No sound on HDMI In after updating firmware

[mkwilson]

Hi All,
For some time I'd been using some pretty old firmware (Nov 2016?), but was forced to update as desperate attempt to get my T4 running again (see other thread). This worked. However, after updating the firmware I find that if I plug my laptop into the HDMI In, it is not recognising the Beyonwiz as an audio device (video is fine). I have no problem with audio or video from Fox IQ4 into the HDMI-In port. I can't change audio device settings on the laptop because the HDMI audio device is not recognised. Any ideas? I can't find anything in T4 setup menus that would appear to help.

BTW, I'm 99% sure that my laptop used to recognise my Samsung TV that was on the other side of the Beyonwiz as the HDMI device (rather than the Beyonwiz).

Cheers
Mark

[Grumpy_Geoff]

I have no problem with audio or video from Fox IQ4 into the HDMI-In port.

That says to me that there's nothing wrong with the T4's HDMI-IN port.

[mkwilson]

I have no problem with audio or video from Fox IQ4 into the HDMI-In port.

That says to me that there's nothing wrong with the T4's HDMI-IN port.

I agree, there can't be a problem with the port. I'm trying to work out whether there's something in the change in firmware that will make my attempts to sort this our a fruitless task, or if there is something somewhere in the setup on Beyonwiz that I can tweak (I lost all settings when updating the firmware and not everything is where I'm used to it being).

[Paul_oz53]

Have you powered off the T4? The hardware can lock up on HDMI IN audio with no sound. Rebooting will not fix this. That the IQ4 works suggests this is not the problem but worth a try.

[MrQuade]

It's because your laptop is now trying to send multichannel audio to the T4. It needs to be forced to send stereo pcm audio.

This has popped up several times recently. It is a result of the improved audio code that came with one of the previous firmware updates. The T4 advertises multichannel capability on its HDMI-in, so devices that Auto negotiate audio will send that instead of stereo.

I don't know why the T4 tells upstream devices this, since the actual firmware update was supposed to just add multichannel downmixing output.

[mkwilson]

It's because your laptop is now trying to send multichannel audio to the T4. It needs to be forced to send stereo pcm audio.

This has popped up several times recently. It is a result of the improved audio code that came with one of the previous firmware updates. The T4 advertises multichannel capability on its HDMI-in, so devices that Auto negotiate audio will send that instead of stereo.

I don't know why the T4 tells upstream devices this, since the actual firmware update was supposed to just add multichannel downmixing output.

Thanks. I did see those other threads, but it wasn't apparent that a firmware change initiated the problem. Now to see how I can force the laptop to send stereo pcm....the existing audio devices (headphones / speakers_ are already set to stereo output.
Cheers.

[peteru]

Get your PC to decode the EDID of the T4 and report back what you find. The T4 uses a custom EDID that only advertise capabilities that it actually supports. It sounds like your PC is giving up because it's not seeing something it expects to see. It could be as simple as the PC insisting on a specific sample rate.

Can your PC output 32, 44.1, 48 and 96 kHz audio over HDMI? Are all of these optional or is your PC configured for only one of these?

[mkwilson]

Get your PC to decode the EDID of the T4 and report back what you find. The T4 uses a custom EDID that only advertise capabilities that it actually supports. It sounds like your PC is giving up because it's not seeing something it expects to see. It could be as simple as the PC insisting on a specific sample rate.

Can your PC output 32, 44.1, 48 and 96 kHz audio over HDMI? Are all of these optional or is your PC configured for only one of these?

Hi Peter,

Thanks very much for following up. How might I go about getting the PC to decode the EDID? Is this done through some thing in the Extended Display settings on the Windows 10 control panel, the Intel Graphics settings or something else?

I'm 99% sure that I've been able to set the PC output to to those frequencies over HDMI...when I've been connected to an appropriate device (can't see one at the moment). The PC is a Dell XPS 15 (top of the line options).

Cheers
Mark

[MrQuade]

Thanks very much for following up. How might I go about getting the PC to decode the EDID? Is this done through some thing in the Extended Display settings on the Windows 10 control panel, the Intel Graphics settings or something else?

You might want to have a look at Nirsoft MonitorInfoView.
https://www.nirsoft.net/utils/monitor_info_view.html

That will at least tell you what the Laptop is getting from the T4 (or "monitor" as it thinks it is).

Not sure how you get the laptop to intercept or spoof that info, or to otherwise force it to use a specific sound format though. That will be driver dependent I imagine.

[mkwilson]

Thanks very much for following up. How might I go about getting the PC to decode the EDID? Is this done through some thing in the Extended Display settings on the Windows 10 control panel, the Intel Graphics settings or something else?

You might want to have a look at Nirsoft MonitorInfoView.
https://www.nirsoft.net/utils/monitor_info_view.html

That will at least tell you what the Laptop is getting from the T4 (or "monitor" as it thinks it is).

Not sure how you get the laptop to intercept or spoof that info, or to otherwise force it to use a specific sound format though. That will be driver dependent I imagine.

Thanks. I will be a little 'time poor' over the next 48 hours, but will definitely have a go and report back.

[peteru]

Getting the EDID info is hardware, driver and operating system dependent. You can get the EDID data using the nVidia settings utility under Linux, if you have an nVidia card.

You could boot using a Linux live DVD / USB to verify that Linux can output to the HDMI port and the Beyonwiz works correctly when fed a signal from the PC using Linux. If that works, you will have established that the problem is with the Windows device drivers.

[peteru]

Just to be sure, once you decode the EDID, you should see that there is an audio descriptor block in a CEA-861F ver 3 extension which advertises the following capabilities:

Format: Linear PCM
Channels: 2
Sample rates: 32, 44.1, 48 kHz
Depth: 16 bit

That is the only audio capability block. The old EDID in previous version of firmware was falsely advertising all sorts of capabilities that were not properly supported. This new EDID only advertises what is known/expected to work.

I guess the trick with your PC is to make sure it is configured to output stereo to HDMI, not multi-channel.

[mkwilson]

Just a (not very useful) update. I have tried downloading / using a number of EDID readers / utilities found on the web (including the one that Mr Quade kindly suggested). Each of these thus far seems to only report attributes of the video capabilities of the 'monitor'. They each also provide a block of 'raw data', comprising a a matrix of hexadecimal numbers that mean zilch to me. I also tried the nVidia control panel on my laptop (it's a GeForce one) but it only had 3D rendering options. I am waiting to get an account at a site the hosts a product that is held out to report audio specs.

I have zero experience with Linux unfortunately. I'm on Windows 10. The big challenge to me seems to be that I can't force HMDI output to stereo until the relevant device is detected. I note that the HDMI audio does default to stereo when plugged into my Marantz AVR (though I can make it play 5.1 without a problem). Will keep trying as and when I get time.

No a huge deal I guess...my main use was recording a few programs from the iPlayer before they expired so I could take them with me when I went for a weekend break in the hills.

Thanks again.

[peteru]

Just paste the hex data block into this site: http://www.edidreader.com/

[mkwilson]

Just paste the hex data block into this site: http://www.edidreader.com/

Ahhh....that's how that site works....lol. Learn something every day.

I've pasted the hex data block from other monitors into that site and some of them do report audio details (e.g. for my AVR

But the data from the Beyonwiz does not contain any details re audio. In the 'Extension Header Info' section, the 'supports basic data' checkbox is not checked.

[peteru]

In that case the drivers on your computer are broken and destroy the EDID from the Beyonwiz. It may be fixable with a driver update.

For your reference, here is the EDID that is programmed into the drivers in 19.3 series firmware:

Code: Select all

	0x00, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0xFF, 0x00, 0x0A, 0xFA, 0x34, 0x54, 0x00, 0x00, 0x00, 0x00,
	0xFF, 0x1C, 0x01, 0x04, 0x80, 0xA0, 0x5A, 0x78, 0x1A, 0xDA, 0xFF, 0xA3, 0x58, 0x4A, 0xA2, 0x29,
	0x17, 0x49, 0x4B, 0x00, 0x00, 0x00, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x01,
	0x01, 0x01, 0x01, 0x01, 0x01, 0x01, 0x02, 0x3A, 0x80, 0x18, 0x71, 0x38, 0x2D, 0x40, 0x58, 0x2C,
	0x45, 0x00, 0xBA, 0x88, 0x21, 0x00, 0x00, 0x1E, 0x01, 0x1D, 0x80, 0x18, 0x71, 0x1C, 0x16, 0x20,
	0x58, 0x2C, 0x25, 0x00, 0xBA, 0x88, 0x21, 0x00, 0x00, 0x9E, 0x00, 0x00, 0x00, 0xFC, 0x00, 0x42,
	0x65, 0x79, 0x6F, 0x6E, 0x77, 0x69, 0x7A, 0x20, 0x54, 0x34, 0x0A, 0x20, 0x00, 0x00, 0x00, 0xFD,
	0x00, 0x17, 0x3D, 0x0F, 0x44, 0x0F, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x01, 0x7A,
	0x02, 0x03, 0x36, 0x71, 0x7F, 0x03, 0x0C, 0x00, 0x40, 0x00, 0xB8, 0x2D, 0x2F, 0x80, 0x00, 0x00,
	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
	0x00, 0x00, 0x00, 0x00, 0xE3, 0x05, 0x1F, 0x01, 0x49, 0x9F, 0x14, 0x20, 0x10, 0x05, 0x04, 0x13,
	0x3C, 0x12, 0x23, 0x09, 0x07, 0x01, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x20

[mkwilson]

In that case the drivers on your computer are broken and destroy the EDID from the Beyonwiz. It may be fixable with a driver update.

For your reference, here is the EDID that is programmed into the drivers in 19.3 series firmware:

Code: Select all

Thanks. Yeh, I've tried updating drivers through several methods (that was my first reaction), but they all insist that I've got the most recent driver. The whole thing was working fine until I updated the Beyonwiz firmware. I've also tried connecting with 2 other laptops (1 brand new, 1 much older) with the same problem. Will try some other driver tricks and let you all know if I have any success.

[peteru]

The old drivers advertised 2 channel and 6 channel AC3, which was not correct as the T4 can not handle compressed audio input on HDMI.

[mkwilson]

The old drivers advertised 2 channel and 6 channel AC3, which was not correct as the T4 can not handle compressed audio input on HDMI.

My HDMI cable is in good nick, but I doubt it is HDMI 1.4 or 2.0. I noticed from the data that I pasted into the EDID reader that part of the CEA data block (Vendor Specific Data Block HDMI 1.4) refers to HDMI 1.4. Is there any chance that the cable could be part of the problem? I'm guessing not.

[peteru]

Very unlikely. You can get HDMI working and the EDID is retrieved, so you know the cable works. This sounds like a software problem on the PC.

[mkwilson]

Very unlikely. You can get HDMI working and the EDID is retrieved, so you know the cable works. This sounds like a software problem on the PC.

OK, thanks. Just to be clear, the drivers/software that we are talking about on the PC are for the Intel Display Audio, and potentially something that might appear in the audio inputs and outputs should the Beyonwiz ever be recognised as providing audio? Or am I missing something? My Nvidia drivers are all current too.

[peteru]

Laptops can be tricky. If you have a dual GPU system (both Intel and nvidia), then, depending on the mobo connector configuration, your HDMI audio could be Intel or nvidia. It will also depend on the BIOS config settings. To complicate things further, for some laptops you can only use the manufacturer drivers (which are often not updated as often as they should be), but for others you also have the option of installing vanilla drivers from Intel or nvidia. Windows is pretty messed up in that regard and makes it almost impossible for a user to fix things up. Linux driver situation is usually simpler. But I digress...

I suggest that you look into the drivers for all HDMI and audio related hardware you have. Probably Intel HDA, (very often Realtek audio too), Intel GPU and nvidia GPU. Plus, whatever Dell have for your particular model. Bear in mind that Dell tend to be pretty bad when it comes to having up to date and bug free drivers for their laptops.

Please post the hex EDID data that your laptop reports. I'll take a look and see if I can figure out what's happening.

[mkwilson]

Thanks Peteru. Below is the hexdata:

0x00 0xFF 0xFF 0xFF 0xFF 0xFF 0xFF 0x00 0x0A 0xFA 0x34 0x54 0x00 0x00 0x00 0x00 0xFF 0x1C 0x01 0x04 0x80 0xA0 0x5A 0x78 0x1A 0xDA 0xFF 0xA3 0x58 0x4A 0xA2 0x29 0x17 0x49 0x4B 0x00 0x00 0x00 0x01 0x01 0x01 0x01 0x01 0x01 0x01 0x01 0x01 0x01 0x01 0x01 0x01 0x01 0x01 0x01 0x02 0x3A 0x80 0x18 0x71 0x38 0x2D 0x40 0x58 0x2C 0x45 0x00 0xBA 0x88 0x21 0x00 0x00 0x1E 0x01 0x1D 0x80 0x18 0x71 0x1C 0x16 0x20 0x58 0x2C 0x25 0x00 0xBA 0x88 0x21 0x00 0x00 0x9E 0x00 0x00 0x00 0xFC 0x00 0x42 0x65 0x79 0x6F 0x6E 0x77 0x69 0x7A 0x20 0x54 0x34 0x0A 0x20 0x00 0x00 0x00 0xFD 0x00 0x17 0x3D 0x0F 0x44 0x0F 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x01 0x7A

Yeh, I've had some fun with Dell drivers over the years....exacerbated by the fact that their 'Dell Support Assist' gave incorrect feedback regarding the status of several drivers. I've had a go at each the drivers to which you refer. Where relevant I use the Intel , nvidia proprietary drivers.

[peteru]

Hmmm. That's only the first half the EDID. It seems to match what the T4 sends. However, that part of the EDID has no information about audio.

The second half of the EDID - the part that describes audio support and the various video modes - is missing.

[peteru]

I'm just having a look at the differences between the old and new EDID in the T4 drivers. Trying to figure out what may be tripping up your particular configuration. The most obvious difference is the upgrade from EDID version 1.3 to 1.4. This adds support for additional descriptors, but it's supposed to be backwards compatible with compliant implementations.

The other difference is that the old EDID had a whole bunch of video and audio modes that were advertised, but not working correctly when actually used. The new EDID no longer offers these modes, but it does add new modes that are known to work. The removed modes are:

Code: Select all

    VIC  3:   720x480    59.940 Hz  16:9    31.469 kHz  27.000 MHz
    VIC  2:   720x480    59.940 Hz   4:3    31.469 kHz  27.000 MHz
    VIC  7:  1440x480i   59.940 Hz  16:9    15.734 kHz  27.000 MHz
    VIC  6:  1440x480i   59.940 Hz   4:3    15.734 kHz  27.000 MHz
    VIC  1:   640x480    59.940 Hz   4:3    31.469 kHz  25.175 MHz
    DTD  3:  1280x720    60.000 Hz  16:9    45.000 kHz  74.250 MHz (698 mm x 392 mm)
    DTD  4:   720x480    59.940 Hz   3:2    31.469 kHz  27.000 MHz (698 mm x 392 mm)
    DTD  5:   720x480    59.940 Hz   3:2    31.469 kHz  27.000 MHz (523 mm x 392 mm)

Notice the last two modes look a bit suspect, especially DTD 5.

They are replaced with the new modes supported in 1.4

Code: Select all

    VIC  31:  1920x1080   50.000 Hz  16:9    56.250 kHz 148.500 MHz (native)
    VIC  20:  1920x1080i  50.000 Hz  16:9    28.125 kHz  74.250 MHz
    VIC  32:  1920x1080   24.000 Hz  16:9    27.000 kHz  74.250 MHz
    VIC  19:  1280x720    50.000 Hz  16:9    37.500 kHz  74.250 MHz
    VIC  60:  1280x720    24.000 Hz  16:9    18.000 kHz  59.400 MHz
    VIC  18:   720x576    50.000 Hz  16:9    31.250 kHz  27.000 MHz

In terms of audio, the following were dropped:

Code: Select all

    AC-3:
      Max channels: 2
      Supported sample rates (kHz): 48 44.1 32
      Maximum bit rate: 0 kb/s
    AC-3:
      Max channels: 6
      Supported sample rates (kHz): 48 44.1 32
      Maximum bit rate: 0 kb/s

Leaving us with only:

Code: Select all

    Linear PCM:
      Max channels: 2
      Supported sample rates (kHz): 48 44.1 32
      Supported sample sizes (bits): 16

That's as vanilla as audio can get.

I'd expect that any laptop that uses hardware based on technology released in 2018 or later should work with that. It's not like this is bleeding edge 8k HDR stuff.

[MrQuade]

That's as vanilla as audio can get.

I'd expect that any laptop that uses hardware based on technology released in 2018 or later should work with that. It's not like this is bleeding edge 8k HDR stuff.

We know the laptop supports some sort of compliant audio format since it was working with the previous T4 firmware. That's unless the laptop was previously sending non-compliant audio and the T4 was still processing it somehow.

Since the laptop doesn't seem to be recieving the EDID audio data, might point back at that 1.3 vs 1.4 difference. Is it just hitting a magic byte in between the video and audio blocks and interpreting that as the end of data?

[mkwilson]

Thanks guys. Just to reiterate...I've tried connecting on 2 laptops (though both are Dell XPS 15s; one from Sept 2016, one purchased February 2020; both fully up to date re drivers according to each product's support page) . Both laptops connect without problems and with audio to my AVR, TV set, and 20 inch monitor. The hex data generated when connecting on each of these laptops is identical (not surprisingly I guess). Under the old Beyonwiz firmware, IIRC the laptop would recognise the Samsung TV that is connected via the Beyonwiz's HDMI out as the connected device. That is certainly what would come up in the Display Settings for Intel HD Graphics controls. Not sure if that is helpful.

[MrQuade]

Is it an ultra hd tv by any chance, and is the T4,/receiver connected to the tv on a tv port that has "HDMI enhanced mode" activated for UHD purposes?

I have found that I had sound issues with some devices when connected like that. It was almost like some of the enhanced mode differences were "leaking" through to the HDMI in.

Saying that, It doesn't sound like this is the case to me though, as the laptop is definitely not seeing the edid info.

[mkwilson]

It is a UHD TV, but I don't believe it is connected through any HDMI-enhanced port.

[peteru]

Under the old Beyonwiz firmware, IIRC the laptop would recognise the Samsung TV that is connected via the Beyonwiz's HDMI out as the connected device. That is certainly what would come up in the Display Settings for Intel HD Graphics controls. Not sure if that is helpful.

That does not sound right. There is no code to somehow copy the upstream sink EDID and for the T4 to advertise it as it's own.

[mkwilson]

Might be because I have a splitter. However, running directly into the T4 makes no difference to the problems discussed above. I've also noted that with the latest firmware the HDMI-in recording quality (from my IQ4) is not great relative to what I got from the Nov 2016 firmware. I've set it at max bitrate, and set the frame rate to both 60 and 50bps (what the IQ4 outputs) but there's a lot of juddering that I didn't get previously. I sort of had it perfect (and hence why I hadn't updated firmware for years). I suspect I might be better off trying to find a copy of old firmware in my random USB drawer.

[peteru]

Might be because I have a splitter.

Well, that makes it a bit more complicated. A splitter can potentially replace, modify or corrupt EDID info. I would not be surprised if the splitter only handled EDID 1.3 and when it encounters 1.4 it drops the extensions. The behaviour may also be dependent on the order in which devices connect and power on.

running directly into the T4 makes no difference to the problems discussed above.

Have you tried retrieving the EDID with a direct connection? Do you get different results?

I've also noted that with the latest firmware the HDMI-in recording quality (from my IQ4) is not great relative to what I got from the Nov 2016 firmware. I've set it at max bitrate, and set the frame rate to both 60 and 50bps (what the IQ4 outputs) but there's a lot of juddering that I didn't get previously. I sort of had it perfect (and hence why I hadn't updated firmware for years). I suspect I might be better off trying to find a copy of old firmware in my random USB drawer.

I'll check the code, but as far as I can remember, there were no changes to anything that would affect recording quality. The importance of matching framerates has always been a factor. The Broadcom hardware and firmware will basically do a decent job converting everything to a 60p mode. However, that means you need to play back in 60p mode. If you record 50p/i content in 60p mode and then try to play back in 50p mode, you will get a mess.

[mkwilson]

Might be because I have a splitter.

Well, that makes it a bit more complicated. A splitter can potentially replace, modify or corrupt EDID info. I would not be surprised if the splitter only handled EDID 1.3 and when it encounters 1.4 it drops the extensions. The behaviour may also be dependent on the order in which devices connect and power on.

running directly into the T4 makes no difference to the problems discussed above.

Have you tried retrieving the EDID with a direct connection? Do you get different results?

I've also noted that with the latest firmware the HDMI-in recording quality (from my IQ4) is not great relative to what I got from the Nov 2016 firmware. I've set it at max bitrate, and set the frame rate to both 60 and 50bps (what the IQ4 outputs) but there's a lot of juddering that I didn't get previously. I sort of had it perfect (and hence why I hadn't updated firmware for years). I suspect I might be better off trying to find a copy of old firmware in my random USB drawer.

I'll check the code, but as far as I can remember, there were no changes to anything that would affect recording quality. The importance of matching framerates has always been a factor. The Broadcom hardware and firmware will basically do a decent job converting everything to a 60p mode. However, that means you need to play back in 60p mode. If you record 50p/i content in 60p mode and then try to play back in 50p mode, you will get a mess.

Sorry, I should have been clearer. The tests I've run over the past few days have involved the HDMI out from the laptop going straight into the HDMI in on the T4 (which then runs straight to TV). I will double check the EDID output, using this direct connection, but am 99.99% sure that's what I used to generate EDID data already posted. Historically, I have have used a switch (sorry, not splitter) to alternate between the 2 input sources to the T4 (IQ4 / laptop). The splitter is actually used between the IQ4 and the switch (one path runs to TV, one to switch).

Re record quality, I'll playback some files recorded under previous firmware to try to work out whether the issues are from recording or a change to playback settings upon installing new firmware. Appreciate your help.

[peteru]

I checked the code and there were changes to the encoding settings, but if anything they should have resulted in quality improvements. The target bitrate was increased from 500,000 to 524,288 and the default maximum bitrate went up from 1,000,000 to 1,048,576. Internally, the transcode engine used to convert at a base frame rate of 30 fps and that was lifted to 60 fps to ensure that there is no temporal quantisation. The conversion rate is not the same as the input or output rate, it's what the transcoder uses for the intermediate stage. Running at 60 fps achieves the best possible quality - this required some tricky memory allocation reshuffling, but I managed to get that working. Old drivers were producing interlaced output, new drivers are progressive. So effectively, the old conversion used 30i, whereas new conversion uses 60p.

The other thing that has changed is the encoding level. Old code would only use 3.1, new code uses 4.0, giving better quality at the same bitrate.

The other improvements were the ability to configure more encode parameters from user space. The old drivers would ignore most user settings and encode everything using hardcoded defaults. The new drivers respect whatever is configured by the software (enigma2) and also ensure that the settings can not exceed transcoder limitations. The new drivers allow the user to turn up the maximum bitrate up to 31,457,280.

So, yes, there are differences in the transcoder settings between old drivers and new drivers, but all of the changes should result in better outcomes. You may need to change the encoder settings in enigma2 to turn up the quality, if enigma2 sets these to something lower than the driver defaults. I currently don't have access to working T4 hardware to test any of this.

[mkwilson]

Thanks a lot, I'll experiment with some short recordings in the next few days, and will continue trying to sort out the laptop connection. I guess it doesn't help that I was using 5 year old firmware previously (but once I found a version that did what I needed consistently I stuck with it).

[mkwilson]

No progress on getting the HDMI audio recognised by the laptop(s). However, my HDMI recording quality issues seem to have resolved (or largely so) by reducing the bitrate in the HDMI recording settings. I had sort of assumed that my 'old' recordings would have been at the maximum 25mbps. However, I checked the file details for a HDMI-in recording made before upgrading firmware (and it plays beautifully) and was surprised that is was something just over 10Mbps. So I switched the max bitrate on the HDMI-in recording settings to 10Mbps and a few test recordings have been pretty good (my earlier recordings sport were unwatchable). So

[mkwilson]

Well, after 2 nights of replacing every suspect driver that I can identify on the XPS laptops, still no joy. However, my partner bought a new Inspiron laptop yesterday...plugged into Beyonwiz, both video and audio recognised immediately. In one of peteru's posts above there's a mention that the Beyonwiz firmware should work with any 'hardware released in 2018 or later'. Certainly my main XPS laptop is older than that (late 2016), maybe the relevant hardware in my new (Feb 2020) XPS has also retained some old relevant hardware? If I decided I to, how far would I have to roll back the Beyonwiz firmware to get to a version that uses EDID 1.3?

[peteru]

Driver version 20180307 or earlier.

[mkwilson]

Thanks peteru.