Brief Odac impressions
May 13, 2014 at 5:50 PM Post #1,921 of 2,018
"Merely" 20 bits of resolution could kill you. 13 or 14 is more than enough. 24 bit DAC can process 24 bit word data lenght, this 20 bit refers to noise floor - if processing 24 bit data it can achieve "only" 20 bits of accuracy (the rest is covered by noise, although it has to be mentioned that one would have to apply crazy gain to make this noise disturbing). You should check sound science forum, there is a thread explaining 16 vs 24 bit.
 
Btw does anyone remember what is the resolution of vinyl discsc translated into bits? Was it something like 10?
 
May 14, 2014 at 7:13 AM Post #1,923 of 2,018
The only thing I found on this is:
 
ENOB stands for Effective Number of Bits and is another measure of a DAC’s performance. No 24 (or 32) bit audio DAC can achieve true 24 bit performance, In fact, 20 ENOB is generally considered the “Holy Grail” of real world DAC performance. The ODAC is just under 19 ENOB and the Benchmark, even referenced to its full 7+ volt maximum output, is 19.3 ENOB. The FiiO E10, even in 24 bit mode, is only 16.2 ENOB.
 
A.
 
May 14, 2014 at 9:37 PM Post #1,924 of 2,018
  The only thing I found on this is:
 
ENOB stands for Effective Number of Bits and is another measure of a DAC’s performance. No 24 (or 32) bit audio DAC can achieve true 24 bit performance, In fact, 20 ENOB is generally considered the “Holy Grail” of real world DAC performance. The ODAC is just under 19 ENOB and the Benchmark, even referenced to its full 7+ volt maximum output, is 19.3 ENOB. The FiiO E10, even in 24 bit mode, is only 16.2 ENOB.
 
A.

 
Thanx, this ENOB explains a lot. Let me ask a naive question. Does ODAC have a 24-bit DAC chip?
Assume it does, then what kind of noise is causing the loss of about 5 bits: generated by the ODAC's  electric power circuitry implementation and parts, ... or some other noise/distortion ... ? - Just curious, and that's all.
 
May 15, 2014 at 7:52 AM Post #1,926 of 2,018
Every electrical component would have it's own noise but in practice, as you see noise specified often below -100dB it's usually a non-issue. Noise can also be induced by poor PCB layout (bascially electromagnetic interference between copper traces, or some capacitance between signal/power trace and ground), resistors of very high value also produce noise... this stuff is important in the design process.
 
When you are buying a ready-made unit all you care about is its total noise performance, which in most modern units is often more than acceptable.
 
You also need to look at this loss of 5 bits in some perspective - not every bit is equally important in digital to analog conversion, as every consecutive bit represents how the volume of the signal changes relative to the bit before, and there is a finite difference in volume that can be achieved (set by the number of bits that the signal was recorded in), any "error" (be it from noise, jitter, or else) in the first bit scews the whole package. If you get the last bit wrong then it's possibly not even audible (meaning the change of volume as a consequence of conversion error).
 
Now, 16 bits represents a dyncamic range of 96 dB if I remember correctly, this is the difference in volume achievable between first and last bit, with 24 bits it can be even more (144 dB). So a 24 bit DAC with 20 ENOB can convert 24 bit digital signal to analog (meaning it will accept data in packages of 24 bits) but the maximum difference between the loudest and the lowest (in volume) signal can "only" be 20 bit, so about 120 dB, any signal that is lower in volume will be buried in the overall noise of the circuit.
 
To put this into perspective a 100 dB signal would permanently damage your hearing, as far as I remember prolonged exposure to anything above 80 dB will have permanent negative consequences on your hearing. There were some studies to determine what is the dynamic range of human hearing and 16 bit standard for A/D conversion was assumed with some headroom included.
 
May 15, 2014 at 10:50 AM Post #1,927 of 2,018
   
Hi All,
 
From the above:
 
"24 bit, with close to 20 bit resolution ~ software volume control becomes viable without loss of detail."
 
This confuses
tongue_smile.gif
me. What does it mean? Is my JDS Labs  ODAC a 24-bit or merely a 20-bit dac unit?
 
I remember that someone  on this site explained in very technical terms - I can't find it - that the ODAC was a 20-bit dac unit.

 
Bitrate dictates the noise floor. There's a point at which higher bitrates physically cannot realize improvements to the noise floor. In a measurement sense, 32-bit DACs top out somewhere near 20-bit performance, just like 24-bit DACs.
 
In other words, 32-bit performance is all marketing. You'll never hear a digital noise floor better than ~20bit, despite its theoretically higher performance.
 
ODAC is a 24-bit DAC. Just keep in mind that 24-bit and 32-bit performance is purely theoretical.
 
May 16, 2014 at 11:53 AM Post #1,929 of 2,018
  I think i've gotten this ODAC issue fixed. You know what's funny?
Thanks to this ODAC i've discovered 3 issues with my setup that i've corrected.
 
First, it turns out that the ODAC needs a ton of power from a USB port or so it seems. This probably would not work on my laptop without a powered usb hub.
 
Next, at one point I had plugged my amp into my $40 Belkin Surge Protector with filters. My first reaction was something like "Ah Hah! There's my nice and full mids".
Only awhile later did I realize it was bloating the bass slightly with my DJ100 and HD-580. Even the soundstage seemed smaller. Since the Q701 is not super bassy, it's sound wasn't changed as much it seems.
I guess the DJ100 with ODAC is good to analyze any power issues
biggrin.gif

 
Some guy wrote a review of the Micro Amp and said that you shouldn't connect it to a surge protector with filtering because it robs it of power. I just laughed at that. I'm eating my words now.
 
Next, most of my albums sounds OK, but something sounded really off. My impression was like that the mids varied too much. They were often very good and full and then the bad stuff sounds dramatically worse than they should. It felt as if the mids were super thin on this stuff and the treble was slightly too much. The treble itself seemed OK on most good music.
 
On my HRT MSII I used Medium Gain with my amp. Sounded perfect. One reason I did this is that there was an article on Headfonia claiming low gain is generally best blah blah blah. I figured since I'm getting to the 40% volume mark with the Micro Amp that medium gain would be perfectly OK, just like on my HRT MSII. I switched back and forth and didn't hear any difference at all.
 
At random I switched to high gain and left it there. There was this voice at first telling me to NOT DO IT because at high gain on my Q701 I need to set the volume at 8-10%!!!! If I accidentally bumped it to volume my ears might blow up.
 
I have no clue why it sounds better. On my HRT MSII medium gain is perfect. Why the difference with the ODAC? My one idea is that maybe the Micro Amp limits the power to 1.21 gigawatts at Medium Gain. Maybe the Q701 needs about 5 Gigawatts. Here's where it gets even worse..I was doing A/B comparisons with the E10 and the E10 sounded absolutely PERFECT on medium with the Q701! What the heck?
 
So previously I felt that bad songs were thinner that they normally should be. It seriously felt like I was listening to the Q701 with bad recordings, but with the DT-880. It felt like the Q701 was too thin and trebly, but only with the bad stuff. It was such a huge distraction. I mean the Q701 isn't this thin and sounding this way on any other dead neutral gear. Again, most warm recordings or good material sounded perfect usually. Maybe a slight extra amount of treble. It did feel that the sound was more fatiguing than it should be with bad recordings.
 
So right now my harsh/fatiguing/bad tracks are still bad, but made to be not quite as bad. It looks like on High Gain the mids really fill in and the Q701 is back to it's old self (like medium gain on the HRT!!). The treble is now perfectly smooth mostly and even some of the fatiguing stuff isn't so bad and quite listenable. Of course you still know it's bad, but doesn't stick out as much now. So basically due to the slightly thinner mids, my brain is what made me think there was more treble. The Q701 with the bad stuff just sounded so much like my Koss A/250 and a DT-880 in the mids! Yet on a very good song it could be like say the HD-600.
 
It's finally sounding like a normal Q701 should. The way i've ALWAYS heard it, but even better. It just didn't make sense why my Q701 would sound slightly off like this.
 
Last night before bed this setup was sounding so amazing I didn't want to sleep!
 
The ODAC now doesn't seem TOO REVEALING at all. It doesn't make my Q701 cold and analytical. I'm soooooo glad about this. I was thinking (and hoping) that no DAC could do that to the Q701 and I guess I was right.
 
BTW I do find it better than the HRT MSII by far. The HRT MSII doesn't actually sound that much warmer IMO. Just slightly. In comparison to the ODAC, the HRT MSII has a much smaller soundstage.
 
Again, the differences in how I heard this previously and now are not super huge. What i'm going to do is give it 3-5 days and then go back and check my earlier impressions. I will edit them if needed if anything is different. I plan on doing an ODAC review in a week. You wouldn't believe what a nightmare this has been! Sounding a lot more musical now
biggrin.gif
Even with the garbage tracks! Nope..the ODAC doesn't suddenly rob my precious Q701 of it's natural warmth (a lot of the time it's in the recording too obviously). Just make sure it has a TON of power or something. Still annoying have to use high gain at 8% volume when medium is fine on any other DAC!! I can't explain this one.

 
I am very glad that I've found your post. Very informative.
The most important for me is this:

 
Next, at one point I had plugged my amp into my $40 Belkin Surge Protector with filters. My first reaction was something like "Ah Hah! There's my nice and full mids".
Only awhile later did I realize it was bloating the bass slightly with my DJ100 and HD-580. Even the soundstage seemed smaller. Since the Q701 is not super bassy,
it's sound wasn't changed as much it seems.
I guess the DJ100 with ODAC is good to analyze any power issues 
biggrin.gif

 
Some guy wrote a review of the Micro Amp and said that you shouldn't connect it to a surge protector with filtering because it robs it of power.
 

 
 
I am not interested in any surge protector (with filters) any more.
 
Thanx.
 
May 16, 2014 at 1:33 PM Post #1,930 of 2,018
The  ODAC will work just fine on 98% of the modern day PC and Laptops out there.....IT was designed to run on USB power...all the measurements were done on PC USB power....dirty or not.
 
I forget where I posted this but my 2 ODACS have been measured for current draw and it was something like 50 milliamps of current (1/4 watt or so).......hardly taxing USB power.
 
Its not a power hungry device at all.
 
A.
 
May 16, 2014 at 1:43 PM Post #1,931 of 2,018
   
NwAvGuy claims ODAC needs just 125mA of the allotted 500mA current available to an enumerated USB 2.0 device. Unless his estimate is way off, power problems are due to out of spec USB ports. Hopefully a USB 3.0 port would provide a steadier stream of current, even though USB 2.0 ports really should be adequate. I'm disappointed to know this is even a concern. Will point NwAvGuy and Yoyodyne this direction.

 
 
So is it 50 mA or 125 mA?  
 
Sorry for being
tongue_smile.gif
 curious.
 
May 16, 2014 at 3:23 PM Post #1,932 of 2,018
Well, the max current I measured was 50.5 milliamps, connected to the O2 Amp and not connected to the amp. The current draw did not change at all with volume changes on the amp.
 
The voltage was 5.03 vdc before the ODAC was powered and 5.03 with the ODAC powered.
 
The PC was an older gateway desktop tower with a USB 2.0 port.
 
The USB power was not even phased with this minmal load.
 
100 ma is the unit spec for a single USB port for 2.0.
 
Alex
 
Jul 4, 2014 at 4:10 AM Post #1,933 of 2,018

 
David and Goliath
wink_face.gif


I was wondering how the ODAC sounds after reading all the praise that some Head-Fi members had given it so I went ahead and bought one with RCA outputs from Walter (top notch service, highly recommended). What a nice surprise, it sounds really good. Well, not on par with the Reference's DAC with 4x PCM1704UK DAC chips but you probably can't buy anything better at this price level.
 
Jul 4, 2014 at 9:24 AM Post #1,934 of 2,018
 
 
David and Goliath
wink_face.gif


I was wondering how the ODAC sounds after reading all the praise that some Head-Fi members had given it so I went ahead and bought one with RCA outputs from Walter (top notch service, highly recommended). What a nice surprise, it sounds really good. Well, not on par with the Reference's DAC with 4x PCM1704UK DAC chips but you probably can't buy anything better at this price level.

 
Comparison done with a strict volume matching right ?
redface.gif
 
 
A Xonar DX/D1 is cheaper and perform as good as ODAC . But you have to deal with C-Media / Asus Driver, and it's an internal sound card , so not as flexible as ODAC .
 
Jul 4, 2014 at 9:45 AM Post #1,935 of 2,018
   
Comparison done with a strict volume matching right ?
redface.gif
 
 
A Xonar DX/D1 is cheaper and perform as good as ODAC . But you have to deal with C-Media / Asus Driver, and it's an internal sound card , so not as flexible as ODAC .

 
Well, I did only very brief comparison so far.
Tried to match volume for both as much as I could.
I was not interested in soundcards as I wanted an external DAC with ESS chip to be plugged into Reference's input for comparative listening.
For my purposes the ODAC was the one
wink_face.gif

 

Users who are viewing this thread

Back
Top