Toward higher end DACs
Dec 11, 2007 at 4:14 PM Post #136 of 142
Quote:

Originally Posted by regal /img/forum/go_quote.gif
The grain in the upper mids on the wadia was very obvious vs. the Dac in both our opinions.


I hear none of the grain you're describing and I've got an 830, but since all that stuff is subjective there's no surprise there and I'd be willing to be that the rest of the audio chain was quite different between your friend's and my setups. And it's fine that you've written off opamps entirely, I have no interest in debating that with you. Like whatever you want.
 
Jan 22, 2008 at 8:57 PM Post #137 of 142
Quote:

Originally Posted by infinitesymphony /img/forum/go_quote.gif
Burr-Brown PCM1702/PCM1704
Burr-Brown PCM1792/PCM1794/PCM1796/PCM1798
Burr-Brown DSD1792/DSD1794/DSD1796



I had a question about the DSD1792, PCM1792, DSD1796 and PCM1796 I'm hoping someone can help me with, as the people here seem to be very familiar with these things.

I know this is a little off topic as I'm not discussing headphones, but if these chips are being used how does the chip handle native DSD conversion when the clock speed doesn't list one that's a multiple of 2,822,400?

The Onkyo TX-NR905 manual states the following:

"In listening modes other than DSD Direct and Pure Audio, DSD sources are converted and handled as PCM."

The Denon 4308CI manual states the following:

"When playing DSD signals in the DIRECT or PURE DIRECT mode, the DSD signals are converted as such into analog signals. When playing in other surround modes, the DSD signals are first converted into PCM format, then into analog signals."

Both these receivers I believe use the PCM1796.

The only chip I found that seems to list a clock speed that would allow for a full conversion is the DSD1700, whic has a clock speed of 11.289MHz.

Can anyone offer any insight on this? If it's not a DSD1700 or DSD1700E does this mean you're not getting a full DSD to analog conversion?

Thanks for any help anyone can offer.
 
Jan 23, 2008 at 11:30 AM Post #138 of 142
Quote:

Originally Posted by dobyblue /img/forum/go_quote.gif
I know this is a little off topic as I'm not discussing headphones, but if these chips are being used how does the chip handle native DSD conversion when the clock speed doesn't list one that's a multiple of 2,822,400?


DAC chips are labeled DSD-capable if they can interface with a DSD decoder chip. Sort of like how HDCD can be done with any DAC as long as the external filter is either PMD100 or PMD200; the DAC chips aren't responsible for the initial decoding of the signal. At least, that's my current understanding.
wink.gif


Implementations of DSD can vary, which is why I don't think it's the best choice for high-res. In many cases, decoding to PCM beforehand has measurable benefits versus the "pure" analog filter + DSD method, so don't necessarily be discouraged if a player or receiver downconverts to PCM.
 
Feb 5, 2008 at 10:09 AM Post #140 of 142
Quote:

Originally Posted by Iron_Dreamer /img/forum/go_quote.gif
I used the proceeds to upgrade my headphones, and have been happier with my system ever since.


I'm happy to see you got rid of those cd3000s.
 
Feb 5, 2008 at 10:32 AM Post #141 of 142
Quote:

Originally Posted by regal /img/forum/go_quote.gif
The grain in the upper mids on the wadia was very obvious vs. the Dac in both our opinions.


That's a very interesting observation you make there. I noticed the same graininess on quite a few DACs and CD players. In many cases this goes away after the unit has warmed up. It is most noticeable on high pitched female voices, especially when you use electrostatic headphones. They seem very good at picking up this error.
 
Feb 5, 2008 at 11:10 AM Post #142 of 142
While I agree that there are some well implemented and good sounding opamp output stages, I also agree that in a $2000 product the least that they could do is a discrete output or discrete opamp output. There have been many who build discrete opamps that sound better(so I have been told), I still prefer a discrete tube, FET, or transformer output. Not the most efficient, but good sound is rarely size or power efficient. A current output DAC with a good I/V stage or good voltage output DAC does not really need anything else to output good sound as long as the input impedance of the preamp mates well with it. I would also prefer the I/V stage to not use opamps, FETs would be better generally speaking. But if nothing is needed, why muddy up the signal? This is the route I will be going with the Opus DAC when they get the new DAC chips. Besides, a voltage output DAC with anything after it is a buffer for the DAC used to help mate it to amps, provide SE outputs from a balanced DAC, or to drive long cables. If this is not the case, why would a Red Wine Audio iMod sound so damn good for an iPod? This is also how people take a cheap player and do a Swensen mod and they sound much better than the original opamp output stages.

However, if gain is needed or impedance matching is needed then a well implemented opamp or discrete output would be used. I believe that the main reason builders use opamps or discrete output stages is to mak their product a more universally compatible device. If not, there would be a lot more stand alone buffers sold. Most people don't care to learn how to mate components impedance wise and suffer a penalty in sound. That is why, IMHO, they add buffers(opamps, tubes, etc.)

With my upcoming build on the drawing table, I see no reason for me to put anything is the signal path after the DAC and send the signal to a preamp with a 10kohm input impedance. If I was going to skip the active pre and use a passive one this would be another story. I hope I am right on this, because after hearing my friend's Swensen modded Toshiba DVD before and after with and without an active pre I am sold on skipping the whole stage all together. My identical 3950 Toshiba sounded OK with a passive and active pre with the nod going to the active approach. His Swensen modded player blew mine out of the water with the active, but sucked ass with the passive. It is all about component matching in the end, IMHO.

If we are looking for the best DAC to be had in the world, you should start with an all discrete DAC and not just a chip. Start building, it will be big! That said, I have read about a few guys who have built their own discrete DAC(not the component, the DAC itself) and with massively impressive results. Besides, if the Transporter, SACD players, and other DACs had perfect output stages, what the hell would we mod and bitch about?
 

Users who are viewing this thread

Back
Top