Why hasn't any company come up with something better than SPDIF?

[chrispy35]

Subject says it all. Not a Rotel-specific question really but a question I think the people here can ponder as effectively as anyone else.

Everyone complains about jitter on digital connections and the fact that the data format of SPDIF is inherently difficult (compared to other possible formats) for clock recovery circuitry at the receiving end.

Why hasn't any company, high-end or other, chosen to go with a low-jitter (high transition-density) data format even though it might be proprietary to their system? The hardware for driving the digital link doesn't need to be changed, just the encoding/decoding of the data on it and possibly the data rate. I realize that some (most?) companies leave this to their IC vendor (like Crystal) so the question equally applies to the IC companies as well.

Do you think it is possible for a design to include two encoding formats with software selection between the two? The default could be SPDIF to reduce interoperability problems for people who don't care about details like this.

I would think that a major change like this would do more to improve jitter seen at the receive end of a digital link than going from modest-priced to ridiculous-priced interconnect would. i.e. Dealing with the problem at the source rather than trying to smooth the way for a problem that already exists.

[Aussie Geoff]

Chrispy35,

There are two competing formats (both standards based).

1) Firewire or IEEE 1394 - Which is used for digital sound only by some processors and recievers (there are some proprietory version of this too link Denonlink).

2) HDMI - which does digital video and sound via the one interface.

Both of these send more in packets of data and allow / require local retiming / reclocking which largely resolves the jitter issue.

HDMI seems to be the medium to long term answer - one plug and cable for your video and sound and at full digital resolution.

Geoff

[aud19]

Ditto what Geoff said. Regarding your question on why companies don't develop somethiing, even proprietary.... Money. It costs a lot of money to develop an entirely new connection etc and if it was proprietary would bennefit a very small amount of people. Meridian is one example of a company that has in fact done this but there equipment is very expensive, likely due to things like their proprietary connections as well as being top shelf and very awesome sounding gear.

So I guess the short answer is that you can have exactly what you asked for but you have to have BIG $$$ to buy Meridian gear :lol:

Jason

[chrispy35]

I see and partially agree both points here. I would say the money issue isn't really that bad for mid-fi stuff since the physical aspects of the connection are already dealt with and a $10 FPGA at each end could probably handle the various encoding/decoding schemes (i.e. SPDIF + something better). I guess it all comes down to return-on-investment.

As for 1394 and HDMI, the same question still applies: in pursuit of better sound, doesn't it make sense to use a superior interface? People will pay lots of money for vibration isolation, cables etc; why would they be less willing to pay for better technology inside the equipment? They're both standards so interoperability should not be an obstacle (HDMI has a specific audio packet, not sure how 1394 does things).

I'm just anxious to see the demise of SPDIF and the corresponding move on to something better (and not just for DVD players).