diff --git a/src/audio/utils.cpp b/src/audio/utils.cpp
index eab46756c..aac02d744 100644
--- a/src/audio/utils.cpp
+++ b/src/audio/utils.cpp
@@ -496,17 +496,6 @@ void int2float(char *out, const char *in, int len)
         }
 }
 
-void short_int2float(char *out, const char *in, int in_len)
-{
-        const auto *ini = reinterpret_cast<const int16_t *>(in);
-        float *outf = (float *)(void *) out;
-        int items = in_len / sizeof(int16_t);
-
-        while(items-- > 0) {
-                *outf++ = (float) *ini++ / SHRT_MAX;
-        }
-}
-
 /**
  * Converts int8_t samples to uint8_t by adding 128 (standard
  * shifted zero unsigned samples).
diff --git a/src/audio/utils.h b/src/audio/utils.h
index 4e8915177..b1dcacffd 100644
--- a/src/audio/utils.h
+++ b/src/audio/utils.h
@@ -127,13 +127,6 @@ void float2int(char *out, const char *in, int len);
  * @param[in] in 4-byte aligned input buffer
  */
 void int2float(char *out, const char *in, int len);
-/**
- * This fuction converts from int16_t to normalized float
- * Input and output data may overlap.
- * @param[out] out 4-byte aligned output buffer
- * @param[in] in 4-byte aligned input buffer
- */
-void short_int2float(char *out, const char *in, int in_len);
 
 void signed2unsigned(char *out, const char *in, int in_len);