48 kHz extension to iSAC.

Test:
-manual test with voe_cmd_test.
-manual test with RTPEncode & NetEqRTPPlay.
-manual test with simpleKenny.
-Bit-exact test of iSAC-swb and iSAC-wb with head revision of trunk. The bit-exactness is confirmed on all files generated by running webrtc/modules/audio_coding/codecs/isac/main/test/QA/runiSACLongtest.txt
Review URL: https://webrtc-codereview.appspot.com/937025

git-svn-id: http://webrtc.googlecode.com/svn/trunk@3226 4adac7df-926f-26a2-2b94-8c16560cd09d
diff --git a/webrtc/modules/audio_coding/neteq/recin.c b/webrtc/modules/audio_coding/neteq/recin.c
index 00c8f81..c2f0d2d 100644
--- a/webrtc/modules/audio_coding/neteq/recin.c
+++ b/webrtc/modules/audio_coding/neteq/recin.c
@@ -380,11 +380,12 @@
             MCU_inst->scalingFactor = kTSscalingTwo;
             break;
         }
+        case kDecoderISACfb:
         case kDecoderOpus:
         {
-            /* We resample Opus internally to 32 kHz, but timestamps
-             * are counted at 48 kHz. So there are two output samples
-             * per three RTP timestamp ticks. */
+            /* We resample Opus internally to 32 kHz, and isac-fb decodes at
+             * 32 kHz, but timestamps are counted at 48 kHz. So there are two
+             * output samples per three RTP timestamp ticks. */
             MCU_inst->scalingFactor = kTSscalingTwoThirds;
             break;
         }