0
@@ -632,4 +632,35 @@ bits:value
0
 0+:the number of wasted bits per sample (minus one) encoded as unary
0
 _
0
 .TE
0
+.TS
0
+tab(:);
0
+| c s |
0
+| c | l |.
0
+_
0
+Subframe Type
0
+_
0
+bits:type
0
+=
0
+000000:SUBFRAME_CONSTANT
0
+000001:SUBFRAME_VERBATIM
0
+00001x:reserved
0
+0001xx:reserved
0
+001xxx:SUBFRAME_FIXED (xxx = Predictor Order)
0
+01xxxx:reserved
0
+1xxxxx:SUBFRAME_LPC (xxxxx = Predictor Order - 1)
0
+_
0
+.TE
0
+.bp
0
+.SUBSUBSECTION "the CONSTANT subframe"
0
+.PP
0
+If all the samples in a subframe are identical, one can encode them
0
+using a CONSTANT subframe, which is essentially a single sample value
0
+that gets duplicated `block size' number of times when decoded.
0
+.SUBSUBSECTION "the VERBATIM subframe"
0
+.PP
0
+This subframe simply stores all the samples as-is,
0
+with no compression whatsoever.
0
+It is a `fallback' encoding method for when no other subframe makes one's
0
+data any smaller.
0
+
0