Prospect of better audio with HDMI V1.3a? Pay attention audiophiles - could be good!

[welwynnick]

I've had a big downer on HDMI (and V1.3 in particular) for as long as I can remember. TrueHD & DTS-HD compatibility made me think twice, though, but something happened today that got me really hopeful for audiophile standard interfacing in the near future. Really not sure where to put this thread, as it's about HD audio quality. HiFi? Receivers? Audio processors? HD DVD? Blu-ray? Cables? Dunno, I guessed it was most appropriate here, going by the usual audience. Well, I was browsing the latest HDMI V1.3a specification today.... What? What 's so funny? Doesn't everyone?
Anyway, I stumbled across a small new section that I hadn't seen or heard of before. This described, in not very much detail, a new and optional facility for the audio clock in the processor / amplifier to control the clock in the source using a CEC Set Audio Rate command back the source. Here's what the spec says:

7.11 Audio Rate Control Overview
The Audio Rate Control feature allows a Sink to slightly and continuously adjust the audio clock rate of the Source in order to match the Sink’s crystal-based audio clock. The Sink controls the Source’s audio clock rate with the CEC <Set Audio Rate> command. See CEC Supplement section CEC 13.16 for details. Source ACR behavior is not affected by Audio Rate Control. When Audio Rate Control is enabled the Source shall continue to generate correct ACR packets that accurately reflect the current (possibly adjusted) audio clock rate.

So what does this mean for us? I'm not entirely sure, as the new capabilty doesn't look closely defined, but to me it reads very much like Sony and Pioneers' descriptions of iLink operation. That's a rate adaptive connection, where the sink (amp) has the master audio clock, rather than the other way round.

It's not a synchronous connection, but I think it's a good thing because the timing information doesn't have to come from the player. That means it's less susceptible to the jitter from (usually) a video clock, a less-than-stellar connection, and an audio clock regenerator in the amp that is slaved to the player. In this configuration, the timing information is generated in the amp, leaving the source to supply only the digital amplitude information. So bits really will be bits for once. Timing info is sent back to the source to maintain flow control.

I don't know where this came from, and I don't know why I never noticed it before. Maybe the audiophile manufacturers, so dismayed by the disappointing jitter performance of HDMI, have been lobbying HDMI for something better. If this is it, then I think we have a great deal to look forwards to. Note that this facility was not incorporated in HDMI V1.3, though, only 1.3A.
Well chuffed, Nick

[ianh64]

Its the reason why no decent audio manufacturer has bought out an HDMI 1.3 solution yet.

Their view is to do HDMI once and get it right. Not come up with some compromise solution.

[peterweg]

Is this the lip-sync functionality described here?
http://www.hdmi.org/pdf/HDMI_Insert_FINAL_8-30-06.pdf

If so, its not new its dated 30th August 2006. I remember reading it about that time.

[welwynnick]

Originally Posted by peterweg

Is this the lip-sync functionality described here?
http://www.hdmi.org/pdf/HDMI_Insert_FINAL_8-30-06.pdf

If so, its not new its dated 30th August 2006. I remember reading it about that time.

That's time control, not frequency control. Lip sync functionality was one of the headline features added in V1.3. Audio Rate Control was quietly added with other tidy-up / bug-fix actions in V1.3a. Here's a summary of the additions for both:

1.3a 2006/11/10 Cable and Sink modifications for Type C (Table 4-20, 4.2.6)
Source termination recommendation (after Table 4-15)
Removed undershoot and max rise/fall time limits (4.2.4).
Modified slope of TP1 and TP2 eye diagrams (4.2.4, 4.2.5)
HDMI cable assembly AC-coupling support required (4.2.6)
CEC capacitance limits changed (4.2.10)
Valid range for RGB video quantization added (6.6)
Added audio sample rate exceptions for ARC (7.3, 7.3.1, 7.3.2)
Added Audio Rate Control Overview (7.11)

1.3 2006/06/22 Significant new features:
- Type C “Mini-Connector” (4.1.9.5, 4.1.9.6)
- Cable Categories 1 and 2 (4.2.6)
- Deep Color [4:4:4] (6.5, 8.3.2)
- Reference Cable Equalizer (4.2.3.2, 4.2.5, 4.2.6)
- Higher-speed single-link (4.1.2, 4.2.3,through 4.2.6, 8.3.2)
- xvYCC Enhanced Colorimetry (6.7.2.3)
- Gamut Metadata transmission (5.3.12, 6.7.3, Appendix E))
- DST audio format (5.3.10, 7.6.3)
- High-bitrate compressed audio formats (5.3.11, 7.2.4, 7.3.3, 7.6.2)
- Auto-Lipsync Correction feature (8.3.2, 8.9)
Updated normative reference from CEA-861-B to –D (1.2, throughout).
Updated Overview for new features (3)
Several minor editorial (throughout)

If I understand the significance of this, there is the opportunity for HDMI to provide an entirely jitterless connection between compatible players and amplifiers. It's all about jitter, and this is my attempt to explain it's significance. HDMI appear to recognise that some manufacturers will prefer have the audio master clock in the processor or receiver, and have the player slaved to that.

Synchronous links between equipment are ultimately preferred (where the phase is controlled directly). Here, the digital audio link seems to be isochronous instead, where the frequency is controlled instead. Similar, but not quite the same. Still, that's a pretty advanced implementation, but the HDMI feature does look very simple to incorporate.

Given that the potential of even 16/44.1 audio is usually limited by jitter (not resolution), the opportunities to make the most of 24/96 audio with this sort of jitter control are very attractive to audiophiles. Otherwise TrueHD would be little better than CD.

Nick

[Avi]

Originally Posted by welwynnick

That's time control, not frequency control. Lip sync functionality was one of the headline features added in V1.3. Audio Rate Control was quietly added with other tidy-up / bug-fix actions in V1.3a. Here's a summary of the additions for both:
If I understand the significance of this, there is the opportunity for HDMI to provide an entirely jitterless connection between compatible players and amplifiers. It's all about jitter, and this is my attempt to explain it's significance. HDMI appear to recognise that some manufacturers will prefer have the audio master clock in the processor or receiver, and have the player slaved to that.

Synchronous links between equipment are ultimately preferred (where the phase is controlled directly). Here, the digital audio link seems to be isochronous instead, where the frequency is controlled instead. Similar, but not quite the same. Still, that's a pretty advanced implementation, but the HDMI feature does look very simple to incorporate.

Given that the potential of even 16/44.1 audio is usually limited by jitter (not resolution), the opportunities to make the most of 24/96 audio with this sort of jitter control are very attractive to audiophiles. Otherwise TrueHD would be little better than CD.

Nick

Nick


Does this mean all HDMI "1.3" devices meet the minimum mandatory spec of "1.3a" or not ?

AVI

[Ian_S]

HDMI has no mandatory spec unfortunately.

Just a variety of versions and a whole set of optional functions. So, you can have an HDMI 1.3a chip fitted and still only use functionality that was available in HDMI 1.1.

Personally I expect to see a lot of this as I'm sure it will make sense to use one chipset/version on everything.

People will need to understand that having HDMI 1.3 does not mean deep colour, HD audio bitstream etc. It just means the HDMI chipset can transmit those things *IF* the rest of the electronics implement that capability.

You need to look at the functionality of the device, NOT the HDMI revision level.

[Timbo21]

Originally Posted by Ian_S

HDMI has no mandatory spec unfortunately.

But at least if it has the ability, then manufacturers are able to implement it, which is what companies like Meridian and Arcam will do, and I dare say Denon and others as well.

T.

[Avi]

Originally Posted by Ian_S

HDMI has no mandatory spec unfortunately.

Just a variety of versions and a whole set of optional functions. So, you can have an HDMI 1.3a chip fitted and still only use functionality that was available in HDMI 1.1.

Personally I expect to see a lot of this as I'm sure it will make sense to use one chipset/version on everything.

People will need to understand that having HDMI 1.3 does not mean deep colour, HD audio bitstream etc. It just means the HDMI chipset can transmit those things *IF* the rest of the electronics implement that capability.

You need to look at the functionality of the device, NOT the HDMI revision level.

I understand that the "implentation" may be optional but my point is that you can't "optionally" provide some features unless the version of HDMI supports it. So my question is do all HDMI "1.3" products have the "option" if implmented to offer all features supported by 1.3a or are there different "HDMI versions" chipsets for 1.3a and 1.3.

AVI

[Ian_S]

Originally Posted by Avi

I understand that the "implentation" may be optional but my point is that you can't "optionally" provide some features unless the version of HDMI supports it. So my question is do all HDMI "1.3" products have the "option" if implmented to offer all features supported by 1.3a or are there different "HDMI versions" chipsets for 1.3a and 1.3.

Different chipsets.

[welwynnick]

I think HDMI may have made a big mistake in not making every piece of functionality compulsory at each version. Allowing manufacturers to pick and chose what they want to implement has caused a great deal of confusion and mistrust. For all the fine talk and the good intentions, I think it's been an example of how not to implement a new standard. I can understand the commercial attraction of drip feeding new capabilty every year of so to generate more hardware sales, but the benefit of a new HDMI version is undermined if what it delivers in practice isn't clear. It doesn't necessarily do what it says on the packet.

V1.3a does look like rather an afterthough, tying up lose ends from the problems with 1.3. Just as it seemded to be with 1.2. No great surprise, though; that's how it happens with complex systems technology.

With regard to Audio Rate Control, although this may be simple to implement at the interface level through the addition of some simple CEC messages, the playback system itself may have some significant changes. It looks to me like a HATS / PQLS (?) type of implementation - i-Link over HDMI if you like. if that is the case, then the whole digital audio playback architecture would have to change. Which is the real objective, to my mind. The audio master clock will be in the DAC / processor / amp, rather than in the transport, and this is just a control interface to support that architecture. So I'd pretty sure that any V1.3 equipment will not support audio rate control.

Nick

[Rasczak]

Originally Posted by welwynnick

I think HDMI may have made a big mistake in not making every piece of functionality compulsory at each version.

To be fair though most normal consumers aren't really going to be that bothered about jitter. And, if the revision works as you have outlined (which I certainly hope it does), then the high end brands will use it on their kit anyway.

[welwynnick]

Originally Posted by Rasczak

To be fair though most normal consumers aren't really going to be that bothered about jitter.

Oh, I dunno - my wife has as much interest in hifi as I do in needlepoint, but when I plugged the i-Link in, even she was drawn to say "that sounds much clearer".

And that was only with CD. With 24 bit audio, there is even more to be gained or lost.

Nick

[Nic Rhodes]

Originally Posted by welwynnick

I think HDMI may have made a big mistake in not making every piece of functionality compulsory at each version. Allowing manufacturers to pick and chose what they want to implement has caused a great deal of confusion and mistrust.

If they had done this it would have saved my afternoon!! I think all these new technologies are making 'mistakes' about making stuff optional, whether HDMI, HD DVD or Blu-ray. Make it all compulsory.

[welwynnick]

I had a few more ideas about digital audiophile sound over HDMI.

Firstly, I had been thinking that HDMI V1.3 might allow improved audio, because, with THD or DTSMA being compressed audio, the timing info (audio clock) would have to come from the DAC / processor / amp. That means that the connection wouldn’t contribute to the overall jitter. On reflection, I doubt this would be realised, because the audio clock would probably still be derived from the HDMI video clock in the same way that PCM audio is. That suggests bit-stream may indeed sound the same as LPCM, so we may as well use the latter over V1.1.

However, use of the Audio Rate Control with V1.3a might still be an advantage. I think the HDMI clock might be going in the wrong direction, though. It’s the sink that needs the clock, not the source. So why not have the clock run back to the source – the line is already there? If rate control can be implemented in CEC protocol from the sink, then why can’t the direction of the clock? The source will be slaved to the sink, and it won’t matter that it’s the video clock – all the source has to do is provide the digital data, not the clock.

This could work with either a disc player or a media player (from a hard drive). But of course, this wouldn’t work with any sort of broadcast source, where the sink would have no control over delivery. However, if Audio Rate Control can work backwards with a clock in the sink, then perhaps it could work forwards with a source clock. This time, the source clock would be responsible only for flow control, but the timing info would still come from clock in the sink, just slaved to the source clock by forward running Audio Rate Control commands. That would be rate control, not phase control, as it’s not synchronous. It wouldn’t be the same as we have now, where the audio clock is decimated directly from the video clock in the source.

Still not sure about all this; just thinking out loud. Maybe all the clever HDMI committees (what am I saying?) have discussed and rejected this years ago.

Regards, Nick

[Rasczak]

Originally Posted by welwynnick

Still not sure about all this; just thinking out loud.

Please keep doing so - it's very interesting to hear your take on this. HDMI is certainly far from the perfect interface. However from an audiophiles perspective won't we just be better off staying with full analogue audio decoding in our players rather than looking to pass decoding to the receiver? I can certainly see the benefits for the budget market to adopt a 'transport only' strategy for their players though.