or Connect
Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › Thoughts on a bunch of DACs (and why delta-sigma kinda sucks, just to get you to think about stuff)
New Posts  All Forums:Forum Nav:

Thoughts on a bunch of DACs (and why delta-sigma kinda sucks, just to get you to think about stuff) - Page 22  

post #316 of 6360

You guys may have seen this, but here's Ayre's take on PCM vs DSD and includes samples from songs in both formats http://www.ayre.com/insights_dsdvspcm.htm . I'm in the "DSD must die!"camp myself.

post #317 of 6360
Quote:
Originally Posted by Maxvla View Post

Sanlitun, I don't mean to direct this at you, but your post brought it to mind...

There has been a lot of talk recently with phrases like "higher dynamics" and "more dynamic" and so on, when the poster actually means it's just louder. This is the opposite of "more dynamic". More dynamic means quieter and louder. Increasing dynamics means making quiets quieter and louds louder, not increasing the overall loudness.

Please choose your words more carefully, people.

 

Yep-and one of the most important aspects of music, and an easy to identify trait (imo) separates good/great gear from the mediocre...

 

-Daniel

post #318 of 6360
Quote:
Originally Posted by Sanlitun View Post

Even though I am not a big fan of DSD I use the X-Sabre as my general DAC as it does everything so well, or well enough. On the X-Sabre I would say hi-res PCM is -much- better sounding than DSD, and pretty much the only time I listen to DSD is if there is no PCM hi-res of the recording. And maybe not even then. The DSD sound I hear through the X-Sabre is characteristic of the DSD sound I have heard through other sources, slightly rounded and rolled off. Converting PCM to DSD in JRiver does not sound better for me, it just sounds a bit smoother and less resolved like it's going through a "two martini" filter.  

I personally don't really have an issue with the Sabre sound, and I arrived at using the X-Sabre after trying DACs like the Gungnir that were just way too punchy for me. There seems to be an effort with some recent DACs to bump up dynamics to achieve a more realistic performance, but it can be unpleasant, and make some vocals too shouty.

I did same experiment on JRiver (on-the-fly conversion from 44.1k PCM to DSDx64), feeding Hilo.   And I observed exactly the same thing:

The sound difference is quite obvious (I wanted to say 'huge').  The converted DSD is smoothed out in the highs, contains less information.     It may sound smoother, but does not give me a positive experience:  the smoothness doesn't let me feel natural or realistic.  it only let me feel it is unreal.

 

This experiment only tells me one or both the following 2 things are true:  1) converting from 44.1k PCM to DSD is lossy, at least true for JRiver.   2) Hilo's PCM implementation is not inferior to it's DSD implementation.

Nothing more than that.  I still need to test DSD vs PCM of the same music.


Edited by yfei - 1/16/14 at 10:35pm
post #319 of 6360

has anyone heard the V800 and compared it to the BDA2 and WFS DAC2 or maybe NadM51? I know the BDA2 would be miles ahead but just wondering about the v800

post #320 of 6360
Quote:
Originally Posted by yfei View Post
 
Quote:
Originally Posted by Sanlitun View Post

Even though I am not a big fan of DSD I use the X-Sabre as my general DAC as it does everything so well, or well enough. On the X-Sabre I would say hi-res PCM is -much- better sounding than DSD, and pretty much the only time I listen to DSD is if there is no PCM hi-res of the recording. And maybe not even then. The DSD sound I hear through the X-Sabre is characteristic of the DSD sound I have heard through other sources, slightly rounded and rolled off. Converting PCM to DSD in JRiver does not sound better for me, it just sounds a bit smoother and less resolved like it's going through a "two martini" filter.  

I personally don't really have an issue with the Sabre sound, and I arrived at using the X-Sabre after trying DACs like the Gungnir that were just way too punchy for me. There seems to be an effort with some recent DACs to bump up dynamics to achieve a more realistic performance, but it can be unpleasant, and make some vocals too shouty.

I did same experiment on JRiver (on-the-fly conversion from 44.1k PCM to DSDx64), feeding Hilo.   And I observed exactly the same thing:

The sound difference is quite obvious (I wanted to say 'huge').  The converted DSD is smoothed out in the highs, contains less information.     It may sound smoother, but does not give me a positive experience:  the smoothness doesn't let me feel natural or realistic.  it only let me feel it is unreal.

This experiment may only tell me 2 things:  1) converting from PCM to DSD is lossy, at least true for JRiver.   2) Hilo's PCM implementation is not inferior to it's DSD implementation.

This is invalid test, you're just upsampling redbook material, it doesn't add any information just stretches existing data, any upsampled material sounds glossy and overly smooth.

What you want is to get original DSD, convert to PCM and compare the results.

post #321 of 6360
Quote:
Originally Posted by Andrew_WOT View Post
 

This is invalid test, you're just upsampling redbook material, it doesn't add any information just stretches existing data, any upsampled material sounds glossy and overly smooth.

What you want is to get original DSD, convert to PCM and compare the results.


I think the guy was responding to Purrin's statements like DSD sounds better on certain DACs and recommendation like "We recommend converting PCM content to DSD" (X-Sabre part in OP)..

 

"Smoother and less resolved" is also what I hear when converting PCM to DSD.

 

By the way, I couldn't see how converting DSD to PCM can be "more valid" than converting PCM to DSD, neither is lossless (or reversible) process anyway.

post #322 of 6360
Quote:
Originally Posted by kn19h7 View Post
 
 

"Smoother and less resolved" is also what I hear when converting PCM to DSD.

 

By the way, I couldn't see how converting DSD to PCM can be "more valid" than converting PCM to DSD, neither is lossless (or reversible) process anyway.

No upsampling which just muddies things up, you start with highest res material and convert it to another high-res format. You can convert DXD to DSD and compare, this is the most valid PCM vs DSD test.

Or just go to http://www.2l.no/hires/index.html and grab their free samples of the same track in DXD (PCM) and DSD128 and compare. 

post #323 of 6360
Quote:
Originally Posted by Andrew_WOT View Post
 

No upsampling which just muddies things up, you start with highest res material and convert it to another high-res format. You can convert DXD to DSD and compare, this is the most valid PCM vs DSD test.

Or just go to http://www.2l.no/hires/index.html and grab their free samples of the same track in DXD (PCM) and DSD128 and compare. 


DSD and PCM are different enough for me not to think the conversion as "resampling".. (btw I don't see anyone mentioned what PCM formats they have tried, not sure why you assume low-res)

 

Higher resolution files may get better results in the conversion, but even so its just that the effect is at less extent. We are talking about the general effect in sound when doing the conversion.

post #324 of 6360
Quote:
Originally Posted by kn19h7 View Post
 
Quote:
Originally Posted by Andrew_WOT View Post
 

No upsampling which just muddies things up, you start with highest res material and convert it to another high-res format. You can convert DXD to DSD and compare, this is the most valid PCM vs DSD test.

Or just go to http://www.2l.no/hires/index.html and grab their free samples of the same track in DXD (PCM) and DSD128 and compare. 


DSD and PCM are different enough for me not to think the conversion as "resampling".. (btw I don't see anyone mentioned what PCM formats they have tried, not sure why you assume low-res)

 

Higher resolution files may get better results in the conversion, but even so its just that the effect is at less extent. We are talking about the general effect in sound when doing the conversion.

Did you do the test I suggested, what's the result, what DAC did you use?

post #325 of 6360
Quote:
Originally Posted by Andrew_WOT View Post
 

Did you do the test I suggested, what's the result, what DAC did you use?


Nope, I am not at home now.

 

Not sure if I will do either, I am lazy.

post #326 of 6360
Quote:
Originally Posted by Andrew_WOT View Post
 

 

What you want is to get original DSD, convert to PCM and compare the results.

What you want is

 

1. Master Analog tape --> Small EQ if needed ---> DSD64 ---> Listen

2. Master Analog tape --> Small EQ if needed ---> PCM 176.4/24 ----> Listen

3. Compare these two, DSD DACs except the Loki play both PCM and DSD.

 

The hard part will be the A/D to have the same tonal characteristics in both formats, good luck Charlie. DSD will win because PCM will always have extra inherent steps for processing.

 

Converting a DSD to PCM what rate, redbook...? and then compare? What on earth for? That's really apples and oranges.

post #327 of 6360
Quote:
Originally Posted by One and a half View Post
 
Quote:
Originally Posted by Andrew_WOT View Post
 

 

What you want is to get original DSD, convert to PCM and compare the results.

What you want is

 

1. Master Analog tape --> Small EQ if needed ---> DSD64 ---> Listen

2. Master Analog tape --> Small EQ if needed ---> PCM 176.4/24 ----> Listen

3. Compare these two, DSD DACs except the Loki play both PCM and DSD.

 

The hard part will be the A/D to have the same tonal characteristics in both formats, good luck Charlie. DSD will win because PCM will always have extra inherent steps for processing.

 

Converting a DSD to PCM what rate, redbook...? and then compare? What on earth for? That's really apples and oranges.

Who said anything about redbook, use highest rate supported by your DAC, even DXD.

post #328 of 6360
Thread Starter 
Quote:

Originally Posted by One and a half View Post

 

 

1. Master Analog tape --> Small EQ if needed ---> DSD64 ---> Listen

2. Master Analog tape --> Small EQ if needed ---> PCM 176.4/24 ----> Listen

 

The hard part will be the A/D to have the same tonal characteristics in both formats, good luck Charlie. DSD will win because PCM will always have extra inherent steps for processing.

 

Please explain the extra inherent steps for PCM processing.
 
BTW, the workflows you stated above almost never happen unless you are at a boutique mastering shop with an analog EQ. Mastering in 2014 is usually done in the digital realm. Ever tried to EQ on native DSD? I thought not.
post #329 of 6360
Quote:
Originally Posted by purrin View Post
 

 

The extra steps with PCM are shown in the figure. More processing = more noise = delays. Snap is from DSD guide 

 

 

This recording from Opus 3 used analog EQ, then direct to a Korg. 

 

This recording was done at the wonderful venue Järna Kultur center. Using a AKG C12VR mic on Yaminas vocals and a stereo Thuresson on the Steinway D grand piano. The mic feeds goes through the famous Opus3 4-channel valve mixer where mild Equalizing and high-pass filters are applied. Then straight into the Korg MR2000S and captured in DSD128. Those files captured by the MR2000S are the once you are downloading. No other after processing is applied. It is essentially as real as it gets. Comparable to direct-to-disc recorded vinyls but without the losses in duplication. In otherwords Direct-To-DSD.

 

Titling the thread "DSD must die" is like saying, let's not make any more chocolate ice cream. It's a flavour I don't like it at all it needs to go, by that rationale, but many others do like chocolate ice cream, so why deny them that choice? There are all the other flavours, it's good enough is it?

 

There are two companies offering DSD EQ at the moment. Perhaps that's a good thing, to force mastering to use as simple tools as possible instead of killing music with gross compression effects. Philosophies....

post #330 of 6360
Thread Starter 
Quote:
Originally Posted by One and a half View Post
 

The extra steps with PCM are shown in the figure. More processing = more noise = delays. Snap is from DSD guide 

 

 

LOL, nice. Maybe this is why I prefer R2R DACs. No S-D modulators. Way go referencing DSD propaganda which avoids any mention of noise-shaping techniques employed in the DSD chain and DSD's inferior dynamic range (bits) in the ultrasonic region. 

 

Quote:
Originally Posted by One and a half View Post

 

Titling the thread "DSD must die" is like saying, let's not make any more chocolate ice cream. It's a flavour I don't like it at all it needs to go, by that rationale, but many others do like chocolate ice cream, so why deny them that choice? There are all the other flavours, it's good enough is it?

 

That is a poor analogy. Saying DSD must die is like saying HD-DVD or Betamax must die.


Edited by purrin - 1/17/14 at 1:52pm
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 › Thoughts on a bunch of DACs (and why delta-sigma kinda sucks, just to get you to think about stuff)