Hi again,

This is our monthly update of things happening in Freesound development! Remember we told you we’d try to write regularly-spaced-in-time blog posts letting you know what’s going on? Well we made it, so once month later here is a new issue 🙂

Here are the new features deployed this month:

• Sound signature. If you go to your settings page you’ll see a new form field with the title Sound signature which allows you to introduce some text. You guessed it right? This text will be added at the end of the description of all of your sounds. If you change the sound signature, it will automatically be updated on all of your sounds. This way you’ll be able to add attribution instructions or other information that you want to appear in each sound. What? You have uploaded no sounds? Maybe it’s time to start!
• Subscribe/Unsusbscribe to forum threads at any time. Those of you who participate in the forums already know that when you post in a thread you have the option to subscribe to the thread and get email notifications when new messages are posted. In these email notifications you have the option to unsusbscribe from the thread as well. Now these two operations can be easily performed from the thread page. This also means that you can receive notifications without actually participating in the conversation.

Besides these new features, we also made some other smaller changes and bugfixes:

• Added minor improvements in the UI of the describe sounds page.
• Fixed a bug which was wrongly reading metadata of some file formats including lossless version of M4a files.
• Fixed a bug which disabled the group by pack option in search page.
• Fixed a bug that caused some pack filters to fail in the search page when pack names contained non alphanumeric characters.
• Fixed a bug in the search page which prevented sorting options other than Automatic by relevance to be used.
• Correction of other smaller errors, significant code clean ups and internal optimizations.

That’s it for now, thanks for reading and stay tuned for the next post!

The Freesound Team

Dear Freesounders,

We’ve been working hard on Freesound lately and have some news to share with you. You’ll remember that earlier this year we started a campaign to promote donations in Freesound. We had a public discussion in the forums where many of you participated (thanks!), and we finally started the donations campaign in July 2017 (see this forum post for details on the campaign). We’re now getting 15 times more donations than we used to get before starting the campaign which is a great start! It moves us closer to our sustainability goal and it has already allowed us to start spending more time and resources on improving Freesound for all of our users. Thank you all!

We have recently released some new user-visible features to Fresound:

• Donations via stripe. Some potential donors didn’t want to use Paypal to send donations to us.  As an alternative, we added support for donations by credit card using Stripe.
• Support for AAC files (.m4a extension). You can now upload sounds encoded using Advanced Audio Coding (AAC). AAC is a lossy audio compression standard designed to be the successor of MP3. Even though we still prefer sounds to be upload using uncompressed formats or non-proprietary formats like OGG, we’re sure that AAC support will come in handy for many Freesound users. [EDIT 20/10/2017: user guyburns reported that lossless m4a files are also supported but that there seem to be problems detecting the number of channels].
• Faster random sound navigation. We’ve always had the ability to randomly browse for a random sound in Freesound. Now this is faster, try it today!
• Sound license history. We now keep track if a sound uploader changes the license of their sound. This is important because you may have downloaded a sound under one license, and when you go back to check its status some months later the license may have changed! We will now tell you what the license was when you downloaded the sound so that you can be sure of your rights and obligations.

We’ve also made a number of smaller changes which are not as visible, but help us improve Freesound for us and you:

• We now keep a record of what the sound of the day was each day. Do you want a throwback to what the sound of the day was on a particular day of the year? We’ll be able to tell you!
• If you forget your password and your email address you can now reset it using just your username, much the same way that you can log in with both your username and email.
• If you’re uploading many sounds at once with a geolocation, we now remember the last location that you added so that you can quickly add the next.
• Tickets now show faster for our tireless team of moderators (thanks everyone!)
• We’re making changes to move Freesound to a full HTTPS setup, protecting you when you enter passwords when logging in.
• We’re fixing small programming errors that have accumulated over years of Freesound development. If you’ve ever got a message saying “You found some part of Freesound that’s broken”, then don’t worry, we get a notification of the error and we’re working to make sure that these errors happen less for our users.

That’s all for now! Hopefully this will be the first post of a series of more-or-less-regularly-spaced-in-time posts that we’ll be doing to keep you updated about things happening in the Freesound dev world. Want some more exciting news? Then wait for the next post 🙂

Thanks for reading and keep on Freesounding!

The Freesound Team

Dear Freesounders,

On Monday 3rd at 10:30 a.m. CEST Freesound will be down for approximately two hours while we perform some maintenance on our servers. We’re sorry for the inconvenience this may cause.

Cheers,

edit: all servers have been updated and we’re back to normality (Monday 3rd 1 p.m.)

EDIT 8/6/2017: email sending is back to normal. We’ve resent emails with activation codes to all users that registered during the period while the email service was not working properly.

Hello,

As some of you reported, Freesound is experiencing problems with emails. We have identified the issue and worked out the fix, but are still waiting for our email provider to do some extra stuff. We hope that email sending will go back to normal in the following hours.

After all email problems have been fixed, we will re-send activation emails to those users who registered after the problems started and could not activate their accounts.

In the meanwhile please be patient 🙂

frederic

Dear Freesounders,

This is post is to let you know about two features that we have just deployed for your enjoyment:

• Login with email: from now on you can login to Freesound either using your username or your email address. Might not be a game changer, but will certainly be helpful for many users.
• Support for Twitter cards: you can now share Freesound sounds on Twitter and these will be displayed in a nice card with a Freesound player. All you need to do is to add a Freesound sound URL in your tweets 🙂

That’s it for now!!
Keep on Freesounding,

frederic

Dear Freesound community,

We would like to give you an update about the current situation of Freesound funding and to get your feedback about some of our future plans regarding its sustainability. Be prepared for a rather long read 😉

As most of you will know, Freesound was started in 2005 at the Music Technology Group (MTG) of Universitat Pompeu Fabra. The initial aim of the project was to create an open database of sounds that could also be used for scientific research. However Freesound became bigger and bigger and turned into a wide and active community of users which grew well beyond any initial goals. Freesound currently hosts more than 4400 hours of audio and serves around 50000 sounds per day. For more statistics, have a look at this blog post.

As you can imagine, the cost of maintaining Freesound has been increasing as the site has grown, and it is not an easy task to find resources (both human and machine) to maintain and keep improving Freesound. Freesound has always been developed and maintained by researchers at the MTG and, as hard as it might seem to believe, Freesound has never had a single person fully dedicated to it (with very few exceptions). This is because Freesound’s current sustainability model is based on maintaining and improving Freesound as a “side-effect” of other research projects.

Our current sustainability model only allows us to maintain Freesound and does not allow us to spend time working on adding new features or improving existing ones. To change that, we need to rethink the current sustainability model and get to a point where Freesound can generate enough to be partially sustained by itself. This means to generate enough regular income so that we can dedicate more resources to Freesound and make it a better place both for its user community and the research community.

Because of the nature of Freesound and the philosophy of openness, sharing and reuse that we promote, we think that the sustainability model that best fits our case is one based on user donations (similarly to Wikipedia). Even though Freesound has been accepting user donations since the beginning, it is something we have never promoted. Currently, we are *only* raising around 2000€ per year in user donations, but we think we could raise much more. Considering the huge amount of users that regularly download content from Freesound, we estimate that with a very small yearly contribution from some of these users, we can collect enough to keep Freesound in good shape. For the sake of total transparency let us mention that, ideally, we want to raise enough to be able to maintain a couple of people dedicated full-time to do research and development with Freesound. We also want to be able to cover some of the expenses related to hosting and server maintenance. We estimate that this could be covered if we raise a minimum of 100000€ a year. This means that we need to increase current donations by a factor of 50. Please note that we are not thinking of imposing any yearly fee for the use of Freesound. What we are suggesting is that if we reach a significant number of donors, donations as small as 10€ a year would be enough to reach our goal.

In order to increase donations we are planning to start a campaign to actively promote them. We will do that by using banners and pop-ups, and occasionally sending some emails. Our main target is to reach those users who regularly download content and try not to impact users who are already contributing to Freesound by uploading sounds, by writing in forums or being part of the core community. We will also publish a regular report on how much money we raised and how we spend it.

Before making further steps with the donations campaign we wanted to share this with you and get your feedback. Feel free to tell us your thoughts in this forum thread that we set up for this occasion.

the Freesound team,

Alastair, Andrés, Bram, Frederic, Sonia, Xavier

Dear Freesounders,

As promised, here are some plots and charts and insights we got after analysing the more than 1,300 survey responses that you filled in some weeks ago. Warning: you’ll see some box plots in this post which might look strange. These are not really hard to interpret. In general, the more concentrated the “box” is in the top scores, the better.

Let’s start with your answers to the question “I’d describe myself as…”:

I’d describe myself as a…

NOTE: this question (and some others below) multiple responses where accepted per participant. The percentage shown in the figure is calculated over total number of participants in the survey (therefore the sum of all bars > 100%). 

Apparently, Freesound user profiles are quite heterogeneous :). A similar observation can be made looking at your responses about what do you use the sounds you download for:

The Freesound sounds I download, I use them…

And, again, we also observe significant heterogeneity in the kinds of sounds you’re interested in finding:

In particular, I’m interested in finding the following kinds of sounds:

The previous chart can be compared with the termcloud that we showed in our previous 2016 in numbers blog post, it seems to be quite well aligned, doesn’t it?

Even though you are interested in a wide variety of different sounds, the survey responses tell us that, when searching in Freesound, users typically end up finding what they’re looking for (or at least some other useful sounds). Also, the quality of uploaded sounds is also generally well valued.

When I search in Freesound…

And how often do you use Freesound? Well, apparently most of you are quite regular users and use Freesound at least a couple of times a month…

I normally use Freesound…

…but the majority of users only download sounds, do not upload:

I  use Freesound…

This is nothing new. In fact, we know from our data that less than 1% of users do upload sounds (way less that what the survey suggests). This kind of balance between uploading/downloading (i.e. producing/consuming) is quite normal in the web.

For the questions regarding Creative Commons licenses, it seems that Freesound users are generally aware of Creative Commons licenses and consider them to be important (including downloaders). Also, there seems to be a tendency of users who would like to be able to use more CC licenses than those currently offered (CC0, CC-BY, and CC-BY-NC), but this does not seem to be a very generalized or strong claim.

About Freesound and Creative Commons licenses

Finally, lets finish this post with a chart with the things you like the most about Freesound…

What are the things you like the most about Freesound?

Apparently you like lots of things about Freesound :), most prominently that fact that Freesound is free to use (as long as their sounds when license requirements are met), but also very importantly you value its philosophy and community of users.

Again, thanks for your participation in the survey, this certainly gives us insights to continue working in the future!

frederic

EDIT 18/4/2017: Thank you all for participating in the survey! After having collected more than 1,300 responses we decided to close the survey form. We’ll analyse the collected data now and get back to you with some conclusions in the coming weeks 🙂

Dear Freesounders,

Some of you will remember that in 2012 we conducted a survey in the Freesound forums with the aim of getting a better understanding of what are your motivations for using Freesound, what are the uses you give to it and, in general, know more about the nature of Freesound itself (you can see the posts here: https://www.freesound.org/forum/freesound-project/32400/). Some of the outputs of this survey were published in this research paper: http://mtg.upf.edu/node/2479. Now five years have passed and we decided to run a new survey to get an updated status of the Freesound community.

So please participate in the survey and help us improve Freesound 🙂
It should not take you more than 5 minutes. After we’ve collected all the data we will make it (anonymously) public for everyone.

Access the survey by clicking here

If you want to share or discuss anything related with the survey with the rest of the community you can do it in this forum thread.

Thanks for your participation!

the Freesound team

Dear Freesounders,

We are experiencing problems with Freesound servers and as a result Freesound is being quite slow. We are working on it and we’ll try to find a solution as soon as possible, we’ll let you know.

Thanks for your understanding!

EDIT: FYI, during the last weeks we made a number of improvements which significantly reduced the number of slowdowns, making them almost inappreciable most of the time. Yesterday (15th March 2017), we were finally able to deploy a new set of changes which should make the slowdowns completely disappear and make Freesound faster.

frederic

Dear Freesounders,

We collected some statistics from last year’s usage of Freesound that we would like to share with you. We’re sure you’ll find this insanely interesting 🙂 Let’s start with the most obvious one: the number of new sounds uploaded during 2016…

35,625 new sounds!

which corresponds to…

588 hours of audio!

Impressive isn’t it? This was achieved at a rate of approximately 100 sounds uploaded every day. As shown in the plot below, most of the sounds uploaded during 2016 were released under the Creative Commons 0 and Creative Commons Attribution, licenses:

All in all, the number of sounds hosted in Freesound is growing at a rate higher than linear, which means that it will take less time to grow from 300k to 400k than it took from 200k to 300k sounds. This is shown in the figure below, which plots the evolution of the total number of sounds since the starting days of Freesound:

In total Freesound now hosts more than 327k sounds, corresponding to more than 4400 hours of audio content. Can you guess when we’ll reach 400k? 🙂 But what about the kinds of sounds which are being uploaded? Well, this is somewhat difficult to answer, but to have an idea you can look at the tagcloud below, built using only sounds from 2016:

• NOTE: To avoid bias in the tagcloud, we have removed tags from a significant number of sounds that were uploaded by us (the Music Technology Group) as part of our Good-sounds project (a ~5000k single note sound collection).

The most used tag is field-recording, which means that the most common type of uploaded sounds are probably field-recordings, did you expect that? We can see anyway that other tags like loop, voice and effect are also pretty much used, which shows how heterogeneous is the content uploaded to Freesound. Did you ever tried playing with the random sound browsing mode? Give it a try, it’s a good fun 😉

This is the ranking of top sound uploaders for 2016:

• NOTE: Similarly as above, we removed our Music Technology Group user from this list which would obviously take the first position with the ~5000k single-note sound collection.

Thank you to all of the above uploaders, and also to everyone else who uploaded a sound in 2016 who doesn’t appear on the list.

And now what about some numbers about the consumption of Freesound content? Let’s  continue with another obvious statistic: the number of sound downloads (including packs) during 2016…

16,000,000 downloads!

Yes, 16 million downloads in a single year! At a rate of 50k downloads per day, Freesound has accumulated a total of 94M downloads since launch(!). In fact, as the figure below shows, the number of sound downloads is growing every year:

These downloads happen after users find relevant sounds when they’re searching. So the next obvious question is: what are people searching for? Again this might be hard to answer, but below you can see a “termcloud” of the most common search terms of 2016:

So yeah, wind, explosion, etc… By the way, this happens at a rate of 150k queries every day! If you want to maximise the number of downloads of your sounds, you now know what you have to do 😉

You’ll also like to know that during this year 2016 you’ve exchanged more than 25k  messages, added 1.2M sound ratings, written 2500 forum posts and more than 44k sound comments! All in all this shows how active and vibrant the Freesound community is. We would sincerely want to thank everyone that participates in the community by rating and commenting sounds, writing in the forums and also searching and downloading content. Let us also take this opportunity to very specially thank all  Freesound uploaders, more than 16k users have contributed at least one sound, and the Freesound moderators team, who do an incredible job on their free time to make Freesound possible and have indeed listened to all of it 🙂

So that’s it, keep on Freesounding and make these stats even more awesome next year!

frederic (on behalf of the Freesound team)