Freesound Forums

I find when I read through the forum posts, some comments are so insightful that I want to immediately go and look at the samples contributed by the poster.

Unfortunately, it doesn't seem to be easy to do that. Clicking on a poster's name takes you to their forum profile, which is different from the profile that lists their samples. It's possible to convert one into the other with some creative URL editing, but that's kind of cumbersome to do every time.

Would it be possible and easy to include a link to the poster's contribution along with each post or unify the forum profile with the contribution profile? Or perhaps that feature is already there and I've again missed it.

Thanks

mab

Hello Mab,

Bram may give you a fuller answer, but apparently this link from the forum to the database is not possible, as the forum code was some kind of imported system and not written by Bram.

What I do is probably what you have discovered. Click on the contributor's handle to go to his forum page, and extract the user id from the end of the url. Then click on your own name to switch to the database section, and paste the userid over your own id= in the url.

Cumbersome, as you say.

Mike

You could probably write a small PHPbb hack/add-on to do the cut-and-paste for you and create a link somewhere on the profile page...

Although I personally just set my homepage as my sample profile.

a quicker way would be to copy the username, go to search on the left hand side and uncheck everything except usernames!

alternatively, click the small freesound icon below the user's name.

- bram

Bram
alternatively, click the small freesound icon below the user's name.

- bram

excellent! thanks

mab

Awesome! Thanks bram.

While I was at it I also added the "Sounds added: XX" on the left. Plagasul and me think I could probably throw away the "Joined: date" entry as it's not that useful. Sounds added + number of posts should be more than a "believability rating", right?

what d'u guys think?

- bram

I like that idea. It's true that "community standing" has nothing to do with join date around here, and everything to do with the quantity and quality of uploads (and post count to a lesser extent.)

If the ratings database was a little bit more consistent (i.e. had more ratings in it) you could even have an "average rating" figure, though that might be a bit unfair to some for various reasons...

Instead of that, what about a download counter?

These probably wouldn't be appropriate for the forums, but might make an interesting addition to the profile page.

I actually regularly look at "Joined date" - I find it interesting to see who are the "old hands", who have joined recently but have made a big impact already, and who has taken a short or long time to discover Freesound, tho' being perhaps a professional sound engineer etc.

By the way, the freesound icon below the user name only appears as a small square with a red cross, on two out of three PC's, all running windoze xp and IE6. The avatars appear on all PC's.

cracking idea bram, although mine looks a bit shamefull, with 0 uploads

i must start recording.....talking of me recording, a couple of replies to this would be handy

Halleck
It's true that "community standing" has ... everything to do with the quantity and quality of uploads (and post count to a lesser extent.)

If the ratings database was a little bit more consistent (i.e. had more ratings in it) you could even have an "average rating" figure, though that might be a bit unfair to some for various reasons...

As it is said that the number of downloads of a sample is a better measure of usefulness than rating, how about a user's total sample usefulness rating something like the total number of downloads that user's samples have had in the last 7 days ... and a second figure: the number of d/l in total for that user's contributions over the last 7 days, divided by the total number of samples they have contributed. This latter being a user's average sample usefulness rating. That should keep the computational side of FS busy!

Edit: If such computations were done weekly, 'twould probably suffice.

I'd say that's more of a "popularity" rating than a "usefulness" rating... there's lots of undiscovered gems on the site that don't get downloaded often.

Halleck
I'd say that's more of a "popularity" rating than a "usefulness" rating... there's lots of undiscovered gems on the site that don't get downloaded often.

You are absolutely right. Recently I've been toying with the 'random sample' feature, and virtually anything can come out! This random feature is a great source of information on the kind, length,quality, format of sounds... That's cool, I can easily imagine someone doing a Ph D on it in the near future. A 20000 item database is a piece of cake for an statician (which I'm not). Best of all, with some programming one could resample the database as many times as needed. This way you don't have to assume anything about the distributions being normal or not (they call this Monte-Carlo methods).

Mike, I dont' think that usefulness can be quantified. It depends on the subject that uses the sound so it's subjective by definition, I guess. Just like the beauty of a sound, the mood that it conveys, or any other subject-specific trait. As for popularity, the number of downloads looks like a fairly good index.

I must confess, what's intriguing for me is that 'popular' samples tend to be downloaded again and again as time passes, whereas 'undiscovered gems' (to quote Halleck) remain forgotten. Can this behaviour be explained on the basis of sample quality only? I think not. I would rather say that the average user is not fond of discovering, but of following others. Could we call this gregariousness? So, our behaviour as users is another valuable source of information we are all building for the future (besides the database itself). Marketing guys, here it is. For free!

Everything about FS is so damn COOL!

dobroide
I must confess, what's intriguing for me is that 'popular' samples tend to be downloaded again and again as time passes, whereas 'undiscovered gems' (to quote Halleck) remain forgotten. Can this behaviour be explained on the basis of sample quality only? I think not. I would rather say that the average user is not fond of discovering, but of following others. Could we call this gregariousness? So, our behaviour as users is another valuable source of information we are all building for the future (besides the database itself). Marketing guys, here it is. For free!

First I thought that users mainly download stuff that's on the front page, but:

searched nr_times
wind 1320
thunder 1200
rain 1200
scream 1090
bell 1081
explosion 1076
gun 1014
car 906
crowd 834
piano 807
water 788
bass 772
sex 743
fart 701
phone 700
guitar 698
drum 669
fire 669
alarm 665
Siren 646
beep 646
telephone 633
train 625
dog 591
voice 579
horror 574
door 572
snare 569
birds 569
crash 564
radio 559
scratch 555
record scratch 520
kick 516
laugh 494
applause 493
drums 484
footsteps 483
storm 482
cat 463
horn 452
bird 448
ring 447
click 445
heartbeat 431
violin 430
trumpet 430
music 405
ambient 404
forest 402
Camera 400
sonar 400
clock 384
glass 380
bells 378
gong 374
sea 374
scary 370
ocean 364
horse 362
bomb 362
space 358
vocal 352
traffic 351
static 349
nature 345
heart 344
loop 332
war 331
police 325
thunderstorm 321
punch 320
chant 319
whistle 315
city 314
metal 306
techno 305

...

These are the searches people do. As you can see, thunder really is popular

- bram

Bram

First I thought that users mainly download stuff that's on the front page, but:

searched nr_times
wind 1320
thunder 1200
rain 1200
scream 1090

...

These are the searches people do. As you can see, thunder really is popular
- bram

Don't take me wrong, I by no means meant that Humphries' thunder isn't worth thousand downloads. But I think there is (perhaps) a general 'pioneer' effect, which renders much-downloaded files appealing to more people. Sort of "if all these guys downloaded that then it can't be bad". Just like in pop music: once a band gets a large audience it's very likely that their next album will be a hit. Well, music is just a type of sound, and FS consumers (downloaders) could well behave in a similar way.

So there are themes clearly 'preferred'... Bram, I'm thinking you have access to lots of info potentially interesting about lots of people... Not as much as Google, but hey...

f*** man, sex is only 13th in the list, something has *Got* to be done about that.

Anton
f*** man, sex is only 13th in the list, something has *Got* to be done about that.

glad to see i'm not the only one who only noticed sex in the rank

I'm getting a bit lost as to where this thread is going ...

A couple of points. So folk think my suggested scoring system is a measure of popularity rather than usefulness - OK - (rather splitting hairs in my book, but never mind, I mean presumably a user downloads a sample because they can see a use for it) - so my question is:- would folk like to see contributors rated in such a way?

Secondly, when I upload a new sample which obtains a reasonable number of downloads in a short time, I find that apart from the faithful downloaders like Schulze (thankyou my friend), the big majority of the downloaders have only joined FS in the last few days. This suggests to me that users join FS, grab what they need at the time, and disappear without trace. Why can't we hold on to more users for longer? Why don't they come back again and again for new stuff?

Mike

I believe that the main reason of Bram's post was to delete the Joined part of the text block below the user picture, because it looks big. We are trying to determinate if Sounds Added + Post is enough as a hint to trust someone, and i believe it is.

Let's delete the "joined" one