If GigaStudio would play MP3 Samples...

[Metuschelach]

...that would be really nice.

Hello,
Why to transform, e.g. a sampled Steinway recorded costly in 24bit high quality into a MP3 Smaple library?
Is this not silly?

Let me say why maybe not. Some giga instruments are so huge, that it is nearly impossible to play them in realtime, especially if you play many instruments at the same time.

With MP3-Compression it would be possible to reduce the Grandioso Sample (about 4 gig) to to 400 MByte for example. If you have a normal good PC you could load then the whole piano into the RAM and play in Realtime to my opinion.

But do you not loose quality?
Yes, of course, you do. But how different is the difference?
If you want to produce a profesional song, you can still use the uncompressed material.

What do you think about this?

[PolarBear]

Hm. You\'d leave out an important issue: Polyphony. You\'d reach a limit at a certain number of simultaneous mp3-streams. Limited by what your CPU and RAM could play. IMO much more CPU power is needed in decompressing mp3-streams than speed is saved with compression - because if you don\'t wanna loose that much quality you have to choose something more than 192kbit per mp3-stream. Remember: 160 voices of polyphony would be equavlent 160 mp3-streams. You\'ll need some horse power... images/icons/grin.gif

Ho do others think?

Hansi

PS: Don\'t know if GSIF-drivers can handle mp3 at all with low latencies...

[Metuschelach]

Hello Hansi,
I think you are right, you need a lot of CPU-Power, but how much exactly?

[Chadwick]

One other thing Metuschelach

As of the current revision, Gigastudio doesn\'t allow the user to load ALL of a sample into ram - only the 64k buffer. You don\'t get much of an advantage in this instance.

Of course, if Giga allowed you to allocate ram to ALL of a sample...

[Simon Ravn]

One other \'interesting\' problem is that when you make an MP3 file, it encodes it into blocks of xxx bytes - so you will get added silence to the beginning and end of your MP3.

[PolarBear]

Originally posted by Simon Ravn:
One other \'interesting\' problem is that when you make an MP3 file, it encodes it into blocks of xxx bytes - so you will get added silence to the beginning and end of your MP3.<font size=\"2\" face=\"Verdana, Arial\">Hey Simon, very interesting. I found this out once I tried to loop an ambience mp3 for a website project. Is there any way to get around this? (thinking of tools that could change mp3 without \"remake\" a wav-file?)

Please explain a bit more if you know more about it. Thanks in advance! images/icons/smile.gif

Hansi

[Simon Ravn]

Polar, I really dont know more than you. I found out the same way - I needed to deliver something loopable, and found out it was impossible in MP3. I don\'t think there\'s any way around it.

[Chad Seiter]

I don\'t know if you\'ve ever had this problem with Winamp (or even use Winamp!), but for a while, when I selected a whole directory of, say, 100 MP3\'s, Winamp opened up a new player for each one and was streaming them all! So when this happened, the processor was so bogged down, the audio was skipping like mad and I could barely click anything... and ended up having to reset after taking three minutes to close three windows! I don\'t think the streaming MP3 compression would work. Of course, you may not need more RAM, but I\'m operating on a 1.3GHz processor and it was rendered totally useless... so I don\'t expect to ever see streaming MP3 technology. I say it\'s better to invest in faster hard drive and RAM technology before ever thinking about doing that.

-Chad