Sound Upload - Cost $10/per 10 sec

Even with the sounds cap to a limit of 10 secs max for each sound we upload, users in SL are still able to cut all their songs in 10 sec each and put then together for playing using a scirpt. Instead of limiting us to 10 sec per sound, it should be $10/per 10sec per sound.

Cause we all know that listening to a looping sound track over and over and over again, causes us to all go in a killing frenzy in the outlands. Now we all don't want that, now do we.

The the way sound works in SL by design (I assume) is not for playing pre-recorded songs.

I, for one, like this.

I was thinking that it may seem the tech is built this way to discourage playing full songs, but that's probably not the case since you can chain files to create full songs anyway. The exact best way to script perfectly seamless playback of chained sounds doesn't appear to be well known at all, but there are a number of scripts out there doing a fair job.

My guess is a lot of the reasons for small sound chunks are technology related, and probably some legaleze. 10 seconds is a manageable size, both for load/playback in clients, for scripting, and for server side storage. And it does have the effect of preventing the SL storage servers from becoming Napster with an interactive 3d UI.

That's my guess anyway. I thought the same thing about pay-per-second (I figured it's a buck a second now anyway just keep the meter running), but as I worked with the sound scripting and came to understand a little of how the routines and clients managed sound, then thought about the distribution headaches a la Napster, 10 second clips started to seem pretty reasonable.

Ok I can understand that we don't want SL to turn into a virtual files networks like Napster or Kazaa, as we already download a lot just to run SL smoothly....*scarcasim* But what is stopping people to use these scirpts. If people are taking the time to work on good scirpts that take hours to make what is stoping them from taking that time to cut apart a song in 10s each file, and then scirpting it together. This may not be something to worry about now in SL, but I believe that it is an important factor as this is one of the first full stream based games of its kind. I can't seem to understand the difference of downloading a one minute song than to download a 1 min song in section of 10 seconds. Ok maybe you download only the sections you want to hear while you are in the area, instead of downloading the whole song. But does this prevent them from taking the file usage on the servers??

I belive that the 10s is good, I am not disagreeing with that at all. It allows the users not to get over lagged when my boom box is playing music. They only download the sounds when they are in the area.... Crap! I just figured it out. They are doing it right. Just took me a lot of thinking about the different "IF" factors to find out that they are doing it the best logical way.

Well it seems this post was a waste of my time. I just figured out my own problem for myself.... *sigh* Not again!!!

Ok delete this post.

Watching a lightbulb go off over somebody's head, even a figurative one, is well worth the words methinks. Welcome to Second Life

I just dont like the fact i have to use wav files

What would you prefer to use?

PCM WAV files have been one of the leading standard audio files for years and years now (along with AIFF). When you buy a sound effect CD, or download sounds from the internet, chances are they're gonna be in WAV format.

The only real type of sound that's really popular in other formats right now is music encoded as MP3s, which 9.99 times out of 10 is illegally copied commercial music, which guess what, isn't allowed in Second Life.

If I were to guess at an immediate reason for why SL didn't accept straight MP3 files, it would be because they're trying to disuade you from uploading copyrighted material which is not allowed to be stored on their servers and could get them into trouble.

But that does stop me from converting mp3 to wav

doesn't i mean, damn my typos, changes my whole sentence

Whether or not it's related, (file under errata), as I understand it, MP3 is a "closed" format, which is why editors will convert it to another format temporarily for editing (the two I use do anyway), then actually resave a new MP3 file. I'm no digital sound expert, but did work heavily with MS streaming media dev for a while (adding script commands and triggers to existing media files and doing conversions of libraries), and one of the big things they used against MP3 was once the file is baked it's basically a locked box, which compromises a lot of what you can do with the file. Wav (and wma) files are very easy to inspect and tear apart.

So what I'm saying is that the reason for everything is that MS is evil.

hahaha..

Yeah, I was gonna say some other stuff similar to what you said. That MP3 is also a pretty useless file format from a coders point of view. WAV is definitely way more flexible and easier to work with.


And Hikaru, nope, doesn't stop you from doing that You'd be surprised at how many people don't realise you can convert back to WAV though. Like I said; more work for the user = more disuasion.

ogg?

Heyevry one, I'm new, just here to say I really would like a pay per second thing, though there should still be a limit, but not *just* ten seconds, more like 4 minutes... five to make it even.... but then again, I'm new, so I don't know whats good for me...

10 seconds is a managable chunk that can be downloaded while the previous 10 seconds is played. I don't want to have to download a 5min sound clip every time I get near someones radio. What happens if lots of people have radios? And I'm just flying along and pass about 10 of em in a 1 minute time span, am I then going to be downloading 50 minutes of audio that I won't even hear?

Realtime streaming audio is not easy on the bandwith meter, and having a lot of them going (which has a high potential) is not going to be fun.

My understanding was that sounds in SL are meant to be for sound effects, birds singing people saying hello, gun shots or fire crackers. If you can make the system designed best to do those things play your music file, then more power to you. Don't change the system around being able to play music though, please.

O the joy of MP3 -> WAV converterz

I totally agree with Ama here. I think the system is fine the way it is. It is still possible to string a whole song together with parts, and while it's not perfect, it's good enough as is. Would I like to see it better? Yes. However, when I think of what people could do with it if it were easy to upload whole songs into the system, I realize that having some form of difficulty barrier makes a WHOLE lot of sense.

With no barrier in place and support of every format under the sun, people will upload all kinds of garbage and noise and just make the world miserable in general.

By supporting a limited sound format range, and by only allowing small sections, that becomes more work than the average casual user will want to go through, leaving whole songs and longer-length audio to those who are willing to work for it (and are generally well-intended).

For the curious...

The SL engine uses Ogg-Vorbis as the audio compressor.

Does that mean we can upload .OGG files?

1) if you can do it chances are someone will
2)read #1

if you can chain song clips into one full length song someone wil ldo it. Most of the .wav files i chose to upload wont even upload. I thik that we should be able to upload at last 30 sec at a time seeing as that is all thet we need anyway. or at least a better chain script that is easy to use and wont cause pausing in between clips should be handed out as kind of a default thing

This may be a dumb question but if we ul .ogg files, would we be able to see them much less use them? .ogg is a compressoin yes? They are not playable in that format? I've had a boatload of .ogg files and went though alot of time hunting through them and converting them to get the sounds i wanted. If there was a way to play .ogg in that format i wish i knew.

Zeb

If it seems interesting to many folks, we can fairly easily add direct uploading of ogg files. We would likely only support a subset of compression options/rates though, to keep the sound quality more uniform.

Comments welcome.

ogg support would be good, even if it does only support a subset of the various encoding settings. it's usually easier to just resample a sound clip, rather than converting it to a different format.

I went to Google and typed in Ogg, and three clicks took me to a page with a list of players.

Philip, sure, I would love to see direct upload of ogg vorbis files supported. And by the way, your use (even awareness) of this non-MS technology increases my faith in Linden Lab enormously!

So, Phillip, when you say SL uses Ogg Vorbis as the compressor, does that mean that you compress the WAVs into ogg format, and then send that to the clients and uncompress for sound on the client side? If so, you could save on server real-estate by allowing clients to upload OGG files.

At any rate, uploading ogg-encoded files wouldn't reduce lag when you play them back, would it?

Good questions...

It would not save server space/CPU, as the compression is done on the client prior to upload.

And, for the same reason, it wouldn't reduce lag - however, if you want to speed the delay between triggering a sound and playback, you have two choices:

1. Use the llPreloadSound() command to periodically cause nearby clients to stream/decompress the audio in preparation for triggering playback (so for example this would be good for preloading game 'effect' sounds to give good quality playback on interaction).

2. reduce the bitrate you choose at upload time, for example to 32Kbps instead of 64Kbps (the default). This will decrease the delay in playback, with some loss in quality. Try it and see how it sounds.