-
Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
DTS DCH Driver for Realtek HDA [DTS:X APO4 + DTS Interactive]
- Thread starter Ferather
- Start date
I've got some of these do you think it's useful ?Then it must of had the right CFG for your device, or some other data you need. You cant really add-mix CFG keys either
Latest realtek from MSI server :
https://pan.hotpe.top/f/ro0HG/9828_UAD_WHQL_NB.zip
Latest intel SST :
https://pan.hotpe.top/f/mbMFg/Intel_SST_RPL_TWL_v10.29.00.12108.zip
https://pan.hotpe.top/f/eXLhp/Intel_SST_RPL_TWL_v10.29.00.12108_OemExtensionInfs.zip
https://pan.hotpe.top/f/R8rCB/Intel_SST_TGL_ADL_v10.29.00.12108.zip
https://pan.hotpe.top/f/xWMfa/Intel_SST_TGL_ADL_v10.29.00.12108_OemExtensionInfs.zip
https://pan.hotpe.top/f/jAmT2/Intel_SST_MTL_ARL_v20.40.12061.0.zip
https://pan.hotpe.top/f/lyAfg/Intel_SST_MTL_ARL_v20.40.12061.0_OemExtensionInfs.zip
Some rtkhdaud files I backed up frome the MSI OEM system
Alpha 17 C7V (AMD Ryzen™ 7000 series mobile processor / GeForce RTX™ 40 Series):
https://pan.hotpe.top/f/296tl/17KK.zip
Vector 17 HX A13V (Intel® Core™ processor (13th gen) / GeForce RTX™ 40 Series):
https://pan.hotpe.top/f/JpAf4/17S1.zip
Raider A18 HX A7V (AMD Ryzen™ 7045 series mobile processor / GeForce RTX™ 40 Series):
https://pan.hotpe.top/f/zW1sg/182K.zip
Raider A18 HX A9W (AMD Ryzen™ 9000 Series / GeForce RTX™ 50 Series):
https://pan.hotpe.top/f/0EAHZ/182L.zip
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
What features do you use from AAF? If its just DTS, then install a working driver and add DTS to it, else use the older AAF?
I don't have an Intel SST, so there is no way for me to help you at my end, seems like I should avoid it.
I don't have an Intel SST, so there is no way for me to help you at my end, seems like I should avoid it.
I'm using ICE , it really sounds nice, all AAF can't install with new SST driver, also after installing AAF 9823 I lost my sound
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
Ok I cant help you with ICE sadly. I suggest you remove all AAF, including any AAF Intel bus drivers, restart the computer.
Make sure everything is working with the basic Microsoft drivers, then install the older AAF.
If you leave the current version bus driver, the older version will not install (its older then current).
----
Just so you know its now Alans fault you need a specific CFG, its also not your fault, but you did just learn something.
Make sure everything is working with the basic Microsoft drivers, then install the older AAF.
If you leave the current version bus driver, the older version will not install (its older then current).
----
Just so you know its now Alans fault you need a specific CFG, its also not your fault, but you did just learn something.
Last edited:
Hummm, not his fault, just intel and MSI , MSI audio drivers really sounds bad, even generic realtek is better than that
This one is using DTS, do you think it's possible to install on other models ?
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
I've said it before about MSI, over the last 3 years or so. Cant say much about Intel SST, as this driver works as it should, but not if you need a custom CFG.
Here you can see the HDA controller used for the S1220A (CS3), which is an AMD Starship/Matisse HD Audio Controller. No driver.
The GPU HDA bus uses an AMD driver.
----
I doubt your device ID will match.
Here you can see the HDA controller used for the S1220A (CS3), which is an AMD Starship/Matisse HD Audio Controller. No driver.
The GPU HDA bus uses an AMD driver.
----
I doubt your device ID will match.
The most stupid thing about SST on msi : I can't disable it or I'll lost the microphone
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
Maybe buy something like one of these: https://www.cubilux.com/ (https://www.cubilux.com/products/usb-headphone-adapter), or better.
Then disable on-board via bios. Permanent issue workaround. Then just add DTS via the guide here.
Then disable on-board via bios. Permanent issue workaround. Then just add DTS via the guide here.
Last edited:
My customers won't accept that (I'm selling MSI)
I want to, but I really don't know where to start
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
You would replace all of the CFG with a working one, then re-sign.
If that works, you will need to trail and error to find exact keys.
My pack can be re-signed using the same certificate, if you look.
----
Feel free to use my pack as a reference and base, to create your own.
If you intend to publish your own pack, create your own certificate.
----
As a side note Crystal Sound 4 could be PAM-X (you never know!).
====
Attachments
Last edited:
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
Video reminder (edited):
GPU: 3840 x 2160 = 8,294,400 (8.3 million) pixels, 1 bit a pixel would be 8.3 Mbits (a frame), x refresh (frame) rate, lets say 1000Hz, 8.3 Gbits/s (1ms).
No change to pulses per bit, since pixels display 16.77+ million colours, its already direct data (bit value).
PAM-X Superposition GPU, could also be called pixel pulse technology.
No encoding, no compression, no binary at all.
====
Example PAM-X SPU main unit specification:
Bit values: 256
Bit rate: 200 Mbits/s
Channels output: 10
Power supply: 600w
Example PAM-X speaker unit specification:
Bit values: 256
Bit rate: 200 Mbits/s
Response: X Hz - Y kHz
Power draw: 60w
====
PAM-X would be using 200MHz +.
Optical not power, no EMI-RFI.
PAM 16,777,216 @ 200MHz.
PAM-X Superposition (RGB).
----
Linear superposition PAM-X.
GPU: 3840 x 2160 = 8,294,400 (8.3 million) pixels, 1 bit a pixel would be 8.3 Mbits (a frame), x refresh (frame) rate, lets say 1000Hz, 8.3 Gbits/s (1ms).
No change to pulses per bit, since pixels display 16.77+ million colours, its already direct data (bit value).
PAM-X Superposition GPU, could also be called pixel pulse technology.
No encoding, no compression, no binary at all.
====
Example PAM-X SPU main unit specification:
Bit values: 256
Bit rate: 200 Mbits/s
Channels output: 10
Power supply: 600w
Example PAM-X speaker unit specification:
Bit values: 256
Bit rate: 200 Mbits/s
Response: X Hz - Y kHz
Power draw: 60w
====
PAM-X would be using 200MHz +.
Optical not power, no EMI-RFI.
PAM 16,777,216 @ 200MHz.
PAM-X Superposition (RGB).
----
Linear superposition PAM-X.
Last edited:
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
If you need to reset DTS and the config cache, delete all the end points {DEVIDE-ID} from DTS\APO (see the image below).
Then shutdown the computer, and disconnect the power (PSU switch, wall switch) for 10 minutes.
Its a bit hard to explain but its like RAM is powered with config data, and its forgotten if fully powered off.
----
Leave the APO key and data, just the endpoints (in the red box).
====
Same 256 bit values 200MHz, direct.
Then shutdown the computer, and disconnect the power (PSU switch, wall switch) for 10 minutes.
Its a bit hard to explain but its like RAM is powered with config data, and its forgotten if fully powered off.
----
Leave the APO key and data, just the endpoints (in the red box).
====
Same 256 bit values 200MHz, direct.
Last edited:
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
T = Transcoder, from power value to colour value, then visa versa (via supply).
It does not matter what power is used, only the bit values and rate.
Reminder, PAM-X Superposition is both a signal and a format.
====
Reminder Optical PAM-X is smaller than binary, even compressed binary.
The last step is variable value bit bank storage, drives.
1 channel 1 sample at 24 bit depth = 24 bits, for PAM-X see below:
8 channels, 2 samples, 18 bits. Could be 256 values, 1024 values, so on.
----
40 channels @ 48 kHz (48 Kbits/s) = 1.92 Mbits/s (any value range).
1.92 x 14400 seconds = 27,648 Megabits, 3.456 Gigabyte file data (4H).
1.92 x 240 seconds = 460.8 Megabits, 57.6 Megabyte file data (4M).
----
Basic 8 bit (value) CPU-DSP example, up to 16,777,216 values.
You could probably power all three LEDs at 1v or around.
----
X positions at Y rate + emitter travel = analogue (to your ear).
Last edited:
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
Addition to above:
2 channels @ 48 kHz (48 Kbits/s) = 96 Kbits/s (any value range).
0.096 x 240 seconds = 23.04 Megabits, 2.88 Megabytes.
Lossless, no compression, no encoding.
----
1 bit per sample with X value (position). 48k samples, 48 Kbits/s.
====
Notice the errors in the fist image, and generated positions.
Image 2, which would reconstruct the most accurate?
Also, which one has the most potential to remove DAC's?
----
High enough resolution, it only needs transcoding (UHD).
----
Speaker SPL, and hardware frequency support varies.
Format dB and frequency are not active specs.
You can do 20 Hz - 20 kHz at various resolutions.
====
Samples should be original, however there is storage limits.
2 channels @ 48 kHz (48 Kbits/s) = 96 Kbits/s (any value range).
0.096 x 240 seconds = 23.04 Megabits, 2.88 Megabytes.
Lossless, no compression, no encoding.
----
1 bit per sample with X value (position). 48k samples, 48 Kbits/s.
====
Notice the errors in the fist image, and generated positions.
Image 2, which would reconstruct the most accurate?
Also, which one has the most potential to remove DAC's?
----
High enough resolution, it only needs transcoding (UHD).
----
Speaker SPL, and hardware frequency support varies.
Format dB and frequency are not active specs.
You can do 20 Hz - 20 kHz at various resolutions.
====
Samples should be original, however there is storage limits.
Last edited:
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
High data rate advantages with filtering (as well as more resolution):
Pulse Code Modulation and Demodulation
----
Here is what PAM-X would look like, except coloured dots (closer).
Colour plotted (dot) wave.
----
I've also made the mistake of 256 levels from DAC's (2^8).
Instead 1 bit RGB is equal to 24 bit PCM, in values.
2 bit PAM-X would equal 48 bit PCM, and so on.
32 bit, full: 4,294,967,296 / 16,777,216 = 256 bit RGB.
----
UV and IR are possible ways to increase the base values.
RGB + IR or UV (if 256 also), 256^4 = 4,294,967,296.
RGB + IR + UV, 256^5 = 1,099,511,627,776.
The first one would be equal to 32 bit full, per bit.
----
Another option is to increase the RGB pulses per data bit.
2 pulses: 16,777,216^2 = 281,474,976,710,656.
Note the data bit rate does not change.
====
Irrelevant side note, open 'Task Manager' then move your mouse.
If CPU usage goes up, buy a mouse with an onboard CPU.
Pulse Code Modulation and Demodulation
----
Here is what PAM-X would look like, except coloured dots (closer).
Colour plotted (dot) wave.
----
I've also made the mistake of 256 levels from DAC's (2^8).
Instead 1 bit RGB is equal to 24 bit PCM, in values.
2 bit PAM-X would equal 48 bit PCM, and so on.
32 bit, full: 4,294,967,296 / 16,777,216 = 256 bit RGB.
----
UV and IR are possible ways to increase the base values.
RGB + IR or UV (if 256 also), 256^4 = 4,294,967,296.
RGB + IR + UV, 256^5 = 1,099,511,627,776.
The first one would be equal to 32 bit full, per bit.
----
Another option is to increase the RGB pulses per data bit.
2 pulses: 16,777,216^2 = 281,474,976,710,656.
Note the data bit rate does not change.
====
Irrelevant side note, open 'Task Manager' then move your mouse.
If CPU usage goes up, buy a mouse with an onboard CPU.
Last edited:
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
2 bits RGB additive: 16,777,216 + 16,777,216.
2 bits RGB multiple: 16,777,216 x 16,777,216.
24 bits multiple would be: 16,777,216^24.
24 bits multiple binary: 2^24 (16,777,216).
24 bit 48k = 1 bit x 48000 = 48 Kbits/s.
2 channels, 2 x 48 Kbits/s (96 Kbits/s).
256 x 256 x 256, is, 256^3 (RGB).
====
An interesting formula:
16 bit, 96 dB | 24 bit, 144 dB | 32 bit, 192 dB | 32 bit float, 1528 dB.
96 /6 = 16 | 144 /6 = 24 | 192 /6 = 32 | 1528 /6 = 254.67.
Closest full number for 32 bit float is 255, close to 256 in bits.
Should be, 2^256, if you did 256 binary.
====
RGB + IR or UV (if 256 also), 256^4 = 4,294,967,296 | 32 bit full.
32 bits multiple would be: 4,294,967,296^32.
32 bits multiple binary: 2^32 (4,294,967,296).
32 bit 48k = 1 bit x 48000 = 48 Kbits/s.
2 channels, 2 x 48 Kbits/s (96 Kbits/s).
----
RGB + IR + UV, 256^5 = 1,099,511,627,776 | 40 bit full.
40 bits multiple would be: 1,099,511,627,776^40.
40 bits multiple binary: 2^40 (1,099,511,627,776).
40 bit 48k = 1 bit x 48000 = 48 Kbits/s.
2 channels, 2 x 48 Kbits/s (96 Kbits/s).
2 bits RGB multiple: 16,777,216 x 16,777,216.
24 bits multiple would be: 16,777,216^24.
24 bits multiple binary: 2^24 (16,777,216).
24 bit 48k = 1 bit x 48000 = 48 Kbits/s.
2 channels, 2 x 48 Kbits/s (96 Kbits/s).
256 x 256 x 256, is, 256^3 (RGB).
====
An interesting formula:
16 bit, 96 dB | 24 bit, 144 dB | 32 bit, 192 dB | 32 bit float, 1528 dB.
96 /6 = 16 | 144 /6 = 24 | 192 /6 = 32 | 1528 /6 = 254.67.
Closest full number for 32 bit float is 255, close to 256 in bits.
Should be, 2^256, if you did 256 binary.
====
RGB + IR or UV (if 256 also), 256^4 = 4,294,967,296 | 32 bit full.
32 bits multiple would be: 4,294,967,296^32.
32 bits multiple binary: 2^32 (4,294,967,296).
32 bit 48k = 1 bit x 48000 = 48 Kbits/s.
2 channels, 2 x 48 Kbits/s (96 Kbits/s).
----
RGB + IR + UV, 256^5 = 1,099,511,627,776 | 40 bit full.
40 bits multiple would be: 1,099,511,627,776^40.
40 bits multiple binary: 2^40 (1,099,511,627,776).
40 bit 48k = 1 bit x 48000 = 48 Kbits/s.
2 channels, 2 x 48 Kbits/s (96 Kbits/s).
Last edited:
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
Note you can only have one volt-current specification at a measured point in time.
A physical emitter can only be in 1 position at a measured time.
PAM-X, 1 bit = 1 position at a measured point in time.
----
200 Mbits/s = 200 million positions per second.
----
Reminder, you can do multiple pulses at a faster rate but keep the bit rate-clock.
For example 2 pulses at twice the rate, represents 1 bit in time.
16,777,216^2 with the same bit rate, and bit clock.
[2:1] - 2 pulses 1 bit = 400 million pulses, 200 million positions | per second.
[1:1] - 1 pulse 1 bit = 200 million pulses, 200 million positions | per second.
16,777,216^2 = 48 bit PCM.
====
[1] - 16,777,216^12 = 4.9732323640978664215538224814682 e+86
[2] - 2^256 = 1.1579208923731619542357098500869 e+77
1 is bigger than 2.
====
Both a signal and a format overall, for storage we need variable value bit banks.
Ultimately 1 bit positions at the rate required to 'analogue' sound.
Note that 48k is 48,000 positions per second.
====
spectrum.ieee.org
Binaries days are numbered (pun intended). Optical Superposition.
====
Example PAM-X speaker specs (updated with PCM equivalent):
Bits: 1 (24, 48 - RGB)
Bit rate: 200 Mbits/s
Pulse rate: 200M - 400M
Response: X Hz - Y kHz
Transcoder: 60w, Max
SPL: 100 dB @ 60w
Bits: 1 (32, 64 - RGBV)
Bit rate: 200 Mbits/s
Pulse rate: 200M - 400M
Response: X Hz - Y kHz
Transcoder: 60w, Max
SPL: 100 dB @ 60w
A physical emitter can only be in 1 position at a measured time.
PAM-X, 1 bit = 1 position at a measured point in time.
----
200 Mbits/s = 200 million positions per second.
----
Reminder, you can do multiple pulses at a faster rate but keep the bit rate-clock.
For example 2 pulses at twice the rate, represents 1 bit in time.
16,777,216^2 with the same bit rate, and bit clock.
[2:1] - 2 pulses 1 bit = 400 million pulses, 200 million positions | per second.
[1:1] - 1 pulse 1 bit = 200 million pulses, 200 million positions | per second.
16,777,216^2 = 48 bit PCM.
====
[1] - 16,777,216^12 = 4.9732323640978664215538224814682 e+86
[2] - 2^256 = 1.1579208923731619542357098500869 e+77
1 is bigger than 2.
====
Both a signal and a format overall, for storage we need variable value bit banks.
Ultimately 1 bit positions at the rate required to 'analogue' sound.
Note that 48k is 48,000 positions per second.
====
Could Nvidia's Revolutionary Optical Switch Transform AI Data Centers Forever?
Nvidia's new optical network switch, announced at GTC, promises to revolutionize AI data centers by drastically cutting power consumption and boosting efficiency. With co-packaged optics (CPO) technology, Nvidia aims to reduce the number of lasers and improve data traffic power efficiency. But...
spectrum.ieee.org
Binaries days are numbered (pun intended). Optical Superposition.
====
Example PAM-X speaker specs (updated with PCM equivalent):
Bits: 1 (24, 48 - RGB)
Bit rate: 200 Mbits/s
Pulse rate: 200M - 400M
Response: X Hz - Y kHz
Transcoder: 60w, Max
SPL: 100 dB @ 60w
Bits: 1 (32, 64 - RGBV)
Bit rate: 200 Mbits/s
Pulse rate: 200M - 400M
Response: X Hz - Y kHz
Transcoder: 60w, Max
SPL: 100 dB @ 60w
Last edited:
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
Note you don't have to use all of the values represented by RGB, you can easily set a custom value range (colours used).
Although it would be pointless, you could do 2 colours 1 pulse, which is equal to binary in total values.
PAM-X audio reduces power used, while increasing signal fidelity and audio resolution.
And as you have already seen, it also reduces digital bitrates and length.
Values^Bits = [Bit Value Range]^[Bits Used], No binary!
----
For external devices such as the DMAS (Digitally Managed Audio System), no internal soundcard is needed.
Instead, the DMAS communicates as a virtual soundcard, receiving a RAW RGB stream.
The PAM-X amp is a bit value changer, position processing.
----
Same again, but a bit more basic.
Although it would be pointless, you could do 2 colours 1 pulse, which is equal to binary in total values.
PAM-X audio reduces power used, while increasing signal fidelity and audio resolution.
And as you have already seen, it also reduces digital bitrates and length.
Values^Bits = [Bit Value Range]^[Bits Used], No binary!
----
For external devices such as the DMAS (Digitally Managed Audio System), no internal soundcard is needed.
Instead, the DMAS communicates as a virtual soundcard, receiving a RAW RGB stream.
The PAM-X amp is a bit value changer, position processing.
----
Same again, but a bit more basic.
Last edited:
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
HDA, limited to ~37 Mbits/s can deal with 8 channels 24 bit 192k, 36.864 Mbits/s.
In PAM-X RGB (24), that would be 8 x 192 Kbits/s, 1.536 Mbits/s.
Note: In PAM-X RGBV (32), that would also be 8 x 192 Kbits/s, 1.536 Mbits/s *.
* If V = 256 and 256^4.
In PAM-X RGB (24), that would be 8 x 192 Kbits/s, 1.536 Mbits/s.
Note: In PAM-X RGBV (32), that would also be 8 x 192 Kbits/s, 1.536 Mbits/s *.
* If V = 256 and 256^4.
Last edited:
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
If you have been following my posts on PAM-X and the corrections made, you should now understand how to use Superposition, colour spectrum.
You should have also established that more original data with audio, reduces the requirements for hardware, including filters.
The data rate, and pulse rate required to transmit audio data is reduced, 1.536 Mbits/s for example.
In case you are wondering how many channels PAM-X can do at 192k and 36.864 (HDA), its 192, so 192 x 192,000 bits = 36.864 Mbits/s.
----
Note: The PC end with an external DMAS will need a mixer, so the output is one stream, this should be handled by the driver-other.
Also note that the PC end will be in RAW mode, so no buffer, no APO's, all features are handled by the DMAS.
====
Note, for spatial audio its 1 bit per sample followed by 1 bit 3D position (RGB - XYZ), so 2 bits linear per sample.
====
Also note, Windows will need to be updated to support more than 8 channels, its certainly not a PAM-X limitation, or USB for that matter.
----
Personally I would buy a DTS DMAS, and then relevant speakers.
You should have also established that more original data with audio, reduces the requirements for hardware, including filters.
The data rate, and pulse rate required to transmit audio data is reduced, 1.536 Mbits/s for example.
In case you are wondering how many channels PAM-X can do at 192k and 36.864 (HDA), its 192, so 192 x 192,000 bits = 36.864 Mbits/s.
----
Note: The PC end with an external DMAS will need a mixer, so the output is one stream, this should be handled by the driver-other.
Also note that the PC end will be in RAW mode, so no buffer, no APO's, all features are handled by the DMAS.
====
Note, for spatial audio its 1 bit per sample followed by 1 bit 3D position (RGB - XYZ), so 2 bits linear per sample.
====
Also note, Windows will need to be updated to support more than 8 channels, its certainly not a PAM-X limitation, or USB for that matter.
----
Personally I would buy a DTS DMAS, and then relevant speakers.
Last edited:
- Joined
- Dec 8, 2020
- Messages
- 3,209 (1.95/day)
- Location
- United Kingdom
Spatial audio is mixed into hardware channels, the sample frame is a long pulse, double width.
Note that 2 white bits will not match 1 double width, white is also max volume.
0 value bits (black) represents 0 power in a speaker.
====
Reminder the DMAS is self controlled, and can assign both hardware and spatial channel count, layout.
The consumer may only require 5.1, therefor it would be 6 hardware + X spatial.
The transmitter end, the PC end, can very easily have a hardware mixer pre DMAS, if needed.
----
Spatial audio channels can have a static (constant) position, and represent missing channels.
It can also be dynamic positions, as in, moving around hardware channels.
----
Another benefit of the DMAS, you don't need to pick a motherboard based on its audio (sound card).
====
Retrofitted optical HDMI AB cable (£15.99 2m), A(1): optical transmitter (Tx), B(2): optical receiver (Rx).
Obviously we don't need 48 Gbits/s (more in PAM-X), 20 (ch) x 200 Mbits = 4 Gbits/s.
----
Windows only really needs to support more channels, and not so much channel layout or purpose.
The driver and related software allow the consumer to setup channels and layout.
Its also possible for the software to configure the DMAS main unit, rather than the unit its self.
Last edited: