Skip to content

Improve performance for wav read sample #117

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
hoxily wants to merge 2 commits into
base: master
Choose a base branch
from
Open

Improve performance for wav read sample #117

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
08eae12
improve performance of Wav.ReadCallback
hoxily Aug 4, 2021
5b7a0f3
ignore .vs directory
hoxily Aug 4, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions .gitignore
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,3 +38,4 @@ UpgradeLog*
*.mpq

.idea
/.vs
21 changes: 17 additions & 4 deletions Assets/Scripts/Diablerie/Engine/IO/Wav.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -59,10 +59,10 @@ private void ReadCallback(float[] data)
var chunk = _chunks[_chunkIndex];
int samplesAvailable = chunk.sampleCount + chunk.sampleOffset - _sampleIndex;
int samplesToRead = Math.Min(data.Length - i, samplesAvailable);
for (int j = 0; j < samplesToRead; ++j, ++i)
{
data[i] = ReadSample(_reader);
}
// When smaplesToRead reaches 35000 or above, reading a float a time will consumes about 3 seconds, which stalls the whole game.
Comment thread
mofr marked this conversation as resolved.
// Use bunch read to spped up.
Comment thread
mofr marked this conversation as resolved.
ReadSample(_reader, data, i, samplesToRead);
i += samplesToRead;
_sampleIndex += samplesToRead;
if (i < data.Length)
++_chunkIndex;
Expand Down Expand Up @@ -130,6 +130,19 @@ private static float ReadSample(BinaryReader reader)
return BytesToFloat(byte1, byte2);
}

private static void ReadSample(BinaryReader reader, float[] sampleBuffer, int startIndex, int sampleCount)
{
byte[] buffer = new byte[sampleCount * 2];
Copy link
Copy Markdown
Owner

@mofr mofr Aug 4, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Could we reuse the buffer array somehow to avoid frequent allocations and excessive garbage collector work?

Copy link
Copy Markdown
Author

@hoxily hoxily Aug 5, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

How about create a reuseable fixed length byte[] buffer when new Wav(), and transfer this buffer to ReadSample method. Works like a Stream's internal buffer.

Copy link
Copy Markdown
Owner

@mofr mofr Aug 5, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Sounds good

reader.BaseStream.Read(buffer, 0, buffer.Length);
for (int i = 0; i < buffer.Length; i += 2)
{
byte byte1 = buffer[i];
byte byte2 = buffer[i + 1];
short s = (short)((byte2 << 8) | byte1);
Copy link
Copy Markdown
Owner

@mofr mofr Aug 4, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Let's reuse BytesToFloat here to avoid the transformation logic duplication, please.

Copy link
Copy Markdown
Author

@hoxily hoxily Aug 5, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Maybe we shall test the performances of the following two solutions:

  1. reuse BytesToFloat method;
  2. directly insert body of BytesToFloat;

After performance test, choose the fastest one.

Copy link
Copy Markdown
Owner

@mofr mofr Aug 5, 2021

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do you expect significant function call overhead? In this case, I like your idea to test both of them. If the inlined version works faster I think it's a reasonable cost to duplicate the code.

sampleBuffer[startIndex + i / 2] = s / 32768.0F;
}
}

private static float BytesToFloat(byte byte1, byte byte2)
{
// convert two bytes to one short (little endian)
Expand Down