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…
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 120k 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)
Hi frederic and team – and congrats on all those healthy stats!
One thing I noticed when I was making a dataset of field-recordings taken from Freesound was that things tagged “field recording” are quite often… not field recordings! Fairly often, they’re simulations of field recordings, or musique conrete created from field recordings. I wonder if this was the case in 2016 too.
Thanks for your comment! This is indeed true, and I guess it is also the case for 2016. In Freesound the term “field-recording” seems to have been adopted to cover these cases you mention. Also other ones like urban recordings are often tagged as field recordings (although in this case it could be easily argued that urban recordings are field recordings too).
Hi freesound team
thanks not just for the interesting stats but in general for investing all your work into this wonderful site!
Yet, since 2 weeks I am experiencing access problems, getting a server error displayed pretty often telling me something like “oops we did something wrong”. Is this just the case for me in Switzerland or is this a general problem?
This is a general problem we are aware off. We’re working on it, it has to do with our server infrastructure. We have a couple of things to try which we think will improve the situation. This will hopefully happen in the coming days and we’ll let you know.
Thanks for pointing that out and for your nice words!
I don’t understand this: the tagcloud shows a HUGE field-recording tag… but the most active uploaders during 2016 obviously were NOT field-recordists. A contradiction?
(for few samples I’m not in the list, ouch! 😉
… except for klankbeeld and AcousticMemory, I mean
The list was top 20, but we removed the MTG user and it remained top 19, so maybe if we kept top 20 you would appear 😉
Re the field-recording question, I think the tag field-recordig is used by many people in Freesound and not always with the strict field recording meaning (see some comments above). However, to better answer your question, I’d say that there does not need to be a strong relation between the most active uploaders and the resulting tagcloud. Take for example the most active uploader with ~450 sounds, this represents 1% of the total number of uploaded sounds. It has an impact yes, but it is not defining the tagcloud. Moreover, if many different users use the field-recording tag, it will always appear the biggest because it is the most common, not because it has the majority of sounds. Top uploaders might be uploading sounds that are not field recordings, but these sounds might be of different nature and don’t have a common tag, therefore they don’t appear as dominant in the tagcloud.
… and producing/uploading 100 field recorded samples require a lot more time than, say, 100 synthetic samples (or so I believe). But thats the way it is, when big heterogeneous numbers come into play… it’s somehow ‘unfair’ to assume that 101 is larger than 100 🙂
Anyway, who cares the ‘ranking’ stuff, thank you for the great job. Long life to Freesound!
Very good blog post.Really thank you! Fantastic. ddbeefckbfabddgg
Thanks Frederic and team for this well written and superb article. These are indeed massive and incredible numbers presented wonderfully in the form of figures and statistics. Freesound has always been my number one platform where I can listen to, acquire and upload sounds. The working force behind Freesound deserve all support needed.
Long Live Freesound.