diff --git a/core/nsfplaycore.cpp b/core/nsfplaycore.cpp
index 037f06b..1f71c71 100644
--- a/core/nsfplaycore.cpp
+++ b/core/nsfplaycore.cpp
@@ -292,7 +292,18 @@ uint64_t nsfplay_samples_played(const NSFCore* core)
 	return 0;
 }
 
-uint32_t nsfplay_render(NSFCore* core, uint32_t samples, int32_t* stereo_output)
+uint32_t nsfplay_render32(NSFCore* core, uint32_t samples, int32_t* stereo_output)
+{
+	NSF_MUTEX_GUARD();
+	if (!core) return 0;
+	NSF_UNUSED(core);
+	NSF_UNUSED(samples);
+	NSF_UNUSED(stereo_output);
+	// TODO
+	return 0;
+}
+
+uint32_t nsfplay_render(NSFCore* core, uint32_t samples, int16_t* stereo_output)
 {
 	NSF_MUTEX_GUARD();
 	if (!core) return 0;
diff --git a/include/nsfplaycore.h b/include/nsfplaycore.h
index 63baa24..930251b 100644
--- a/include/nsfplaycore.h
+++ b/include/nsfplaycore.h
@@ -249,7 +249,9 @@ uint64_t nsfplay_samples_played(const NSFCore* core); // samples since song_play
 // - stereo_output can be NULL if the output isn't needed
 // - returns number of samples rendered, may be less than samples if song is finished (will zero fill unused output samples)
 // - see shared/sound_convert.h for conversion examples to various formats
-uint32_t nsfplay_render(NSFCore* core, uint32_t samples, int32_t* stereo_output);
+uint32_t nsfplay_render32(NSFCore* core, uint32_t samples, int32_t* stereo_output);
+// nsfplay_render32 with a conversion to the most common 16-bit stereo format for convenience
+uint32_t nsfplay_render(NSFCore* core, uint32_t samples, int16_t* stereo_output);
 
 // manually trigger render buffer allocations if needed
 // call after the desired NSF is loaded and all settings have been made, but before the first render or emu_init/run