Windows Media Audio (WMA) / ASF Data Structure

[mc84excel]

I'm reading a wma file as a 1D array of bytes. I want to know which item in the array is the start of the music data.

I understand that this can be calculated using the header data but I can't work it out. I've tried reading up on the WMA file structure (e.g. the first 128 bits being the Object ID, the next 64 bits being the Object Size and so on) but despite a lot of time spent on this, I'm still lost.

Can anyone please end this headache?!

[Norie]

Eh, why?

This sort of encoding is usually proprietary, and to be honest I've no idea why you would want to decipher it.

[mc84excel]

Thanks Norie.

The Data Structure for ASF (which includes WMA) is publicly available. Download the PDF from: http://www.microsoft.com/en-us/downl....aspx?id=14995

I'm not seeking to decipher the stream data itself. I only want a VBA function which will return where the file property data ends/music data starts (where the argument is the wma file as a byte array)

[mc84excel]

The annoying thing is that I know that the header object size can be calculated. But I haven't learnt python so I can't convert the code to VBA (I've tried).

[mc84excel]

Eh, why? ... to be honest I've no idea why you would want to decipher it.

I want to checksum wma files while ignoring any file property information etc.

If I can determine the start/end of the raw music data, I can run the checksum calculation on that portion of the file only.

I have this working fine for mp3, I now want it for wma.

[mc84excel]

I'll try another approach:

Is it possible to convert C++ code into VBA?

[mc84excel]

To help anyone who is minded to help out on this thread, I am providing code which converts a filename string into an array

You can test it on the WMA files in the Sample Music folder.

I need to know which array element marks the start of the raw music data. (IIRC this happens immediately after the end of the file properties data. So if you can return the file header object size from the ASF properties, the thread is solved?)

Please Login or Register  to view this content.

[mc84excel]

Hmmm. It looks like I can't solve this by myself.

There is C++ code which I can probably use part of it to give me what I need. Except. I don't know C++ so I can't convert C++ to VBA. (I've tried but it isn't easy converting from one language to another. Particularly when you don't know the original language!)
And there are no bilingual (C++ & VBA) programmers on ExcelForum (or if there are - they are in hiding!)

There are various Python scripts out there (which I found by googling) that I MAY be able to use part of their code to get the variable I'm seeking.
Except (you guessed it) I don't know Python. So I can't convert Python to VBA (I've tried but etc. - same problem as converting C++ to VBA).

I don't suppose there are any Python/VBA programmers on ExcelForum? (Although the C++ code looks far more promising)

[FDibbins]

Sorry no-one else has offered help on this yet (bumping it for you)

[shg]

(shg converted this code from VB, thanks shg)

I disavow that code -- it's terrible.

[mc84excel]

I disavow that code -- it's terrible.

I thought it was you who originally ported this VB code to VBA for me. My mistake. Sorry shg, I did not intend to offend you. (I have now removed the credit)

[mc84excel]

I'm sure it is easier then what it seems. I just can't make that final connection.

I quote from page 6 of the ASF Specification pdf (link provided in post #3):

3.1 Header Object (mandatory, one only)

The Header Object is a container that holds any combination of the standard objects listed in the
following sections. A File Properties Object, a Header Extension Object and at least one
Stream Properties Object are required to be present. Implementations shall ignore any objects
that they do not understand.

The Header Object is represented using the following structure.

Field name Field type Size (bits)
Object ID GUID 128
Object Size QWORD 64
Number of Header Objects DWORD 32
Reserved1 BYTE 8
Reserved2 BYTE 8

The fields are defined as follows:

Object ID
Specifies the GUID for the Header Object. This field shall be set to ASF_Header_Object.

Object Size
Specifies the size of the Header Object. This includes the sum of 24 bytes plus the size of the objects contained in
the Header Object. Valid values are at least 30 bytes.


OK so I open the WMA file in a Hex Editor (or import the file into VBA as an array of bytes). Now I try to follow the above.

The first field name according to the PDF is "Object ID". It has a size of 128 bits. 128 bits = 16 bytes. So I can ignore the first 16 elements in the array as they are useless for my goal.

The second field name is "Object Size". I am not 100% certain (so correct me if you know) but I'm guessing that obtaining this value is the start towards returning the value I need.

The Object Size field has a size of 64 bits. 64 bits = 8 bytes. So this means the byte values (000010 to 000017 in Hex Editing, elements 16 to 24 in byte array (base 0)).

I'm testing this on a file and it gives me the following values:

A6 14 00 00 00 00 00 00

Now what? How do these bytes give me the size of the Header Object?

[mc84excel]

Ha Ha Ha Ha Ha. I've cracked it!

From the previous post, I have the following hex values from a test file:

A6 14 00 00 00 00 00 00

If I read these in REVERSE ORDER, I get: 00 00 00 00 00 00 14 A6

And 14A6 appears to be the start of the music stream when hex editing this file!


Now all I need is a VBA Function that takes an argument of a variant array and returns the start of the music stream as a long value.


So. It would need to read the 16th to 24th elements (assuming base 0) of the array, reverse the order of these elements and then convert this to string and from string to long (converting from Hex to Dec)

[mc84excel]

Please Login or Register  to view this content.