Which is the best-sounding Realtek driver so far?

[qubit]

Lol you guys who claim drivers don't sound the sound of audio don't know much..
There are in-built Highpass/Lowpass (even EQ) on some (most) audio drivers.
Other forums are C-RAZY about new releases, will null-test them and use their ears. You can read semi-complete descriptions all you want. It's like in Mastering. They have the whole truth and partition it out in semi-truths or even deliberately sneer in on false info.. so you will pay pay pay for something that isn't better than what existed on WinXP with free plugins (JUCE framework and Steinenberg?¿ - lol). They nerf the code and functional 'purity' that powers our tech since the hardware speccs are becoming increasingly better..
Now you know.

I'd love to see an objective review about this. It's possible, I guess, but the driver writers would have to believe that there's a financial benefit in it for them to add extra processing just to reduce the sound quality.

 

[Operandi]

Lol you guys who claim drivers don't sound the sound of audio don't know much..
There are in-built Highpass/Lowpass (even EQ) on some (most) audio drivers.
Other forums are C-RAZY about new releases, will null-test them and use their ears. You can read semi-complete descriptions all you want. It's like in Mastering. They have the whole truth and partition it out in semi-truths or even deliberately sneer in on false info.. so you will pay pay pay for something that isn't better than what existed on WinXP with free plugins (JUCE framework and Steinenberg?¿ - lol). They nerf the code and functional 'purity' that powers our tech since the hardware speccs are becoming increasingly better..
Now you know.

Well..... nevermind I guess. I know drivers have high, low pass filters and EQ, I didn't know thats what being addressed. Personally I just want the best audio stream out of the PC streamer, or whatever I happen to be using as a source, hence my recommendation to not use SPDIF and to make sure the OS isn't resampling the output. I would never use any high / low pass filters or EQ unless it was something like Roon (I don't even use the bass and treble controls on my amp) but yeah if you are using drivers for that, then I have no doubt there would be differences.

 

[emanresu]

I'd love to see an objective review about this. It's possible, I guess, but the driver writers would have to believe that there's a financial benefit in it for them to add extra processing just to reduce the sound quality.

Just go to other audiophile forums and see how they have favourite release versions "because of X, Y & Z audible reasons". Wherever Mokichu went with his releases and some russian alternatives will open your eyes on this, if you still don't understand..

Why do you think bcdedit with platformclock, dynamictick, tscsync and such exists?
How do these settings correlate with Fidelizer, Audiophile Optimizer and other tweaks?

You talk about resampling, which is a good thing to mention.. if you take something short (~0.1s) and "boomy sounding"/"wave-looking" in Audacity and then resample it (maybe from original 48kHz to 44.1kHz) you'll be amazed how f*cked up the end result looks like. There's no sense in that we don't have a solution.. since.. maths!

I'm a bit worried that the very fundamentals behind what we are using every day to communicate will be lost. In a neat pre-requesite covered UAD DCH .appx box.
And it will sound like shit compared to what was already in the 90s in CD quality.

Did you know that CD (44.1/16) is 44100 SAMPLES per second, and this is somehow translated also to 44100 Hz. One sample per Herz. Which is logical. What do you think the sample buffer / latency figure is on (fake) Wasapi for 44.1 / 16, that we are all using, on MS Win11? 1/100 of that CD player rate. Totally insane. Our senses will be unifyingly dull in one generation because of this.. what HiFi lol.

Well..... nevermind I guess. I know drivers have high, low pass filters and EQ, I didn't know thats what being addressed. Personally I just want the best audio stream out of the PC streamer or whatever I happen to be using, hence my recommendation to not use SPDIF and to make sure the OS isn't resampling the output. I would never use any high / low pass filters or EQ unless it was something like Roon (I don't even use the bass and treble controls on my amp) but yeah if you are using drivers for that, then I have no doubt there would be differences.

I have learned that you want what the original author intended. A shared experience that they created to us. Or.. if you can OBJECTIVELY (to the whole world) improve it. So be it ! But all these injections of functions in-between source material (like Roon EQ/upsampling options) and what is the final output, will create mathematical assumptions that should not be there. And just because MS started with DRM in Vista and ditched the Kernel drivers (why?) we enjoyed in XP (hello beautifulest start-up sound ever! Nice hills!) all these extreme solutions have appeared to band-aid what could be a 'non-problem'. They do not solve the underlying issue with more false math... If considered the 'whole' as a perfect equation. The people behind some 'fixes' are innovative as fuark. But once that fidelity bug bite you, and you realize something is wrong from the get-go (you are now not trying to improve, you are actually mending something and you don't know where it's bleeding) - one can snap IRL. Because of audio. It's not worthy.

 

[kiriakost]

Not sure if this was a joke reply or not but its correct. Drivers don't have any impact on quality, the software stack itself does (the OS) and Windows resampling isn't the best but even this has minimal impact compared to everything else going on.

Since the beginning of time, Realtek = poor audio hardware, software EQ and many more functions in software, large footprint of CPU usage in gaming.
True, drivers did not had impact on quality, at Creative X-Fi products.

I am thinking to activate ALC889A+ too, just for gaming headset and VoiP.
Because at my current Game, the game thinks X-Fi mic input structure as hard to activate.
And I am after to test if ALC889A+ this has any better compatibility.
The vote of sound quality and quad speakers playback, this goes to X-Fi.

Within the past 15 years, THE standard for the test of Audio drivers timing performance, this is no other than steinberg - cubase, and to load on it, several MIDI plug-in.

 

[qubit]

Just go to other audiophile forums and see how they have favourite release versions "because of X, Y & Z audible reasons". Wherever Mokichu went with his releases and some russian alternatives will open your eyes on this, if you still don't understand..

I said objective tests, as in from proper reviewers without an agenda. Hard to find, I know. The opinions of self-styled "audiophiles" who often think that marker pens applied around the outside of a CD make it sound better and that expensive HDMI cables give a better picture have no credibility.

Did you know that CD (44.1/16) is 44100 SAMPLES per second, and this is somehow translated also to 44100 Hz. One sample per Herz. Which is logical.

No, it's not logical. The highest possible frequency, as per Nyquist's theorem is only half the sample rate, so 22050Hz. Feed it an analog signal above that and you get aliasing. It's all here:

Nyquist–Shannon sampling theorem - Wikipedia

en.wikipedia.org

 

[Kapone33]

Since the beginning of time, Realtek = poor audio hardware, software EQ and many more functions in software, large footprint of CPU usage in gaming.
True, drivers did not had impact on quality, at Creative X-Fi products.

I am thinking to activate ALC889A+ too, just for gaming headset and VoiP.
Because at my current Game, the game thinks X-Fi mic input structure as hard to activate.
And I am after to test if ALC889A+ this has any better compatibility.
The vote of sound quality and quad speakers playback, this goes to X-Fi.

Within the past 15 years, THE standard for the test of Audio drivers timing performance, this is no other than steinberg - cubase, and to load on it, several MIDI plug-in.

I agree with this but the new 1220 driver (codec) is actually better than the Asus Xonar XE. it has a greater sound stage range and vs the 897 has a much higher signal to noise ratio.

 

[emanresu]

I said objective tests, as in from proper reviewers without an agenda. Hard to find, I know. The opinions of self-styled "audiophiles" who often think that marker pens applied around the outside of a CD make it sound better and that expensive HDMI cables give a better picture have no credibility.


No, it's not logical. The highest possible frequency, as per Nyquist's theorem is only half the sample rate, so 22050Hz. Feed it an analog signal above that and you get aliasing. It's all here:

Nyquist–Shannon sampling theorem - Wikipedia

en.wikipedia.org

I don't know if the 44100 is per channel or Stereo 2.0 - but I guess the latter?
Nyquist's >theorem< - in my experiments I smashed that to the window, to the wall!

 

[bug]

I don't know if the 44100 is per channel or Stereo 2.0 - but I guess the latter?

He gave you the Wikipedia link, it's all there.

Nyquist's >theorem< - in my experiments I smashed that to the window, to the wall!

Nothing easier to disprove/dismiss than what you don't understand.

 

[kiriakost]

Nyquist's >theorem< - in my experiments I smashed that to the window, to the wall!

If Nyquist has the chance to live and a second life, and use 12 bit digital storage oscilloscopes of our times, then he would be able to improve the first theorem by adding field test examples at YouTube too.
Sampling resolution this is always per channel.

--------------------------------------------------------------------------------------------------------
I did activate ALC889A+ at my Gigabyte GA-Z87X-UD5H.
Last official driver release REALTEK: June 25 2015 ... 6.0.1.7541
(Gigabyte logo + Realtek HD audio manager panel shown at my screen).
--------------------------------------------------------------------------------------------------------
First official hardware release:
Shanghai – April 02, 2008 Realtek Semiconductor Corp.will demonstrate its latest R&D achievements at the Spring 2008 Intel Developer Forum
High Definition Audio Codecs with Embedded Class-D Amplifier
ALC889

The ALC889 is a 7.1+2ch DAC/6ch ADC High Definition Audio Codec offering quality audio performance (DAC 108dB SNR, ADC 104dB SNR).
The multi-channel DAC/ADC is the ideal choice for home entertainment PCs, providing support for multi-streaming to enable enjoyment of 7.1-channel audio/movies/on-line gaming, etc, and simultaneous 2-channel VoIP.
The ALC889 complies with the latest Microsoft Windows Vista Premium requirements, and offers DSD (Direct Stream Digital) format support and Full Rate Blu-Ray/HD DVD playback capabilities.
--------------------------------------------------------------------------------------------------------

The person behind the unofficial drivers, he does not have the knowledge to keep the Realtek HD audio manager panel active (after his own add on driver tweaks).
And so I was forced to downgrade from HDA_6.0.1.8186, back to the original work of Realtek.

Realtek Semiconductor Corp. High Definition Audio System Software Ver: R2.79 Installation and Setup

 

[qubit]

I don't know if the 44100 is per channel or Stereo 2.0 - but I guess the latter?
Nyquist's >theorem< - in my experiments I smashed that to the window, to the wall!

Yes, it's 44100 per channel and 16-bit for CD audio.

Digital sampling for audio and video is a fascinating subject, so perhaps have a look at some reference stuff on it.

That Wikipedia article is a bit too technical with the maths, but there's loads of sites that explain the principles in plain language, along with YouTube videos. Check them out.

 

[emanresu]

Unfortunately as it is, much is a subject of interpreting what has been already implemented, when developing an add-on or alike - and then trying it. So there are a lot of mismatching behind the curtains that we really don't know of. I don't believe in Nyquist's theorem. I can tell the difference between 44.1 / 48 / 96 (88.2 is difficult) and 192 very clearly. Also bitdepth. It's a matter of knowing what to listen for, which cues. But I on the other hand have blown 32bit floating point ("infinite headroom") to back to da blaze with bass. It gave the classical Ozone v1 sound (a fizz). And this was with 32bit fp all the way. It didn't even sound the same on playback. What you guys tend to forget is that there is information in the header, and that the .wav format extension only tells the program what to do with it. Some things fly by depending on the player. I'm at the end of my extensive mindbreak of a journey. Installing Win XP x32 today with a tweak to make it handle 4GB+ files and RAM helllooooo kernel drivers! No need for pre-emphing something that shouldn't need to.

Btw, the Realtek chips.. I'm not sure about this, but if the Brazilian forum guy I mentioned managed to extract the "super-codec" from ASUS drivers and then apply it on an older chip.. it's pretty telling.

 

[qubit]

I don't believe in Nyquist's theorem.

What? It's not a matter for debate or opinion, but established fact, hard science.

 

[bug]

What? It's not a matter for debate or opinion, but established fact, hard science.

Yes, but believing in hard science is so passé.
Ask him how he concluded he can hear the difference between 44.1 and 48kHz sampling rates, I dare you

 

[emanresu]

Lol

 

[kiriakost]

Yes, but believing in hard science is so passé.
Ask him how he concluded he can hear the difference between 44.1 and 48kHz sampling rates, I dare you

I have read claims over the past 25 years, that people can sense the difference between different copper wires, and or banana plugs.

At the field, I have see many deaf DJ, them to blow hundreds of Watt of speaker-system too ( I did the repairs).
And even, hardcore guitarist to think fake Nakamichi (plugs, connectors) as to be the holly grail.
When Nakamichi - Japan, never ever sold any connectors.

At 2023 the only alive Japanese R&D this can be found at CANARE L-4E6S cables.

 

[emanresu]

It's easy.

44.1kHz is rhythm-friendly. You can switch your beat in-between instruments, as the producer(s) intended. So called dynamic 'focus'.

48kHz can be a bit more square/edge in the heard audiospectrum (listen to where the instruments 'end' in headphones). This gives it punch. But due to a wider sound that is utilized. It is also flatter than 44.1kHz since there is no dynamic focus, all instruments can take equal part of the soundscape.

192kHz is soothing (not necessarily clean or HiFi:esque), but too slow. Especially when re-sampling has occured. Lost Magic Inc.™

16bit is like a flat line in-between your ears.

24bit has Z depth 'from you'/forward (headphones).

32bit also has Z depth 'backways'/towards you.

The problem with digitalizing analog is due to that each step must consider the 3Dness in an 3D environment. Analog usually has objects/instruments that are perceived as 3D in this 3D environment. When ReMastering occurs either the objects or the environment is totally flattened out. Then they do even more bad stuff. I noticed this after listening to a 'bitperfect' rip of the original Dark Side of the Moon (Japan, CD quality). Then I also had a listen to a counterpart, a vinyl 192kHz 24bit rip. The dude unfortunately had applied noise reduction, which removed the 3D space of the show. It's there for a reason and fills out the nothingness in-between instruments / separation.

You cannot undo these operations.

Thanks for asking.

 

[kiriakost]

Back -> on topic
Official page and or final work, PC Audio Codecs > High Definition Audio Codecs Software

PC Audio Codecs > High Definition Audio Codecs Software - REALTEK

www.realtek.com

The dude unfortunately had applied noise reduction, which removed the 3D space of the show. It's there for a reason and fills out the nothingness in-between instruments / separation.

Even at vinyl times, the dudes in UK, they discover the sound mastering procedure.
The long story in sort, all our sounding material (music) this is processed and far and away from the original recording.

 

[Operandi]

What? It's not a matter for debate or opinion, but established fact, hard science.

Its the Nyquist theorem, not the Nyquist law. In simplest terms it states you need at least 2x the samples to accurately reconstruct a sign wave, the key phrase being "at least". The thing to aware of with digital audio is "accurate" is not a hard fixed point as even Red book audio is encoded at 44Khz to avoid the aliasing distortion appearing in the audible spectrum. This is all following a theorem that in general terms has proven to be highly accurate but where that ultimate point of accuracy is debatable. Thats why we have different approaches to oversampling and reconstruction filters, HD audio, and completely different DAC topologies. Things may all measure the same or sometimes even worse in aggregate on a test bench but human hearing is highly non-linear and what constitutes realistic natural sound from digital sources dosn't neatly fit within the definition of the Nyquist theorem.

 

[emanresu]

Yes OR that 44.1kHz is split into 2 channels of 22.05kHz? Hence "twice the sample rate"..
You know Dolby PL and their witchcraft of channel splitting and such.. Stereo 2.0 goes 4 channel.. and yet the 2 channels carry that info, but how?

Even at vinyl times, the dudes in UK, they discover the sound mastering procedure.
The long story in sort, all our sounding material (music) this is processed and far and away from the original recording.

Yes.. many people confuse 'bad' Mastering with just not being able to play what they are hearing 'pure out'. Try Kernel KS (or WDM-KS) as output, and you'll be surprised. Not ASIO. Not Wasapi-BS. Nor MME (2nd best, but too open). Thanks for the links !

Yao'll Google them 4 Bcdedit Platformclock Platformtick Dynamictick and Tscpsync.. you will be surprised.

 

[Operandi]

Yes OR that 44.1kHz is split into 2 channels of 22.05kHz? Hence "twice the sample rate"..
You know Dolby PL and their witchcraft of channel splitting and such.. Stereo 2.0 goes 4 channel.. and yet the 2 channels carry that info, but how?

Its 44.1Khz per channel of 20-20,000Khz of PCM. All the extra headroom in the spec is to avoid the aliasing artifacts the Nyquist theorem addresses.

No clue about how any Dobly stuff works or any of that, I'm strictly all about two channel stereo.

 

[kiriakost]

Its the Nyquist theorem, not the Nyquist law. In simplest terms it states you need at least 2x the samples to accurately reconstruct a sine wave, the key phrase being "at least".

There is also another battleground, in there they also debating about required samples so one digital storage oscilloscope this to accurately reconstruct a sine wave over it screen.

Anyway, my own battleground ended today after of me wasting again several hours in research, it appears that Realtek it loved Win7 64bit up to the version R2.81
Realtek HD Audio Manager: Loads successfully
Realtek CPL loads at control panel
Realtek icon loads successfully at taskbar, mixer appears too.

The talk about ultimate sound fidelity this never ends, when I am feel thirsty, I just use my Japanese (90s) heavy weight of sound system, and my special 2x150W pair of three way drives these designed and made in Holland.
When and if the Europeans will return at sound speakers development, then our youth will have back the appropriate tools so them to listen quality.

 

[emanresu]

There is also another battleground, in there they also debating about required samples so one digital storage oscilloscope this to accurately reconstruct a sine wave over it screen.

Anyway, my own battleground ended today after of me wasting again several hours in research, it appears that Realtek it loved Win7 64bit up to the version R2.81
Realtek HD Audio Manager: Loads successfully
Realtek CPL loads at control panel
Realtek icon loads successfully at taskbar, mixer appears too.

The talk about ultimate sound fidelity this never ends, when I am feel thirsty, I just use my Japanese (90s) heavy weight of sound system, and my special 2x150W pair of three way drives these designed and made in Holland.
When and if the Europeans will return at sound speakers development, then our youth will have back the appropriate tools so them to listen quality.

Wingland Caudiommonwealth of Great Hitannia!

 

[qubit]

Its the Nyquist theorem, not the Nyquist law. In simplest terms it states you need at least 2x the samples to accurately reconstruct a sign wave, the key phrase being "at least". The thing to aware of with digital audio is "accurate" is not a hard fixed point as even Red book audio is encoded at 44Khz to avoid the aliasing distortion appearing in the audible spectrum. This is all following a theorem that in general terms has proven to be highly accurate but where that ultimate point of accuracy is debatable. Thats why we have different approaches to oversampling and reconstruction filters, HD audio, and completely different DAC topologies. Things may all measure the same or sometimes even worse in aggregate on a test bench but human hearing is highly non-linear and what constitutes realistic natural sound from digital sources dosn't neatly fit within the definition of the Nyquist theorem.

You're splitting hairs with that highlighted statement. And it's worth remembering that a theory has lots of evidence supporting it, ie it's not an unverified hypothesis or idea that someone can simply choose to disagree with. Therefore, if someone "doesn't believe" in it, they have no credibility, especially after it's been explained to them.

The finer points of sampling, DAC topology etc all flow from this root theorem.

 

[lordmogul]

Been a while since I had to seriously look into the quality of Realtek audio, got a Xonar DGX with the UniXonar drivers running here, and it still does a perfectly fine job.
But there are driver "patches" available to change things on the driver, so the idea that different Realtek drivers can have different results doesn't sound too weird. No idea if there is any "golden" driver available and which one that would be, but if there is, I feel like that would be a question for audiophile forums. (Just don't get caught into buying fancy new stuff when all you want is finding the best driver for your chip and use case)

 

[Super XP]

Been a while since I had to seriously look into the quality of Realtek audio, got a Xonar DGX with the UniXonar drivers running here, and it still does a perfectly fine job.
But there are driver "patches" available to change things on the driver, so the idea that different Realtek drivers can have different results doesn't sound too weird. No idea if there is any "golden" driver available and which one that would be, but if there is, I feel like that would be a question for audiophile forums. (Just don't get caught into buying fancy new stuff when all you want is finding the best driver for your chip and use case)

Are you using Windows 11 or Windows 10? Because my Xonar DGX broke once Windows update updated to Windows 11, no matter how many times I would uninstall and reinstall the drivers. So I ended up with the Creative Sound Blaster Z SE. But in order for that one to work with Windows 11 I had to install an older driver software. Now that works great.

One would think that Windows 11 and Windows 10 are driver similar but they are not. Micro$oft messed that part up quite heavily.