Picander ha detto:
Audiovero ha centrato il punto, meno male che è entrato in discussione (convoluzione partizionata come brutefir, partizioni assimmetriche come jconv) così stanno prendendo la strada giusta. Molto molto interessante, le premesse per un ottimo risultato ci sono tutte
Intanto le cose vanno avanti a velocità sostenuta (and counting):
http://yabb.jriver.com/interact/index.php?topic=68828.150
Convolver features as of build 17.0.68:
Convolver__JRivolver
yes_______ yes______Supports Convolver config file format Config Format
no________yes______All processing is 64bit
no________yes______FFT/iFFT evaluation is lazy (only run when necessary)
no________yes______Pink noise RMS output automatically normalized to -6 dB below input
yes_______yes______Any number of paths, targetting any input or output channel
no________yes______Filter files (impulse response) can be in any format supported by Media Center (.wav, .ape, .flac, .mp3, etc.)
yes_______yes______Partitioning is used to reduce latency
no________yes______Partition size will automatically adjust with the sample rate
no________yes______
Latency is handled automatically for lip-sync (not including filter delay)
no________yes______Filter files can be any sample rate (one filter can be used for all sources)
no________yes______Handles flushing nicely so the tail of the last song isn't heard when playing a new song
yes_______yes______Handles volume attenuation for clip protection automatically (JRiver DSP clip protection used)
no________yes______Convolution uses SSE3 in the convolution kernel
yes_______yes______Output delays supported
yes_______no_______Input line delays
yes_______no_______Multiple output channels in a path
yes_______yes______Multiple input channels to a path
yes_______no_______Output channel weights
yes?______yes______Bit perfect (limited testing by Uli Brüggemann)
Puoi abbassare il FI se vuoi ora Picander
