eARC Firmware Updates for HDMI 2.0b class devices

There's no reason for receivers that support object based audio and eARC to not support object based audio through eARC.

Mandatory lip sync is the fly in the ointment and the latencies for audio decode will vary depending on the format, channels and bitrate. HDMI has offered a technical solution for lip sync since 1.3 but its implementation has only ever been as good as the weakest chain in the hardware setup.
 
That should be easy to solve by adding a delay value in settings, where the display delays showing the frame by however many specified ms/frames. You'd only need enough RAM to buffer a few decompressed frames.
 
That should be easy to solve by adding a delay value in settings, where the display delays showing the frame by however many specified ms/frames.
One device is decoding video, another device is decoding audio (and only gets audio through eARC). Which device is the master device and how does that device know what the latency is for the other device, possibly the other two devices if the master device is the Blu-ray player and/or receiver? The latencies are very unlikely going to the the same.
 
You set it on the TV as an 'audio sync' option. The TV then sends audio to the decoder (either internal or external) as soon as the audio is decoded, and updates the video with the user-selected latency so the video updates when the audio decoder is ready with the audio stream.
 
You set it on the TV as an 'audio sync' option. The TV then sends audio to the decoder (either internal or external) as soon as the audio is decoded, and updates the video with the user-selected latency so the video updates when the audio decoder is ready with the audio stream.
What about for people who using audio not routed from the TV?
 
Makes no difference. The TV still displays with a user selected latency to compensate for latency differences between display and audio. The only time this won't work is when audio is ahead of the picture. Is that ever the case? If so, compensation can be done in the player in software, just delaying the streams.
 
I never realized just how bad ARC was until I read the following:

What this means is that the bandwidth that currently limits HDMI-ARC outputs in the connection between TVs (output) and receivers (input) will see a bandwidth boost from the previous 1 Mbps in ARC to a whopping 38 Mbps in eARC.

ARC is somekind of S/PDIF over HDMI with all its limitations (DD 5.1, DTS 5.1 and PCM 2.0), some newer TVs can carry DD+ through ARC (upto 7.1 and Atmos), but not DTS-HD HRA (still not enough bandwith). https://www.avsforum.com/forum/39-n...olby-digital-plus-dd-atmos-over-hdmi-arc.html
 
Mandatory lip sync is the fly in the ointment and the latencies for audio decode will vary depending on the format, channels and bitrate. HDMI has offered a technical solution for lip sync since 1.3 but its implementation has only ever been as good as the weakest chain in the hardware setup.

It was optional, though. eARC will have mandatory lip sync. HDMI.org has a table comparing S/PDIF to HDMI-ARC and HDMI-eARC
earc.jpg
 
Makes no difference. The TV still displays with a user selected latency to compensate for latency differences between display and audio.

If the user has to fix it manually, this is not automatic lip sync.

The only time this won't work is when audio is ahead of the picture. Is that ever the case? If so, compensation can be done in the player in software, just delaying the streams.

Which player's softtare? The Bu-ray player, the A/V receiver? How does the player know what the latency is for the TV to decode, perhaps upscale, and add processing and output an image to the panel and the same for the audio system so as to delay one of the streams? What if the user changes the TV image settings and that changes?
 
I'm just talking about how lip syncing can be solved. It doesn't need any fancy-schmancy handshaking or SMTPE synchronisation or anything. Obviously the ideal is systems that can automagically coordinate, but if the systems can't do that because the hardware makers can't get their act together, it'd be easy in the player (BluRay or streaming box) or in the case of audio pass-through, the TV, to adjust the audio so it lip-syncs.
 
I'm just talking about how lip syncing can be solved

Not really, you saying it can be solved but not providing any solutions to the actual problem.

It doesn't need any fancy-schmancy handshaking or SMTPE synchronisation or anything. Obviously the ideal is systems that can automagically coordinate, but if the systems can't do that because the hardware makers can't get their act together, it'd be easy in the player (BluRay or streaming box) or in the case of audio pass-through, the TV, to adjust the audio so it lip-syncs.

If it's "easy" you can surely explain how to do it because A/V sync has been a problem for more than a decade and trickier in this new implementation where one device may only get video and another only gets audio and neither talk to each other and yet both introduce variable latencies. As for audio passthrough, that - as the name implies - just passes on the audio stream. The device passing on the audio has zero knowledge of what latency is introduced by the device doing the decoding on the other end.

You either have the answer to a very complex problem that has plagued the A/V industry for years or you're not understanding the problem.
 
AFAICS, lip sync is a problem for a rather small part of the population who own older TVs and/or receivers that induce a disproportionate amount of lag.
My guess is TV/AV makers nowadays prefer to just use higher-clocked transcoders/codecs that don't induce discerneable lag than to implement more expensive ICs with buffers and needs to constantly pull info from CEC (which isn't all that reliable to be honest) just to get the same result.

I for one have no lip sync issues for 99.9% of the usage I have with my TV + Receiver connected through ARC, which is PS4 games and Netflix through PS4.
When I use the system with my HTPC I just use a dedicated HDMI connection for sound to the AV Receiver and another HDMI 2.0 to the TV, so I can get 24bit 192KHz 5.1 LPCM or lossless codecs.


What I do have issues is with the fact that my HDMI 1.4 receiver can't get lossless sound when I'm using the PS4 Pro connected to the 4K TV.
I refuse to upgrade my decent receiver for HDMI 2.0, but HDMI 2.1 should be much more future-proof thanks to eARC.
 
Not really, you saying it can be solved but not providing any solutions to the actual problem
Huh? It's really not complicated.

Lip-sync issues occur when the audio plays with a time differential to the video, so when the display is showing frame n, the audio is playing the audio for frame n+x, where x is the delta (+ve or -ve) of the latency difference between the display and the audio. So the video source takes synchronised audio and video data, outputs them to the display and the speakers, both of which may introduce an amount of lag between getting the data and outputting it. There's only one way to fix that and that is to add a delay to whichever of the two devices has lower latency, so the latencies are the same for both. If the audio plays 48 ms after the video, add a 48 ms to the video display. If the audio plays 29 ms before the video, add a 29 ms delay to the audio.

Where this compensatory lag adjustment takes place can be anywhere in the device chain. The obvious place to do that is the player which has full control over delivering the audio and video steams, so can delay them however much it wants. The display and audio system don't need to know anything as they just get the data to work with. If you're passing audio through a TV and on to a receiver, you could pass the combined signal to the TV and have it apply a latency to the audio when passing it on to the receiver. Or you could have adjustment on the audio receiver to delay playback for laggy video. Of course, that can only adjust for when the receiver is ahead of the video - you can't play audio before you've received it from the combined stream. Which is where the adjustment on the player is the smartest place to have it. Not hard to implement at all and it doesn't need anything clever. Other solutions like device synchronisation using SMTPE would make it all seamless, but add complexity.

As to why the industry hasn't done this, who knows? The world is full of half-arsed and inept solutions, and common problems that are easy to fix but not fixed. As far as the industry is concerned, every component can blame the other. "It's the display manufacturer's fault for introducing lag - they should fix it." "It's the audio receiver's responsiblitiy to handle audio; they should fix it." Perhaps the cost to the players to support enough frames of buffered video to allow enough lag compensation is deemed too much? Although I'd be surprised if audio latency was higher than video and I'd have thought you just need compensation on the audio side, which is cheap(er). If we see receivers with audio latency compensation, perhaps the player makers just assume that'll take care of it and not bother to put in settings in the players?

However, the problem is easily solved with the introduction of user-selectable latency adjustment in the player or elsewhere in the chain.
A very quick Google says receivers and some soundbars already have audio compensation settings...
Most AV receivers have a setting where you can alter the lip-sync delay so it matches the video perfectly. Some sound bars also have the option.

And VLC has audio adjustment so you can adjust lip-syncing mid-stream with keyboard shortcuts.
 
Huh? It's really not complicated.

And yet, you have failed to provide a solution to a problem which has existed for more than a decade and which you describe as "easy" and "not complicated" but where your posts only ever describe the problem and not a solution. It's like saying "cancer is bad, we should solve that". Yes, yes we should. :yep2: But it's more complicated than just saying we should solve it.

A very quick Google says receivers and some soundbars already have audio compensation settings...
And VLC has audio adjustment so you can adjust lip-syncing mid-stream with keyboard shortcuts.

Exactly. eARC requires that the HDMI system solves the problem rather than the user. Putting a latency slider somewhere in the mix is not the desired solution here. That 'solution' has existed for as long as the problem and like the problem, that manual solution requires continual change because latencies for both screen and audio are indeed variable. :yes:
 
And yet, you have failed to provide a solution to a problem which has existed for more than a decade and which you describe as "easy" and "not complicated" but where your posts only ever describe the problem and not a solution.
Dude it's right there in the post:
-> Delay the source that is coming up earlier
 
That 'solution' has existed for as long as the problem and like the problem, that manual solution requires continual change because latencies for both screen and audio are indeed variable. :yes:
What do you mean 'continual change'? During a specific video, or any existing combination of courses and receivers, it should remain constant. Set it up for your TV and your receiver, perhaps with settings for different input channels, and you're done. Worst case, you pick up the remote and push the 'adjust lip-sync +/-' buttons to fix it the same as adjusting the volume. That to me is a solution. An automagical one would be nice, but I consider the problem solvable without needing complex protocols that CE manufacturers can't settle in and implement properly. In the case where you're using pass-through audio, the TV can determine from its settings how much latency it's adding and adjust the audio pass-through latency on top of a user specified constant. So if you set the audio correction +60ms in game mode and the audio is delayed 60ms before sending, and the TV is switched to movie mode adding an additional 120 ms of processing, the audio can be adjusted for 180 ms latency.
 
Setting aside what is a possible fix for the issue, the correct solution is for it to be automatic. Then a product either has a working implementation, or it doesn't and can be evaluated as such.
 
Dude it's right there in the post:
-> Delay the source that is coming up earlier

That's what you need to do, not how. How does the device serving the video and audio know how much to slow either stream to whatever device is faster?

What do you mean 'continual change'?

If the user changes the audio output device, say TV speaks to/from an A/V receiver and surround system, the latency will certainly change. Behind the scenes the playing device may actually switch audio formats. If the user changes the display settings, perhaps switches from the 'stadium mode' they were using for watching footie over to 'cinema mode' which is heavy on post-processing, that will alter the latency of the display device.

The HDMI 2.1 eARC system requires this just work.

Worst case, you pick up the remote and push the 'adjust lip-sync +/-' buttons to fix it the same as adjusting the volume.

That's not the worst case, that's how things are now. eARC is supposed to solve this.
 
Back
Top