XXHighEnd

Wonderful improvement over my M-Audio Audiophile USB for Spdif conversion. GerardA what once was harsh in those hi frequencies, female voices and violins is all very smooth and proper sounding.

But I still can't get Adaptive or Special mode to work. Each time I try I get a message that there are too many buffer errors. When I retry, I then get a message that, "The Device is allocated but cannot play, check dac settings". Once this message has appeared I have to restart the computer cause xx quits working. After restarting four or five times I went back to normal mode with the settings below. Normal mode, for now, is fine for getting how cr*ppy my previous spdif signal was. Damn, gotta love this sh*t. :)

Those messages occur with xx set for 24 bit only or 32 bit. Also with the wasapi driver, engine 3,it plays for 2 or 3 seconds and then gives a message that the buffer is too large!

So Peter, or anyone else who is sucessfully using the Hiface in Special or Adaptive mode, what settings work for you? 

Dear B.,

Let this be a not so much serious post with an undertone of truth though ...

If I were you, I would take a day off and invite your wife to the zoo, find a nice restaurant for after that, have some beautiful drinks going along with it, and then bring up the subject of needing a new PC.


Everything just works. The HiFace is a very generous device IMO. I used it with 22 samples at Special Mode (but it depends on the PC whether it can cope with that), and Q1 = 4 for Engine#3. Normal mode just ran, and Adaptive Mode I never tried.


To me this looks a nice 100% the same as your PC's inability to change formats. Before we could blame the soundcard, but now we can't anymore.


I never heard of the message, hence it doesn't come from me. Can you post a screen copy of it please ?


In any case, glad you have one (again) at last. Now let's hope it doesn't bring you misery ...
:)
Perer

Hi Boleary, Peter,

It's to cold to go to the zoo here, so I went to my work.
Not really in the right mood to give you a serious answer too...

My hiFace arrived two days ago. First on the HTPC, but this freezes because of a AMD :evil: SB700 chipset.
Then on the centrino laptop it works (special mode) , sounds impressive / strange like more upfront and sounds appearing unexpectedly like in real live.
Also no noise at all like I got from mobo-spdif before.
Sometimes music stops/drops out for a second. Second album don't start. Reboot PC then first song ok, second in fast-motion.
Then messages about everything, like DAC endpoint, DAC samplerate, etc...
(Laptop is not tweaked to death yet, too )

So we really have to wait for new drivers to get happy!
(On Head-fi there is a thread with up to date answers)

Of what I've heard there are NEWER drivers which DO NOT WORK.

I work with the ones John kindly provided here : Re: XXHIGHEND-M2TECH-WIN XP (http://www.phasure.com/index.php?topic=1105.msg9567#msg9567). I strongly advice to try those, so I can at least judge everything regarding the same driver version.

I really didn't notice a single problem (after I made it working for XXHighEnd of course) ...
But keep in mind that at this moment I did not try Adaptive Mode. I suppose though, you can go as low as 32 for buffer size (@ Q1=1).

Let me know what happens with the above referred to drivers !
Peter

Guys,

I've had the Hiface working in my office system (with Atom PC) for over a week now. The DAC here is only up to 96KHz capable, but DAP in KS 'Adaptive mode' works perfectly. I can't remember (I'm working out of the country right now), but I think I have buffer=1024 and Q1=1. IIRC, buffer=128 and Q1=2 also works fine... I can check this when I'm back this weekend.

In any event, the Hiface definitely works fine with XXHE.

Mani.

Well somewhere between the murkats and the mongeese it occurred to me change core appointment so I switched from scheme 3 to scheme 1. In special mode with the buffer set to 384 and Q1 at 5 all is well. If i reduce the buffer to 256 I have to increase Q1 to 7 or I get distortion. Does this mean that the proper buffer setting is 384? Can't rely on my ears right now cause everyone is sleeping.

Adaptive mode works with the same special mode settings: 384/5. If I go below Q1 at 5 I get the "too may buffer errors" and "device allocated but cannot play messages," which then require a reboot.

Haven't tried engine 3 yet. Am off to work now but first I gotta go to the zoo....forgot my wife was still there.... :)

:rofl:

I come to 2048 and Q1=1 (or higher) for the HiFace. This is W7.

Note : The "guide" for finding the buffer doesn't seem to work here; if I follow the guide (Tooltip on the KS Mode combobox) I end up at 256, but next sliding Q1 (from 0) to 1, and ... quite some Buffer Errors.

Hope this helps.
Peter

I must add something to this :

1. This was with the HiFace used at DAC Needs = 24 bits (this may matter, but I didn't check it).
2. In Normal Mode I accidentally tried Q1=5, but it didn't work well. Next tried Q1=20 an no Buffer Errors around. Notice that the Set Device Buffer Size does nothing for Normal Mode.

For the above too, I used DAC Needs = 24 bits.

For your conveniency, 0.9y-8c will show the Set Device Buffer Size and Q1 value (in samples for Special Mode) in the X3PB log file.

Well, let me just say the wait was worth it, though I'm glad it's over. Been listening in Special Mode the last couple hours and, ah, thank you Peter; the sound here is magnificent. Never heard anything like it coming from a stereo system.

Wait a bit more and you'll have two. Now, *that* will be something ... :)

With my Atom PC and the Hiface feeding an old Musical Fidelity X24 DAC (24/96 max), I've arrived at the following settings:

DAP
KS Adaptive (Special not tested)
buffer=1024
Q1=1

I'm getting no buffer errors (apart from a single one right at the beginning due, I suspect, to my integrated video driver trying to keep up with XXHE going into Unattended mode). This X24 DAC has been stored away for the best part of 10 years. I hated it when I bought it - way too bright and forward. However, it now sounds pretty listenable with XXHE and the Hiface.

Mani.

Of course, if you're going to use a Hiface, the spdif cable is going to be critical.


I'm currently using this cable between the Hiface and the X24 DAC. I'm not sure if it's down to the Hiface or the cable, but I've never heard this DAC sound so smooth... It could well be the cable, as this is exactly what other people have described when talking about the effect this cable has on the sound. I have a pretty expensive Cardas spdif cable knocking around somewhere and will do a comparison at some point.

Mani.

Thanks Mani. I've got a cable on order. Not being in a position to audition them, I am willing to defer to an Audiodidact  :).

Regarding yours and Peter's settings, I can't duplicate them. 1024/1 induces too may buffer errors. Also, I can't use core appointment scheme #3; #1 works for me. Don't know why this is so. What rationale did you use in settling on 1024/1? Just wondering.

I just followed Peter's instructions... sort of! I set Q1=1 and started with the lowest buffer=32. With the 'log activity' box checked, I continued increasing the buffer size until there were no buffer errors in the X3PB log file. Now of course, I could reduce the buffer size from 1024 and increase Q1 from 1 and probably find a number of other values that work. But buffer=1024 and Q1=1 sounds good.

Incidentally, scheme 3 works fine for me on my Atom PC.

Mani.

I didn't forgot you,
Will build this week and one for myself too, I still have a six-pack WBT next-gen lying around.
So I will mod my juli@ with a WBT spdif output and use coax instead of Toslink.
I always have used this cable myself, but since the juli@ has no decent RCA connection.
so therefore I stayed on Toslink connection.
I think the change will be pretty dramatic.
Hope you will experience the same and that your next digital cable will be an I2S ;)

Roy

Gotta save my nickels and dimes for that Dac.......

Am a little confused here.....nothing new really. Did more testing this morning and "achieved" the following regarding Device Buffer size, Core appointment schemes and Q1 value of 1,, which was the setting for each of the schemes below:

Scheme   Buffer Setting

1. No Appmt           2048
2. Scheme 1           4096
3. Scheme 2           2048
4. Scheme 3           4096
5. Scheme 4           2048

In establishing these values I used the exact procedure described in the tool tip. Today is the first day I can get a Q1 value of 1 without "too many buffer errors" occurring.  :wacko: Now others, Peter and Mani, set Q1 to 1 with a buffer size of 1024. Tool tip and release notes make it clear that you must establish the correct device buffer for Adaptive Mode to work properly. How can the same device, the Hiface, have a different buffer size depending on the Scheme chosen or depending on the type of processor in the particular computer? It's early here so I can't really test for sound quality between these settings. Just wondering. 

Uhm ... mine is equal to this :


or IOW same as yours ...
(see earlier in this topic)

Oops, you are, of course, correct; at 5 a.m., keeping the facts straight is sometimes like herding cats. So, theoretically, should one of the settings I posted above sound better than the others? Again, its still too early here to turn up the volume.

Just wondering, though, I do understand that, in my case, knowing "just enough" sometimes proves to be dangerious.  :)
 

I guess it doesn't matter much whether it is 5am or 1pm when I wrote my last post, because I got confused with your sequence numbers and Scheme numbers. So, I come up with 2048 at Scheme-3 ! (and not 2 as I said before).

Why can it be different per system ? because it largely depends on the system itself. The ultra low latency requires processor power, and (merely) responsiveness (which is something different). This is why you see the Schemes matter.

What sounds best ? I really don't know in the HiFace case. I would still pick te lowest, but it isn't so low (because the buffer size doesn't seem to be). Now, in that case I'd hop over to Special Mode, because (I know) that can go low. Or at least I used it at 22 samples, and didn't hear strange things during the two tracks I tried it. Keep in mind : for KS Special Mode 22 samples = 22 samples, and the Device Buffer Size does nothing else than allowing you to select the 22 (or not).

Apart from whether you may use a lower latency with Special Mode, I'd me (more) interested in knowing whether it sounds really different from Adaptive Mode (remember, I don't have a decent DAC to use the HiFace with, so I never could/can try).

Peter

In the "Thoughts On Sound Quality"  thread I posted the following under 9y-8:


So far, I'm still preferring Adaptive but its a close call!

Finally, my HiFace is working super too!
The higher buffervalue (2025) makes XXHighEnd work flawlessly on the laptop.
I was not really convinced untill I replaced the rca-to-bnc adapter with a real bnc-plug on the spdif-wire.
Now the BNC-hiFace connects through bnc with Belden 1505 wire to BNC to DAC.
It is a big  difference, before I was missing the depth which is now back together with the detailed highs.
And perfect PRAT!
I guess an adapter close to the source is worse what you can do.

Now I'll wait to see if the ASrock pcie-to-usb3.0 will make the AMD-htpc work too.

Btw, listening to special on y9-8c, sounds better then never before!

GerardA,

Did you compare integrated motherboard USB controller to one of those PCI-to-USB controllers, especially on PCI-Express?
I use Belkin's PCI-to-USB adapter and this way I get much better SQ out of my M-Audio Transit (compared to 4 different motherboards and PSUs). Now I consider upgrading to PCI-Express controller. I wonder if there is any sonic difference...

Marcin,

I have to wait for the card to arrive this week...
Because I only have one slot available it had to be pci-e.
Since USB pcie cards are very rare the prices are high, 50 euro for a Belkin and then it's not real pci-e but pci converted to pci-e.
That's why I decided for a USB3-card for 20 euro. (Alternative would be to buy a new mobo).
I read the controllers on these cards are much better then the mobo-chipsets like the amd-southbridge.
So I hope it will work and sound good.
You make me already very happy with your message that pci sounds better already!
When I have it working I'll report again!

Good review of the Hiface in 6moons: http://www.6moons.com/audioreviews/m2tech/hiface.html

Good review here as well by Mike Galusha.  http://www.enjoythemusic.com/magazine/equipment/0210/m2tech_hiface_usb.htm

Also, for those wanting to take the HiFace to a higher level of performance, John Keny (jkeny alias) who posts on the DIY, CA and this forum as well (!), has modded the HiFace to: (1) pull off i2S to feed directly to his DAC and (2) use 2 lithium batteries to provide the critical 3.3V to the clock and output circuits versus feeding them via the onboard regs which are fed 5V from the USB port.  I'm going to buy one from him to try with my BADAC.

Another interesting experiment would be to directly connect the HiFace to the BNC port of your DAC using a USB cable from your PC/MAC AND compare that to plugging the HiFace directly into the USB port and using a SPDIF cable.  Anybody tried this?

How will the modded HiFace connect to the BADA, how ugly does it look and how much does the modded HiFace cost?


Won't this experiment just vary with your choice of cable, whether USB or SPDIF?

P.S. - I thank boleary and earflappin for the links to the excellent HiFace reviews.  I can't think of a better way to spend $150 on computer audio.

Received my ASRock PCIe to USB 3 card yesterday and after installing, updating the bios and finding new drivers still not working.
So I read the nice small box and it says it only works with ASRock mobo's.
No information about this on the internet or the supplier....
So now I'll try a sharkoon. Good thing is there are a lot of different brands now, although with the same NEC-chip.

About the upgrades by JKeny better read what he has to say on Headfi and diyaudio.
It seems he does not like the sound of I2S-out anymore. Still a better (battery) powersupply can be nice to try.

I'll report again when I have news.

Robin Hood, the modded HiFace will connect to the BADA via S/PDIF BNC (the BADA has a S/PDIF BNC connector in addition to AES/EBU and toslink).  Here's a link to pictures of jkeny's mods: http://theartofsound.net/forum/showthread.php?t=4880&page=5.  If you don't do the i2S mod (I will not being doing that) then there will be no hole in the case).  So basically there is just a white plastic clip out of which the wires go to the lithium batteries.  He uses two batteries - one for the clock and one for the output circuitry.  He charges 50 euros for the mod.

Regarding the cables I'm not sure.  I hope jkeny has experimented and can let us know.  The question is what degrades the signal more - a USB cable or a SPDIF cable.  It will be easy enough to experiment with once I get it.

I found a PCI to USB2 card with NEC chip here for 10 Euro: http://www.vedt.nl/cgi-bin/FS?longshine&Longshine+is+VET+met+DT&Top&shop&Content

And guess what? It works!

So USB from AMD-chipset: No
PCIe to USB3 (Nec) : No
PCI to USB2 (Via) : Yes but distorted.
PCI to USB2 (Nec) : YES!

And it really sounds better then anything before.
With the hiFace from the laptop it did not really sound convincing in the end.
From AMD-htpc there is a huge improvement.
The htpc is W7/tweaked and seperate USBcontroller, laptop is Vista, not tweeked...

Keep it up GerardA and you might just convince my to give up my laptop....something Peter has been unsuccessful at getting me to do!

Now, if you could only convince your spouse, hehe ;)

GerardA,

I told you, that a good quality USB controller matters, especially that your interface uses 5V from USB, no magic.

Yes! And what about a PCIe to PCI adapter?
You already tried it?
Makes a seperate Power supply for the PCI card possible.. ;) ;)

No, I haven't, but now thanks to you it's on my shortlist :D

Anyone here had a chance to listen to a JKeny modded Hiface? How's it sound? Any replies greatly appreciated.

I've had one on order for a few weeks.  Hopefully not too much longer.  I'll let you know.  I'm enjoying a stock HiFace now and in many ways prefer it to my Lynx AES16 card. 

Thanks Danny. Look forward to hearing your impressions of the mod.

Well, there's always the next "best thing". Check this out:
http://www.computeraudiophile.com/content/Halide-Design-Bridge-Review

Excuse me ? USD 450 and only 24/96 ? What would be better about it than your HiFace ?

True, but Chris's impression is that the sound beats the Hiface.

Stock Hiface. That kind of devices (powered from USB) will play differently with each PC. (PSU, motherboard power section, cpu, overall performance, etc not to mention OS and software...)  I think Chris overlooked this factor, which has great impact on SQ, even when it comes to S/PDIF only.

Without degrading anything of the reviews or expressions otherwise, I never understand them (hence which makes it worthless to me).

The last thing you will hear me say is that they are wrong or off per definition, but all there is in another context. It is not wrong, and probably consistent within itself. But different from, say, here.

If it is better, well, it just is. But especially for a device like this I would like to know why.


PS: Like Marcin just gave an example.

I can tell that there is a huge difference between device A connected to system X and device A connected to the same system X, but underclocked, undervolted and highly optimized. I didn't believe until I heard myself and got a jaw drop.  :teasing:

My system is now running at four cores @ 1GHz and 0.825V (default is three cores, 1.325V), 2x2GB DDR3 (1.5V, CL5), no fans (even in a PSU), SSD for win 2008R2 (leightweight 64bit "server edition" of Windows 7), 2x1.5TB eco drives for data, no optical drive. That plus XXHighEnd and some minor os optims is a good way to start computer playback. It's not enough to buy a fancy DAC or soundcard and forget about everything else. Good power equals ~50% of overall SQ (IMHO).

Quite some more even. IOW, totally agreed !

also agreed :)

Just reporting that I got my jkeny-modded HiFace a couple days ago and think it is an appreciable and worthwhile improvement.  There is significantly enhanced clarity which results in a much more spacious presentation of the sound, and there is a lot more “air”.  The music really sparkles!  I haven’t A/B’d this yet, but I’m confident that the modded Hiface will easily trounce me Lynx card. 
I opted for John’s new configuration which houses the battery and the HiFace in a black plastic box with the USB connecter in the rear and the BNC and 3.3V power connecters and on/off switch in the front.  It’s definitely more professional looking than what I envisioned I’d be able to string together with batteries and wires piled up beside my computer.   I have yet to get the final 5% (sonically, per John) working and that’s to supply 5V battery power to the HiFace as well.  John and I are scratching our heads as to why this isn’t working.  I think the modded USB cable he provided is bad, but the ball’s in my court to try John’s next debugging step.  For now, the 95% improvement is certainly good enough for me! 

I think John is working on a new bigger black box that will incorporates the 3.3V and the 5.5V power supply.  That will really be a big improvement from a usability standpoint.  Okay, I’m not the most savvy guy when it comes to DIY electronics, but I had a ton of questions to get this far and since I’m not over the finish line yet, I obviously have several more to get the 5V working.  Though my little black box is a big step up from a hacked HiFace with a battery by it’s side, John won’t be done until all of the stuff is in a single “set it and forget it” box.  I guess my only real question back to John will be if he’s willing to take mine back for a trade-up!   :veryhappy:   It certainly sounds good enough that it’s worth staying the course.
 

Here's an attempt for a small contribution from my side ...

What John did, is cutting the ground from the PC to the output of the HiFace. Thus, the officially needed ground from i2s is not there. This can work, but officially will impeed for jitter because of current flowing where it shouldn't. And, with too much jitter on the i2s lines you will have no sound at all. So, when things seem to be really difficult, I would really try to connect that ground again. Just for trial.
Also notice that removing the ground may be a good idea for when you suffer from ground loops towards the PC, but nothing tells in advance you suffer from that. And if not, it really is better to have the ground there. Also, it may be so that with 3.3V and 5V you have two potentials, and i2s may struggle with that.

For what to reconnect to which you have knock on John's door.

To John : I hope this isn't inappropriate because you told this yourself in the open. But if you think it is, let me know as soon as possible, and this post will be gone.

Anyway, I hope this helps.
Peter

One of the biggest issues the hiFace has is that it doesn't have a wordclock input - it cannot be slaved to the DAC. You can optimise everything else as much as you want, but it's performance will always be compromised as a result.

M2Tech are addressing this with their 'EVO': http://www.computeraudiophile.com/content/M2Tech-Evo-and-DAC-details-and-photo

BUT... this requires a DAC with a wordclock output!

One thing I will never understand about the Berkeley Alpha DAC is the omission of a wordclock output. In the Pacific Microsonics Model One/Two, they went to such great lengths to get the wordclock right. Why on earth didn't they use this knowledge in the design of the Alpha?

Mani.

Mani,

I'm hoping the new interface that Berkeley Audio allegedly has under development will allow the DAC and the PC to have their clocks master-slaved.  In that case, the HiFace, even its modded incarnation, will join my growing pile of obsolete interfaces. But in the meantime it's at least a fun and relatively inexpensive detour along the road to audiophile nirvana.

Dan

I may have experimented a long while ago with having no ground on the I2S tapped from the Hiface but I abandoned this as bad engineering & of no sonic benefit anyway.

This has nothing to do with what Danny is talking about which is a USB cable with the USB +5V cut & an external +5V supplied. I haven't cut the USB ground from the PC - I just spliced into it with the new ground from the external 5V supply. I tested all these cables before I sent them out.

Anyway, as I said the major 95% sonic improvement is what Danny is happily hearing.

BTW, I have synched the local clock from a ESS DAC back to the Hiface & it sounds excellent - this is my next product to release so I think your concern that the modded Hiface is compromised is premature :)

PS. It will be interesting to compare the modified Hiface to the new EVO and I'll be doing so as soon as I can. From what I know of the EVO I expect the modded Hiface to still sound better :) So the range might be: stock Hiface entry level; EVO for more outputs & flexibility - I believe my modded Hiface will offer best sound! 

Why? Because the EVO still has regulators on board :)

Dan,
You're so fickle - you only just got the modded Hiface & you're looking at you next device :) I thought you would be a life-long customer :)

Thanks for your impressions of the product - nice summary.

I'm keeping all this modular so I'm not yet putting it all in one box - I'll be doing an add-on box which will hold the charger, 5V supply & possibly an ADUM USB isolator but I'm not sure it will work with the Hiface. So you don't need to exchange your box for another :)

Hi John,


Assuming you addressed this to me, I didn't say that ! nor do I think it by any means. If you got the impression something like that was in my head, I am very very sorry.
Man, I think the contrary ! (and only tried to help out).

I hope this is settled now ... :)
Peter
 

No Peter, it was not addressed to you but rather it was a reply to Manishandher who said:
The first part of my email was addressed to your comments about me cutting the ground.

So no problems. Peter
 

I'll bite, John. Whats a UDUM USB isolator?

Hi Danny (is this Danny Doxan by any chance?)
It turns out that the ADUM USB isolator will not work with the Hiface according to Marco so it's a non-runner!

What it is, is a way of electrically isolating the PC from the unit connected through USB cable (Hiface). Why? Because there is a possibility that noise & ground issues can get to the Hiface. The Hiface is already isolated on it's output through the SPDIF transformer so this is not really an issue - it was just an experiment really.