Hi. I'm in a bit of a quandary here, since I'm not sure how to proceed in using an ADV212 J2K compressor for my particular implementation. We're trying to build a system that can select from one of six "old school" NTSC (analog) video inputs, digitize that video, and then pass it to a 212 for compression. While our eventual implementation will stream J2K data to an FPGA for formatting and eventual transmission to storage, we're trying to work up a proof of concept configuration first that simply takes the J2K data stream into a high speed reconfigurable digital I/O card, which then stuffs the succession of compressed images into a PC memory via the PCI bus interface on the card. Then code in the PC will operate on the images to decompress and display them on a monitor. My point of confusion is that this all needs to happen real-time with no loss of image data no matter what compression parameters are chosen, and I'd like to keep the interface between the 212 and that I/O card as simple and straightforward as possible. My first thought was that JDATA would be the easiest implementation, and I can just use a clock from the 212 along with valid data handshaking to fetch and synchronously clock out the J2K data when it is available. However, I've seen documentation somewhere here that says that JDATA is a relatively slow way to transfer data out of the 212. "Relative" to what? More importantly, is it so slow as to prevent what I'm trying to do here - a real-time pipeline of that video? Can you guys offer any advice about how to approach this problem before I proceed too far based only on my own best guess? Thanks!