Audio-gd Digital Interface - Page 41

Originally Posted by Currawong 

Leeperry: maybe someone with a DI will try one. There is no need to argue so seriously about it. :)

I wasn't being so serious about it ;)

We all know how the Empirical gear is hyped to death(its +50yo manager boasts about hearing 1ps jitter), I'm surprised ADUM4160 wouldn't work at all on it?! Saying that your gear is the best is easy, every manufacturer does it("the Goldpoint Headphone Pro sounds clearly better than most other headphone amplifiers")...hearing what the fuss is all about is another story altogether.

ADUM4160 w/o WM8804 would be missing its elder brother IMHO.

Quote:

Originally Posted by Newk Yuler 

Quote:

I have the diyparadise Ultravox version of the ADUM. It has a painfully obvious degrading affect on an Empirical Audio Off Ramp.

Thanks for sharing with us your personal experience with the ADUM.

I was also suspecting that it might actually degrade the performance of very good converters. Since I noticed relatively big improvements on my worst converter and far less improvement with a "tweaked" one, it is easy to imagine a situation like yours where it actually degrades the performance.

This is similar to reclockers: when using a low end transport, using a recloking device (such as the Ref-3) might help. When using a high-end transport (such as the CD-7, Esoteric transports...), a reclocker will degrade the performance. Kingwa said to me once that putting his Ref-3 reclocker following the CD-7 will sound worse than the CD-7.

Many people in the audio world don't understand between a specified intrinsic jitter of a digital receiver (50ps for the WM8804 for example) and the actual jitter and phase clock noise that passes through the receiver into the DAC.

Believing you can solve all the problems in the digital world by means of galvanic isolation and a "50ps" digital receiver/reclocker is a very simplistic way to view digital audio.

On Digital devices such as the DI and even more so on the Empirical Audio units, there is a REAL understanding of how digital audio works (besides putting together off the shelf chips without any consideration to the layout...).

What you have heard makes sense to me on 2 levels:

1. It makes sense from a technical point of view

2. You reported a personal and subjective listening experience: it should be respected as such and not doubted, especially from someone who has zero experience with your converter (or with the Digital Interface, which is the title of this thread).

Originally Posted by Currawong 

maybe someone with a DI will try one.

to be fair, there's many ADUM4160 dongles available and their effect on the Tenor chip are well known...but the whole idea would be to feed it from the DPS. The current solution to do that is to build a TREAD PSU and this board, quite cumbersome :/

I think a Jameco Linear / Regulated power supply would do a good job...would a TREAD be that much better?

Originally Posted by Dynobot 

I think a Jameco Linear / Regulated power supply would do a good job...would a TREAD be that much better?

You could look up its ripple figure in its datasheet. A properly built TREAD has 0.00x mV ripple AFAICR.

Comparo – Digital Interface and External Power Supply

Packaging

Both Audio-gd’s Digital Interface (DI) and External Power Supply (EPS) arrived in a sturdy plain cardboard box. The trip to AUS had taken five days. Inside the box, the products were protected from shock by a plastic foam cradle. There is no purpose-made packaging nor was there an instruction manual. “Extras” incl. power lead & usb cable.

Build quality

Build quality of DI and EPS are good. The EPS especially has some heft. The hardware looks good: nice gold fittings and functional plastic switches. I am not in a position to measure it, but hopefully the BNC socket is true 75ohm. The omission of rubber feet is a pity since the units are meant to stack (one on top the other). I opened up the DI to look inside; everything is neat and tidy. I have not opened the EPS. Aesthetically, I would have preferred the USB/COAX switch on the front panel (of Version A) to have been on the rear panel instead. One niggle is the stock USB cable seems a bit oversize so that it requires a bit of force to get it in/out.

Format

Why do we need separates? True, it allows some flexibility but it is a nuisance in that additional cabling is required as well as valuable shelf space.  Why not incorporate the DI into the DAC itself? Perhaps Kingwa can answer..

What kind of product is this?

Where does the DI fit in the world of DIs? It is not asynchronous, it does not require specialist drivers; it is a reclocker with DSP.

My set up

My system is listed in my sig. it has some strengths and weaknesses. I am not into cans. Maybe one day.

DAC-19DSP settings: PLL inactive; All other settings: Default

I like a “near field” listening experience. If I go to a concert I want to be in the front row. Same with music playback. I know that PLL inactive brings things forward a little. I have gone to some trouble to push the noise floor down (undervolting + linear PS) on my computer. All these things allow me to hear details. This is my preference.

USB cables (a short diversion): The stock USB cable shipped with the DI is fairly good (sound wise). There are audiophile USB cables out there but this is an area that I have not ventured since I am of the opinion that USB is an inherently flawed interface that is nevertheless convenient. I once read a post by Gordon Rankin (of Wavelength Audio) where he states that the number of errors over USB is proportional to cable length. I have conducted my own listening tests with generic USB cables over various lengths and can confirm his finding: less is more. The best sounding USB cable I have is a humble 12in cable that was shipped with my Musiland. I compared this to the stock 60in DI cable; I found the 12in to sound better. No surprise.

I play music exclusively from music files stored on my hard disk. My music files are created from CD using EAC. I rip one CD to create one .WAV. I don’t use any form of compression because I don’t want my undervolted CPU to do anything it doesn’t have to during playback. My objective is to keep RF as low as possible.

Comparo

I began the comparo with the USB input on the DAC19-DSP (Tenor TE7022 USB input chip). My impression of this interface is that it is not particularly good. It is not dynamic and voices are not distinct. Soundstage is restricted. The general sense I have is that everything is reduced to less than it could be (compared to other, more revealing interfaces). Read on..

My Musiland 01USD has been modified with Vampire 75ohm BNC and linear PS. I run the Musiland with 1.0.8.4 driver (not the latest) which I think sounds best.

The Musiland tends to overemphasise the upper mids to the point of being trebly. I also find the bass isn’t as tight as it should be. There is a sense of detail and dynamics (which is good) but vocals are not weighted properly and backgrounds are not as black as they could be (which is bad). This last point is something I attribute in part to its power supply. The Musiland takes its power from the USB 5V line. As a mod, I have snipped the 5V USB line and power the Musiland with an external linear PS. This has quietened things down a bit.

In sum, the Musiland has some good attributes and some bad. Soundstage is good. However, there is a sense of overemphasis and the feeling that things are a little jigged and “enhanced”. But still and all, for the money, it is pretty good with the mods that I have made.

Empirical Audio (EA) Freeway2 (USD640). My impression of this unit is that it has a very clean and neutral sound that borders slightly on the clinical. To its credit, vocals are distinct and bass is tight. To the contra, emotional involvement can be a bit low. BTW, though the Freeway2 is now discontinued, it is somewhat equivalent (if price is any indicator) to the newer Off-Ramp (USD699). Of course, EA would probably say that the current Off-Ramps have been improved. For this reason, I wanted to include in this review a comparo with an EA Off-Ramp with Superclock (USD999). (Superclock is a USD300 option for the Off-Ramp.)

I found the DI sans external EPS to be a little slow. Whereas the Musiland moves along with tizzed up highs, the DI seems to move along in a bit of a plodding, bass-centric fashion. The music is rhythmic and reasonably dynamic with nice highs and a passable if lazy bass. Also, instruments are not as authentic as the Freeway2. This being said, I think the DI is a satisfying listen which is to say that I don’t think it is inhibited in its ability to convey the essential character of the music. The addition of EPS to the DI improves the sound. Things are tightened up a notch for sure but the unit still lacks speed and authenticity.

In sum, I think the DI is a good listen that is improved by the EPS. It has a good soundstage. However, I think it feigns a “natural” sound by its skew toward the lower registers. I think it is ahead of the Musiland but not at the level of the Freeway2. As confirmation of this finding, I took the DI over to a friend’s place for a comparo with his EA Off-Ramp + Superclock. (His system, BTW, is stellar: Playback Designs DAC, EMM Labs Pre, Pass Labs mono blocks..) The result here was the same except that four ears now confirmed the DI’s shortcomings.

A note: As I say, I am not a “can man”. However, I do have an A-gd DAC-19 + C-2 combo. And, hidden away in the back of my toy-box is a pair of Senn HD490s. (Sorry if the thought of this offends anyone out there!) Anyhow, I did have a listen through these “pseudo-cans” and report that I could no longer hear a difference between DI with/without EPS. Take whatever you want from this.

What was not tested

I have not tested the upsampling function on the DI.

I have not tested the S/PDIF input on the DI.

Issues

USB cable connector is a bit oversize so that it requires a bit of force to get it in/out of the DI.

Tenor TE7022 USB input chip does not support 88.2kHz. This is going to be an issue if, for example, you use 88.2 hi-rez material.

Also, when the computer fires up, the TE7022 sets it self at half volume which is its default setting. For me, this is a nuisance because it means that I have to always go into the Control Panel to reset the level to max (as this is the level I like to run the unit at).

Is the DI bit-perfect? Had I a DTS receiver I would have done a test. But I do not.

Conclusion. Do I keep the DI?

What has this comparo shown? Not very much to be truthful! I have compared the DI to.the USB input of my DAC-19DSP and my modded Musiland and thought the DI was better. I have compared the DI to an EA Freeway2 and Off-Ramp + Superclock and thought the DI lost that contest. At no stage did I compare the DI to something near its price point. Also, I have not experimented with DIP settings nor I2S connection to DAC. So, there may be more performance to wring out of the DI yet! I do think the DI is a good listen and will be keeping it. I just don’t think it is amongst the best devices of its type in the marketplace.

USB polling tweak: http://www.audioasylum.com/cgi/vt.mpl?f=pcaudio&m=78994

That was a nice reviews,thanks for posting your views.Considering we have a similar setup.

The only difference is that I hear from headphones and not speakers from this rig.

Quote:

Originally Posted by Audio Bling 

Comparo – Digital Interface and External Power Supply

 

Mine arrived a few minutes ago....all seems good so far. Just one question - I am using iTunes with a MacBook Pro and CIA VDA-2 DAC; when I switch to 24 bit 96000hz in Audio setup, the sound stops - any ideas what I am doing wrong?  Cheers

Quote:

Originally Posted by Audio Bling 

Comparo – Digital Interface and External Power Supply

 

Very nice write-up Audio Bling!

Since I haven't received the DI yet, I am curious about the following points (if you don't mind):

- Did the unit need any burn-in?

- Did you try the DI + DAC19DSP with the PLL turned ON? Except when using the Jkeny's modded Hiface, I prefer to always leave the PLL on. In fact, even with Jkeny's Hiface turning the PLL off just throws away part of the DAC19DSP coherency (to my ears).

- What drivers/media players are you using? (With the Teralink X2 which uses a similar Tenor chip, I got better results when using Teralink X2 own asio drivers than using regular DS over Win XP, so I am wondering if the drivers are going to be an issue with the DI...).
 

Quote:

Originally Posted by leeperry 

 

You could look up its ripple figure in its datasheet. A properly built TREAD has 0.00x mV ripple AFAICR.

Wow....0.00 is low.

As always how does that relate to human hearing, even the Super hearing capabilities of Audiophiles I don't know....

Originally Posted by Dynobot 

Wow....0.00 is low.

 

As always how does that relate to human hearing, even the Super hearing capabilities of Audiophiles I don't know....

 

I made a thread about it a while back: http://www.head-fi.org/forum/thread/496161/

and the TREAD was measured at 0.060 mVRMS.

This IS a problem with the TE7022, I found that even at comparable volume levels there still seems to be just a tad bit of something missing.

If you connect both SPDIF and USB to your source.  Listen to SPDIF at X-volume level per your pre-amp/amp....then go back and use USB making the necessary source adjustments to get the same volume level.  I could say that the USB seems to have a darker background compared to Spdif.  But there still seems to be a bit of something missing from the upper mids which to me gives the illusion of a darker background.  To me USB seems to be a bit thick sounding compared to Spdif.  Kind of like Cardas cables vs Audioquest.
 

Quote:

Originally Posted by Audio Bling 

Issues

 

Also, when the computer fires up, the TE7022 sets it self at half volume which is its default setting.

Slim.a..

>“Did the unit need any burn-in?”

Yes, but I didn’t find much improvement after, say, four days. Just to be clear, I ran the DI + EPS 24/7 for about 5days then for a number of hours per day thereafter. I did quite a bit of listening over the course of two weeks. It is possible that it may improve further. I will report if it does.

>“Did you try the DI + DAC19DSP with the PLL turned ON?”

Yes, I did. I just seem to lose a degree of detail (and coherency) when I do this. I will try PLL ON again as I always like to re-validate preferences and settings from time to time and particularly after I make a change (as I just have done with the USB polling edit). Regardless, my opinion of the DI will not alter with the PLL setting as I have listened to the DI both ways.

>“What drivers/media players are you using? “

I use a cMP2 box which is to say I use cPlay (latest version). I use Asio4all to both DI and Freeway2. I use Foobar (with Kernal Streaming) from time to time again for the purpose of re-validating my settings. I didn’t listen to the DI with Foobar but will do so and report if my opinion changes.

Quote:

Originally Posted by Audio Bling 

Slim.a..

 

>“Did the unit need any burn-in?”

 

Yes, but I didn’t find much improvement after, say, four days. Just to be clear, I ran the DI + EPS 24/7 for about 5days then for a number of hours per day thereafter. I did quite a bit of listening over the course of two weeks. It is possible that it may improve further. I will report if it does.

 

>“Did you try the DI + DAC19DSP with the PLL turned ON?”

 

Yes, I did. I just seem to lose a degree of detail (and coherency) when I do this. I will try PLL ON again as I always like to re-validate preferences and settings from time to time and particularly after I make a change (as I just have done with the USB polling edit). Regardless, my opinion of the DI will not alter with the PLL setting as I have listened to the DI both ways.

 

>“What drivers/media players are you using? “

 

I use a cMP2 box which is to say I use cPlay (latest version). I use Asio4all to both DI and Freeway2. I use Foobar (with Kernal Streaming) from time to time again for the purpose of re-validating my settings. I didn’t listen to the DI with Foobar but will do so and report if my opinion changes.

Thanks for your very informative and thorough answer! It gives a better understanding on where the DI stands up. According to Kingwa's own admission, the Ref-3 is better than the DI and the CD-7 is better than the Ref-3.

Since even the dac19dsp can benefit from something like the EA units, I really think Kingwa should make a higher end usb converter to match his mid-level and higher end DACs. Maybe he prefers to wait until he can makes things as good as the CD-7 ... who knows.