From f8c5e54962cfc615c50f1310eb4053d5760a9fc7 Mon Sep 17 00:00:00 2001
From: Jamie Pine <ijamespine@me.com>
Date: Mon, 26 Jan 2026 19:39:09 -0800
Subject: [PATCH] Add audio input entitlement and enhance audio sample
 extraction logic

- Added the `com.apple.security.device.audio-input` entitlement to the Entitlements.plist for improved audio capture capabilities.
- Refactored the audio sample extraction logic in macOS to handle both interleaved and planar audio formats, improving sample processing and interleaving of channels.
- Updated the Assets.car file to reflect changes in the audio capture implementation.
---
 tauri/src-tauri/Entitlements.plist         |   2 +
 tauri/src-tauri/gen/Assets.car             | Bin 3847048 -> 3847048 bytes
 tauri/src-tauri/src/audio_capture/macos.rs |  56 ++++++++++++++++++---
 3 files changed, 51 insertions(+), 7 deletions(-)

diff --git a/tauri/src-tauri/Entitlements.plist b/tauri/src-tauri/Entitlements.plist
index 46f67566..0d041c2c 100644
--- a/tauri/src-tauri/Entitlements.plist
+++ b/tauri/src-tauri/Entitlements.plist
@@ -8,5 +8,7 @@
     <true/>
     <key>com.apple.security.cs.disable-library-validation</key>
     <true/>
+    <key>com.apple.security.device.audio-input</key>
+    <true/>
 </dict>
 </plist>
diff --git a/tauri/src-tauri/gen/Assets.car b/tauri/src-tauri/gen/Assets.car
index db1bac230a2c2bcbb656dd283089c4be47d6e1f5..7b81cf777f0b01ca6474699403672548c5e84f92 100644
GIT binary patch
delta 864
zcmZwGzfTlF6bJA<&XeOeDj=L7pyKa=nRzp_GZfs-&di3=lG4u3SXx+^SR5_6#7cV-
z8xxA9^grMqu(0!Dj6oYM{SuDq*yNL!*L>f;+1dT?KWh8%`D9~dD$BA-n1FGZfoT|n
z=F#tslbHkF=$-0L@3~p)eL>}@Z_RS;>(NK^XQXXI5<G>#jFAOXPM6FZA(*hFc%;a^
z)J*4+i1D*WQ&isiQYgFm>g9`r_13$->-Jp@MxYK?paEB5wCcNNYP3_bDTu<ef&%75
zvaE=?V&no9)_G6rk~t<y8_(w~SNrYe`}$6i>oO%`DLLd&p0lEGmXRSPt>@f|bjkZ_
z;>AVI>(HF9wWSDoS#WGx7J@0sOJ<!(EQlln+1$Bchl5;MMt>31kIkieJ6WTB(2~ir
zESa)FF&C0$B2w;2hF}Ue6y&)hPv=34-M!r=jKMfez$8q;G|a#(T!ZUy1Lj~J7T_kF
zWh_ApZox9Fz-?HCHCTr`a2M{u2Hb}S@DLus<7#jB$-lolp_JYP&zwy;b|S+HgYzbe
zR%u~^;lqdHgwlMlQ>%N}c?t|RVGB42kU*e7SBD)_{c2m6VhSZODRB=}t|LBcAZBEW
zxT4BMJv`6`a{1tTw7SPzIaqMj@z!?zjabMs{S80D2<j>7T$Y*H^WLlt`p%$#hPqh&
J{IvaH^gp>2=CA+&

delta 864
zcmZwGId2m|7zW_AO^9<kfD@A75C;hN#B(1hV$ZRwl$0sy=ny3Z1rmkbMwY0ck3@r%
zV)HM;PoTp>2n8erC2wR>nUx-Wnvv%H=Fkqle60-N<LUOsOqOMJXuu*Yz&y;s?D3E7
z)5am48n$(B_({>>Q{5h(p}y9w%CF=1`e(JvL(V+QIbl&rBCVt(u%t=wMd1p~nThAq
zTp2>oKSlH4f@CSkofpra9d50^8M*$*RiFwrxCB#h8K%pTtLIbQz+|9?azd3vU^0q$
zOb}X_m>c0~6!~Ny&SkLUKCN=lt-r19DW?mpz<xc&0?mjs*tD=g=1Li_6`RZvPEdD|
z^D@?#D_yE><UvFd1xIJ?6R|#M!Z{VHNGb*YpMwjoW4s6&rTSW}YmIa{=Yg=C6$G0|
z#HZwmr94GtQVe`DC@K~h9|x`W5BBRY1G6v(^RNJm(10bl0$1T0G+`N5;5wXVtid|m
zfEH}PP1uAsY{4zK4R>H0?!rB|4-epBd9eTJ&+k3WW0cAgZK5Dj;dD3?BO+PPr3^x8
zJ9*2FnwZ9CTJInA9s_|L=l}%<90W*E<xx+U-@DGbB$YvEG<ZKD@EV*tVwKHvXHr4w
z<Ru2BbqE&?PdcW5(y`#mlg@7K6>Dah`ivi;ih72+l4W{nJl1X0Rg8Q8#!pb2<+l&J
H@239(2z2H_

diff --git a/tauri/src-tauri/src/audio_capture/macos.rs b/tauri/src-tauri/src/audio_capture/macos.rs
index d929e65a..ef5451d0 100644
--- a/tauri/src-tauri/src/audio_capture/macos.rs
+++ b/tauri/src-tauri/src/audio_capture/macos.rs
@@ -163,25 +163,67 @@ fn extract_audio_samples(sample_buffer: CMSampleBuffer) -> Result<Vec<f32>, Stri
         .audio_buffer_list()
         .ok_or_else(|| "Failed to get audio buffer list".to_string())?;
 
-    let mut samples = Vec::new();
+    let buffers: Vec<_> = audio_buffer_list.iter().collect();
+    let num_buffers = buffers.len();
+    
+    if num_buffers == 0 {
+        return Ok(Vec::new());
+    }
 
-    // Iterate through audio buffers
-    for buffer in audio_buffer_list.iter() {
-        // Get raw bytes and interpret as f32 samples
+    // ScreenCaptureKit on macOS provides audio in Float32 format
+    // The audio can be either:
+    // - Interleaved (1 buffer with L,R,L,R,... samples)
+    // - Planar (2 buffers, one for L channel, one for R channel)
+    
+    if num_buffers == 1 {
+        // Interleaved stereo or mono in a single buffer
+        let buffer = &buffers[0];
         let data_bytes = buffer.data();
         let num_samples = data_bytes.len() / std::mem::size_of::<f32>();
         
         if num_samples > 0 {
             unsafe {
-                // Interpret bytes as f32 samples
                 let data_ptr = data_bytes.as_ptr() as *const f32;
                 let data = std::slice::from_raw_parts(data_ptr, num_samples);
-                samples.extend_from_slice(data);
+                return Ok(data.to_vec());
             }
         }
+    } else {
+        // Planar format - separate buffer for each channel
+        // We need to interleave them: L0, R0, L1, R1, ...
+        let mut channel_data: Vec<Vec<f32>> = Vec::new();
+        let mut max_samples = 0;
+        
+        for buffer in &buffers {
+            let data_bytes = buffer.data();
+            let num_samples = data_bytes.len() / std::mem::size_of::<f32>();
+            
+            if num_samples > 0 {
+                unsafe {
+                    let data_ptr = data_bytes.as_ptr() as *const f32;
+                    let data = std::slice::from_raw_parts(data_ptr, num_samples);
+                    channel_data.push(data.to_vec());
+                    max_samples = max_samples.max(num_samples);
+                }
+            }
+        }
+        
+        // Interleave the channels
+        let mut interleaved = Vec::with_capacity(max_samples * num_buffers);
+        for i in 0..max_samples {
+            for channel in &channel_data {
+                if i < channel.len() {
+                    interleaved.push(channel[i]);
+                } else {
+                    interleaved.push(0.0); // Pad with silence if needed
+                }
+            }
+        }
+        
+        return Ok(interleaved);
     }
 
-    Ok(samples)
+    Ok(Vec::new())
 }
 
 fn samples_to_wav(samples: &[f32], sample_rate: u32, channels: u16) -> Result<Vec<u8>, String> {