DTS DCH Driver for Realtek HDA [DTS:X APO4 + DTS Interactive]

[Ferather]

Again you would install the other 'DTS-X Ultra' pack (the one after this thread), following the guide (I added a quick guide below):




----

Quick guide

Install the 'DTS-X Ultra' pack, copy the Interactive.reg file (setup folder) and then paste-rename it (for example: HDMI-DTS).
Edit the .reg file with the GUID from FX Config, save then restore it. Use FX to restore the .reg if needed.
Restart the PC, select DTS as the format, disable the device then restore again, then enable.

The third step sets DTS as the default format.

----

5.1 Speaker Test (use Edge): https://www2.iis.fraunhofer.de/AAC/ChID-BLITS-EBU.mp4

Edit an Edge shortcut adding the following: --try-supported-channel-layouts --force-wave-audio
Now open Edge using the shortcut and try spatial to multichannel via YouTube.

Listen to the full audio, while moving the object around.

----

 

[fidoboy]

Again you would install the other 'DTS-X Ultra' pack (the one after this thread), following the guide (I added a quick guide below):

View attachment 376303
View attachment 376305

----

Quick guide

Install the 'DTS-X Ultra' pack, copy the Interactive.reg file (setup folder) and then paste-rename it (for example: HDMI-DTS).
Edit the .reg file with the GUID from FX Config, save then restore it. Use FX to restore the .reg if needed.
Restart the PC, select DTS as the format, disable the device then restore again, then enable.

The third step sets DTS as the default format.

----

5.1 Speaker Test (use Edge): https://www2.iis.fraunhofer.de/AAC/ChID-BLITS-EBU.mp4

Edit an Edge shortcut adding the following: --try-supported-channel-layouts --force-wave-audio
Now open Edge using the shortcut and try spatial to multichannel via YouTube.

Listen to the full audio, while moving the object around.

----

The first thing I do after each Windows install is to uninstall Edge, so I don't have it to test. Any case the problem is that I haven't the DTS format to select it, the same that I already have Dolby Digital.

The main issue for me is that the PC could do the PCM 5.1 to DTS 5.1 compression in real time, to send it through HDMI to the TV and then with the optical out to decoder.

I want to note here, for the record, if it helps other readers, that it's very important also that the image scaling should be done on GPU so the TV doesn't have to do it. This way there is no delay between image and audio.

If the scaling is to be done on the TV and it's not a native format then it will be a small delay. So, just in case of using a external decoder for the audio, it's important to select in Windows that the image scaling it's done by the GPU.

My objetive now is to get DTS compressed format, 5.1, to send it as two channels bitstream to the TV. And I've tried all methods and for the life of me, I can't get it working.

I managed to have Dolby Digital, with a little of work, using the APO Driver (it doesn't work with any other) but I can't achieve the same with DTS. Can anybody with some experience with this help or share his experience?

 

[fidoboy]

Again you would install the other 'DTS-X Ultra' pack (the one after this thread), following the guide (I added a quick guide below):

View attachment 376303
View attachment 376305

----

Quick guide

Install the 'DTS-X Ultra' pack, copy the Interactive.reg file (setup folder) and then paste-rename it (for example: HDMI-DTS).
Edit the .reg file with the GUID from FX Config, save then restore it. Use FX to restore the .reg if needed.
Restart the PC, select DTS as the format, disable the device then restore again, then enable.

The third step sets DTS as the default format.

----

5.1 Speaker Test (use Edge): https://www2.iis.fraunhofer.de/AAC/ChID-BLITS-EBU.mp4

Edit an Edge shortcut adding the following: --try-supported-channel-layouts --force-wave-audio
Now open Edge using the shortcut and try spatial to multichannel via YouTube.

Listen to the full audio, while moving the object around.

----

Wich I don't like from your driver is that there is no wizard to install it so it can't be easily removed between tests as it happens with AAF Optimus and with APO Driver. Any case, after I installed yours there is no DTS format available to select it on the drop down selector in Windows panel. Should I use FX configurator to achieve it? Also, in my case I'm only using the HDMI output (so there is no speakers connected to PC or headphones), so How is the device configuration that I should have into the device manager? Why is needed the stereo mixer device method described here? Is it only to keep the output always opened the same as when using SoundKeeper app?

It could be of great help if the installer wizard could have a step to select the intended sound output and desired format configuration. This way the installer could avoid the need of using FX Configurator and other hassles. It's not clear for me when it's needed the usage of UWP apps like DTS Unbound or DTS:X because I don't really understand what they do. Years ago, with only the soundcard panel all of this was configured very clear, with no hassle, and you can select the desired format for all the outputs (like DTS or Dolby) but now it seems that you need to get a mastering to get same results.

As many computers doesn't have optical outs, like in my case, I should do it with the HDMI wire, but this results into additional problems with formats, because the TV can't manage the compression task to convert the sound from PCM into a valid format for SPDIF optical (wich only have two channels and limited throughput). With a PC you can convert from DTS HD or Dolby TrueHD to standard and old DTS or Dolby formats but there is no drivers to ease the task.

Another issue that comes into my mind is: Why can't I get both formats available at the same time into the drop down selector as before? Now as I've noticed from the forums about the AAF Optimus or APO Driver you can only have one, but not both. So If you want to change from Dolby to DTS you should uninstall everything, remove the driver, clean all and remove ghosts to have it working. Why there is no way to have both formats available into the drop-down selector?

 

[Ferather]

If you follow the steps given the format will show, its not really possible to implement a device selection and programmer with command prompt.
Its sort of doable, using txt file output and then rename, but its a bit complicated, much easier to use FX, sorry about that.

====

Here is an updated DMAS image, I removed PCM and replaced it with PAM throughout, RGB optical in and out, RGB optical speaker in.
Base specification is 256 bit (256 colour values), 200 Ms/s (200 million samples a second), per channel (200 Mbits/s).



The PAM amp simply changes bit value (up, down), basic on logic.

----



----

256 (R) x 256 (G) x 256 (B) x 256 (UV) x 256 (IR) = 1,099,511,627,776 bit values.

 

[fidoboy]

If you follow the steps given the format will show, its not really possible to implement a device selection and programmer with command prompt.
Its sort of doable, using txt file output and then rename, but its a bit complicated, much easier to use FX, sorry about that.

====

Here is an updated DMAS image, I removed PCM and replaced it with PAM throughout, RGB optical in and out, RGB optical speaker in.
Base specification is 256 bit (256 colour values), 200 Ms/s (200 million samples a second), per channel (200 Mbits/s).

View attachment 376857

The PAM amp simply changes bit value (up, down), basic on logic.

----

View attachment 376858

----

256 (R) x 256 (G) x 256 (B) x 256 (UV) x 256 (IR) = 1,099,511,627,776 bit values.

Very nice

 

[Ferather]

Just so you know if you transcoded a traditional PCM file, 24 bit 48k to RGB Optical, it would be 48 Kbits/s, which means 10 channels would be 480 Kbits/s.

Also note, a DMAS main unit would need to add pseudo samples in order to get up to 200 million, essentially become a DAC.
This is best avoided by using as much 'original' information as possible, and not add pseudo samples.

----

Lossless (not changed), not compressed and no transmission encoding.

----

"The most accurate DAC and amp, is no DAC and amp."

 

[Ferather]

Note that the power stage is within the speaker unit, its called a transcoder because its not a converter, instead it transcodes light pulses into voltage output.
Ultimately its still PAM data, simply converted to voltage values from colour values, changing fast enough to work like alternating current.

There is no power stage amplifier, there is however a PAM amplifier, which changes bit values up or down per sample.

----

The DMAS main unit will always know bit 256 (256 bit max) is max volume, regardless of the power stage, which could be 12v, 14v, so on.

 

[fidoboy]

Note that the power stage is within the speaker unit, its called a transcoder because its not a converter, instead it transcodes light pulses into voltage output.
Ultimately its still PAM data, simply converted to voltage values from colour values, changing fast enough to work like alternating current.

There is no power stage amplifier, there is however a PAM amplifier, which changes bit values up or down per sample.

----

The DMAS main unit will always know bit 256 (256 bit max) is max volume, regardless of the power stage, which could be 12v, 14v, so on.

Thanks a lot for all the clarifications but I already know that when going from PCM to Dolby Digital or DTS Digital means a lossy compression. I can also understand that there is also a 48Khz limit (something that is not very relevanrt because the human hear can't make a difference). When mastering audio, into the studio, there is, of course, because it's needed a grade of detail beyond the human ears capabilities, but it's not needed for just listening in normal conditions. Anything beyond the 48Khz is useless.

So my purpose is not to have a lossless digital stream but only to encode all the 5.1 channels into two to send it as a raw stream to the TV and then, untouched, to the digital decoder with a optical wire. That's why I'm looking for a driver that can do that through HDMI connection. It works flawlessly with Dolby Digital, but it doesn't seems to work the same for DTS Digital or Dolby Digital Plus and I can't understand why, because is technically feasible.

 

[Ferather]

Oh sorry I was talking about another topic (an audio unit I am working on), unfortunately I not able to make multiple posts, so I have to join posts together. I would not worry about the lossy DTS encoder.
People have tested DTS Digital Surround and DTS-HD MA (latest), and can not tell the difference with current hardware, so I would say its lossy but more like unnoticeable filtering.

Interestingly you would be more likely to notice the difference between two DAC's and-or different speakers than the DTS encoder.


The way audio specifications work at this moment you should not be able to hear above 20khz (40khz samples), but thats because they use it differently.
They also do the same with bits and volume, when instead BOTH specifications can increase fidelity with the same sine wave.

With traditional PCM, increasing the sample rate to 192k generally just increases the recordable sound frequencies, which will be pseudo (added) if you started with 48khz.


Bits and sample rate are best used to accurately reproduce a sine wave, opposed to volume and frequency specifications.
32 bit float is a good example, it is used to increase fidelity certainly not go up to 1528 dB output!

====

Here is an example of a programmable digital filter, the sample rate stays the same (48k), but the frequencies are reduced, frequencies we can not hear.
You can argue that indeed that filter is lossless, as long as it does not change the audible frequencies, still original data.

 

[fidoboy]

Oh sorry I was talking about another topic (an audio unit I am working on), unfortunately I not able to make multiple posts, so I have to join posts together. I would not worry about the lossy DTS encoder.
People have tested DTS Digital Surround and DTS-HD MA (latest), and can not tell the difference with current hardware, so I would say its lossy but more like unnoticeable filtering.

Interestingly you would be more likely to notice the difference between two DAC's and-or different speakers than the DTS encoder.


The way audio specifications work at this moment you should not be able to hear above 20khz (40khz samples), but thats because they use it differently.
They also do the same with bits and volume, when instead BOTH specifications can increase fidelity with the same sine wave.

With traditional PCM, increasing the sample rate to 192k generally just increases the recordable sound frequencies, which will be pseudo (added) if you started with 48khz.


Bits and sample rate are best used to accurately reproduce a sine wave, opposed to volume and frequency specifications.
32 bit float is a good example, it is used to increase fidelity certainly not go up to 1528 dB output!

====

Here is an example of a programmable digital filter, the sample rate stays the same (48k), but the frequencies are reduced, frequencies we can not hear.
You can argue that indeed that filter is lossless, as long as it does not change the audible frequencies, still original data.

View attachment 378245

Oh sorry I was talking about another topic (an audio unit I am working on), unfortunately I not able to make multiple posts, so I have to join posts together. I would not worry about the lossy DTS encoder.
People have tested DTS Digital Surround and DTS-HD MA (latest), and can not tell the difference with current hardware, so I would say its lossy but more like unnoticeable filtering.

Interestingly you would be more likely to notice the difference between two DAC's and-or different speakers than the DTS encoder.


The way audio specifications work at this moment you should not be able to hear above 20khz (40khz samples), but thats because they use it differently.
They also do the same with bits and volume, when instead BOTH specifications can increase fidelity with the same sine wave.

With traditional PCM, increasing the sample rate to 192k generally just increases the recordable sound frequencies, which will be pseudo (added) if you started with 48khz.


Bits and sample rate are best used to accurately reproduce a sine wave, opposed to volume and frequency specifications.
32 bit float is a good example, it is used to increase fidelity certainly not go up to 1528 dB output!

====

Here is an example of a programmable digital filter, the sample rate stays the same (48k), but the frequencies are reduced, frequencies we can not hear.
You can argue that indeed that filter is lossless, as long as it does not change the audible frequencies, still original data.

View attachment 378245

But do you know about any driver wich works with HDMI output to use the DTS Digital format and/or Dolby Digital Plus? I've managed to add something called "DTS Connect" to available formats (in addition to the typical list with bitrates and frequencies) in my HDMI output using FX Configurator. But then when I select this as output format for the device it doesn't works and it fails when doing the test in Windows. A message with "The device can't play the selected audio" (or something similar) appears and it doesn't work. I've tried the @AAF Optimus driver and also the APO Driver from Alanfox but without any luck. My sound chip is a Realtek ALC233. I suppose that this however doesn't change anything

 

[Ferather]

Sounds like something did not setup correctly, I can use RustDesk to setup the pack I made for non-Realtek on your HDMI if you want, but will be tomorrow.
The version I put together includes DTS's latest technology DTS:X processing, so technically its DTS:X core, more modern.

Message me a good time for after 3pm UK (GMT).

====

Another DMAS post, adding to yesterdays post. Even though the sample rate is 200 million (200Mhz), the frequency data is 20hz-20Khz (human range).
Aliasing should not be an issue, most certainly not with 256 bit 200Mhz, the only thing to do is filter beyond 20Khz.

The end resulting 256 bit 200Mhz is spent on accuracy for 0-20Khz, 48k x 4166 = 199,968,000.
Analogue voltage changes more frequently than 48k, hence a DAC.

0.001v - 0.002v - 0.003v is more accurate and more rounded than 1v - 2v - 3v, steps, with the same total 12v (example).


Reminder, for RGB optical, bit value defines the voltage, the sample rate defines voltage at a given time.
This can also be written as position and positions per second, or audio resolution.

 

[Ferather]

Here is some extra information from the Class-D design. In the case of the DMAS, we are starting with 256 bits 200Mhz (200 Ms/s, 200Mb/s) from the input (source).
There is no PWM, instead its PAM, there is a filter but generally that would be at the microphone end, and not strictly the main unit.

We can also define much more bit values using RGB optical, starting with ~16.7 million with RGB alone.



Also note optical is immune to EMI, and transmits longer distances.
And to reiterate, the transcoder is in the speaker unit.

----

I would vouch for DTS processing (CPU), which will be mostly user and-or config changes, such as crossover, speaker correction, so on.

What is True Sound? The concept explained | Stuff

 

[Ferather]

DTS Digital Direct would be an interesting name for their product, SoC. Additional objects would add a channel to transmission (main input).
For example a 8 channel DMAS would receive its normal 8, and anything after that is regarded as extra channels or objects.

It is also possible to always add 10+ command values to transmission, for example 266 total colours, 256 audio values + command.


The hardware ultimately uses one main clock, the sample rate (which is per second), and bits per second.

 

[Ferather]

In terms of bandwidth and channel limits there are none, instead you work with a standard, like SPDIF, for example: consumer at 32 channels, professional unlimited.
The main input should be capable of 40 channels of data, this allows +1 sample bit (header) per sample, which marks the start-stop of samples.



32 channels would be equal to 33 channels in transmission as there is +1bit per sample, 33 x 200 Mbits/s = 6.6 Gbits/s.
40 x 200 = 8 Gbits/s. Ideally changes to the standards should have a version, like 1, 1.1, 2.1, so on.

SPDIF has been upgraded-changed a fair amount of times, but completely lacks a version number to work with.

The speaker outputs are mono, and will not need much more than the 200 Mbits/s sample rate.
Note the illustration is 8 channels, 2 samples, 9 channels of bandwidth in total.

----

Also note you can increase the bit values from 256 to 512+ and not change the required bit rate, same sample rate.
In general, unless for example volume up is held down, command-program bits will not be constant.

Most commands would probably not need the full 200 Mbits/s rate, instead in can be reduced to every X sample.
In addition, the full 200 Mbits/s could be used to produce additional commands (e.g 4 x 50 Mbits/s).

A basic example would be A:50 million times followed by B:50 C:50 D:50 = outcome at 200 Mbits/s (1 channel).

----

Stereo (2 channel) PCM, 256 bit @ 200 Ms/s = 102.4 Gbits/s, 51.2 Gbits/s mono. Now lets use 512 total values.
Excluding program bits, we can do 256 for audio, 256 for internet, so 1 channel can do 51.2 Gbits/s.

Note the actual data rate of the internet channel would be 200 Mbits/s in RGB.

To replace HDMI in the full, add a RGB return (equal to input, 2 lanes) to the main input cable.

 

[Ferather]

In terms of a multicore SoC, its also possible to use one optic core to address each SoC core with a set bit range (256 bit, 512 bit, so on). Since there are millions to billions of available colour values.
For example, core 1 could be 256 bit, and a range of colours, the next core and 256 bit another colour range, following the previous (now a total of 512 colours), and so on.

You can also do one core per channel, and distribute the incoming audio using bit position to channel number, as given in the illustration above.
The reserved bit specifies the sample, the clock, you can add channels but will always increase the required bitrate.

----

Its also possible to use more than one channel for say video in order to get more bitrate than one channel. Bit value (colour) range defines its audio, video or data.
Potentially audio, video and data bits can be in a random position in the sample-clock, but still be linear in channels (1, 2, 3, so on).

----

If clock timings are an issue, the video and data can use their own optical core, making the main cable multi-fiber in-out.
Two channels should be enough to equal the DAC in a GPU, 2 x 200 = 400 in linear form.

Note that RGB optical is digital, but does 'not' use binary bits (0 or 1) to produce values (in a long length).

 

[Ferather]

Off topic post, but I was watching a video the other day on YouTube, and tried the test for myself. Occasionally I get a fast response, but mostly in the 140-160ms area.
I would guess the time it turned green and the refresh rate of my monitor and other factors managed to line up to a more instant response.

Cant really explain it, Human Benchmark - Reaction Time Test