Upsampling DACs v2
post-143150
Thread Starter
Post #1 of 25

Nezer

Antibacterial soap... kills bacteria... bad karma?
Joined
Dec 6, 2001
Messages
1,273
Reaction score
13
Joined
Dec 6, 2001
Posts
1,273
Likes
13
I want to start another posts to answer the original poster's questions and NOT stray into a technical pissing match.

From the original post
Quote:

I am thinking of adding an external DAC to my CD player (Musical Fidelity X-Ray). I am consideirng one of the latest upsampling DACs such as the Bel Canto DAC 2.0, Perpetual Technologies P1A/P3 or Musical Fidelity A3-24 DAC. Would be very interested to hear the views/reviews taht anyone out there has.


I'd like to expand this one step further. I would be instrested in hearing how these external DACs stand up to DACs integrated into CDPs such as the Cary 303/200 or the 306/200 or the Wadia 301.

Particularly do these external upsampling DACs remove or smooth out the awful digital artifacts that can be heard in 44.1khz sampled audio?

I'm planning on upgradeing some sources here in the near future and I am *very* intrested into hearing peoples thoughts on this.

Let's *please* refrain from getting into technical arguments here and stick to topic!!
 
     Share This Post       
post-143179
Post #2 of 25

kelly

Herr Babelfish der Übersetzer, he wore a whipped-cream-covered tutu for this title.
Joined
Jan 1, 2002
Messages
5,435
Reaction score
11
Joined
Jan 1, 2002
Posts
5,435
Likes
11
I think my views are known by now, but I'll post again anyway in case I ever get a balanced source and need a Blockhead loaner.

The absolute best digital source I have ever heard with no discernable digital artifacts to my ears was the Wadia 270 transport with the Wadia 27ix DAC. This combination retails for close to $15000.

Closer to my monetary reality are the Sony XA777ES and Perpetual Technologies P3A. I evaluated these two units in comparrison and found that while both achieved a remarkable sense of smoothing the digital artifacts on the frequencies extensions, both were left with what could only be called a "digital signature." That is to say with some effort, one could easily that the source in question was a compact disc player.

In my experience thus far, I've found that the XA777ES and P3A (as well as the Linn Genki and Linn Ikemi that I auditioned) are far better at eliminating digital artifacts than entry level players. The most noticable and bothersome appearance of digital artifacting to me surrounds the sibilants (exntended T and S sounds that are particularly prominent with high pitched vocalists and accentuated by certain microphones--Sting is a good example of a sibilant vocalists whose recordings demonstrate these artifacts). While these artifacts are virtually gone on these sources with exceptional recordings such as those on the Chesky label, they can still be picked out readily in the less audiophile-centric digital recordings. One perfect example can be found on the Bruce Dickinson compilation released just last year. While it's obvious that the most prominant of these artifacts originated in the studio, it's frustrating that the player doesn't seem to compensate for them as the Wadia does so easily.

The ART DI/O at a much lower price tag in it's "88.2khz" mode does seem to produce less digital artifacts than other entry level players and is chief among the reasons why I am so quick to recommend, despite the fact that it does not handle the problem nearly so neatly as the above described DACs.

Second to the artifact issue, a benefit that seems to be shared among "upsampling" DACs of late is the perceived depth of soundstage. The ability to hear one musician placed on the z plane behind another. Almost always paired with this is I think a set of terms that all basically mean the same thing--tonal accuracy, transient speed, inner detail. This seems to be resultant of a combination of two factors -- one being the ability of the playback system to reproduce finer gradients of volume and another that has to do with delivering the seperate parts of the audio frequency at a pace such that the instrument seems to arrive and begin playing all together and at once. This seemes to effect both the realness of the instruments tones itself as well as the texture or air or whaetver audiophile term you prefer that surrounds the instrument at the moment of play--that is, not only the plunk of the piano key but the rub of the wood against the string and the slight refraction of the note as it decays in the piano's wooden case and the space around it. While this second quality is often attributed to the transducer, from what I have heard I also believe it to be very influenced by the source compnent. In these areas too, I feel that the stock ART DI/O, stock XA777ES/modded P3A and Wadia 270/27ix represent three distinct tiers of accuracy.

Let's not convert this thread into the argument of why these qualities exist. Rather, I'd be interested in discussing which DACs and players do it best for their respective price points. If they happen to not call themselves "upsampling", I'm sure we'd all be just as eager to hear about them so long as they do a great job of reproducing the music.
 
     Share This Post       
post-143186
Post #3 of 25

Nezer

Antibacterial soap... kills bacteria... bad karma?
Joined
Dec 6, 2001
Messages
1,273
Reaction score
13
Joined
Dec 6, 2001
Posts
1,273
Likes
13
Quote:

Originally posted by kelly
Let's not convert this thread into the argument of why these qualities exist. Rather, I'd be interested in discussing which DACs and players do it best for their respective price points. If they happen to not call themselves "upsampling", I'm sure we'd all be just as eager to hear about them so long as they do a great job of reproducing the music.


Amen!!!

You had mentioned your thoughts with regards to CDPs in my earlier thread about upgrading sources.

It's such a shame that the Wadia 27ix costs $9000!

I haven't looked into the P3A in detail yet.
 
     Share This Post       
post-143304
Post #4 of 25

pigmode

Truck-Fi
Joined
Jun 26, 2001
Messages
5,371
Reaction score
96
Joined
Jun 26, 2001
Posts
5,371
Likes
96
Quote:

While it's obvious that the most prominant of these artifacts originated in the studio, it's frustrating that the player doesn't seem to compensate for them as the Wadia does so easily


You want your source to compensate for a sonic trait that resides on disc!?!? If your source is toning down sibilence, what else is it toning down?
 
     Share This Post       
post-143318
Post #5 of 25

kelly

Herr Babelfish der Übersetzer, he wore a whipped-cream-covered tutu for this title.
Joined
Jan 1, 2002
Messages
5,435
Reaction score
11
Joined
Jan 1, 2002
Posts
5,435
Likes
11
Quote:

Originally posted by pigmode


You want your source to compensate for a sonic trait that resides on disc!?!? If your source is toning down sibilence, what else is it toning down?


I imagine it's toning down any of the distortion that matches the fingerprint the Wadia filter is looking for. I imagine anything that looks like that artifact sounds more like an artifact than it does a piece of the instrument, no matter where in the chain that artifact originated.

But I don't really know. These are questions better left to Wadia's engineers. I can only offer you what I heard. I recommend auditioning the Wadia rather than taking my word for it. You may think the Wadia is "toning down" some important aspect of the music that other digital sources are not toning down. Maybe you can hear this and I can't. Your opinion would be appreciated.
 
     Share This Post       
post-143332
Post #6 of 25

pigmode

Truck-Fi
Joined
Jun 26, 2001
Messages
5,371
Reaction score
96
Joined
Jun 26, 2001
Posts
5,371
Likes
96
I doubt I will ever listen to a Wadia. I'm one of those audio Ostriches that won't listen to something I can never justify affording. (sigh)

However, I am becoming intrigued with the Bel Canto DAC2, and am saving for a player in the price range of the Cary 306/200. I think my AH! 4000 is the weak link ims.
 
     Share This Post       
post-143344
Post #7 of 25

Nezer

Antibacterial soap... kills bacteria... bad karma?
Joined
Dec 6, 2001
Messages
1,273
Reaction score
13
Joined
Dec 6, 2001
Posts
1,273
Likes
13
If you can afford a 306/200 you can certainly afford a Wadia 301!
 
     Share This Post       
post-143347
Post #8 of 25

pigmode

Truck-Fi
Joined
Jun 26, 2001
Messages
5,371
Reaction score
96
Joined
Jun 26, 2001
Posts
5,371
Likes
96
Quote:

Originally posted by Nezer
If you can afford a 306/200 you can certainly afford a Wadia 301!


Sorry, I meant the 303. (shrugs, droops shoulders, looks down and languidly kicks an imaginary object on the floor)
 
     Share This Post       
post-143541
Post #9 of 25

Nezer

Antibacterial soap... kills bacteria... bad karma?
Joined
Dec 6, 2001
Messages
1,273
Reaction score
13
Joined
Dec 6, 2001
Posts
1,273
Likes
13
If memory serves me correct the Cary 303/200 runs for about $3500 USD as does the Wadia 301. The are priced so close I think you owe it to yourself to listen to both of them.

As for what else might be getting tossed out... Nothing... at least nothing that matters if properly implemented.

At 44.1k samples over about 14-15khz are poorly represented*. I believe this is the effect I'm hearing. I have a Meier Analoguer and while it does help it doesn't eliminate it and it does, in fact, throw some of the baby out with the bath water.

A more proper solution would be implemented in the digital domain. The effect can be described mathematically (I found the formula somewhere that explained this effect) and thus, can be compensated for across the frequency band in a DSP.

Now, for this to be useful you *HAVE* to upsample! Up sampling in and of itself doesn't fix the problem as the piss-poor representation existed in the lower-sampled source. What upsampling does give you is the room to fix it and shove the new high-frequency effects out-of-band (i.e. at a 96k sample rate the high-frequency nasties are pushed out to about 28-30k well above my hearing and most other people I know, not to mention all my headphones).

By upsampling and then applying a DSP filter to compensate (not remove) this effect I believe it's possible to get it 95% back to what it was had you have sampled at the higher-rate to begin with.

This is just one possible solution I'm sure. I don't care *how* a DAC does it's job, I just want to hear impressions of what others think about these mid-priced DACs in particular how the smooth-out the tope end.

If you can't hear this effect consider yourself lucky and I don't recommend you listen for it because you aren't getting rid of it for cheap.

Kelly, for example, has nailed what I'm hearing on the head and I would be intrested in hearing from people that also hear this and what they think of the DACs listed above.

=== Footnote
* http://headwize.com/projects/showpr...=meier4_prj.htm
* http://www.taddeo-loudspeaker.com/Flicker.htm

If you want to argue these points please start another thread to do so.
 
     Share This Post       
post-143659
Post #10 of 25

kelly

Herr Babelfish der Übersetzer, he wore a whipped-cream-covered tutu for this title.
Joined
Jan 1, 2002
Messages
5,435
Reaction score
11
Joined
Jan 1, 2002
Posts
5,435
Likes
11
Quote:

Originally posted by Nezer
At 44.1k samples over about 14-15khz are poorly represented. I believe this is the effect I'm hearing. I have a Meier Analoguer and while it does help it doesn't eliminate it and it does, in fact, throw some of the baby out with the bath water.


Nezer
I've waited so long to hear exactly that opinion.

Have you tried using the analoguer with mp3s? The reason I ask is that I find mp3s particularly frustrating in that they don't seem to have the benefits of high frequencies but still manage to produce the digital artifacts. It seems that a controlled attenuated rolloff without phase distortion would be ideal for dealing with mp3s--which is what I've come to understand the Analoguer basically is?

Another product that claims to do what the Analoguer does is the Digital Antidote. I'm very curious about this product as well--and it comes in two varieties, a passive one and an active one. The active one is considerably larger and costs a lot more. I don't currently understand the workings of them but both operate in the analog domain, not in the digital domain.

By the way--I'm burning in the Sony XA777ES right now and decided last night to give their "digital filter" a try. Instead of sounding like InstaSonyWadia, it sounded murky and flattened--like some of the tubes in tube amps I liked least. Ah well, "STANDARD" mode it is.
 
     Share This Post       
post-143668
Post #11 of 25

AndreYew

500+ Head-Fier
Joined
Jun 21, 2001
Messages
586
Reaction score
230
Joined
Jun 21, 2001
Posts
586
Likes
230
Nezer,

"At 44.1k samples over about 14-15khz are poorly represented."

If you want to stop the "technical pissing match", as you put it, stop making ridiculous technical claims like this. Make all the subjective claims you want (eg. "The Wadia sounds smoother to me than the Sony."), but when you bring in technical claims, you're bringing it all in.

--Andre
 
     Share This Post       
post-143672
Post #12 of 25

kelly

Herr Babelfish der Übersetzer, he wore a whipped-cream-covered tutu for this title.
Joined
Jan 1, 2002
Messages
5,435
Reaction score
11
Joined
Jan 1, 2002
Posts
5,435
Likes
11
Andre

I think that quote is a little too "no duh" to qualify as a technical pissing match. I don't think anyone would disagree with the undersampling problem. Call it what you want but the effects of that are the sounds that Nezer and I do subjectively describe.

I guess what I mean is... even I understood it. It can't be THAT technical.
 
     Share This Post       
post-143685
Post #13 of 25

Nezer

Antibacterial soap... kills bacteria... bad karma?
Joined
Dec 6, 2001
Messages
1,273
Reaction score
13
Joined
Dec 6, 2001
Posts
1,273
Likes
13
Quote:

Originally posted by kelly
Have you tried using the analoguer with mp3s? The reason I ask is that I find mp3s particularly frustrating in that they don't seem to have the benefits of high frequencies but still manage to produce the digital artifacts. It seems that a controlled attenuated rolloff without phase distortion would be ideal for dealing with mp3s--which is what I've come to understand the Analoguer basically is?


I have used the analoguer with MP3s and I find the exact same effect as with my CDP. What I hate started doing is running the lowpass-filter in LAME to 21500 which ends the polyphase at 22050 (Nyquest limit). I find this adds some of the life and sparkle back into the MP3s and, to my ears, I can't tell it apart from the original (assuming a good encoding method and psycho-acoustic simulator was used).

The analoguer throws some of this benefit away but it's a necessity when listening to Sting, Natalie Merchant and some Pearl Jam acoustics.

Quote:

Another product that claims to do what the Analoguer does is the Digital Antidote. I'm very curious about this product as well--and it comes in two varieties, a passive one and an active one. The active one is considerably larger and costs a lot more. I don't currently understand the workings of them but both operate in the analog domain, not in the digital domain.


These guys certainly are talking my language:

http://www.taddeo-loudspeaker.com/Flicker.htm

Thier passive device is relativly cheap and I don't think one could gow wrong trying it out. It would be intrested to check into thier patent to see how they are acomplishing this feat. Dr. Meier is quite open about how he is doing it with the Analoguer (though I don't understand anything past the definition of the problem). These guys are a little short on technical information (I like pictures) so I'm left to wonder, but the passive device is so cheap...

Quote:

By the way--I'm burning in the Sony XA777ES right now and decided last night to give their "digital filter" a try. Instead of sounding like InstaSonyWadia, it sounded murky and flattened--like some of the tubes in tube amps I liked least. Ah well, "STANDARD" mode it is.


Have you considered any mods to the 777ES? Like maybe upgrading the clock or power caps? It may be an ES line but it's still mass-produced and that means cost does factor into the equation. I await your total review of the 777ES. I'm guessing any DAC/CDP that satisfies your ears will likly satasify mine as well. I wish I had more money right now so I could start expermenting.

I have a query into Bel Canto about a dealer in my area. Hopefully someone is nearby so I can sample this black-box. I have find a Cary dealer and he has a 303/200 in-stock that I plan to audition one day soon. Also, another dealer should be getting a Wadia 301 in stock within the next few weeks. I'm going to ask that they burn it in so I can sample it with my gear. As far as Wadia gear goes, the 301 is about all I'm willing to spend. $9000 for a DAC (27ix) is outrageous if you ask me!

If all else fails, I will write a Cooledit filter to try and compensate for this shortfalling. The problem here is I suck at math but fortunatly my girlfriend doesn't and she can program in C/C++. I'm planning on using a computer as a massive jukebox with Monkey's Audio anyway. Monkey's Audio compresses a 192KHz/16-bit file very well especially considering I'll have nothing over Nyquist anyway. This way I don't sacrafice much (if any) storage space and I still have that extra-bandwidth to shift this crap out-of-band. Now, finding a reasonably-priced sound card with a digital output capable of 192K and an associated DAC might be a chore but I can always go with 96K. I'm a firm believer in going overboard.

Unfortunatly, Monkey's Audio doesn't seem to like 24 or 32 bit audio files.
I'll just let the DAC deal with pulling that extra detail out anyway as it will probably do a better job than the up-sampler in Cooledit.

Do all the outboard DACs listed in the initial post support at least 96k and does SPDIF have enough bandwidth to handle a 192k stream?
 
     Share This Post       
post-143691
Post #14 of 25

Nezer

Antibacterial soap... kills bacteria... bad karma?
Joined
Dec 6, 2001
Messages
1,273
Reaction score
13
Joined
Dec 6, 2001
Posts
1,273
Likes
13
Quote:

Originally posted by AndreYew
Nezer,

"At 44.1k samples over about 14-15khz are poorly represented."

If you want to stop the "technical pissing match", as you put it, stop making ridiculous technical claims like this. Make all the subjective claims you want (eg. "The Wadia sounds smoother to me than the Sony."), but when you bring in technical claims, you're bringing it all in.

--Andre


This isn't a pie-in-the-sky claim, it's a cold fact about digital sampling:
http://headwize.com/projects/showpro...meier4_prj.htm

So if you want to argue the point, don't do it here. Argue with Dr. Meier who knows infinitly more about this than I do.

And please don't troll with this flamebait.

If you want to discuss it further with me lets take it to a new thread or in PM but I know nothing more thatn what Dr. Meier explains in the above post.
 
     Share This Post       
post-143696
Post #15 of 25

kelly

Herr Babelfish der Übersetzer, he wore a whipped-cream-covered tutu for this title.
Joined
Jan 1, 2002
Messages
5,435
Reaction score
11
Joined
Jan 1, 2002
Posts
5,435
Likes
11
Quote:

Originally posted by Nezer
I'll have to hunt these products down. I would bet though that the Analoguer is a better value based on Meier's track record. :)




I of course have a great deal of respect for Jan also, but I'd not be so quick to dismiss the Digital Antidote. There's not a lot of talk about it but most of what's out there is positive. Carlo seemed to like it, too.

Quote:


Have you considered any mods to the 777ES? Like maybe upgrading the clock or power caps? It may be an ES line but it's still mass-produced and that means cost does factor into the equation. I await your total review of the 777ES. I'm guessing any DAC/CDP that satisfies your ears will likly satasify mine as well. I wish I had more money right now so I could start expermenting.


I bought the XA explicitly to have modified by Modwright. I'm only using it right now to get through the initial burn in and to make sure there aren't defects with the player. I'll post my initial impressions with it before I send it off. The mods will in the very least include the AudioCom Superclock which should make a big difference. I'll probably be doing quite a bit more depending on cashflow.
 
     Share This Post       

Users Who Are Viewing This Thread (Users: 0, Guests: 1)

Top