search

johnahartman / alembic

Automatically exported from code.google.com/p/alembic
Other
0 stars 0 forks source link

Support reads of a different precision #137

Closed GoogleCodeExporter closed 8 years ago

GoogleCodeExporter commented 8 years ago 
It should be possible to hint to a I*Property to read it's data with a 
different precision than what is written on disk.

You should be able to instantiate a IDoubleArrayProperty for something that is 
really an IFloatArrayProperty.

Issues with the key will be averted with these changes:
http://code.google.com/p/alembic/issues/detail?id=124

There will probably be some imposed limitations, (things stored as strings 
can't be read as floats) I'll detail any implementation limitations thoroughly 
after my first pass.

Original issue reported on code.google.com by miller.lucas on 7 Feb 2011 at 10:09

GoogleCodeExporter commented 8 years ago

Original comment by ard...@gmail.com on 10 Feb 2011 at 5:05

GoogleCodeExporter commented 8 years ago 
Moving to backlog, maybe this will be in 1.1?

Original comment by miller.lucas on 30 Jun 2011 at 4:56

GoogleCodeExporter commented 8 years ago

Original comment by ble...@gmail.com on 7 Jul 2011 at 10:22

GoogleCodeExporter commented 8 years ago 
I just want to chime in that I have run into this problem, and being able to 
read in a different precision sounds like a solution.  For example, when 
pairing the maya AbcExport and the prman AlembicRiProcedural plug-ins, 
arbitrary geometric properties get written by AbcExport as double precision, 
but the prman procedural only wants to read floats, so they get skipped over.  
I've been wondering what to do about this.

Original comment by kat...@gmail.com on 25 Aug 2011 at 5:15

GoogleCodeExporter commented 8 years ago 
AlembicRiProcedural currently skips doubles because the RenderMan interface 
doesn't support them either.

Ideally, two things could happen to remedy this:
1) AlembicRiProcedural could recognize double GeomParams and convert them to 
float on output
2) AbcExport could have a way to specify that the data be written as float.

The first one should be easy to do and we can track that as a separate issue.

Original comment by steve.lavietes@gmail.com on 25 Aug 2011 at 5:23

GoogleCodeExporter commented 8 years ago

Original comment by ble...@gmail.com on 25 Aug 2011 at 10:28

GoogleCodeExporter commented 8 years ago

Original comment by ble...@gmail.com on 26 Aug 2011 at 11:53

GoogleCodeExporter commented 8 years ago

Original comment by ble...@gmail.com on 28 Oct 2011 at 11:34

GoogleCodeExporter commented 8 years ago 
First pass is here:
http://code.google.com/r/millerlucas-dev/source/detail?r=aec8ba7db0f37ee36526acb
aa98e1723952c6d4f

string, wstring can't be read as any other type.
We'll probably need to impose this on float16_t as well, since the conversion 
to float32_t or int32_t is incorrect.

These changes also need to be propagated to Abc.

Original comment by miller.lucas on 12 Dec 2011 at 8:15

GoogleCodeExporter commented 8 years ago 
float16_t was excluded here:
http://code.google.com/r/millerlucas-dev/source/detail?r=d24e39cdd98cf49bb79e81a
10c2a1386366d9903

Original comment by miller.lucas on 12 Jan 2012 at 10:14

GoogleCodeExporter commented 8 years ago

Original comment by miller.lucas on 24 Jan 2012 at 1:40