Creative SXFI Amp

[VSG]

A word of caution to anyone ordering this that Creative's list of incompatible devices is incomplete. I was stoked about this product from reading the reviews so when the UK pre-ordes opened I pulled the trigger. The SXFI app has shipped and arrives tomorrow. I got an email from Creative instructing me to install the SXFI app via the Play store but I couldn't find it. I then search the Play Website and find it, only to discover my Moto Z (Android 8.0 and USB-C) isn't compatible with the app, despite not being in Creative's list of problematic devices. I now have a £145 SXFI amp paperweight arriving tomorrow that I will have to return as unfit for purpose as I can't set it up. No idea how widespread the compatibility issues are with other handsets but it's pretty disappointing. I can't get the app to run on my Nvidia Shield tablet (Android 7) either. For anyone thinking about buying one I'd recommend installing the app first to avoid wasting your time and money.

Refer to this post and see if you can get a refund: https://www.avsforum.com/forum/378-...x-fi-headphone-technology-3.html#post57380932

 

[Ferrum Master]

if the AK4377 support native DSD up to 256 ... why it is not mentioned, did they choose to not use it?

It doesn't matter what you see going through the device from windows sound control panel or similar and what really goes on in between the Creative in house ASIC GPIO's to the DAC. In what mode it does drive them. It is a closed source in house info. There are two layers.

But in my experience Creative does usually right with that(unlike cmedia), their issues could be power limitations, heat and latency. Balancing all those things out ain't no simple task, there could be sacrifices, could be not. Only getting my hands on the device could answer me that. DSD mode ain't the holy grail of anything also. Each device has their own sweet spots, and usually it is not maximum mode as drawbacks crawl in.

But most importantly... this is true hardware audio again... they are back in their sandbox. The accent is on the DSP part, not the end DAC part. That makes it special.

 

[GreiverBlade]

But most importantly... this is true hardware audio again... they are back in their sandbox. The accent is on the DSP part, not the end DAC part. That makes it special.

well the previous user report on incompatibility did seal it ... that thing is too restrictive/picky/closed

but ... to me X-Fi was a gimmick until now ... and will probably not change after they did go "super" ... no matter the "WOAAAAHHH" effect they got on the CES

DSD mode ain't the holy grail of anything also.

did i wrote that? nope ... tbh, i don't use DSD audio that often, but i am happy to have the compatibility with my current DAC/AMP

But in my experience Creative does usually right with that(unlike cmedia), their issues could be power limitations, heat and latency. Balancing all those things out ain't no simple task, there could be sacrifices, could be not. Only getting my hands on the device could answer me that.

i agree on that ... tho for me getting my hand on the device would end in a utter incompatibility with my S70 (not sure i could install the app on it ) and probably using it as a USB-C soundcard for my main rig


oh well, i understand your point, no worries.

 

[Ferrum Master]

but ... to me X-Fi was a gimmick until now.

The best thing that chip did is hardware resampling algo, it well respected and top notch even to todays standarts. With the Audigy[EMU10Kx] they had to use odd clocking, only E-MU had two proper oscillators driven by dedicated xilinx ASIC. But that's a vintage approach. Creative had to have such processing power on board to handle this feature actually. That's the reason why cmedia[most Xonars] sucks, their clocking, few scalers. It is dino age. It starts to act funny when doing non native modes also that's why the drivers are not stable, because of the chip itself being stubborn and buggy. XMOS is async and pretty much with DSD does what X-FI did internally on their cards for years... it took years to catch up Creative.

But the DSD mode really depends of the DAC itself, if to use it all... even the old and still my favorite PCM1794A does even support DSD (year 2003?), at start it was unofficial, but nobody really drives them using that mode. Pretty much experimenting and measuring with each individual device.

The other like positional audio acceleration just was a pleasant byproduct, that could be used with the sound card CPU, and they did. Well that was 2005, that's pretty much dark ages also. As we know HW audio was killed by M$. They had to have the processor of such scale to handle the DSP job, so it wasn't just for the gaming. It's an old habbit of theirs. VR could trigger need for really low latency audio. As each effect does apply latency, especially in windows audio stack, HW mode seems to have a reason to live again. Console gaming, handhelds doesn't even have options to that yet. There could be money in it for sure.

Listening to plain old audio... Stereo... nah... we don't need that for sure. Let's pick up a simple DAC and call it a day.

Also... I hate USB with my guts... I don't do audio using that, due to noise, and latency reasons. For me SPDIF optical is still sufficient for me. Galvanic isolation as it is optical and works everywhere. SPDIF out still is present in most devices.

Sh**t I need to work also today... lol.

 

[bug]

Also... I hate USB with my guts... I don't do audio using that, due to noise, and latency reasons. For me SPDIF optical is still sufficient for me. Galvanic isolation as it is optical and works everywhere. SPDIF out still is present in most devices.

What do you mean, noise? USB is digital, even the signal for 0 or 1 isn't perfect, it still gets reconstructed as 0 or 1.

 

[Ferrum Master]

What do you mean, noise? USB is digital, even the signal for 0 or 1 isn't perfect, it still gets reconstructed as 0 or 1.

EM noise. Ground and power rails from PC is polluted as F*. Joining any two voltage sources adds up their distinctive noise spectrum, the logic 1 level is also with an voltage level, that's gets terminated in the end receiver device, it sucks everything in. That's pretty much random, each silicon has their own characteristics rejecting ripple noises, mechanisms etc it ain't so simple. Each device also could add noise to the rail. It is very very hard to suppress it completely. Other methods are used.
It is all fine in the PC doing calculations having only 0 or 1. They can have the distinct noise and it is permitted... it works fine, the problem starts with our ears and analog sections we don't like actually hearing for example mouse movements, as it causes CPU interrupt thus the power section feeding the CPU screeches causing noise in the whole system and it passes on everywhere.

The noise sneaks in the latter analog section of the amplifier and compromises noise floor, THD, it lingers there, some stupid device whistling. Very funky things actually. That's how a bad design with proper testing differs from some cheap devices, yet the specs and endparts on the paper seem the same. Modern DAC's have filters and higher rejection ratios, but it still have their limits.

 

[bug]

EM noise. Ground and power rails from PC is polluted as F*. Joining any two voltage sources adds up their distinctive noise spectrum, the logic 1 level is also with an voltage level, that's gets terminated in the end receiver device, it sucks everything in. That's pretty much random, each silicon has their own characteristics rejecting ripple noises, mechanisms etc it ain't so simple. Each device also could add noise to the rail. It is very very hard to suppress it completely. Other methods are used.
It is all fine in the PC doing calculations having only 0 or 1. They can have the distinct noise and it is permitted... it works fine, the problem starts with our ears and analog sections we don't like actually hearing for example mouse movements, as it causes CPU interrupt thus the power section feeding the CPU screeches causing noise in the whole system and it passes on everywhere.

The noise sneaks in the latter analog section of the amplifier and compromises noise floor, THD, it lingers there, some stupid device whistling. Very funky things actually. That's how a bad design with proper testing differs from some cheap devices, yet the specs and endparts on the paper seem the same. Modern DAC's have filters and higher rejection ratios, but it still have their limits.

How on earth does it do that, if the input is digital?

More to the point: a DAC takes a 0 and translates it into some voltage level, takes a 1 and translates it to another. If that 0 is made up of 25% noise, it's still a 0 and translated to the right voltage level, not to that level +25% added noise.

 

[Ferrum Master]

How on earth does it do that, if the input is digital?

More to the point: a DAC takes a 0 and translates it into some voltage level, takes a 1 and translates it to another. If that 0 is made up of 25% noise, it's still a 0 and translated to the right voltage level, not to that level +25% added noise.

You didn't get that your digital data gets converted to analog audio you hear at some point at the end device. It is not jitter. It is EM noise, hooking up two devices with different power sources causes random issues. Read up mate. I am finished.

 

[bug]

You didn't get that your digital data gets converted to analog audio you hear at some point at the end device. It is not jitter. It is EM noise, hooking up two devices with different power sources causes random issues. Read up mate. I am finished.

So you cannot think of a mechanism that takes EM noise from the digital phase and carries it over to analog. Neither could my teacher in college.

 

[Ferrum Master]

So you cannot think of a mechanism that takes EM noise from the digital phase and carries it over to analog. Neither could my teacher in college.

Prff, what? Are you joking badly?

Draw Kirchoff law in each section, see where the loops are... it is not digital part of it, you imagine it like open joint while solving it?. It will sink through the inner termination DAC components, they are not galvanically isolated. As real part it has resistance, capacity and is simulated like that.

https://www.analog.com/en/technical-articles/power-supplies-high-speed-adc.html
https://www.analog.com/en/analog-di...ock-jitter-on-high-speed-dac-phase-noise.html

And at last... take an oscilloscope. And measure. See and please your self seeing what happens when hooking up the PC to each device. The noise floor will be polluted with various spikes coming from many devices. PC is one of the worst. Hooking up two devices and measuring? Sure.... Try for yourself.

 

[bug]

Prff, what? Are you joking badly?

Draw Kirchoff law in each section, see where the loops are... it is not digital part of it, you imagine it like open joint while solving it?. It will sink through the inner termination DAC components, they are not galvanically isolated. As real part it has resistance, capacity and is simulated like that.

https://www.analog.com/en/technical-articles/power-supplies-high-speed-adc.html
https://www.analog.com/en/analog-di...ock-jitter-on-high-speed-dac-phase-noise.html

And at last... take an oscilloscope. And measure. See and please your self seeing what happens when hooking up the PC to each device. The noise floor will be polluted with various spikes coming from many devices. PC is one of the worst. Hooking up two devices and measuring? Sure.... Try for yourself.

The digital interface will be covered first and happens to be easiest to treat. ... As a result, in this product line, the interface is not of concern, notwithstanding those spurs that may need attention depending on system requirements

 

[Ferrum Master]

Do measuring, and talk again. Or, actually no. It is better to live without knowing it, if you don't notice it.

 

[bug]

Do measuring, and talk again. Or, actually no. It is better to live without knowing it, if you don't notice it.

Yeah, well, my rule for sound is: if you need to measure the difference, there is no difference
I might think differently if I was mastering sound, but I'm not.

 

[Ferrum Master]

Yeah, well, my rule for sound is: if you need to measure the difference, there is no difference
I might think differently if I was mastering sound, but I'm not.

The rule works for your and your cherry picked equipment and closed environment, that a recording studio is. Not the wild zoo of consumer devices not following any rules. But as i said, If you don't feel it, it is even better like that. Just psychologically. On practice, exactly consumer, enthusiast DAC devices are made wrong. They are tested in ideal environment, using ideal supplies. On the field the things do work different. Just even why USB has ferrite chokes then in data lines? Power supply coupling, strict routing rules? If it is really that immune?

I am sorry for the derail.

 

[Fergutor]

To VSG
Can you clarify something to me?
I couldn't find time to read the whole article/review yet but for what I saw and searched in the article, it seems that there's no emphasis in Atmos (or DTS:X, or other similar), but in 7.1 simulation in headphones (I know they also have this stereo speakers simulation in headphones), and I remember (and I just checked again) that in the CES 2018 show the whole thing was to show Atmos (speakers around especially above) simulated or recreated in headphones. But I see now no mention of it, just 7.1 (and the stereo speakers simulation in headphones I saw a Creative representative in an interview talking about, and also in the Creative page). And that is no special at all (except the way it's done) in the sense that it already exist in many forms, and not only from Creative. CMSS3D, THX TruStudio Pro/SBX Pro Studio Already worked perfectly for that. Sure it isn't personalised as this new tech and surely this one is better for that, but is still not what was shown in CES right? Atmos is not 7.1, and 7.1 doesn't have speakers above...
So they downgraded it (to something already existing because the stereo simulated in headphones was there as well...)?
If not, what about the software needed for it to work, can I for example use my favorite player, Potplayer for example, play an atmos (the player support object based but in the output speakers section doesn have something in accordance so I don't know)file (the player support object based but in the output speakers section doesn have something in accordance so I don't know) and expect it to work properly?

 

[VSG]

The rule works for your and your cherry picked equipment and closed environment, that a recording studio is. Not the wild zoo of consumer devices not following any rules. But as i said, If you don't feel it, it is even better like that. Just psychologically. On practice, exactly consumer, enthusiast DAC devices are made wrong. They are tested in ideal environment, using ideal supplies. On the field the things do work different. Just even why USB has ferrite chokes then in data lines? Power supply coupling, strict routing rules? If it is really that immune?

I am sorry for the derail.

On the contrary, I am learning as I read on. It is relevant enough, and the conversation has been fine.

To VSG
Can you clarify something to me?
I couldn't find time to read the whole article/review yet but for what I saw and searched in the article, it seems that there's no emphasis in Atmos (or DTS:X, or other similar), but in 7.1 simulation in headphones (I know they also have this stereo speakers simulation in headphones), and I remember (and I just checked again) that in the CES 2018 show the whole thing was to show Atmos (speakers around especially above) simulated or recreated in headphones. But I see now no mention of it, just 7.1 (and the stereo speakers simulation in headphones I saw a Creative representative in an interview talking about, and also in the Creative page). And that is no special at all (except the way it's done) in the sense that it already exist in many forms, and not only from Creative. CMSS3D, THX TruStudio Pro/SBX Pro Studio Already worked perfectly for that. Sure it isn't personalised as this new tech and surely this one is better for that, but is still not what was shown in CES right? Atmos is not 7.1, and 7.1 doesn't have speakers above...
So they downgraded it (to something already existing because the stereo simulated in headphones was there as well...)?
If not, what about the software needed for it to work, can I for example use my favorite player, Potplayer for example, play an atmos (the player support object based but in the output speakers section doesn have something in accordance so I don't know)file (the player support object based but in the output speakers section doesn have something in accordance so I don't know) and expect it to work properly?

It never was about an Atmos-type setup, even at CES 2018. This was always about getting the soundstage of a 7.1 surround speaker set up around you into a pair of headphones.

 

[Fergutor]

It never was about an Atmos-type setup, even at CES 2018. This was always about getting the soundstage of a 7.1 surround speaker set up around you into a pair of headphones.

?!
Maybe you didn't understand my comment.
And in the CES 2018 conversion from atmos to headphones was precisely the thing (the main thing)
https://www.avsforum.com/best-ces-2018-creative-super-x-fi-headphone-technology/
https://www.cnet.com/news/creatives-new-super-x-fi-audio-tech-is-frigging-mind-blowing/

https://www.pcworld.com/article/3333606/consumer-electronics/creative-super-x-fi-review.html
And now is like "that didn't happended SXFI is about something else" something that already existed and for some reason pretend is something new (only some details are). Even this guy from creative

and not even the host made any comment about atmos (that I remember) only about 7.1 to headphones (that already exist beautifully as part of cmss3d, thx tru studio, sbx studio pro and others) and the emulation of stereo speakers through headphones... And precisely that was the core of my question, why suddenly this amnesia...
So your...whole comment. I thought you could clear some things for me but...forget it!!

 

[TheDocho]

If I understand correctly, the headphones you selected in the SXFI android app get saved on the SXFI amp itself? Hence, when you connect it to another device, let's say a windows pc, the SXFI amp still 'knows' the it should be optimized for those headphones.

However, is this also true for the graphic equalizer settings, does the SXFI amp retain the equalizer settings when connecting to that windows pc?

 

[VSG]

If I understand correctly, the headphones you selected in the SXFI android app get saved on the SXFI amp itself? Hence, when you connect it to another device, let's say a windows pc, the SXFI amp still 'knows' the it should be optimized for those headphones.

However, is this also true for the graphic equalizer settings, does the SXFI amp retain the equalizer settings when connecting to that windows pc?

Yes to the former, the latter is on the app itself and not an amp profile saved on the device.

 

[kpmn]

https://forums.hardwarezone.com.sg/...eative-super-x-fi-amp-6-34pm-5908001-308.html

http://www.sharejunction.com/sharej...e=Creative files lawsuit against Nvidia in US

 

[VSG]

https://forums.hardwarezone.com.sg/...eative-super-x-fi-amp-6-34pm-5908001-308.html

http://www.sharejunction.com/sharej...e=Creative files lawsuit against Nvidia in US

Was there anything you actually wanted to say or just share those links?

 

[EarthDog]

Edit: Ah, you are referring to the DAC and amp in here. I see what you mean, and we can explore this further via our actual audio editors. I wrote this up because I was one of two TPU editors at CES experiencing this, which you absolutely had to in person, and neither of us are the designated audio editors.

Thanks for taking the time... though I am a bit confused if this is a CES article (news bit) a review (without audio numbers?) or an unboxing. It feels like 1/3...

Looking forward to some legit testing of the device itself! I appreciate reading about your experience.

 

[VSG]

Thanks for taking the time... though I am a bit confused if this is a CES article (news bit) a review (without audio numbers?) or an unboxing. It feels like 1/3...

Looking forward to some legit testing of the device itself! I appreciate reading about your experience.

I don't have the means to quantitatively test DACs/amps, and the USP here is the Super X-Fi tech which is pretty much a subjective thing. Don't expect any more additions to the review here, you might want to check elsewhere if you want what you feel is legit testing of the device itself that is not done here.

 

[EarthDog]

Will do. Thanks again!

 

[VSG]

Will do. Thanks again!

Least I can do