Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[fitbrit]

In the PCs I make, we use galvanically isolated AES/EBU connectors coming from the S/PDIF headers of certain motherboards. These sound fantastic and my company has many awards as a result. The problem is that the Realtek drivers meant for the audio chipsets result in completely garbled/distorted output in MC. They work in other apps that use exclusive mode, but seem to die with WASAPI in MC. Using the default Windows drivers rectifies the problem. However, it is hard to ensure that the Realtek drivers are never installed.
Is there any way this can be looked into?
I would be willing to loan a system my company has built for testing purposes.

A very important addendum: The drivers are absolutely ok when using the PC's standard outputs in the i/o panel - they just don't work with the AES or BNC from the motherboard's SPDIF header - in MC only.

[Spike1000]

I do hope someone picks this up. . . It sounds like a really 'interesting' problem. Good luck in getting it fixed.

Spike

[leezer3]

How technically advanced are your company (And to a certain degree, how much are they willing to spend)?

My immediate thought would be as follows:
Change the PCI device / vendor ID for the Realtek chip in the BIOS to something unique.
Create a modified version of the Windows drivers that only have this device / vendor ID combo.
Sign the drivers yourselves.

If you've got access to the hardware manufacturing tools or a decent OEM (A manufacturer, not just someone punting white label boxes), changing the device IDs would be trivial.
It would also likely be possible to mod the device IDs in an existing BIOS, but that's harder, and with modern UEFI stuff might well require messing with signature verification and stuff of that nature.

[JimH]

Nice post, leezer3.

fitbit,
Did you try a search (from Google)?

I just did and found this thread:

https://yabb.jriver.com/interact/index.php?topic=119106.0

[fitbrit]

Nice post, leezer3.

fitbit,
Did you try a search (from Google)?

I just did and found this thread:

https://yabb.jriver.com/interact/index.php?topic=119106.0

Thanks to all who responded so far.

Yes, Jim, I have been Google searching for this issue for about 3 and a half years. I too was excited when I saw the topic you linked to when it was posted. If you read the thread, you will see the solution was that there is no solution so far, and just not to use the Realtek drivers.

What we have been doing so far:

1. Ship each system with the sound device hardware ID set to not update drivers in GPEDIT. We make sure Realtek drivers are not installed, and the Windows native ones are.

Problem: With each semi-yearly big Windows update, MS ignores this GPEDIT entry and removes the Windows native drivers. THEN it honours the GPEDIT entry and does not reinstall any drivers for the sound device. So twice a year I could have 50-100 customers suddenly getting no sound from their systems.

2. To let Realtek install its drivers and then "Roll back drivers" in Device Manager.

Problem: This is kind of a "fingers-crossed" scenario where I just hope that a Windows update will not install the drivers again, or when a new version of Realtek drivers are issued. Again, I could potentially have lots of customers who get distorted gibberish from their systems.

Roon and Tidal both work fine with the Realtek drivers in Exclusive mode. However our company was built around JRiver's MC and should continue to use MC for the foreseeable future, and I will always use it personally. Therefore it would be nice to not have to spend tens to hundreds of man-hours each year fixing driver issues for customers. If we could set and forget Realtek drivers, and know they would work with all apps, that would be ideal.

Again, we only get this issue using the internal SPDIF header of the Gigabyte motherboards we use. And only in MC and only with the Realtek drivers. All of these conditions need to be met in order to cause the problem, which is why I named the topic a niche issue even though it could potentially affect 100-200 people that I know of.

[fitbrit]

How technically advanced are your company (And to a certain degree, how much are they willing to spend)?

My immediate thought would be as follows:
Change the PCI device / vendor ID for the Realtek chip in the BIOS to something unique.
Create a modified version of the Windows drivers that only have this device / vendor ID combo.
Sign the drivers yourselves.

This sounds intriguing. I am not sure how to do this, and we do not have the means to do this - we use Gigabyte motherboards.

As an aside, I am having trouble finding out anything about what it would take to get a custom motherboard design made, and we are a very small company. With some growth, we would sell just 100 units a year! But this might be a needed action because even Gigabyte is beginning to eliminate the SPDIF header from the motherboard real estate in favour of LED controllers for flashy lights.

If you've got access to the hardware manufacturing tools or a decent OEM (A manufacturer, not just someone punting white label boxes), changing the device IDs would be trivial.
It would also likely be possible to mod the device IDs in an existing BIOS, but that's harder, and with modern UEFI stuff might well require messing with signature verification and stuff of that nature.

[leezer3]

The driver alterations are a piece of cake (other than the signing, & that isn't too difficult), it's the hardware one that would be more difficult.

Basically, the PCI device / vendor ID are a pair of strings which inform the BIOS and operating systems what a piece of hardware is, and thus what drivers are to be used etc.
I don't know where this is stored for the onboard sound chip (it may be in the BIOS blob, or be on a separate EEPROM).
Gigabyte are unlikely to help you, they're not niche enough.
You might get some results posting on one of the major BIOS mod boards.
Dunno, and to a certain extent might be whether / how much you'd be willing to pay for someone to dig into this for you.
(Not me! I only understand the theory and have done a little playing with BIOS whitelisting)

[Hendrik]

I can try to run my SPDIF through optical into my receiver soon and see if I can reproduce any issues. But I would think we might have heared about this before if it was a common problem.

[fitbrit]

The driver alterations are a piece of cake (other than the signing, & that isn't too difficult), it's the hardware one that would be more difficult.

Basically, the PCI device / vendor ID are a pair of strings which inform the BIOS and operating systems what a piece of hardware is, and thus what drivers are to be used etc.
I don't know where this is stored for the onboard sound chip (it may be in the BIOS blob, or be on a separate EEPROM).
Gigabyte are unlikely to help you, they're not niche enough.
You might get some results posting on one of the msjor BIOS mod boards.
Dunno, and to a certain extent might be whether / how much you'd be willing to pay for someone to dig into this for you.
(Not me! I only understand the theory and have done a little playing with BIOS whitelisting)

Thanks very much - you're helping me understand the problem better.

[fitbrit]

I can try to run my SPDIF through optical into my receiver soon and see if I can reproduce any issues. But I would think we might have heared about this before if it was a common problem.

Hi Hendrik. The issue is when one uses the SPDIF header on the motherboard, which very few people do. See the attached photo. The output from the motherboards internal SPDIF header goes through our daughterboard for galvanic isolation and outputs via the BNC and AES/SPDIF connectors shown on the attached picture.
I don't think you will reproduce the issue via the optical output.

[Hendrik]

Well, then you are screwed!
I would think those two things are pretty similarly handled by a driver though. If you can test with the built-in SPDIF port if it also happens that may help, otherwise, not sure what we can do without reproducing it.

[fitbrit]

Well, then you are screwed!
I would think those two things are pretty similarly handled by a driver though. If you can test with the built-in SPDIF port if it also happens that may help, otherwise, not sure what we can do without reproducing it.

I did offer to send you one of my systems. 
I will try an optical output to see if we can reproduce it in the meantime.

[DocLotus]

Don't know if this is related or not but my Huawei MateBook X Pro laptop, which uses Realtek Hi Def Audio, has always had a problem with sound when any device is connected to any USB port including... Docks, Dongles, Hubs & individual USB audio devices. It would play just fine for a few minutes, a few hours or a day or two. Then it would do one of several things... provide the dreaded Blue Screen of Death, start a load buzzing sound, loose the audio completely, or simply freeze up. The only thing that would clear all of them was a reboot of the computer.

The reason I mention this is a few hours ago I ran the Huawei PC Manager (which I run weekly) & it updated the Realtek drivers. The information provided said the update was to improve stability. So far so good, no problems yet (but it is early). Am thinking there may be Realtek updates for all Windows computers.

Might be worth a look-see for updated drivers at this time. Just a thought.

[JimH]

Or older drivers.

[fitbrit]

Thanks, but I am not sure this is the issue in our case. We've used the latest drivers since 2014, so you'd think that one of all the drivers released in that time would've fixed the issue.

Don't know if this is related or not but my Huawei MateBook X Pro laptop, which uses Realtek Hi Def Audio, has always had a problem with sound when any device is connected to any USB port including... Docks, Dongles, Hubs & individual USB audio devices. It would play just fine for a few minutes, a few hours or a day or two. Then it would do one of several things... provide the dreaded Blue Screen of Death, start a load buzzing sound, loose the audio completely, or simply freeze up. The only thing that would clear all of them was a reboot of the computer.

The reason I mention this is a few hours ago I ran the Huawei PC Manager (which I run weekly) & it updated the Realtek drivers. The information provided said the update was to improve stability. So far so good, no problems yet (but it is early). Am thinking there may be Realtek updates for all Windows computers.

Might be worth a look-see for updated drivers at this time. Just a thought.