Re: AUSampler Bug
Re: AUSampler Bug
- Subject: Re: AUSampler Bug
- From: "T.J. Usiyan" <email@hidden>
- Date: Wed, 04 Apr 2012 16:42:10 -0400
Thanks to Douglas Adams pointing out that I was not swapping byte order, I fixed my issue.
An important bit of info about how to swap was using the pattern
// into Swapped
lHostFloat = CFConvertFloat32HostToSwapped(value);
aStructPtr->mBaseNote = *(Float32 *)&hostFloat;
//from Swapped
Float32 value = CFConvertFloat32SwappedToHost(*(CFSwappedFloat32 *)(&instrumentChunkStruct.mBaseNote))
because CFSwappedFloat32 is a struct. Maybe that would be obvious to some, but that threw me completely for a long while.
On Wed, Mar 14, 2012 at 5:01 PM, Douglas Scott
<email@hidden> wrote:
You need to swap the bytes for the instrument chunk before writing it into the CAF file. The CAF spec says that all data must be big-endian.
-DS
On Feb 26, 2012, at 9:25 AM, TJ Usiyan wrote:
> AUSampler ignores the mBaseNote field in the Instrument Chunk. It does this both in AU Lab and programmatically. I can read and write to the Instrument chunk with expected results, and the other fields are detected and populated perfectly well so it does not seem to be an error in writing to the instrument chunk. If anyone has found a workaround (other than a DLS or aupreset) I would greatly appreciate some advice. For now I will use aupreset, but I would still prefer loading the fils as an array of URLs.
> TJ
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Coreaudio-api mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden