WhoSampled Remixes Mobile App With Music Recognition Feature: Exclusive
WhoSampled rolled out an updated version of its mobile app on Thursday featuring Shazam-like music recognition technology to identify samples in surrounding audio.
WhoSampled rolled out an updated version of its mobile app on Thursday (June 28) featuring Shazam-like music recognition technology to identify samples in surrounding audio. The advancement builds on the app and website’s proprietary crowdsourced database of samples, remixes and covers, making use more accommodating to users’ needs and expectations.
Founder and CEO Nadav Poraz calls the newest version of the app a “one-stop shop” for music discovery and tells Billboard he views the recognition feature as “fundamental,” describing how users previously had to first “use a Shazam or SoundHound to identify a song and then immediately go to WhoSampled to find out what the sample is.”
The feature has been in development for some time, says Poraz, and it is thanks to iOS and Android technology he and his team have now been able to expand WhoSampled’s service beyond its website offerings and into the mobile space and give those top companies a run for their money. “Generally, as we’ve been developing this and testing it, we find that we no longer use any other music recognition app because this is much more useful,” he says. Pricing for the ad-free iPhone app is $3.99 and the ad-supported Android app is free, both of which offering unlimited music recognition for a $9.99 annual subscription.
Trending on Billboard
Billboard recently spoke with Poraz about the new version of the app, the future of WhoSampled and the challenges in expanding the platform while remaining true to its core idea.
With this new version of the WhoSampled app, what are the specific differences between your website and mobile products?
The app’s been out on iOS for six years and on Android for four. What we’re trying to do with the app, generally, is really leverage all those features that you can get on a phone that are different to what you can achieve on a website. For the first few years, it’s been about personalization. Especially on iOS, we allow users to look up their Spotify or Apple Music accounts to the WhoSampled app. Then it gets very personalized, in that the app scans all their libraries or their playlists and searches all of the samples, covers and remixes for tracks that they already know — the music that you already have in your playlist, which is something that really takes a lot of the legwork out. It is quite a different experience than what you get on the website, where you can’t really do that.
Obviously Shazam’s been out for a while and SoundHound and all of the other music recognition apps that tell you what the song that is playing around you is, but we take it to the next level. We also tell you what the samples in it are, or who sampled that track and whether it’s a cover version of something. We surface all of that WhoSampled richness based on the music that’s around you. The app is not designed to replace the website, even though to an extent it provides similar information. It’s about really leveraging all the unique capabilities that you can get on a mobile app and to amalgamate the core WhoSampled experience with those types of features.
As the service continues to expand, do you see algorithms ever becoming involved in the database-creation process?
Yeah, absolutely. It’s definitely something we’re looking into. There is some overlap between signal-analysis technology and what we do with crowdsourcing. Technology can identify some samples, but the reason why the foundation is crowdsourcing is the fact that there’s so much that is still beyond the reach of technology. That includes things like interpolations, when things don’t use the original master recordings, but they’re rather re-sung or replayed. Or instances where a sample was heavily manipulated, or chopped and reversed, and all of those things that all of our contributors and moderators can identify using the human ear. We find the best solution really is to use both at the end of the day. To have robust technology as well as crowdsourcing is probably the way forward. This is what we’re investing in right now and see that as the future of music recognition.
Are the contributors vetted in any way before they are allowed to add to the database? How are moderators appointed and what is the ratio of moderators to contributors?
Anyone can become a contributor, but anything that you submit to the database gets double-checked and verified before it gets published. Our moderator pool is about 40 people. They are super-users of the site. They help a lot with the verification. We then have a small in-house team that oversees all of those processes and the quality assurance, and we keep on fixing the errors if there are any.
Everything that you see on WhoSampled is always verified by a qualified person. The moderators are people who we think are up-to-scratch in their ability to vet other people’s submissions. We have a lot of strong mechanisms in place to ensure that we get high quality stuff and we reduce the amount of noise.
Is there a space for social interaction between users on the app? Are users able to communicate back and forth, or see each other’s cred?
You can participate in the discussion that’s attached to any sample or artist or track. Users can comment in the comment section. It is a bit rudimentary at present, but still there’s some good discussions going over there and you can also see user’s profiles and see what they’ve submitted as well. It’s not a full-on social network and such at all. We’re hoping to ramp up the social features in the future, but at least there’s room for discussions around the actual content.
What have been some of the challenges, if any, of expanding WhoSampled over the past few years, while remaining true to its original idea?
One challenge is to keep on scaling the database whilst retaining the highest quality. That certainly hasn’t been easy. There’s always difficult decisions that we need to make around the content. There’s a lot of gray areas, where we need to make a call as to whether something is valid or not. That’s been a challenge.
Probably what’s the biggest challenge for us as a business is to reach more people — to break out of our niche. We are out there and pretty well known, but compared to the biggest services, we’re still quite niche. That’s been a little frustrating, in that we know that if you present the content that we have to even the most casual music fans, they’ll really engage with it. Even my mom is fascinated by the fact that some song is somehow a cover version of something that she wasn’t aware of, or those sorts of things. If it’s something you didn’t know, you have a mind-blown moment, where you want to know those things even if you’re not a music producer or heavy crate-digger or something.
We’ve been working on finding more and more ways to put this in front of more people — to reach more and more music fans and to get them to appreciate the art of sampling and the beauty of it and to discover music in this way. We are aiming to achieve that by more and more partnerships. We’ve been doing projects with all three major labels and we’ve been doing a few brand partnerships. We’re hoping to work more and more closely with both the music industry and the leading streaming services. That’s our way forward. We’d like to think that this way of music discovery eventually wouldn’t be confined to WhoSampled, our own website and app, but would live in many other places.