GDF 2.0

From: Alois Schlögl <alois.schloegl ***AT***>
Date: Fri, 26 Aug 2005 17:30:40 +0200
Subject: GDF 2.0
Newsgroups: tu-graz.biosig
Organization: Technische Universitaet Graz
Message-id: <>
User-agent: Mozilla Thunderbird 0.9 (Windows/20041103)

The GDF format has shown to be very useful. Compared to other
dataformats its supports many more features. The main advantages are:
+ GDF combines the advantages of different formats while avoiding the
+ GDF is an open format and has a simple definition.
+ GDF supports different sampling rates, channel specific scaling.
+ GDF supports not only INT16 numbers but also other numeric formats.
+ GDF supports efficient encoding of event information.
+ GDF header uses binary data, no rounding errors of the scaling factors.
+ GDF addresses the Y2K issue.
+ GDF enforces overflow detection.

However, the progress in EEG and BIOSIGNAL research does not stop. In
order to address new challenges, further information needs to be
supported. For this reason, there are plans to develop GDF 2.0.

It is proposed to address the following issues in GDF 2.0:
- include of gender and age (birthday) information
- electrode impedances
- position of EEG electrodes
- size of head (distance nasion-inion, A1-A2, circumference)
- recording location (using world coordinates ???)
- replacing the Lab-identification by the IPaddress
- The data/time format will be changed from a 16byte ASCII to an 8byte
binary representation.

The discussion of these plans is now open. Please comment on these plans.

After a discussion period, the new format should be fixed and,
hopefully, GDF 2.0 can be announced.

Best whishes,