ERDAS Apollo Protocol Comparison
Dataset: Projection:

ECWP

  • Requires negligble Server CPU
  • Network pass through only
  • Server memory block cache
  • Client decompression
  • Client warping/reprojection
  • Client caching to minimise network bandwidth
  • Client cache permits offline availability
  • Supports only JPEG2000 & ECW formats
  • Requires browser plugin or application support
  • Persistent connection can cause issues with corporate firewall systems
  • Browser plugin support for IE, Firefox, Chrome on Windows platform only

JPIP 1 2

  • Requires negligble Server CPU
  • Network pass through only
  • Server memory block cache
  • Client decompression
  • No warping/reprojection available
  • Client cache permits offline availability
  • Supports only JPEG2000 format types
  • Performance will vary according to JPEG2000 format type, layout, quality layers etc.
  • Requires browser plugin or application support
  • Complex protocol, requiring more server resources than ECWP to "do the same thing"
  • Large datasets beyond 10 GB become problematic to stream and requires significant amount of memory
  • Persistent connection can cause issues with firewall systems

WMS 1

  • Simple stateless HTTP protocol
  • Mature OGC standard with wide adoption
  • Server memory block cache
  • Caching extremely difficult unless served to a controlled client
  • Server resource intensive to decompress / reproject / produce the map output

WMTS 1

  • Simple stateless HTTP protocol
  • Enforced tiled grid to match common tilematrixsets
  • More difficult to implement than WMS
  • Allows simple integration into commercial mapping API's
  • Fixed resolution, fixed projection limits flexibility. Multiple tilematrixsets must be configured in advance
  • However rigidity means client caching simple to implement
  • Server memory block cache

WMTS via OTDF cache 1

  • Simple stateless HTTP protocol
  • Enforced tiled grid to match common tilematrixsets
  • More difficult to implement than WMS
  • Pre-rendered tile cache format typically 3-4x larger than equivalent ECW
  • Allows simple integration into commercial mapping API's
  • Fixed resolution, fixed projection limits flexibility. Multiple tilematrixsets must be configured in advance
  • However rigidity means caching simple to implement
  • OTDF is designed to match the output tilematrix
  • This 1:1 matching requires no server side decoding or processing. High performance & High scalability