Hi everyone,

2023 is here and that means it is time for us to grab our year in numbers calculator and show you some statistics about last year’s Freesound activity. As usual, we start by showing some general statistics similar to those shown in previous years’ posts, and at the end we focus on a different aspect of Freesound, which this year has to do with the relation between text queries, sound plays and sound downloads. But let’s start from the beginning. The number of new sounds uploaded during 2022 was…

49,153 new sounds!

which corresponds to…

961 hours of audio!

This number is not as high as in 2021 (which reached a record of 56k new sounds), but it is still 3,000 sounds more than in 2020 and even more than in years before so the number of uploaded sounds is still growing strong (more on that below)! The average duration of sounds has not changed much compared to the previous years, and sits at 70 seconds per sound.

Here is the Creative Commons license distribution of these newly uploaded sounds:

In 2021, we observed that the distribution had shifted a bit in favour of using CC-BY sounds (and in detriment of CC0), but this year we see how the use of CC0 goes back to what it used to be before 2021 and it represents two thirds of the licenses used for new sounds.

With the new additions from 2022, Freesound now currently hosts an amazing total of 581,369 sounds, for a total audio length of 384 days and 9 hours. Here is the evolution of the total number of sounds since the beginning of Freesound, and the prediction for the future that we made last year:

As you can see, we are very slightly below the prediction, but that is to be expected as last year’s prediction would be biased by the record number of newly uploaded sounds. If all goes as expected, we should be closing 2023 with about 650k sounds and reach 700k in 2024.

Here is a tag cloud of the tags of the sounds uploaded during 2022:

The usual big tags are still big (field-recording, music, loop, synth, …), but this year it looks like noise has made a strong entrance in the top tags, partially because of the contributions by Hewn.Marrow who seems to have been exploring noise quite deeply (e.g. see this pack):

Here is the classic chart of the users who have contributed the most sounds in 2022:

Thanks everyone (not only those appearing in the table) for all the contributions! It is absolutely incredible to see that many new and high-quality sounds being uploaded every year 🙂

And what about downloads? The number of sound downloads (including packs) during 2022 was…

23,688,268 downloads!

This is even more than in 2020, when we observed a significant increase of downloads which we attributed to the increase of visitors due to COVID-19 lockdowns. All in all, users have downloaded more than 215M sounds and packs from Freesound!

The term-cloud below shows the most common query terms that have been used when searching in Freesound during 2022:

The top 10 terms are the same as in previous years, with slight variations in the ordering: rain, wind, explosion, music, footsteps, piano, fire, thunder, whoosh, and woosh. Approximately a thousand users search for the sounds of rain and wind every day! Also interesting is the use of both whoosh and woosh. According to the dictionary, the first spelling is used as a verb and the second one as a noun. Both are therefore correct, but Freesound does not use intelligence to somehow “link” the two terms. Maybe this is something to improve in the future!

Now some extra general statistics: In 2022, you sent 17k messages, wrote 1.2k forum posts, added 174k sound ratings, and made 49k sound comments. These numbers are decreasing with respect to the previous years in which we saw a big increase that we attributed to COVID-19 lockdowns. This behaviour was somehow predicted last year and now we see that these “community activity” indicator are close to what we observed in 2019.

And now, to conclude this blog post, we will provide some statistics about sound plays (i.e. every time you listen to a sound while browsing Freesound), sound downloads and their relation to user queries and data transferred. If we take the total number of sound downloads in 2022 shown above we can easily see that this means a rate of 2,700 sounds downloaded every hour. But have you ever wondered how many sounds are being played every hour in Freesound? Well, this varies a lot depending on the time of the year and it is actually difficult to estimate even for us, but after some number crunching we have estimated that 8,600 sounds are being played every hour. This could be interpreted so that for every sound downloaded, approximately 3 sounds have been previewed before. What might be more surprising however is that if we look at the number of user queries per hour, we see that about 6,500 text queries are made every hour, which means that only 1.3 sounds are played per user query, and also means that at least a 60% of the queries result in no sounds being downloaded. All in all, considering all sound downloads and sound plays, our severs stream about 2TB of information per day. If you look at the blog post from last year (2021 in numbers), it was actually briefly mentioned that the amount of data transferred was 1TB per day, but our estimate for this year has doubled. This could be due to having wrong estimates in the previous year, but will also be most likely related to our changes in the downloads infrastructure which allowed for much better download speeds and might have therefore resulted in overall higher amount of data transferred every day.

That’s it for this year’s post, thanks for reading and we hope you enjoy a 2023 full of sounds 🙂

frederic, on behalf of the Freesound Team

[Guest Blog post by Mehrdad Pourzaki, Lead Movement Communications Specialist at Wikimedia Foundation]

Dear Freesounders,

The Wikimedia Foundation, the global nonprofit that hosts Wikipedia and other Wikimedia projects, including our audiovisual repository the Wikimedia Commons, is hosting a global contest for The Sound of All Human Knowledge. Submissions are open now through 10 October. The Wikimedia sound logo will be used to identify Wikimedia content across a wide range of uses and smart devices, including personal voice assistants. You can learn more about the contest on wiki, including campaign materials and how-to videos. There is also a humble prize for the winning sound. The Freesound community is primed to participate since you already have a knowledge of and appreciation for audio and free licensing. As the organizers, we warmly invite the Freesound community to take part in this global contest. We’d love to hear from you and please help us reverberate the call.

Wikimedia sites receive 750 million page views a day and that knowledge is further reused by search engines, other platforms, and applications that don’t consistently inform their users that what they are learning comes from Wikimedia and the thousands of volunteers around the world who contribute. Voice assistants are on the rise everywhere in the world and while there is a lot of expertise in our movement when it comes to visual logos and photo competitions, we are excited to be exploring new parameters when it comes to sound and audio production. Play your part in our global search. On September 29 we had a drop-in clinic with our expert partners MassiveMusic, you can find some information in the link above.

Mehrdad Pourzaki,
on behalf of the Wikimedia sound logo project

___

Do you want to write a guest blog post for the Freesound blog? We like that! Please send a proposal using our contact form

Dear Freesounders,

Today there’ll be intermittent downtimes between 13:30 and 16:00 CEST due to maintenance tasks. We’ll let you know when all maintenance tasks are done!

Cheers,

edit: all maintenance tasks finished and we’re back to normality

[EDIT: new Terms of Use, Privacy Policy and Cookies Policy have just been released and are accessible through the links below]

Hi everyone,

This post is to let you know that have been working on an updated Privacy Policy and Terms of Use of the Freesound website, and are also starting the process to migrate old Creative Commons licenses to newer versions. The changes will be effective on April 4th 2022, and all registered users browsing Freesound will be asked to provide consent about the updated terms before continuing to use Freesound. In addition, users that have uploaded sounds will also be given the option to update the licenses of their sounds (more on that below).

The updated terms of service and privacy policy are intented to clarify your rights as users, the ways in which we process your data and, in general, to protect you and us against incorrect uses of the site. This update does not alter the Freesound philosophy in any way, it is just a necessary step that we need to make in order to better comply with regulations such as the GDPR.

Please, take some time to read our updated Privacy Policy and the Terms of Use. Also, it is worth mentioning here that we have rewritten Cookies Policy to make it more clear so you might also want to check it out. We prepared a brief summary of contents of the Privacy Policy and the Terms of Use which we include here:

• Uploaded sounds are released under one of the CC licenses you choose. Other content (metadata, tags, comments, etc…) is provided as CC-Zero.
• Your activity in Freesound is public and can be seen by all users (except your private messages).
• You are responsible for your posted content.
• We (Freesound) keep your data confidential and apply security measures to keep this confidentiality.
• We (Freesound) don’t warrant that all the data on the site is correct. We are not liable for information not posted by us on this site.
• We (Freesound) may remove accounts of users that do not comply with the terms of use or who do not behave correctly (spam, offensive comments…).
• We can use anonymized data collected from Freesound (sounds, statistics about sounds, tags, comments, downloads…) for research purposes.
• You agree to receive email notifications in accordance with your email preferences.

Beyond providing consent for the updated terms, starting on the 4th of April users with sounds will also be given the opportunity to upgrade current Creative Commons licenses from 3.0 to 4.0. The main difference between these two is that Creative Commons licenses 4.0 are more user-friendly and more internationally robust, but the general terms are the same. More information can be found here. The option to upgrade licenses is only offered to users who have uploaded sounds with any of the “old” CC-BY or CC-BY-NC licenses, and will also be available in the user settings page as long as user has sounds with old licenses.

Please let us know if there’s anything you’d like to discuss about these updates by commenting to this post. Thanks everyone!!!

frederic, on behalf of the Freesound Team

Dear Freesounders,

Welcome to our 2021 Sustainability Report! Here we give an update on how 2021 went for Freesound in terms of sustainability, and present our plans for 2022. You’ll see that most of the information is very similar to previous years’ posts, but still, we think that this will be interesting for you. As usual, the report is split in a number of sections discussing specific aspects that contribute to the sustainability of Freesound, and a final section with a summary, conclusions and future perspectives.

Sound uploads

Sound uploads are an essential part of the sustainability of Freesound. In 2021, 56,783 new sounds were uploaded to Freesound. This number is even bigger than in previous years (see more stats here). We saw this year that a significant number of the uploaded sounds (~15%) come from professional sound design studios that decided to upload part of their catalog and release it under a Creative Commons license. This is interesting because it further highlights that Freesound is a relevant tool not only for hobbyists and amateurs but also for sound professionals, and it also promotes the use of Creative Commons among professionals. All in all Freesound continues to be doing very well in terms of sound uploads.

User donations

User donations fell a bit in 2021 compared to 2020, but 2020 was very exceptional due to the COVID19 lock-downs which happened around the world, resulting in a 30% increase in donations. With respect to 2019 (pre-pandemic times), 2021 still represents a 25% increase in donations, and we received almost 59,000€ from 6,600 individual donations. We’ll have to see whether this trend continues and 2022 gets closer to pre-pandemic times, or whether the amount of donations is maintained with similar numbers. Thanks to the boost of donations in 2020 we’ve been able this year to spend a lot of efforts working on the implementation of the new user interface and also on the new infrastructure to which we have been (and still are) migrating. Let us take this opportunity to thank again everyone who donated to Freesound!

Contribution from UPF

Freesound is an initiative of the Music Technology Group, a research group of the Universitat Pompeu Fabra (UPF) in Barcelona, Spain. In 2021, the UPF contributed to Freesound similarly to previous years, however this year the change of the infrastructure has also resulted in more involvement from university’s IT support staff. We have moved to an infrastructure based on a Kubernetes cluster maintained at the UPF level. This has allowed us, among other things, to make Freesound faster and to significantly increase download speeds. We now estimate a daily consumed bandwidth of around 1TB of data. Beyond IT staff, researchers from the university also dedicate time to Freesound related activities (either research, development or administration) and are paid by the university. As you can see, the contribution from UPF is huge, and it is only thanks to the combination of the different sustainability streams discussed in this post that Freesound is sustainable.

Contribution from research grants and Freesound-related research at UPF

Research is at the very core of the Freesound philosophy and, in fact, it is where it all started. We have carried out lots of research activities around Freesound (see some details below), but, similarly to 2020, in 2021 we did not receive any new big research grant with a primary role for Freesound (like the AudioCommons project that we coordinated a few-years ago). As explained in our previous sustainability report, at the end of 2020 we received a grant from the Grant for the Web call to experiment with the application of Web Monetization technologies in the Freesound Licensing project, and the work done for this grant was carried out during 2021. We were able to experiment with some ideas and build a proof of concept for Freesound Licensing, but the project has not advanced much further and we are still in the process of assessing viability from a legal perspective.

As a quick summary, these are the research activities that we carried out at the MTG in 2021 and in relation to Freesound:

• Further development and maintenance of Essentia, the audio analysis library that powers Freesound sound analysis.
• Research on methods for automatically classifying audio events and development of artificial intelligence models that we plan to deploy in Freesound during 2022.
• Research on methods for interpretable machine learning in audio classification.
• Further research on clustering methods to be potentially applied to Freesound search results. We have advanced a lot on the implementation of these methods in Freesound and are quite ready to make them available to the public, but some more efforts are still needed.
• Research on hardware interfaces for accessing Freesound content (the SOURCE sampler that was announced in the forums)

If you’re interested in learning more about the research that happens around Freesound (i.e. using Freesound data) not only at the MTG but also around the world, be sure to check the papers section of the Freesound Labs website.

Commercial usage of the Freesound API

Freesound has an API endpoint which allows third parties to develop applications that incorporate Freesound content. Usage of this API is free for non-commercial purposes, while commercial use of the API requires a commercial license. In this way we make sure that commercial applications using Freesound also contribute back to the community. Note that this is independent from the license of the sounds themselves, which need to be respected regardless of the API usage agreement. In 2021 we increased the number of license agreements and also re-negotiated some of them (based on usage) and were able to increase the yearly income to reach ~6,000€. We spent this money in the same development efforts described in the User donations section above.

Summary and perspectives for 2022

As it is shown in this report, 2021 has consolidated the already existing trends in terms of sustainability and we have been able to translate this into many improvements in the website. We have not been able to further advance with potential new sustainability sources like the Freesound Licensing project and also we have not obtained new research grants, but nevertheless we have been able to carry out research which is more focused on Freesound and can more easily be transformed into actual features for the website. In 2022 we expect again to spend many development efforts in Freesound, with the main goals being to finish the new UI and to deploy research outputs into the website (we have already done significant work in that direction during the first months of 2022).

We’d like to finish this post by saying thank you to everyone who contributed to Freesound during 2021, in particular to those who donated and those who uploaded and moderated sounds. We’ll let you know how things go next year in the 2022’s sustainability report!

frederic, on behalf of the Freesound Team

Hi everyone,

One year more has passed and here we are with our classic year in numbers post in which we give you some statistics about last year’s Freesound activity. We start by showing some general statistics similar to those shown in previous years’ posts, and also extend a bit on a specific topic which, in this year’s post, is about a very commonly asked question: how much disk space do all the sounds in Freesound take?. But let’s start at the beginning. The number of new sounds uploaded during 2021 was…

56,783 new sounds!

which corresponds to…

1075 hours of audio!

About 10,000 more sounds were uploaded in 2021 than in 2020 and 300 more hours of audio! In last year’s post you’ll see that the increase of sounds was not that high, and the average duration of the sounds had significantly decreased (most probably due to the upload of a large short sounds collection), but this year we’re back to the usual average sound duration (which is 66 seconds, by the way), and therefore the hours of audio is big again. We’ve never had this many sounds uploaded in a single year!

What about the license distribution for these 56k newly uploaded sounds? Here it is:

The distribution has shifted a bit compared to previous years’ distribution and we can see that the use of CC-BY has increased, in detriment of CC-BY-NC and CC0. We will see in the future if this consolidates into a trend, but it is most probably just part of the yearly variability in the usage of licenses.

With the new additions from 2021, Freesound now currently hosts an amazing total of 536,763 sounds, for a total audio length of 345 days and 22 hours. Here is the evolution of the total number of sounds since the beginning of Freesound, and our prediction for the future:

As you’ll remember, in May 2021 we reached 500k sounds, and if our prediction is accurate we should be very close to 600k by the end of 2022.

Here is a tag cloud of the tags of the sounds uploaded during 2021:

The usual big tags as still big (field-recording, music, loop, synth, …), but tags like nature, ambience, hand, and drum seem to be much bigger this year when compared to the past year. The exceptional high usage of the tags hand and drum seems to be because of the contributions by stoltingmediagroup, who has uploaded more than 11k sounds this year (see below), including various packs of hand percussion instruments and drum sounds. The increased usage of the tags nature and ambience is probably due to the contributions by klankbeeld, who has also uploaded many more sounds than usual this year. Here are some examples of these sounds uploaded by stoltingmediagroup and klankbeeld:

Do you want to see a chart of the users who have contributed more sounds in 2021? Here it is:

Thanks everyone (not only those appearing in the table) for all the contributions! It is absolutely incredible to see that many new and high-quality sounds being uploaded every year.

And what about downloads? The number of sound downloads (including packs) during 2021 was…

20,948,083 downloads!

As you can see in the plot below, that number is closer to 2019 than to 2020, when there was a significant increase of downloads (related to the increase of visitors due to COVID-19 lockdowns). All in all, users have downloaded more than 191M sounds and packs from Freesound!

For those curious about what is people searching for in Freesound, here is the cloud of query terms that are used by Freesound users when searching for sounds:

As opposed to the tag cloud that we showed before which can be influenced by particular actions like a user uploading a large collection of sounds, the cloud of query terms rarely changes as it is hard to influence the statistics of the millions of queries made every year. The top 10 queries include: wind, rain, music, explosion, piano, footsteps, birds, whoosh, woosh and fire.

Some extra general statistics: In 2021, you sent 43k messages, wrote 1.9k forum posts and made 63k sound comments. All numbers have gone up significantly with the exception of sound comments, which is very similar to last year (but that had a big increase last year). Again, very similar numbers to those of 2018. The number of sounds ratings has also been quite high, 208k sound ratings in 2021. This is a bit less than in 2020, but 2020 had a huge increase compared to 2019 so we’re doing very well. Last year we observed an increased activity in Freesound which we attributed to the increased visitors due to COVID-19 lockdowns. This year we see that some indicators are getting back to normal but we are still reporting very high numbers which makes us thing that many of the users that joined last year must have liked Freesound and stayed.

Finally, to conclude this blog post we would like to put our focus onto the disk usage of Freesound over the years. This is a question that we get often: how much disk space do you need to store all Freesound sounds? Well, here are some answers! Below you’ll see two plots, the first one showing the new disk space that is taken every year (that is to say, the amount of extra Gigabytes that we need every year to store the newly uploaded sounds and related files). The second one shows accumulated statistics for the total disk space taken by Freesound sounds and related files. The related files include the images that we generate for the sounds, the output of our audio analysis and the file previews that are used by the sound players in the website.

As of today, all Freesound sounds and related files take about 7 Terabytes of data. Surprised? Normally people say that they thought it would have been a lot more! But actually this is quite a lot 🙂 Note that many uploaded sounds are very short, and also that some are uploaded in compressed formats. So even if we’re hosting more than half a million sounds, the requirements in terms of disk space are not crazy. This can be seen as all the information that has been uploaded to Freesound. If we look however at the amount of information that we transfer every day mostly in the form of downloads, we see that we transfer around 1 Terabyte per day to users downloading sounds, packs and playing sounds from the browser. Oh wow!

That’s it for 2021 in numbers post! Thanks for reading and we hope you enjoy a 2022 full of sounds 🙂

frederic, on behalf of the Freesound Team

Hi everyone,

Finally! As some of you will know, we have been mostly silently working on a new UI for Freesound during the last years. Even though it is not completely finished yet, we are extremely happy to announce that, from today,  you can start using the new beta UI and give us some feedback so we can make it better, finish it, and eventually replace the current UI.

Just as a brief bit of context, in mid 2017 we contacted product designer Marc Ruaix and invited him to design a new Freesound UI to give it a fresher look. What initially was only going to be a small update of the UI, ended up becoming a full rebuild of the Freesound frontend and parts of the backend. We have now ported most of Freesound pages to the new UI and now ready to open it for for beta testing. While we call the current UI (now old!) Nightingale (NG), the new UI is called Beast Whoosh (BW). We expect to continue working in the new UI during the first half of 2022, porting the missing pages and incorporating your feedback, and we expect to completely replace NG with BW at some point in 2022.

Do you want to participate in the beta? Everyone can do it, right now! You’ll find instructions in this forum post. Please read the instructions carefully, and keep in mind that you’ll most likely find some things not working as expected. Please use this forum to provide feedback.

We hope you enjoy Freesound with the new fresher look. I’ve been already using it for a while and I’m really enjoying it 🙂 Also, let me take this opportunity to wish you all a happy new year!!!

frederic
on behalf of the Freesound Team

The Music Technology Group and Phonos will offer the workshop ?Composing with Freesound? with Laura Llaneli & Frederic Font, from 12 to 16 July 2021 online.

In this workshop, participants will explore in depth the possibilities of Freesound as a creative tool for music composition and sound design. The workshop will cover the basics of free licenses, downloading sounds, recording, editing and uploading sounds, as well as advanced features unleashed with the Freesound API, including text and similarity advanced searches. Also, the workshop will show the potential of the Freesound API with applied examples in audio mosaicing and loop generations and cover other advanced applications towards machine intelligence. Sessions explaining theoretical concepts will be followed by sound and music composition labs. At the end of the week, students working in groups are expected to prepare a radio show using Freesound content, that will be aired on teslaFM during the summer.

Duration of the workshop: 15 hours (Mon. to Fri. 14?17h CET)

Format: Online (zoom + slack)

Language: English

More information and registration: https://www.upf.edu/web/phonos/workshops#freesound

This is one of the workshops of the Phonos Summer Camp, a series of workshops on music and technology for musicians, composers and creators in general.

Find more information on the Phonos Summer Camp here: https://www.upf.edu/web/phonos/workshops

[EDIT: we’ve been able to stop some bots that were abusing the Freesound search and now search is working for everyone again. We’re implementing some measures to stop this from happening again.]

Hi everyone,

Yesterday we had some issues with the search server that were fixed at some point, but last night the issues started again and still persist. We’ll work to fix these as soon as possible and let you know when done. Thanks for your understanding!

Meanwhile you can still to some browsing by navigating user profiles and/or using the random sound feature: https://freesound.org/browse/random/ 🙂

frederic

Hi everyone,

In our recent 2020 in numbers blog post we predicted that we would reach 500k sound uploads at some point in 2021 and… it has just happened! At some point during the past weekend, the number of sounds which are available for download in Freesound has surpassed half a million. Note that by sounds which are available for download we mean sounds that have been already moderated and approved, and here I am not counting sounds that were deleted in the past. Huge congratulations to the whole Freesound community! This is a remarkable achievement, and we could have not imagined at the beginning of Freesound that one day it would host more than 500k Creative Commons sounds uploaded by thousands of people around the world (and downloaded by millions!).

To celebrate this achievement, we are also very happy to say that, as part of the changes in the Freesound infrastructure we’re making these weeks, we’ve managed to significantly increase Freesound speed, including downloading speed. In fact, our monitoring tools indicate that Freesound is now very fast, and our subjective experience using the website also says so. We have just barely started all the infrastructure changes that we’ll carry out in the next months, so many things still need to be changed and improved and you can expect some issues here and there. But the most processing intensive part of Freesound has already been moved to a new environment of the University IT infrastructure that we can scale much easily and will allow us to adapt much better to the future needs of Freesound. Hurray!

We could not be happier to bring you these news 🙂
We hope you enjoy browsing more than 500k Creative Commons sounds with the new super-fast Freesound!

frederic, on behalf of the Freesound Team