It is currently 12 Nov 2019, 01:16

All times are UTC [ DST ]

Search found 1 match

Author Message

 Jump to forum   Jump to topic

Posted: 12 Dec 2017, 07:16 

I realize I've been quiet about the duplicator board progress, so here is a quick update:

I've completely rewritten the FPGA code for the project to make everything more robust and to increase the level of system testing and the error checking. The design now runs the FX3 (USB 3) communication at 64 MHz (double the capture speed) to allow better buffering and 'catch-up' when the PC slows down for any reason - the non-realtime nature of a PC performing the capture to disk causes bursting of data via the USB3; this is now much better implemented.

I've also written a Linux (Ubuntu) GUI front-end application in QT which monitors and controls the capture process. The latest github build is a fairly simple GUI, but contains a robust USB and disk buffering implementation that has been tested successfully with capture runs of over 60 minutes (generating in excess of 250Gbytes of capture). By making an open-source Linux application specifically for the duplicator I can now expand the functionality at will to make the capture process far more user-friendly. Also the whole 'chain' of the capture and decode process can run on a single Linux machine (before the capture relied on the Cypress FX3 Windows test applications for capture).

I have now received 3 Pioneer LD-V4300D laserdisc players and these players will be the reference hardware for testing the Domesday Duplicator board. The new players are both NTSC and PAL capable, so they represent a much more universal approach than the Sony PAL player I was using previously. To get the best results the players need to be correctly calibrated and set-up. I've ordered the required Pioneer 8" test disc (from PacParts in the USA) and some service remotes - I also have the service guides and the test equipment required.

I've also just ordered the parts to make up a couple more Domesday Duplicator boards; these should be ready by the new year and will be used to expand the testing scope (one will go to Mr Happycube and the other will be a secondary test card for my own rig).

I'm also planning on implementing what I call "Player Integrated Capture" into the Linux application - this will provide automatic control of the LD-V4300D from the Linux GUI. Full and partial disc capture will be controlled automatically between the Linux app and the player using the RS-232-C player interface. This should take out the hassle of capturing disks and allow additional things like multiple captures of a range of frames where the GUI pauses on a frame automatically, checks the player is where it should be, captures, and then moves on. Once the VBI frame position information is exposed via ld-decode, this should allow for really smart capture techniques - especially when only one source laserdisc is available. Basically you will be able to specify a frame or range of frames in the app and it will all happen automatically. My plan is to implement this in 'layers' so others can easily add in support for additional laserdisc players.

Once the new LD-V4300D machines are running and I have an end-to-end tested capture, I'll push a Github release (hopefully in the next 2 weeks). After that, everything is just 'enhancement' work :)
Page 1 of 1 [ Search found 1 match ]

All times are UTC [ DST ]

Jump to: