Permonic
1000+ Head-Fier
- Joined
- Dec 12, 2003
- Posts
- 1,178
- Likes
- 10
My entire CD collection is transfered into lossless Monkey's Audio (.ape) format. The usage of my CPU (Pentium-M 1.6 MHz) is usually about 30%. I use Foobar2000, 24 fixed-point, 44.1 kHz and ASIO format as an output. The soundcard is M-Audio Transit.
Recently I discovered that when I play MP3s or even a .wav format, the CPU usage drops significantly to 4-12% range. I am curious about this phenomenon, I tried different output format (Directsound, Directsound 2.0, waveOut; Kernel Streaming doesn't work with M-Audio Transit), using 16 fixed-point, experimenting with resampler.... However in the end, the result is always the same: .ape format uses CPU about 3x more than MP3 and 6x more than .wav format.
The sample rate of .ape format is 44.1/16 bit, compression circa 40-70%.
The question is, whether this is caused by an algorithm of Monkey's Audio format itself or by a Foobar2000 Monkey's Audio plug-in? And why? What do you think about that? Do you experience the same with other soundcards?
Recently I discovered that when I play MP3s or even a .wav format, the CPU usage drops significantly to 4-12% range. I am curious about this phenomenon, I tried different output format (Directsound, Directsound 2.0, waveOut; Kernel Streaming doesn't work with M-Audio Transit), using 16 fixed-point, experimenting with resampler.... However in the end, the result is always the same: .ape format uses CPU about 3x more than MP3 and 6x more than .wav format.
The sample rate of .ape format is 44.1/16 bit, compression circa 40-70%.
The question is, whether this is caused by an algorithm of Monkey's Audio format itself or by a Foobar2000 Monkey's Audio plug-in? And why? What do you think about that? Do you experience the same with other soundcards?