Omnia Audio Omnia.9/XE - Manual (Page 106)

HE-AAC (AAC+) in SBR Mode
Technical Details
SBR mode recreates the high end from a frequency-doubled version of the original frequencies. The base
AAC encoding is performed at one-half the specified sample rate while the SBR encoding is done at the
full sample rate.
Advantages
Standard SBR mode offers the most “bang for the buck” for lower bit rate encoding (64 kbps stereo and
below), but standard AAC or AAC+ in Oversampled SBR mode are better choices for higher bit rates.
When an AAC+ stream is presented to an encoder that can decode standard AAC but not AAC+, it will
still play the base AAC audio eliminating compatibility issues.
Disadvantages
Using SBR causes a lack of tight peak control in material that has been heavily processed and often
requires the Final Output Level to be reduced to compensate.
While the compatibility with players than can decode AAC but not AAC+ can be seen as an advantage (at
least they will play something) the base AAC encoding is done at a lower sample rate, which yields audio
with a narrower frequency range and less high end.
Finally, because of the way in which the harmonics created in the SBR process correlate to their
corresponding fundamental frequencies, the audio can take on a “metallic” sound not found in true AAC.
Recommended Settings
A bit rate of 64 kbps stereo and a sample rate of 44.1kHz can deliver “broadcast quality” audio, but the
Final Output Level should be set to -1.0dBFS to prevent decoder-side clipping.
48 kbps stereo can be adequate for reproduction on smaller devices (mobile phones and tablets, for
instance) with a Final Output Level setting of -1.5.
At bit rate of 32 kbps, we recommend setting the Channel Mode to Parametric Stereo. This creates a
filtered mono mix that won’t be as pleasant as true stereo but represents the best possible use of the
available bits and is a good compromise. The Final Output level should be set to -1.5.
HE-AAC (AAC+) in SBR Oversampled Mode
Technical Details
Like in the standard SBR mode, the base AAC encoding is performed at one-half the sample rate of the
SBR encoding. However, in the Oversampled mode, base AAC encoding is performed at the specified
sample rate while SBR encoding is accomplished at two times (double) the specified sample rate, thus
the term “oversampled”.
Advantages
Oversampled SBR can yield a frequency response all the way out to 20kHz at a sample rate of 128 kbps.
In comparison, a standard AAC stream encoded at the same sample rate would be frequency limited to
16kHz.
This is a benefit in and of itself, there is an additional advantage: AAC players that cannot decode the
SBR portion of the stream and pass only the base AAC portion will yield audio all the way out to 16kHz.
Choosing an Encoding Format and Recommended Settings
106