Multiplexed online log or normal diskgroup?

hi,

On a very high transaction database  4 LUNs presented to the ASM to be used as REDO diskgroup.

since availability is a big concern too, we use two separate storage systems (EMC1, EMC2), from each presented 2 LUNs to avoid downtime in case of storage system crash.

Which configuration seems to have better performance and availability based on described situation :

     Config 1 : Use 4 disks to build a NORMAL redundancy diskgroup and place online logs on single diskgroup. let the ASM provide the redundancy.

     Config 2 : Use 2 disks to build REDO1 diskgroup, and 2 disks to build REDO2 diskgroups, with EXTERNAL redundancy on both diskgroups. multiplex redo logs on REDO1, REDO2 disks , each log group have 2 members.