Up until recently, microlense arrays were usually constructed as homogeneous arrays, meaning that every microlense within the array was identical to the others. That means that every microlense results – more or less – in one pixel in the final constructed image. In other words, the final resolution of the image is only a fraction of the sensor resolution.
This seems to be the case with the Lytro LightField camera: Even though it uses a 11 Megaray Sensor, the final resolution is only 1.1 Megapixels (1080 x 1080 pixels).
German LightField specialist Raytrix recently published information on a new improvement for LightField technology, the “multi-focus plenoptic camera” (MFPC).
Using a mix of several kinds of microlenses in one array, Raytrix managed to extend the depth of field while still allowing for 3D rendering and other LightField features. Using this technology, the authors managed to increase the image resolution to up to 25 % of the sensor resolution.
From the paper:
A new type of plenoptic camera is introduced here, called the multi-focus plenoptic camera (MFPC), whose micro lens array consists of micro lenses with different focal lengths. A MFPC can therefore be regarded as a number of interlaced single focus plenoptic cameras. It is shown that a MFPC extends the depth of ﬁeld compared to a single focus plenoptic camera and still allows eﬀective lateral resolutions at up to one quarter of the sensor resolution.
You can find more information on the new invention in the original paper: Single Lens 3D-Camera with Extended Depth-of-Field