does DAC really make alot of difference?
Feb 5, 2012 at 4:48 PM Thread Starter Post #1 of 84

7Karma

New Head-Fier
Joined
Jan 22, 2012
Posts
13
Likes
0
I have only recently been into the whole audiophile headphone and I been using Xonar essence with beyerdynamic dt 990 600 ohms and I'm enjoying tons of music right now.
 
I was browsing through this site call headroom and they had this micro DAC and such... it was around 300 dollars and I was wondering are these digital-to-analog coverters really worth that much? I mean does it improve the quality of music that's worth 300 dollars? what are the benefits of going digital to analog? and why does going analog improve quality of the sounds?
 
Feb 5, 2012 at 5:08 PM Post #2 of 84


Quote:
I have only recently been into the whole audiophile headphone and I been using Xonar essence with beyerdynamic dt 990 600 ohms and I'm enjoying tons of music right now.
I was browsing through this site call headroom and they had this micro DAC and such... it was around 300 dollars and I was wondering are these digital-to-analog coverters really worth that much? I mean does it improve the quality of music that's worth 300 dollars? what are the benefits of going digital to analog? and why does going analog improve quality of the sounds?

Audio is stored in a digital form (zeros & ones), we hear analog (wave) audio, so audio has to be changed from digital to analog.
So the better the DAC (Digital to Analog Converter), the better the analog audio sounds to our ears.
 
 
 
 
Feb 5, 2012 at 5:18 PM Post #3 of 84
Quote:
Audio is stored in a digital form (zeros & ones), we hear analog (wave) audio, so audio has to be changed from digital to analog.
So the better the DAC (Digital to Analog Converter), the better the analog audio sounds to our ears.


Technically, the better the DAC the more accurate the analog audio is to the original analog signal. It implies nothing about how good it sounds.
 
And diminishing returns kick in so fast with DACs that you'd need a pretty bad one to really change the signal enough to matter.
 
After hearing (or rather not hearing) the differences myself between DACs of a few price ranges, I wouldn't pay more than $100, maybe $200, for a good one depending on the features it offers. That is, until I win the lottery.
 
Feb 5, 2012 at 10:34 PM Post #5 of 84
Small markets demand high margins, and there's lots of people who will buy into malarkey. Your Asus card has good DAC chips on it, they are already doing what you need. I think that HeadRoom product probably predates the Xonar though, to give you some perspective. As far as what I consider a reasonable value; it better do a whole lot more than take stereo PCM in and spit analog out if it's going to cost more than $50. Unfortunately, in recent years it seems that stand-alone DSPs aren't as popular (I blame HDMI, and with good reason - note that I'm not "against" HDMI). If you want a little box to sit on your desk and convert digital to analog outside of your PC, Fiio makes a little box called the D3. I think it's around $30. 
 
Quote:
wow it's that bad? i wonder why they cost so much then...



 
 
Feb 5, 2012 at 10:53 PM Post #6 of 84
The Dac does make a difference but it is usually small enough that you would never pick up on with normal computer speakers as they aren't revealing enough. What you pay for with better DAC's is for the very subtle things that usually take very revealing equipment to bring out. That said I like the sound that I get from properly setup Burr brown DAC's over say the Cirrus Logic DAC's. Better dynamics & less conjestion in the sound. More open soundstage. These are all things that normal computer speaker lack the ability to expose. I use modified biamped studio monitors that are extremely revealing.
 
Feb 5, 2012 at 11:00 PM Post #7 of 84


Quote:
wow it's that bad? i wonder why they cost so much then...



As you're just getting started here you'll have lots to read about and you'll find that a $300 DAC is on the very low end of the price scale. Some are well into the 5 figure range. Granted, most of those really expensive ones are more geared toward home audio, but they could be used with your computer.
 
Keep an eye on your bank account. It can be depleted fast here.
 
Feb 6, 2012 at 1:40 AM Post #8 of 84


Quote:
Technically, the better the DAC the more accurate the analog audio is to the original analog signal. It implies nothing about how good it sounds.

Yep, you worded it better.
 
 
 
 
Feb 6, 2012 at 2:04 AM Post #9 of 84
Seven days of listening and I can confidently conclude that the addition of an M-DAC via optical out has led to a clear and noticeable improvement over the analog out of a Xonar ST with an LME49720HA in the buffer. Most notable has been a drop in background noise, more defined bass and smoother mids. The greater the gain, the greater the the ability to engage the improved sound quality. The insertion of a Music Hall 25.3 with a cryo treated tube into the playback chain performed similarly. So does an external DAC really make a lot of difference? If your speakers, or headphones are revealing enough, absolutely yes. At the very least, the sonic signature of playback is changed. 
 
Feb 6, 2012 at 9:26 AM Post #10 of 84
I think it depends on how BAD the sound card you're used to is.
If yours were as bad as my Toshiba, ANY DAC would sound like a maiden caressing your ears, even the ones that cost next to nothing,
 
Feb 6, 2012 at 10:34 AM Post #11 of 84
Most dac's are transparent , unless there flawed or made to colour the sound your Xonar essence should sound transparent and you have nothing to gain from upgrading it.
 
Even more amusing is the people who claim the chips sound different, chips that measure disortion in the 0.0000X range , signal to noise in the 110db range , dead flat frequency responces, where are the audiable differences?, I guarantee if the chips weren't labeled no one would know which was which.
 
Feb 6, 2012 at 10:50 AM Post #12 of 84
Quote:
Most notable has been a drop in background noise


The M-DAC has similar noise level to what Stereophile measured for the Xonar STX (in fact, the RCA output is more comparable to the cheap Xonar D1). Then again, the actual performance of the card may vary due to PCs with very noisy components, problems related to the power supply or motherboard, ground loops (especially with an external AC powered amplifier), or incorrect usage. Of course, if the differences are not measured, or proven with blind tests, they may not even exist.
 
 
Feb 6, 2012 at 12:56 PM Post #13 of 84
All single element in the path is important, but DAC as "the source" has tremendous importance.
1) DAC converter (chips) is important: there are many types, but IMHO only 2 groups: old plain multibit, and sound processing delta-sigma/single bit. IMHO multibit are way better, and old and cheap chip like TDA1543, or not so cheap but quite old PCM1704 could easily beat the newest and "the best" sigma-delta.
2) Existance (or not) of oversampling (digital filtering). IMHO non overampling (NOS) it way better.
3) Type of I/V converter: active or passive (passive is simpler to be implemented in proper way).
4) Existance of analogue filter. Very simple or not a filter at all is better.
5) Existence of op-amps and how many of them. IMHO no-op-amp solution is the best, because every one op-amp is overcomplicated net of transistors, which behave in unpredictable way with the real dynamic acoustic signal.
6) Clock should be stable, because weight of every sample depends on digital value of the sample and width of... clock tick.
7) Power supply, path of the mass at the PCB, quality of the resistors, existance of capacitors in the path - all aspects are important.
 
 
Feb 6, 2012 at 2:28 PM Post #14 of 84
1. Theres nothing wrong with delta sigma dac's, i've seen them used in some higher end gear there disadvantage is the way they deal with unwanted noise so there signal to noise ratios aren't quiet as good as 16bit/24bit chips but there still more than good enough for 16bit audio and not really at any audiable disadvantage you couldn't tell them apart, this explains it better http://www.beis.de/Elektronik/DeltaSigma/DeltaSigma.html.
2. Can cause audiable differences but your wrong oversampling is way better, NOS dac's don't filter above the audiable range about 22khz oversampling dacs do filter so a NOS dac sends high frequency rubbish into you headphone/speaker amp which then gets amped to your headphones or speakers this will likely cause some kind of audiable defect, So no tweeters or amps are blown most NOS dac's use high frequency roll off the non-OS algorithm rather than a filter but this also rolls off the audiable range about 15khz the roll off becomes audiable, most dac's which are labeled NOS aren't actually NOS as they use a digital filter like oversampling dac's once again to stop equipment being destroyed by the high frequency rubbish.
 
All the others fall under signal to noise or disortion which is measurable and most dacs can handle 16bit and most measure around 0.000x disortion.
 
 
 
Feb 6, 2012 at 2:51 PM Post #15 of 84
In theory you are right (I know theory too), but in practise not.
 
Sigma-delta DAC are inferior in most (not every) cases to old multibit, but probably Sabre DAC (ESS) are exceptions. I mean sonically inferior, not by useless parameters.
 
OS and analogue filtering is not required - I have build few DAC and headphone amps myself - almost all NOS, all without any piece of analogue filtering - it sounds transparent, flawless, without any sign of digital rubbis or auditable deffects. Of course amp should be fast enough and has low amplification ratio in open loop to be completly immune to this rubish - so current buffers are the best (taking in consideration only solid state solutions).
In case of speaker tweeters there is real danger, so simple filter is recommended, but not OS. OS causing a lot of damages to the sound - maybe there are better modern implementation of this, but these buildin in sigma-delta DAC not confirm is superiority.
Voltage amping is also not required (especially with low impedance headphones) - current A-class buffer is a way better choice.
 
Measurable distortions at these levels tells nothing about the sound, so shouldn't be consiered, because less distortions means in many cases in pracice more complicated solution, more unmeasured dynamic distortions and in the result worse sound. Bad example is FiiO E9, E5 and amp in E7 - these equpements are unlistenable to me.
 

Users who are viewing this thread

Back
Top