or Connect
Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › USB to SPDIF converters shoot-out : EMU 0404 USB vs. Musiland Monitor 01 USD vs. Teralink-x vs. M2Tech hiFace
New Posts  All Forums:Forum Nav:

USB to SPDIF converters shoot-out : EMU 0404 USB vs. Musiland Monitor 01 USD vs. Teralink-x vs. M2Tech hiFace - Page 92  

post #1366 of 1712
Thread Starter 

@ upstateguy,

Are you sure there is a 3db difference in your system? Did you play pure test tones to level match the transports?

I did some RMAA measurement a while back ago with the Hiface vs. Musiland driving 2 different DACs and at no time there was any measured the difference either in the frequency response nor the loudness level. Whatever the usb to spdif converter, I had the same measurements at the analog output of the DAC in the following things: Frequency response, THD and volume level.

By the way, an increase about 3db is very big and it will make many compressed recordings (see loundness war) horribly clip. (Digital clipping is very nasty sounding).

The reason the Hiface sounds louder in comparison to other converters (as I indicated in my review) is because the sound is a lot clearer, the individual notes are more delineated and transients are snappier. The sound is more focused in the time domain which makes instruments and voices seem louder in big dynamic changes.

 

@ haloxt,

Did you try Foobar v1.0 vs cPlay/CMP?

I agree with your findings that cPlay/CMP has a more natural top end than Foobar 0.9.x. But I have found that what whatever "bug" gave the 0.9.x foobar grainy highs has been corrected in the newer version.

post #1367 of 1712

I haven't compared them directly with my main equipment, but have tried 1.0 for several hours. Sounded fine, maybe more balanced across the spectrum, but judging media player differences is not something I can do well unless it's something like comparing foobar 0.9 asio which I can tell by its fatiguing sound :p.

post #1368 of 1712

I did an indepth comparison of the BNC  Hiface , PCI EMU0404, and TerralInX1,  all veified bitperfect.   Frankly the stock Hiface is a bit harsh.  I prefer the Terralinx1.   But the stock Hiface IMO i suspect is like a stock Buick grand National,  decent 15 second quarter mile times but with a $5 needlevalve mod it turns into a sub 13 second supercar.   So off to buy batteries.   The design is kind of flawed running master clocks off USB power.  Isochrous is better powering of USB I think.

post #1369 of 1712


 

Quote:
Originally Posted by slim.a View Post

@ upstateguy,

Are you sure there is a 3db difference in your system? Did you play pure test tones to level match the transports?

I did some RMAA measurement a while back ago with the Hiface vs. Musiland driving 2 different DACs and at no time there was any measured the difference either in the frequency response nor the loudness level. Whatever the usb to spdif converter, I had the same measurements at the analog output of the DAC in the following things: Frequency response, THD and volume level.

By the way, an increase about 3db is very big and it will make many compressed recordings (see loundness war) horribly clip. (Digital clipping is very nasty sounding).

The reason the Hiface sounds louder in comparison to other converters (as I indicated in my review) is because the sound is a lot clearer, the individual notes are more delineated and transients are snappier. The sound is more focused in the time domain which makes instruments and voices seem louder in big dynamic changes.

 

@ haloxt,

Did you try Foobar v1.0 vs cPlay/CMP?

I agree with your findings that cPlay/CMP has a more natural top end than Foobar 0.9.x. But I have found that what whatever "bug" gave the 0.9.x foobar grainy highs has been corrected in the newer version.


Hiya slim

 

I know 3db is pretty big, but it wasn't my number.  I picked it up from the Legato Vs Hiface Vs Hagman  review on Computer Audiophile. 

 

My observation was that the HiFace plays louder than either my Blue Circle Thingee and the native USB implementation in my Constantine+.  I have yet to try it with my Stello, but IIRC, the Thingee plays at the same volume as optical out of my music computer (when they both were connected to the Stello).

 

USG

post #1370 of 1712


 

Quote:
Originally Posted by regal View Post

I did an indepth comparison of the BNC  Hiface , PCI EMU0404, and TerralInX1,  all veified bitperfect.   Frankly the stock Hiface is a bit harsh.  I prefer the Terralinx1.   But the stock Hiface IMO i suspect is like a stock Buick grand National,  decent 15 second quarter mile times but with a $5 needlevalve mod it turns into a sub 13 second supercar.   So off to buy batteries.   The design is kind of flawed running master clocks off USB power.  Isochrous is better powering of USB I think.


Have you solved the "charger" problem yet?  Which charger and where to get it?

 

USG

 

Edit to add this:

 

Quote:
Originally Posted by haloxt View Post

I haven't compared them directly with my main equipment, but have tried 1.0 for several hours. Sounded fine, maybe more balanced across the spectrum, but judging media player differences is not something I can do well unless it's something like comparing foobar 0.9 asio which I can tell by its fatiguing sound :p.

 

halo, were you able to find a src for the new foobar?


Edited by upstateguy - 5/22/10 at 9:59am
post #1371 of 1712


See below:
 

Quote:
Originally Posted by regal View Post

I did an indepth comparison of the BNC  Hiface , PCI EMU0404, and TerralInX1,  all veified bitperfect.   Frankly the stock Hiface is a bit harsh.  I prefer the Terralinx1.   But the stock Hiface IMO i suspect is like a stock Buick grand National,  decent 15 second quarter mile times but with a $5 needlevalve mod it turns into a sub 13 second supercar.   So off to buy batteries.   The design is kind of flawed running master clocks off USB power.  Isochrous is better powering of USB I think.


What DAC are you using?

 

The DAC is more likely the cause of harshness, not the converter.  Once the jitter is low, many DACs show their real clothes, which is sometimes unpleasant.

 

In general, if you are hearing more detail, this is the right direction to go in.  If it is becoming harsh, I would recommend to look in other areas of your system for the weakness.  This type of interaction is fairly common.  High levels of jitter can often mask bad components, and can even make the playback sound more "smooth".  It's not real though.  You have to be smart enough to see through this smokescreen.  Although, sometimes "Hardness" or "Sterileness" can occur due to deficiency in the clocks being used by the converter.  In this case, go with another converter or change the clock.

 

Steve N.

Empirical Audio


 

post #1372 of 1712
Quote:
Originally Posted by upstateguy View Post

halo, were you able to find a src for the new foobar?


There's an option for it in Preferences>Playback>DSP Manager, called resampler.

post #1373 of 1712
Quote:
Originally Posted by haloxt View Post


There's an option for it in Preferences>Playback>DSP Manager, called resampler.


That's not secret rabbit code though is it?

post #1374 of 1712



Same DAC SlimA was using and the next model up.  I've heard jitter added files (even spectrum) and know what you are rfering to,  this is different.  I'm sure the Jitter spectrum of the Hiface is concentrated,  it may be low overall but spikes are more audible.   I'm not evaluating it any further until I fix the design.

 

I think it is pretty conclusive thoughout this thread that Isosynchrous lin regulated powered USB is superior to USB SM powered asynchrous .   When you think about how it works it makes perfect sense.

 

 

 

Quote:
Originally Posted by audioengr View Post


See below:
 


What DAC are you using?

 

The DAC is more likely the cause of harshness, not the converter.  Once the jitter is low, many DACs show their real clothes, which is sometimes unpleasant.

 

In general, if you are hearing more detail, this is the right direction to go in.  If it is becoming harsh, I would recommend to look in other areas of your system for the weakness.  This type of interaction is fairly common.  High levels of jitter can often mask bad components, and can even make the playback sound more "smooth".  It's not real though.  You have to be smart enough to see through this smokescreen.  Although, sometimes "Hardness" or "Sterileness" can occur due to deficiency in the clocks being used by the converter.  In this case, go with another converter or change the clock.

 

Steve N.

Empirical Audio


 

post #1375 of 1712
Thread Starter 
Quote:
Originally Posted by regal View Post



Same DAC SlimA was using and the next model up.  I've heard jitter added files (even spectrum) and know what you are rfering to,  this is different.  I'm sure the Jitter spectrum of the Hiface is concentrated,  it may be low overall but spikes are more audible.   I'm not evaluating it any further until I fix the design.

 

I think it is pretty conclusive thoughout this thread that Isosynchrous lin regulated powered USB is superior to USB SM powered asynchrous .   When you think about how it works it makes perfect sense.

 

 

 


 


I am sorry to say this regal but your findings were kind of predictible. Even before listening to the hiface, you said many times that you didn't think that async usb devices could sound good if they drew their power from the usb port.

You didn't reach your final conclusion after long term and objective listening, but you pre-conditioned your ears to hear a "harsh sounding" Hiface weeks before getting your hands on it. In audio, no one is exempt from the placebo effect. That is why it is important to listen for days or weeks in different states of minds before making definitive judgement on a component. So could you be honest and tell us for how long you listened to the Hiface?

Also, did you notice any burn-in with the Teralink X? If the answer is yes, did you allow the hiface to "mature" as much as you did with the Teralink X?

 

The intent of this post is not to be harsh or offensive. I just find it weird that you draw the following conclusion "Isosynchrous lin regulated powered USB is superior to USB SM powered asynchrous" from a limited comparison of one single isochronous usb device vs. one single asynchronus usb device. You didn't a sample big enough to draw conclusions from a listening perspective and the reasoning is flawed from a designing point of view.

Yes, usb power is bad ... but it can be corrected/regulated and if it is bad in the first place it is because the whole power in the computer is bad.

So how can isochronous clocking from a poor clock of a computer be good while the async be bad.

The clocks of the hiface, even when they are powered from the regulated usb, are better than an adaptive usb chip such the CM-108 used in the Teralink X. If you had ever seen the jitter measurements of the CM-108 (which I posted before), you would have seen how bad it is.

 

Have a look at the m2tech Hiface thread, and you will see that there is dozens of people who find the hiface smooth sounding. They find it an improvement over different soundcards (including professional ones with true audio clocks), CD players and even some good usb implementations such as the one in the Lavry DA11.

Saying that you use the same DAC as me is a bit misleading. First of all, I am currently using a slighlty different/newer version (the DSP). But most importantly, there is more to it than just the DAC: if you ask Kingwa (the one who designed the DAC for those who don't know) he will tell that the choice of the digital cable, power cable and power filter and interconnects are essential to the sound of his DACs. I remember that for one his transports (CD7) he clearly stated it should be used exclusively with power filters (i.e they shouldn't be plugged directly to the wall outlet).

Personally, I use 2 different power filters: one for the notebook (which uses SMPS of course) and another for the DAC and AMP (which have linear power supplies). I am also using an Essential Audio Tools Noise Eater (parallel filter) to reduce the contamination between components. I am also using heavily shielded power cords (with high capacitance) to avoid any EMI/RFI contamination.

We have a lot of "noisy" equipments either in our audio system or the rest of the house. If we don't take proper care to filtrate and isolate the components from each other, we are not hearing the equipment at the full potential.

So yes, we might be using similar DACs but it doesn't mean they perform the same way.

I invite you to read the following paper: http://www.nordost.com/downloads/New%20Approaches%20To%20Audio%20Measurement.pdf They measured the output of a CD player with "stock" power cords and feets and then measured the output of the same CD player after using an "audiophile" power cord and support and they found an improvement when using music signal (and not test tones). These are some very new findings (the measurement method is not very simple to do) but confirm what "audiophools" have been saying all along...

post #1376 of 1712

No offense taken,  you have defended the hiface for 92 pages.  And back on page one you were using the same DAC I am.   I know my system well.  Right around 3-4khz the hiface has an issue. No big deal. 

If I still don't like it I'll sell it, before I spend $200 on a cable.   Hell it was hard enough for me to spend $30 on a 75ohm BNC-BNC Mogami 2964.  I wonder somtimes why you have defended a 3" plastic 1 ounce transport so fervently,  I think sometimes you may be suffering placebo from reading the Hiface White Paper,  no offense meant.  Isn't psychoacoustics fascinating?   Charts, graphs, measurements don't mean squat when it comes to human perception.  We are dealing with one of the 5 senses.  You don't see people arguing over which vintage wine tastes better with graphs and measurements.  Same thing here just a different sense. 

post #1377 of 1712
Thread Starter 
Quote:
Originally Posted by regal View Post

No offense taken,  you have defended the hiface for 92 pages.  And back on page one you were using the same DAC I am.   I know my system well.  Right around 3-4khz the hiface has an issue. No big deal. 

If I still don't like it I'll sell it, before I spend $200 on a cable.   Hell it was hard enough for me to spend $30 on a 75ohm BNC-BNC Mogami 2964.  I wonder somtimes why you have defended a 3" plastic 1 ounce transport so fervently,  I think sometimes you may be suffering placebo from reading the Hiface White Paper,  no offense meant.  Isn't psychoacoustics fascinating?   Charts, graphs, measurements don't mean squat when it comes to human perception.  We are dealing with one of the 5 senses.  You don't see people arguing over which vintage wine tastes better with graphs and measurements.  Same thing here just a different sense. 


regal,

 

Sorry to repeat this, but you didn't answer some of the questions I asked:

- how long did you listen to the hiface?

- did you notice any burn-in with the Teralink X? If the answer is yes, did you allow the hiface to "mature" as much as you did with the Teralink X?

- didn't you already imply the following (Isosynchrous lin regulated powered USB is superior to USB SM powered asynchrous) before even listening to the hiface?

 

I don't say, I am immune to placebo (nobody is). However, my testing methodology included long term listening including different DACs with different technologies (discrete R2R, opamp based sigma delta...). 


Again, I am not saying the Hiface is perfect (as jkeny's mod proves it can be further improved). However, in comparison to a Teralink X, the Hiface is a big jump in sound quality. There is simply more details and more of everything. I have done the comparison countless of times and achieved the same results. Frankly, I don't care about specs and white papers like you said. When I was the first to say that the adaptive CM-108 Teralink X sounded better than the async Musiland I talked about what I heard and not what about the supposed superiority of the Musiland. Later on, I found out some technical facts about the poor building quality of the Musiland that explained what I heard. Again, when I found out the Hiface to sound better than my other converters, I spoke my mind. Then later on, the only measurements we had about the hiface showed that it was on the same level as the Pro PCI LynxTwo card and better than some CD players and Nagra recorders.

After that a limited number of people complained about the sound of the Hiface (vs. the majority who loved their hiface). So you (regal) were among the first here to hypothesize that there must be some weird thing going on with the jitter. You first said that the 1ns jitter figure was a poor figure because it was power from the computer. Later on, when I showed you that the 1ns (in the digital domain) was an excellent figure, you said that there must be something wrong with the jitter spectrum. At the time, you didn't even liste to the Hiface in your system (if my memory is correct). So, because you had the idea that an async device that was powered from the usb port HAD to sound bad. So, you tried to find an "excuse" for the hiface to sound bad (hence your theory on a suspected jitter spectrum).

The weird thing is that your findings now that you have listened to the hiface match exactly the suspicion and preconceptions you had before listening to it. I am not saying it is impossible to draw conclusions from jitter spectrum but you have to fullfil 2 conditions:

1. you have to actually measure the device being commented on

2. you have to actually have the skills and understanding to comment on such a graph (only a handful of people might be able to do that like Dan Lavry or Steve Nugent).

 

So here is a thought: who is more affected by placebo: someone who used the unit for months or someone that knew it was going to sound bad before listening to it?

 

Also, I have to partilally disagree with the following statement: "Charts, graphs, measurements don't mean squat when it comes to human perception".
I agree that looking at a limited or wrong bunch of charts, graphs and measurements, you can draw the wrong conclusions. However, saying that they are totally meaningless is pure BS.

Of course that we have to trust our ears, but if we understand what graphs to look for and what measurements have to be done, it is even better as it is far more general.

Basically, there are 2 areas of performance for audio gear: the frequency domain and the time domain. Here is a simplified example: If you look at the measurement only at the frequency domain, all transports measure the same. If you look at the time domain measurement, you start seeing differences. This is true for everything else in audio: many engineers have been to focused on the frequency domain performance (frequency response, THD...) and didn't realize that the time resolution was more important. This explain why all speaker cables will have a linear frequency response on the 20hz-20khz frequency bandwidth but they would have different measurement on square waves and impulse responses (I believe that Steve Nugent measured some of his speaker cables that way).

But the best way to measure the real performance of a component is through a real complex musical signal. This is exactly what Nordost has done. They have found that a DCS converter obviously follows the signal better than a $250 cd player but they also found out that the performance of the cheap cd player could be further improved by applying "tweaks" (power filtration and vibration control). So in this case these measurements not only make sense but prove that what "audiophools" have found out using only their ears is relevant.


My point is that if there is a lack of relevant measurement we should of course trust our ears. However, if we relevant measurement we should use it in order to better understand where the origins of a problem exist in our system.

Say for example, my chain is composed of only 2 components A1 and B1. I find that my audio chain is harsh sounding. Then I try to look at the available measurements of both component to try to infer which is one is more likely to be the culprit. If there is no measurement, I usually choose the component that shows me the greater difference and variation between different albums being played even if the system doesn't sound right. My point is that the most transparent device will have the most details and will affect the sound less.

Say A1 is the most transparent but the overall sound is harsher than when I use the alternatives A2 and A3. I keep it anyway, and I try to find a different component to replace B1 later on with something more suited with my taste. That way, at the end, when I would have replace B1 with B2 for example, the new set-up (in this example A1+B2) will end up being not only tonally balanced but it will have more resolution than if I had settled on A2+B1 for example.

Unfortunately, a real world audio chain is more complex than that and everything matters and it is difficult to maintain a good balance between resolution and smoothness. It is even more difficult to achieve when people have preconceptions about things and are not willing to experiment with different interconnects, digital cables, power filtration (cords and filters) and resonance control. People who never tackle the power and vibration issue will most probably have a hard time achieving a system that is both high resolution and natural at the same time. Most choices will end up choosing component X over component Y because this one sounds smoother (or whatever else) than the other.

 

(As a side note, I have read that the conceptor of Lamm Industries which is known to make some of the most musical preamps and amps in the world don't even listen to his gear before lauching them into production. His psyachoustics models are good enough that measurements are sufficient to make excellent sounding equipment).

 

So if you read the above or read any of the many other reviews, you will that the hiface represents only a small fraction of my system. I am not stubborn and if I find anything better than the Hiface (at a reasonable price), I won't hesitate to sell the hiface and get the newer thing.

If you don't like the hiface in your system, that is good information for people to know. However, saying something like "I think it is pretty conclusive thoughout this thread that Isosynchrous lin regulated powered USB is superior to USB SM powered asynchrous" just shows how little knowledge and lack of deep understanding you have about usb converters.

 

A few months ago, jkeny was saying the same thing about me, that I was defending the hiface and he thought his modded Musiland sounded probably better. However, I admire jkeny because despite his preconceptions about the hiface, he recognized that it sounded good and chose it as a new base platform for his mods. That is probably also why he did a good job with his mod, as he probably listened with his ears and not with his pre-conceptions. As the audiophile grows up, he starts building a set of beliefs: R2R dacs are better than sigma delta, discrete is better than opamp, tubes are better than transistors for amplification... However, one has to keep their mind open. While it is good to have a few beliefs to avoid purchase mistakes, it is good to listen from time to time without preconceptions (which lead to placebo). While I am firm believer that discrete designs are the way to go, I am also open minded. When I was trying the Audio-gd FUN for example, I ended up favoring the IC LME49720 opamp over the Earth opamp. If I were close minded (like some people), I would have automatically chosen the discrete Earth opamp. On the other hand I prefered the Earth discrete opamp over the IC OPA604. So I am very pragmatic and use the 80/20 rule. You can get 80% of the performance of your system applying logic and the rest by actually trying different things even if they don't fall with the personal beliefs and preconceptions.

 

That is to say that accusing me of relentlessly defending a 3" plastic 1 ounce transport so fervently is a simplistic way to see things. During the 92 pages, you will find a lot intersting data that go beyond the hiface or even usb transports. I have done a lot of research, read many articles and talked about a lot of things. Implying that I have been only defending the hiface during the 92 pages is a little bit reductive and far from the truth. I could be wrong but I don't see many other threads on head-fi discussing theory of jitter, actual listening experiences with different transports, digital filters, upsampling, different medial players... 

post #1378 of 1712

You have to understand that Jitter can't be summed up in one number.  Also there still isn't proof that Jitter is the only difference in transports.   What it all comes down to is perception,  and perception is subjective,  as far as I know science hasn't developed a model of consciousness at least not when I went to engineering school.    I'm going to mod the Hiface,  no one can convice me that it is better than my PCI EMU.  My ears, my brain, my gesalt I prefer the old EMU0404.  I have been listening to Jerry Garcia's guitar for 20+ years and know when it is accurate.  At least with my computer,  who knows maybe I have a bad USB controller from my 3 year old ABit motherboard.

 

What I am amazed is my three computer transports are bit-perfect yet they sound different.   Its fascinating and I think we are just a few months away from some really good low cost commercial computer transports. 


Edited by regal - 5/23/10 at 5:23am
post #1379 of 1712
Thread Starter 
Quote:
Originally Posted by regal View Post

 

What I am amazed is my three computer transports are bit-perfect yet they sound different.   Its fascinating and I think we are just a few months away from some really good low cost commercial computer transports. 

Well, at least we agree on one thing, these little devices have great potential. Good CD transports can be very expensive ($10,000+). It is amazing that we can achieve today such a performance from such little devices.

I personally suspect that an "extreme" mod of something like the hiface (galvanic isolation from the computer, battery power, Superclocks, output stage) can be kept under $1000 and compete with the best CD transports out there.

Jkeny's mod is a very cost effective approach as he adressed the most pressing issue (the power supply of the clocks) with a direct supply from the batteries. It is not the most user friendly device I own (as I have to keep remember to recharge the unit from time to time), but it is worth it.

post #1380 of 1712
Quote:
Originally Posted by slim.a View Post



Well, at least we agree on one thing, these little devices have great potential. Good CD transports can be very expensive ($10,000+). It is amazing that we can achieve today such a performance from such little devices.

I personally suspect that an "extreme" mod of something like the hiface (galvanic isolation from the computer, battery power, Superclocks, output stage) can be kept under $1000 and compete with the best CD transports out there.

Jkeny's mod is a very cost effective approach as he adressed the most pressing issue (the power supply of the clocks) with a direct supply from the batteries. It is not the most user friendly device I own (as I have to keep remember to recharge the unit from time to time), but it is worth it.



Yes, great little devices, I agree!

 

The 'extreme' mod of the Hiface I have tried

- Can't do galvanic isolation 'yet' with existing devices like the ADUM

- I have changed clocks for better, lower jitter ones - not much difference in sound

- I have sent the local DAC clock back to the Hiface in place of one of the existing clocks - yes improvements when using I2S, haven't tried this with SPDIF yet. Some more work to do here though.

 

Sorry to hear it's not user-friendly, slim? Are you using a LiFePO4 battery charger? This can stay plugged in permanently & when the unit is turned off , the batteries are recharged & charger should turn off when full. No need to remember to charge battery. 
 

PS. yes we did have a spat earlier in this thread, didn't we? :)

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Dedicated Source Components
This thread is locked  
Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › USB to SPDIF converters shoot-out : EMU 0404 USB vs. Musiland Monitor 01 USD vs. Teralink-x vs. M2Tech hiFace