoundCloud is that I often come across music on YouTube that has "download at SoundCloud" links. However, the SoundCloud "download" button always seemed to be a kind of a lottery. Sometimes it was there, sometimes you needed to log-in, more often than not it was nowhere to be found.

I understand that it's not in SoundCloud's interest to offer downloads and these days I just use youtube-dl. But apparently (at least at some point in time) it gave musicians the impression that it will allow their music to be downloaded.

The musician chooses whether or not to allow downloads

I would pay a premium price to stream in better quality and I cannot for my life figure out why no one has ever suggested this in any of the many recent SoundCloud threads or why SoundCloud hasn't thought of this themselves. But I guess that the two are connected – and that I really am the only person who would pay for this.

SoundCloud uses an encoding algorithm that is extremely bad even for the lousy 128 Kbps they are streaming in. Why should a dedicated music service feature worse audio quality than YouTube? Of course I would pay to improve this during the six hours I spend on SoundCloud each and every working day.

Speaking as a music creator, I like using Soundcloud (for example) to share a new mix with colleagues / clients, who can easily comment at specific times on the recording. Or to share A/B comparisons between different settings. And all of this can be on private URLs shared with only the intended parties.

I don't use it to find new music to "just" listen to, and I never have. But I would happily pay to use it for what I do get out of it. I haven't paid for it yet, because their free tier has so far covered everything I needed...

Is there any business being the middleman among artists? Could have a musicians-only option, similar to IMDBpro, be monetized in SoundClound platform?

A lyricist auction their lyrics among some artists, musicians sell their music to advertisers who are producing a commercial. Musicians selling to filmmakers. Dj buying the rights to remix some songs. I know there are already several businesses doing this, many separate. But SoundClound could dominate this market by having a worldwide scale. And for being the first choice of niche artists.

I am taking this event as reminder to go back to my own self-hosted website with an RSS feed. It doesn't have the complete social factor offered by Soundcloud, but hey, it's my own!

Get copyright laws changed to allow free sampling, as long as the new work has a creative identity that is distinct from all the samples it contains.

Isn't this already true? You are allowed to flaunt copyright under certain conditions, including if the portion copied is sufficiently small, or if the new work is considered transformative. I would imagine that many instances of sampling would be protected because of this.

SoundCloud's responsibility should be to defend the rights of their users. If a DMCA claim is filed in a way that SoundCloud deems to be unfair, then they might offer to litigate on behalf of their user. As it stands, the majority of artists will capitulate to studio even in cases where the law is on their side, because they haven't the means to defend themselves.

Yes and no. The limit is something like 4 bars I believe. That may be a sensible limit for the traditional 3:20 minute song on the radio, where such a snippet can often capture the central motif of the song.

But most of what makes SoundCloud different are the 3-hour DJ sets that can draw on hundreds of different sources, including other songs, or movie dialog, or noises recorded in nature.

There's no absolute truth how to value each and every contribution to something like that, and it may well be fair to keep some royalties flowing to many of these sources. But as far as I understand it, the current regime is somewhat lopsided in the original creators' favour.

I have been using SoundCloud almost exclusively for the better part of 7 years. I have playlists (eg., Personal Favourites for <Month> `<Year>) dating all the way back to early 2011. It's the /only/ online music service I subscribe to. Zero interest in Spotify, don't care for Apple Music, and even other services (eg., Mixcloud, Bandcamp, Audiomack, etc) miss the point of being driven by a community.

It's a service where, at this point, I cannot part ways with, simply because there are zero alternatives that either emulate its flow (twitter like feed), or house the same blossoming artists that go on to become big (eg., Post Malone, Don Monique, Bryon Tiller, Chance the Rapper, Joe Kay, etc) names.

But, SoundCloud has its issues, and its Go+ service (subscriber since day 1) is a missed opportunity. On its service, there are a many issues with SC.

1. SoundCloud (not a developer, but assuming) seems to have chosen poor engineering choices from the top. SoundCloud just chose to not solve it once it became too hard on their database, including when they famously had issues with being early adopters of Mongo. Their most recent frontend HTML5 rewrite chews up CPU usage, and it just hasn't changed in recent years.

2. SoundCloud Go and Go+ were blatant cash grabs. I purchased it purely to bypass ads, but also because I believe in the /concept/ of the service, but have lost faith in the company some time ago. Content Producers would have paid more if only discovery hadn't gone downhill, which may have heightened

3. Despite the iOS app being updated every other week, it seems as though it /actually/ hasn't been updated in years. You cant click "Like" and then "Repost" within 1-2 seconds of each other before the app just decides to crash. The app even lacks basic controls and makes it impossible to queue up songs /unless/ it's put into a playlist, but then you have to discovery > listen > add to playlist > flip to playlist... it's asinine

4. Artist Engagement is a /joke/. Just to "buy" or "download" a track means you're walking into a wall of questionable sites, and a plethora of social media logins. "LOGIN TO DOWNLOAD! NOW RETWEET! NOW LIKE! ALMOST THERE! NOW SNAP AT US". This isn't SoundCloud per se, but they implicitly allow this to happen. They could have easily required you to follow the artist before downloading directly to solve at least part of the problem. Especially considering you can just use a Chrome Extension to download the song...

5. Music discovery is pathetically trash. The further you drive down the rabbit hole of discovery, the slower the experience becomes. Discovery is the /core concept/ of SoundCloud since there is so much underground talent just waiting to be discovered. This was the entire point of SC, and they failed at it.

6. Copyright issues plague people to the point of marring them from becoming 'djs' or potential artists. It also seems like their ability to work with artists in general is just utter garbage.

7. The UI is /garbage/. It needs to be updated. It's almost like 10 teams at SC worked on it after a night of drinking and throwing up. The simplicity of discovery should be the focal point, but it seems like it's a bug when you want to find new music.

8. The power of SC is the community. It's a feed of new music, even music people you follow are reposting. While "The Upload" is a great idea, it's buried deep behind a sluggish UI.

9. Queueing took far too long to release. It's especially stupid considering that if you dive into an artists page, while having an album queued, playing their song will wipe the playlist in favour of the artists. Come on.

I can go on for /hours/. But, I hope SC stays around, a new CEO comes in, and the product is not only rebuilt, but redesigned from the ground up.

Please don't die. I have nobody else.

To download from SC, I just use youtube-dl.[1] It's free, open source, well respected, and works great.

In fact, I don't listen to SC (or watch youtube) any other way.

[1] - https://rg3.github.com/youtube-dl/

> SoundCloud just chose to not solve it once it became too hard on their database, including when they famously had issues with being early adopters of Mongo

Do you have more info on the Mongo angle? I worked in SoundCloud engineering for a while (long ago) so I'm curious how that happened :D

> Please don't die. I have nobody else.

Yeah I share that sentiment :(

I have been wanting something like Youtube Red for Soundcloud, where my listens directly correspond to payments for the creators. I had hoped that Soundcloud Go would be like this but I believe most of my money goes to record labels.

Pardon my ignorance but why would someone clone a business that tried for years and failed? Wouldn't it be like cloning Groupon?

At least they had tons of users and an alternative would get a lot of attention right now. If you have a solid plan to get the finances to work out...

Volume control on embedded player.

Reply on news.ycombinator.com