i find in the mpeg-4 aac standard (ISO14496-3) Table 4.5.1 that there are several tables for different sampling frequencies. Does it imply that it is not encouraged to use frequencies other than those with tables (e.g. 96000, 88200, 64000, 48000 Hz etc)?

And, it said "If the sampling rate is not one of the rates listed in the right column in the table, the sampling frequency dependent tables (code tables, scalefactor band tables etc) must be deduced in order for the bitstream to be parsed". Does it mean that if I want to encode at 14000Hz, I should use the spec for 16000Hz?

I think this is defined quite accurately in the standard:

uint8_t get_sr_index(uint32_t samplerate)
{
if (92017 <= samplerate) return 0;
if (75132 <= samplerate) return 1;
if (55426 <= samplerate) return 2;
if (46009 <= samplerate) return 3;
if (37566 <= samplerate) return 4;
if (27713 <= samplerate) return 5;
if (23004 <= samplerate) return 6;
if (18783 <= samplerate) return 7;
if (13856 <= samplerate) return 8;
if (11502 <= samplerate) return 9;
if (9391 <= samplerate) return 10;

return 11;
}

Where the returned index is used to read the following table:

static uint32_t sample_rates[] =
{
96000, 88200, 64000, 48000, 44100, 32000, 24000, 22050, 16000,
12000, 11025, 8000
};

And that shows which data to use during encoding/decoding.

Menno