I'm using a Sony SDP EP9ES for capture. I also have a small Yamaha demodulator but I ran into the same issue, so I just used that one when watching AC-3 discs, rather than capturing them.

I have a Pioneer RFD-1, and I also have to re-start the capture after a side change, pretty annoying. Maybe your Sony has some additional circuitry to prevent the signal from dropping (which would also be annoying on some receivers, I know my SPDIF-modded SNES causes the audio to drop out on my Onkyo receiver when the sound chip isn't producing audio), not sure.

Was doing a little research on this, and it looks like someone figured out how to decode the RF signal to AC-3 for an FPGA project. That code was then ported to the ld-decode library, and Windows binary versions are available as part of vhs-decode. There's an example here of using ffmpeg and sox to get it in the correct format for decoding:

https://github.com/happycube/ld-decode/t...rocess-ac3

Edit: thought it could be captured as audio, but the RF output of the AC-3 signal is at 2.88MHz so it would need to be captured by a software-based radio receiver of some sort.

Windows downloads here:

https://github.com/oyvindln/vhs-decode/releases/

Does this mean you don't even need an AC3-demodulator anymore and you can simply capture the AC3 as WAV and decode it later, just like it has to be done for DTS?

Yeap, for offline processing at least. The original author of the code was looking to use it in an FPGA device for real-time decoding/playback, but looks like he may have moved on. The VHDL files are there however for someone else implementing a hardware solution someday.

So an analog capture of the noise and then convert that to digital data? That’s wild. You’d need a non AC3 player to do that as the analog out would be muted on an AC3 unit.

Oh wow, I think I already have all my AC3 LDs captured but that's awesome.

I hope that real-time playback is possible someday as well.

(2024-07-26, 06:15 PM)borisanddoris Wrote: So an analog capture of the noise and then convert that to digital data?  That’s wild.  You’d need a non AC3 player to do that as the analog out would be muted on an AC3 unit.

Ah yeah, that would make things difficult for those with AC3 compatible players. I guess capturing the player's AC3 out wouldn't work either.
Maybe there's a way to force the player into outputting the right analog channel in service mode or something.

(2024-07-26, 06:15 PM)borisanddoris Wrote: So an analog capture of the noise and then convert that to digital data? That’s wild. You’d need a non AC3 player to do that as the analog out would be muted on an AC3 unit.

Isn't the AC3 RF out jack the same thing? I think it may work with either, but if someone has a disc I'm sure we can figure it out

Yes the AC-3 RF output is the raw FM signal from the player which the demodulator filters and converts to a bitstream signal. I think that using the right channel analogue audio output (assuming it isn't muted) would result in the signal being FM demodulated which could corrupt the data, which is QPSK modulated, but couldn't say for sure. It's certainly an interesting development!

Borisanddoris, going back to your initial problem, could you elaborate? It sounds like you're saying when you run a LD through your demodulator into your A/V receiver it locks instantly whereas when running into a capture device there's a delay?