[gorkish]

It’s technically not an all or nothing deal with hdmi/dvi. That is to say that bit errors do indeed manifest as image artifacts, though normally imperceptible. I think some people understand that error correction is normally present in digital audio, so they naturally assume that video would be the same. But that is not quite true. For one thing there weren’t cheap chips that could do that at gbps data rates when DVI standard was first created. It was not until DisplayPort 1.4 that they added optional FEC. This is required because a bit error in a compressed stream would manifest as an entire macroblock busting, which affects potentially a large pixel area and multiple frames.

All that being said it’s unusual to find a cable that is both clean enough to do the handshake and keep sync but noisy enough to give you visible snow. So it’s still quite true that practically speaking, yes, it’s usually an all or nothing deal. Cable quality can and does matter though. I was a BlueJeans customer for a long time before the brief Monster spat, but it endeared them to me, and I still try to buy from them when I need to buy a cable I need to be absolutely sure of.

[PaulDavisThe1st]

> I think some people understand that error correction is normally present in digital audio

The only such error correction I'm aware of is when reading data from a CD, which at this point is a tiny part of digital audio. Is there something I'm missing?

[adgjlsfhk1]

most digital signaling has error correction. pcie,SSDs, Ethernet etc

[PaulDavisThe1st]

Not much of which is directly connected to digital audio as a specific thing, however.

[gorkish]

Error detection is present in S/PDIF PCM (including when transported on hdmi) and is also an inherent byproduct of most audio codecs when a digital bitstream is being used, which is normally the case today.

FEC and other types of error correction or recovery is ubiquitous in wireless audio and communications applications including phone calls, Bluetooth, VoIP, wireless microphones, and digital radio. Responsibility for the error correction is sometimes part of the underlying transport mechanism and sometimes incorporated directly into the codec. Encryption & privacy requirements for audio also mean that we solved these problems long ago. IIRC that the WWII SIGSALY encrypted telephone between the US and UK required and implemented error correction.

[PaulDavisThe1st]

I guess being too close to the DAW-space, I tend not to think about codecs. Digital audio to me is conceptually "pure" PCM (or DSD), and most things that deals with that format do not do error correction that I can think of. S/PDIF is good counter-example, and possibly (for similar reasons) ADAT might be as well.

By contrast, most audio-over-IP formats do not (they rely on the IP-level checks).

Anyway, thanks for pointing out the rather important world filled with codecs that we actually live in.

[gorkish]

In your DAW world, AES/EBU transport parity bit corrects most single bit errors as well. It’s a testament to the comprehensive handling of the issue that you as a professional do not need to do much thinking about the problem. Point is still that audio bit errors are historically accounted for due to the obvious consequences of a discontinuity. This persists, often with layers of redundancy, despite that they rarely occur. Video bit errors, not so much