Head-Fi.org › Forums › Equipment Forums › Computer Audio › Audiophilleo 1 and 2 USB to S/PDIF transport
New Posts  All Forums:Forum Nav:

Audiophilleo 1 and 2 USB to S/PDIF transport - Page 56

post #826 of 1172

I see that you have the Vaunix. Can you try it first without, and then add it back into the system? Would love to hear your impressions. 

Quote:
Originally Posted by sridhar3 View Post

Welp, just bought the PurePower upgrade and boxed my AP1 to send out tomorrow.  Hopefully the turnaround time isn't as brutal as it was when the upgrade was first introduced.

post #827 of 1172
Quote:
Originally Posted by Rdr. Seraphim View Post

I see that you have the Vaunix. Can you try it first without, and then add it back into the system? Would love to hear your impressions. 

 

Sure, I don't see why not.  I can drop in and give some brief insights, but you might have to wait a bit if you're looking for detailed impressions though.  Time is short for me until mid-August.

post #828 of 1172

I am planning to order the Vaunix and run it from batteries - only thing is whether to use a voltage regulator between battery and hub as according to Vaunix noise is lowest when the unit is fed 12V.  What becomes a problem though is that if one i using a linear regulator one should chose a battery with over 12V output (say 18V) so this is not so straightforward.

post #829 of 1172

First impressions are fine. Details can come "whenever." Unless, of course, after hearing the difference you can't help yourself! wink_face.gif

Quote:
Originally Posted by sridhar3 View Post

 

Sure, I don't see why not.  I can drop in and give some brief insights, but you might have to wait a bit if you're looking for detailed impressions though.  Time is short for me until mid-August.

post #830 of 1172

Uh, that sounds serious! 

Quote:
Originally Posted by drez View Post

I am planning to order the Vaunix and run it from batteries - only thing is whether to use a voltage regulator between battery and hub as according to Vaunix noise is lowest when the unit is fed 12V.  What becomes a problem though is that if one i using a linear regulator one should chose a battery with over 12V output (say 18V) so this is not so straightforward.

post #831 of 1172

Interesting, those of us that use the 2Stepdance and want good clean portable power use the Energizer XP8000 battery pack (puts out 5-20V) + 12 volt regulator.  The also sell 15V and 19V regulators. 

 

http://www.xpalpower.com/us/products/xp8000/

http://www.shop.energizerpowerpacks.com/xpal-power?product_id=73

http://www.shop.energizerpowerpacks.com/xpal-power?product_id=81

 

 

I use the XP18000 for charging my laptops on flights.  Never leave home without it.

 

Quote:
Originally Posted by drez View Post

I am planning to order the Vaunix and run it from batteries - only thing is whether to use a voltage regulator between battery and hub as according to Vaunix noise is lowest when the unit is fed 12V.  What becomes a problem though is that if one i using a linear regulator one should chose a battery with over 12V output (say 18V) so this is not so straightforward.


Edited by WNBC - 7/16/12 at 6:10pm
post #832 of 1172
Quote:
Originally Posted by Sid-Fi View Post

Compared to my setup sounding awful while the AP2 was away getting modded, it sounds glorious right now.
I'll have to post my AQVOX in the for sale forums tomorrow and save some lucky chap an arm and a leg in shipping from Germany. It cost me $75 to ship the dumb thing from Germany this last Christmas :0.

 

What did you use as a source while waiting to get the unit back and pm me regarding the aqvox if still for sale.


Thanks!

post #833 of 1172
Quote:
Originally Posted by WNBC View Post

Interesting, those of us that use the 2Stepdance and want good clean portable power use the Energizer XP8000 battery pack (puts out 5-20V) + 12 volt regulator.  The also sell 15V and 19V regulators. 

 

http://www.xpalpower.com/us/products/xp8000/

http://www.shop.energizerpowerpacks.com/xpal-power?product_id=73

http://www.shop.energizerpowerpacks.com/xpal-power?product_id=81

 

 

I use the XP18000 for charging my laptops on flights.  Never leave home without it.

 

 

Thanks for the links, that battery pack looks pretty sweet all-in-one solution.  By the time you buy battery, protection circuit, voltage regulator, power adapter for charging etc the XPAL looks pretty reasonable.  Does the "willy cable" step down the voltage?

post #834 of 1172
Quote:
Originally Posted by drez View Post

Thanks for the links, that battery pack looks pretty sweet all-in-one solution.  By the time you buy battery, protection circuit, voltage regulator, power adapter for charging etc the XPAL looks pretty reasonable.  Does the "willy cable" step down the voltage?

I use this to power my JH3A during playback (I have never ran out of since).
post #835 of 1172

My understanding is yes, it does step down the voltage.  I have never measured the voltage out of the Willy Cable so not sure what the +/- voltage that is allowable.  

 

Zilch0md has a nice write-up.  I'm using his configuration.

http://www.head-fi.org/t/507835/meier-audio-stepdance-and-2stepdance-discussion-and-impressions-thread/1155

 

 

Quote:
Originally Posted by drez View Post

Thanks for the links, that battery pack looks pretty sweet all-in-one solution.  By the time you buy battery, protection circuit, voltage regulator, power adapter for charging etc the XPAL looks pretty reasonable.  Does the "willy cable" step down the voltage?

post #836 of 1172

First off, the PurePower IS just a "battery with extras"; it taps off its charging power from the USB line but doesn't do anything interesting to the data (you'll note that the USB data can go straight to the AP via a separate USB line if you like).

 

Second, I'm not sure if the USB asynch standard itself allows data re-tries but it doesn't matter for audio because, by the time the receiving device were to detect a data error, it would be too late to re-send the data because the whole system operates in real-time. (I don't know if the AP buffers a full USB data frame, but I doubt it. When USB is used for data - like with a hard drive or stick - it doesn't operate in real-time, so it has plenty of time to request and wait for a retransmission if necessary.)

 

As for the signal itself.....

 

The USB data signal is just that... data. On the receiving end of things, unless something is screwed up pretty badly, the data will arrive in one piece. This really isn't, and never was, a big problem - which is why nobody much talks about it. USB data drives work at very much higher data rates than audio, which makes them a lot more likely to occasionally lose a bit or two, which is why they use a mode that supports verifying the data and asking for a retransmission (bulk mode). USB audio, in contrast, operates at a pretty sedate speed, which makes loss of data very unlikely - which is just as well since, because audio operates in real-time, it would be too late by the time an error was detected to retransmit the data anyway. (Some devices may actually include a buffer - and I have no idea if the AP does or not - but it really isn't necessary.) I think it's mostly a made-up problem looking for a solution (which plays better for certain vendors than admitting that a $5 wire will work fine.) Devices like the AP are designed to solve problems with jitter and with power-supply isolation - which don't matter to USB data drives but are a problem with audio systems.

 

On the AP, the power to run the receiver circuitry is taken directly from the USB line - from the computer - and its good enough to get the job done. On the stock AP, the output section runs on isolated power, which the AP generates via a circuit which gets ITS power from the computer power. The normal way of doing this, which is probably how the AP does it, is to have a tiny switching power supply inside which gets its power from the USB input. The switcher runs through a transformer, which isolates its power output from its input, and that isolated output is used to run the output section of the device. This works pretty well, but the isolation isn't perfect, and the output of the switcher itself has some noise (like any power supply). Using a battery eliminates this last bit of noise, and using a relay scheme to charge the battery (like the AP-PP), provides "perfect" isolation - better than the other way.

 

Now, as for the player software being bit-perfect or not....

If you think about it for a few seconds, this has NOTHING whatsoever to do with the AP (or other device) itself - or shouldn't. This is simply a matter of whether the operating system and the player itself choose to deliberately modify the data (music) before sending it out - which many of them do for a variety of reasons. If Windows (or MacOS) decides to fiddle with the bits, there's nothing the interface can do about it. HOWEVER, the driver that goes with the device could make matters worse by failing to be bit-perfect itself, or by forcing settings that "encourage" Windows to change the bits. (The old V-Link had drivers that told Windows XP that it could ONLY accept 96/24; if you played a 44/16 file, this forced Windows to convert it to 96/24 on the way out so it would "match the requirement"; which resulted in the overall chain NOT being bit-perfect. As far as I know this was an interaction between some versions of Win XP and the driver, and may not  happen with Windows 7.) The AP-1 has a test mode which verifies bit-perfect operation, and is pretty handy since it's pretty well impossible to know for sure unless you have some sort of actual way to verify it.

 

So the short upshot of this is that the battery option may make the AP sound better because it provides better isolation and cleaner isolated power for the output stage (which is rather like what Philip has always claimed :) ). HOWEVER, the difference will be slight, and will depend on how sensitive your audio system is to that last little bit of remaining noise.... and it will make the USB cable matter even less (the technical facts, and my ears, have always suggested that once you go to a good asynch USB implementation it already doesn't matter anyway). 

post #837 of 1172

Something interesting I pulled up on the original HiFace is that it used bulk mode transfer - ie that it can resend data that fails to be transmitted over the USB cable, what is also interesting is that I found the Hiface sensitive to computer side variables apart from bit perfection.  If one were to go purely from theory one would not expect to hear the difference between a Hiface and an Audiophilleo purely on the jitter numbers, as even though they are vastly different the distortion they produce would probably be classified as below the threshold of human hearing, but that is another matter altogether.  I feel I am drifting off topic here but to try to bring my thoughts back into focus - in my experience there is not a very close fit between what engineers expect and what people hear, and an even less close fit between what people perceive and what they hear but that also is another issue altogether.  My third concern from the start is the way in which jitter is measured especially with the AP2 in that is it purely a measure of the clocks and SPDIF output, or is it a measurement of how well the Audiphilleo can take audio data stream coming from a computer and turn it into an SPDIF signal?  

 

What puzzles me in particular is why M2Tech abandoned bulk mode transfer for the Hiface 2 in preference for the asynchronous XMOS system, especially when other manufacturers are developing ever more expensive devices based on the Hiface 1 USB receiver system?  For me I truly hope the Audiophilleo 2 with purepower is the be-all-and-end-all of transports (as I now own one) but I can't help but feel that we are not reading the whole story here.  Not to worry anyway as the system sounds good to my ears for now and delivered the improvments I was hoping for in terms of timing accuracy and clean transients.

 

Just tonight I have tried an audio tweak suggested on another forum to do with CPU scheduling granularity which apparently reduced the time the CPU can direct to any one task.  I tried the setting, compared against the two settings windows generated, and very quickly realised that the windows setting giving priority to background tasks produced the result I considered most accurate, while the value suggested in the tweak was now what I would consider Ideal in my system and to my preferences (too much warm tonal glow around musical notes).  Hopefully the differences with the Vaunix will be just as clear cut.  this is not a judgement I would have expected purely from looking at the values (tweak value had smaller variable which might expect to be more analytical but this is not the case)

 

On the Vaunix though I just had a thought - I wonder if it works by galvanically isolating the power and ground while letting the digital lines straight through?  If not I might try this tweak if the Vaunix is not to my taste.


Edited by drez - 7/18/12 at 8:19am
post #838 of 1172

About "Willy Cables" and other interesting gadgets.....

 

I have to say I find this thread amusing - to say the least.

 

Yikes! We're here discussing whether a battery pack with relay isolation will sound better than a really well-designed isolated power supply module, and you're talking about using a consumer battery pack that puts out multiple voltages (using who-knows what type of regulation). You do realize that most consumer gear uses nasty, noisy, and not at all well designed switching supplies to do this sort of thing, right? Now, I could be giving it a bum rap, and the Energizer battery pack could be the exception, but I know which side of the bet I'd be taking....... Alkaline batteries also don't have impedance characteristics that are nearly as good as Li-Ion or NimH or lead acid types, so it starts at a serious disadvantage. The whole idea of using a battery is to either use the battery DIRECTLY (if the voltage agrees), or with a really good regulator...... I would trust the excellent isolated supply in the AP a lot more than some cheapo consumer battery pack (especially one not even claimed to be designed for audio applications).

"Battery" is NOT some sort of "magic word" that automatically helps whatever you connect it to........
 

DO bear in mind that if you use a "splitter USB cable" so you can use a different power source to power your AP (or other USB converter), you are only changing the plus power supply; you are NOT making it galvanically isolated (because the ground is the same - and MUST remain connected). Since the AP THEN creates its own galvanic isolation later anyway, odds are that you're not going to produce any difference whatsoever. (What you're trying to do is to re-invent the wheel - and do a better job than Philip did - which is doubtful :))

 

If what you're looking for is galvanic isolation, and you don't want the PP option, (or are using some other sort of USB converter) you can buy a "USB isolator" which will do that for you. Here's one with pretty good specs...  (NO, I haven't tried it).

http://www.electronics-shop.dk/?id=1038&username=&currency=USD&pollanswer=1

NOTE: This and most USB isolators support "full speed" and NOT "high speed" modes (of course, full speed mode is 12 mbps and USB audio tops out at about half of that, but, since jitter isn't specified, there could be issues - especially with non-asynch converters and inputs - although it's hard to imagine how it could be worse than a typical PC.)

 

Here's what I would suggest (from an engineering point of view):

 

Get your galvanic isolation (of the ground) FIRST - with this device or something similar.... (and see if you hear a difference).

If that goes well, THEN use a splitter cable (downstream) and add a battery to supply the "USB +5" power to the DAC's input section. (And use a decent battery - with a good low noise regulator if necessary.)

 

PC OUT ------> GALVANIC ISOLATOR ------->SPLITTER CABLE WITH BATTERY +5 ---------> DAC OR CONVERTER INPUT

 

 

I would expect this to help any converter that doesn't already have its own galvanic isolation....

(But, since the AP already has that, I wouldn't expect to hear any difference :) )

post #839 of 1172

Bulk transfer mode is not real time; therefore it requires a buffer.

 

From an engineering standpoint, you are sort of conflating issues here.... There ARE only two characteristics of an S/PDIF audio stream - the numbers and the timing. So, unless the numbers are actually incorrect, the only factor remaining is the timing (which includes the possibility of actual speed errors and of jitter). All clocks these days are so good in terms of basic speed accuracy that, even with the cheapest one, that is a non-issue. Therefore, with the AP, or anything else, the measure of how clean the S/PDIF is in terms of jitter actually IS the total measure of how good a job it's doing (unless it introduces actual data errors- which isn't going to happen). The only remaining thing is the actual measure of galvanic isolation - which is very difficult to measure in terms of audio effects, and whose effects vary depending on your other equipment.

 

From a marketing point of view, I suspect that "asynch USB" is considered by so many people to be "the ultimate solution" that the HiFace simply abandoned their buffer-based circuit for the one that was more popular rather than fight the crowd. I also suspect that, possibly, their original model didn't support 192/24. With the new receiver chips, you get that thrown in as part of the package. So, from a product point of view, they moved to the features they needed to offer the easiest way - by switching to a new asynch-USB receiver chip.

 

As for "galvanic isolation" - ALL signals must be referenced to something (even each other). In principal, the differential signal in a USB cable is independent of ground. In practice, you can't just float the receiver, and so it has to be referenced to ground somewhere.

 

We should, however, be clear on our expectations. The claimed "direct effect on sound" of jitter is a vague blurring of the soundstage - which could also possibly be interpreted as "lack of space around instruments" or your "vague glow". The effect of LACK of galvanic isolation is going to be background noise (we're talking about the situation where you hear a murmur through the speakers when your Windows screen changes). This could also, in turn, theoretically cause jitter at the receiver by dirtying up the signal.

 

NONE of these things are going to directly change the frequency response, nor make the sound "warmer" or "brighter"......

At most, they will produce jitter or other noise that we wrongly interpret as such (just like hiss can make things SEEM like they have more high end).

 

We should also remember that many devices themselves don't have especially low jitter. Some of the differences we hear are probably just the result of the remaining jitter (what IS the jitter of the old and new HiFaces?) There are also different types of jitter, so it makes sense that, IF the jitter is high enough to hear, then the different types may sound different.

(Of course the solution is to remove all audible jitter.)

 

I have an AP1, and I will probably be sending it out for its PP conversion shortly.

Honestly, I don't expect to hear a difference but, as you say, it is "the be-all and end-all" and that way I don't have to

waste time fiddling with other battery solutions - or wondering if I should have.

 

But, just to be clear, once and for all......

IF the AP (or anything else) is "doing its job" 100.0%, there should be no difference whatsoever no matter what you do at the computer or cable as long as the input remains bit-perfect.

IF there were two perfect such devices, or two perfect interconnects for that matter, then they WOULD sound identical.

IF you can hear differences with computer changes, or can REALLY hear cable differences with the AP, then those differences define how far it is from perfect.

 

[Now I'll backtrack and add a kudo..... if you don't have some way to verify bit-perfect, then asynch-USB is NOT a guarantee that your data IS bit-perfect. Even though asynch USB data is clocked by the receiver, it will "glitch"  if the computer just plain drops or changes a sample. If that happens enough then, of course, things will sound different.]

 

I suspect, though, that a lot of what people "hear" is due to more prosaic issues. As I mentioned in another post, I used to use a V-Link (the original). I was playing 44/16 files on a laptop running Win XP Pro set not to upsample and I was rather impressed when adding the V-Link "made the audio sound better". The funny thing was that, as it turns out, the V-Link driver told Windows it only accepted 96/24, so Windows was automagically upsampling to 96/24. That was, of course, why it sounded different..... but there was no "overt indication" that it was happening. (Luckily my DAC happens to show the incoming sample rate in the window..... )

 

I honestly don't think that most people understand "controlling for one variable", nor are they willing to bother (and I'm usually not either).... but this makes any claims by anyone of subtle differences somewhat dubious.... and I cheerfully include myself in that group. I can say, though, that I'm not going to even bother unless there's legitimate engineering justification for hoping for a change...... There's simply too much snake oil and "magic" out there to taste or try all of it and, in the end, I've never heard something that sounded better where there failed to be a legitimate engineering explanation for the difference lurking about somewhere.

 

 

Quote:
Originally Posted by drez View Post

Something interesting I pulled up on the original HiFace is that it used bulk mode transfer - ie that it can resend data that fails to be transmitted over the USB cable, what is also interesting is that I found the Hiface sensitive to computer side variables apart from bit perfection.  If one were to go purely from theory one would not expect to hear the difference between a Hiface and an Audiophilleo purely on the jitter numbers, as even though they are vastly different the distortion they produce would probably be classified as below the threshold of human hearing, but that is another matter altogether.  I feel I am drifting off topic here but to try to bring my thoughts back into focus - in my experience there is not a very close fit between what engineers expect and what people hear, and an even less close fit between what people perceive and what they hear but that also is another issue altogether.  My third concern from the start is the way in which jitter is measured especially with the AP2 in that is it purely a measure of the clocks and SPDIF output, or is it a measurement of how well the Audiphilleo can take audio data stream coming from a computer and turn it into an SPDIF signal?  

 

What puzzles me in particular is why M2Tech abandoned bulk mode transfer for the Hiface 2 in preference for the asynchronous XMOS system, especially when other manufacturers are developing ever more expensive devices based on the Hiface 1 USB receiver system?  For me I truly hope the Audiophilleo 2 with purepower is the be-all-and-end-all of transports (as I now own one) but I can't help but feel that we are not reading the whole story here.  Not to worry anyway as the system sounds good to my ears for now and delivered the improvments I was hoping for in terms of timing accuracy and clean transients.

 

Just tonight I have tried an audio tweak suggested on another forum to do with CPU scheduling granularity which apparently reduced the time the CPU can direct to any one task.  I tried the setting, compared against the two settings windows generated, and very quickly realised that the windows setting giving priority to background tasks produced the result I considered most accurate, while the value suggested in the tweak was now what I would consider Ideal in my system and to my preferences (too much warm tonal glow around musical notes).  Hopefully the differences with the Vaunix will be just as clear cut.  this is not a judgement I would have expected purely from looking at the values (tweak value had smaller variable which might expect to be more analytical but this is not the case)

 

On the Vaunix though I just had a thought - I wonder if it works by galvanically isolating the power and ground while letting the digital lines straight through?  If not I might try this tweak if the Vaunix is not to my taste.

post #840 of 1172
Quote:
Originally Posted by drez View Post

What puzzles me in particular is why M2Tech abandoned bulk mode transfer for the Hiface 2 in preference for the asynchronous XMOS system, especially when other manufacturers are developing ever more expensive devices based on the Hiface 1 USB receiver system?  For me I truly hope the Audiophilleo 2 with purepower is the be-all-and-end-all of transports (as I now own one) but I can't help but feel that we are not reading the whole story here.  Not to worry anyway as the system sounds good to my ears for now and delivered the improvments I was hoping for in terms of timing accuracy and clean transients.

 

On the Vaunix though I just had a thought - I wonder if it works by galvanically isolating the power and ground while letting the digital lines straight through?  If not I might try this tweak if the Vaunix is not to my taste.

There are a couple of reasons for using XMOS. One big one is Linux support. The XMOS supports native Class 2 audio, which means no OSX drivers to worry about, and native Linux support as well. The M2Tech interface doesn't support class 2. Second, the original Hiface was honestly a mess. It's design was pretty textbook on how not to design a converter. John Kenny was able to get it to perform at the level of the Audiophilleo by throwing most of the parts away, and cutting off the USB power supply entirely. 

 

The reason why the PP is effective is that USB power is garbage. DC ripple is in the hundreds of mV, and precise clocks don't like garbage power. The original MK1 Hiface mod replaced only the power feeding the clocks with a battery, the S/Pdif output stage was still fed by the USB bus and the internal regulators. Just by giving the clocks nice clean DC as opposed to crap, the performance improved considerably. The more you can get away from computer supplied USB power, the better.

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Computer Audio
Head-Fi.org › Forums › Equipment Forums › Computer Audio › Audiophilleo 1 and 2 USB to S/PDIF transport