sound file optimizer

Ferin

I'm wondering if any1 knows of little app, like imageoptim for graphic files, that does the same for music/sound files? I'm looking for something that will squeeze file size a bit and produce results that work well with GameSalad.

I'm not familiar with sound related tasks, they give me quite a hard time ;p.
What size i should aim for when creating samples like shooting, jumping etc. and what size should maximally be background tune(s)? I will appreciate any help guys.
In particular i'm curious what is your experience and how sound file size impact performance
of the game?

One more thing, is better to import to GS .wav files or it doesn't matter? Can i import .mp3 or other acceptable file fomrats and GS will take care of the hard part? Or it's better to optimize files b4 importing?

Thanks in advance.

Fogelberg

Hi!
I use 32 khz mono WAV for sound effects. If you go lower than that you will hear a drop in the sound quality. If you use wav or aiff, GS will automatically convert when you import.

SlickZero

I used to optimize the files myself with Audacity, but here lately, I just export an .mp3 at 96 kHz from Garageband, and let GameSalad handle the conversion. Don't have any sound issues doing that anymore.

Ferin

Thanks for response guys.

@Fogelberg
To optimize files i'm using audacity too, but according to html5 publishing recommendation i was using 22khz mono, i'm not sure about quality loss below 32khz, need to check that.

@SlickZero
I'm trying to 'create' something in garageband too but after 'composing' is finished i import to audacity then to GS. I try too keep file size to minimum because i prefer using bigger number of samples instead of higher quality in smaller quantity. But as i said i'm in musical way almost deaf, maybe that's why its not so important for me .

When (if?) i publish my game i'll ask what you think about sound-side of my creation, i may end up creating everything again or using some1s stuff ;p.

perfectanswer

Best practice for me is to compose or generate a sound in full quality (44khz, stereo).
Than I process the files in Audacity. To shrink the file size don't use mono, better join the channels and export the file as a *.ogg. In the save Dialog is a option box, there you can change the compression level of the file. To get a feeling for the right compression, save it at several ratios and compare filesize vs. quality. The difference between 22khz and 44 kHz is very small in file size, but you can save some bytes anyway.

Ferin

Thanks perfectanswer, i certainly will try that, sounds like good advise from some1 that knows its trade. I'm just wondering what you said about .ogg format, is it a temporary conversion just for audacity use or do i import files in that format into GS?

perfectanswer

.ogg files are the same kind like mp3 files. You can import it directly in GS.

Chobbiface

@perfectanswer. What happens when you join the channels in audacity? Do you still get stereo sound effect but with reduction in file size like mono? Cheers

perfectanswer

There are two ways that a person can feel as stereo signals. On the one hand by the fact that the left and right ear signals arriving differ in their frequency, or by the same signal shifted in time reached the ears.
Starting from 2000Hz the delay is more important for the ear in order to form a stereo image than the frequency differences.
In any case many redundancies come on - i.e. the left and right signals together and have many frequencies at the different time zones, they are not identical (maximum phase shifting).

There are several stereo types in which there may be a MP3 file:
True stereo, mid / side, intensity stereo or joint stereo.

Joint Stereo:
An Mp3 file consists of frames. The encoder can now decide for each frame, whether he is using intensity stereo or M / S stereo.

Intensity stereo:
The encoder keeps track of certain frequency bands and different sound levels. Hence, the decoder interprets the stereo signal. The frequency content in the left and right channel is then equal to M / S: These notes represent the encoder (L + R) and what the data includes stereo (LR)

The short version of the answer is ,yes. It's not stereo, but sounds likes stereo (with reduced file size).

Chobbiface

Well that was a perfectanswer haha. Thanks, learn something new everyday. I'll have an experiment and work out how to join channels next time in audacity.

old_kipper

The really cool thing is that we can tell the height of an incoming audio signal and I understand this is done using the shape of our ears to change the phase relationship of the high frequency content of the signal. Low frequency sounds are much harder to locate.

Clouds

@old_kipper

"Low frequency sounds are much harder to locate."

And near impossible to hear on something like an iPad/iPhone.