Audio Compression

Hi all,

I've been looking through the manual and such but can't find an answer - if planning to make a game intended for Windows desktop, Mac, Linux etc and not targeted at any mobile or web-based platform (even though a desktop version is technically running via node webkit in a browser window) then is it possibe to circumvent the compression of .wav files to .ogg upon importing into Construct? I'd much prefer to keep my audio files lossless to maintain quality but can't find anywhere that this is mentioned.

Cheers

Sorry, this is not supported, but you can choose the highest quality encoding when importing or encode very high quality AAC/Ogg yourself outside of Construct 2 and import the .m4a and .ogg yourself. Beyond a certain quality limit the difference is inaudible, IIRC AAC at 256kbps is indistinguishable from uncompressed audio in ABX listening tests.

Actually, aac and ogg are ABX transparent at 160-192kbps for the vast majority of content.

Certain sounds can be very demanding on encoders -- eg. clapping is something of a torture test. Also, some synthesized sounds have very complex harmonics that can be smeared even at high bitrates.

If you find yourself having a problem, you can still manually encode, then import into construct. Sometimes custom encoding flags can improve quality for a troublesome sample. Google's your friend. Just remember: aac and ogg are both superior to mp3 in quality; in many cases near-transparency can be achieved at 128kbps.

That is unfortunate, but thanks for the quick response and the info

I'm going to assume there are no plans to implement support for uncompressed formats, or even FLAC at any point? I'm a bit picky about audio quality being completely unmolested in the same way that it'd be like forcing the use of jpegs over PNG, especially when real time processing will be used

Incidentally this is one of the more exciting aspects of C2 and I've wanted to use real time audio effects for ages - I modded the ghost shooter example to filter the higher frequencies based on distance and it works really nicely! Out of curiosity, does the use of a lossy/compressed format with a smaller filesize speed up the processing of audio when effects are used or are they uncompressed in memory before they are processed regardless? I'd imagine something like convolution reverb would have a lot of CPU overhead, especially when you've got a few things playing at once

Sorry to bump this, but will this feature be added at any point? Since WAV PCM is supported across browsers (with the exception of IE, which is hardly surprising) and given that it's important to maintain the integrity of all frequencies when appying effects, it seems a bit of a disadvantage not to have the option of at least one lossless sound format available.

Cheers

untune - not many platforms support lossless formats, it's complicated enough with two formats, lossless audio bloats the download size massively, and the difference in quality cannot be perceived with high quality AAC/Ogg, so there's really no point in supporting it.