I wrote-
>>I suspected that long blocks would also have better frequency resolution
>> than short blocks (by a factor of 3).
Monty wrote-
>Yep, that's exactly it; better frequency resolution means better energy
>compaction of strong tones so that they're generally represented in finer
>resolution. It *also* means that the longer time block can reproduce
>lower fundamentals as tones more accurately (rather than pieceing
>together the bass and low midrange from what looks mostly like DC
>offsets)
� Okay, so does that mean that in encoding short blocks (given that there's
enough bits), there would be a frequency resolution of 229Hz for 44100Hz
sampling? Oh my! It doesn't seem to me like frequencies below, perhaps 2kHz
(or more?) could be used for calculating masking without introducing
serious low-frequency artifacts.
� Having noticed the poor frequency resolution, does MP3 format still have
the capability of storing frequency/volume information about frequencies
that are not whole multiples of the number of blocks per unit time? Or can
it only store those specific frequencies?
� "--noshort --resample 24 --lowpass 9.6" is looking increasingly more
useful for encoding Enya's music. Unfortunately 160kBit/sec is sometimes
not enough for 24kHz sampling (but it's close), with or without the --noshort.
Shawn
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
