It's a mp3 file. Sample rate=44.1khz. Bitrate=128kbps.
Clock=12.288MHz, CLOCKF=0xf800, SPI_MODE=0x0804.
vs1003b pitch/tempo shifting
Re: vs1003b pitch/tempo shifting
- Attachments
-
- using algorithm (normal speed).m4a
- (408.99 KiB) Downloaded 5 times
-
- using algorithm (1.5×normal speed).m4a
- (1 MiB) Downloaded 7 times
-
- using algorithm (1.2×normal speed).m4a
- (1.3 MiB) Downloaded 6 times
-
- using algorithm (0.99×normal speed).m4a
- (847.07 KiB) Downloaded 7 times
-
- using algorithm (0.9×normal speed).m4a
- (865.33 KiB) Downloaded 6 times
-
- using algorithm (0.8×normal speed).m4a
- (693.24 KiB) Downloaded 7 times
Re: vs1003b pitch/tempo shifting
It confuses me a lot that speed-down playing is worse than speed-up playing, that's really amazing.
Here is the original song.
Here is the original song.
- Attachments
-
- Original song.mp3
- (4.61 MiB) Downloaded 5 times
Re: vs1003b pitch/tempo shifting
It sounds like that's what that specific algorithm can accomplish.
The difference between speeding up and slowing down is that when speeding up, the playback plays everything only once (uses windowing to play two portions at the same time to "compress" the playing time), while when slowing down the playback needs to play some portions of the output twice.
Although slowing down also uses windowing, it won't sound as good, because for example you'll get transients twice if they appear in the windowed portions. When speeding up any transients only appear once, although their order might sometimes be swapped.
We have some better approaches, but they are not for vs1003.
The difference between speeding up and slowing down is that when speeding up, the playback plays everything only once (uses windowing to play two portions at the same time to "compress" the playing time), while when slowing down the playback needs to play some portions of the output twice.
Although slowing down also uses windowing, it won't sound as good, because for example you'll get transients twice if they appear in the windowed portions. When speeding up any transients only appear once, although their order might sometimes be swapped.
We have some better approaches, but they are not for vs1003.
Visit https://www.facebook.com/VLSISolution VLSI Solution on Facebook
Re: vs1003b pitch/tempo shifting
Thanks for your explanation. By the way, which of your products are capable of performing it well ? What are them ? 
Re: vs1003b pitch/tempo shifting
A better version of the pitch shifter is available for VS1005.
The original version: http://www.vsdsp-forum.com/phpbb/viewto ... =13&t=2092
And later integrated into the regular distribution (see the link in the second post of that thread).
The original version: http://www.vsdsp-forum.com/phpbb/viewto ... =13&t=2092
And later integrated into the regular distribution (see the link in the second post of that thread).
Visit https://www.facebook.com/VLSISolution VLSI Solution on Facebook
Re: vs1003b pitch/tempo shifting
Thanks a lot, Pasi !