[r6] Add 4 files, some small corrections

This adds 4 mono files. 60 is just a mono file, 61-64 have short bursts of high-frequency sine waves added to speech to trigger predictor overflow in decoders that do not use 64-bit ints when appropriate File 26, 27 and 55 had small changed to their metadata as I came across a decoder (Symphonia) that was much more strict than libFLAC and warned me about some things that were not strictly compliant with the specification
ietf-wg-cellar · Feb 1, 2022 · 7bce37a · 7bce37a
1 parent 82b92f4
commit 7bce37a
Show file tree

Hide file tree

Showing 8 changed files with 10 additions and 0 deletions.
diff --git a/subset/26 - variable blocksize file created with CUETools.Flake 2.1.6.flac b/subset/26 - variable blocksize file created with CUETools.Flake 2.1.6.flac
diff --git a/subset/27 - old format variable blocksize file created with Flake 0.11.flac b/subset/27 - old format variable blocksize file created with Flake 0.11.flac
diff --git a/subset/55 - file 48-53 combined.flac b/subset/55 - file 48-53 combined.flac
diff --git a/subset/60 - mono audio.flac b/subset/60 - mono audio.flac
diff --git a/subset/61 - predictor overflow check, 16-bit.flac b/subset/61 - predictor overflow check, 16-bit.flac
diff --git a/subset/62 - predictor overflow check, 20-bit.flac b/subset/62 - predictor overflow check, 20-bit.flac
diff --git a/subset/63 - predictor overflow check, 24-bit.flac b/subset/63 - predictor overflow check, 24-bit.flac
diff --git a/subset/readme.txt b/subset/readme.txt
@@ -125,3 +125,13 @@ with valid but rather unusual or extreme metadata.
 - File 57 has a PICTURE with mimetype image/png
 - File 58 has a PICTURE with mimetype image/gif
 - File 59 has a PICTURE with mimetype image/avif
+
+---------------------------------------------------------------
+                       Files # 60 - # 63
+                Miscellaneous, later additions
+---------------------------------------------------------------
+- File 60 is mono audio
+- File 61, 62 and 63 are signals with rather extreme
+    characteristics that might trigger overflow if a decoder
+    uses 32-bit integers to calculate the predictor where 64-bit
+    integers are appropriate