[maverickronin]

I meant the $200 as a general ballpark figure to get something rather good.  Not as some kind of dogmatic statement about the minimum price needed to get good sound.  Sorry if I came off like that.  I was actually going for the opposite, since $200 is cheap compared to some of the DACs that people praise around here.  Not much of my own gear is very expensive, but I'm perfectly happy with it.

As for the amps, let me try to restate my position a little more clearly.  In my experience, the price/performance ratio for expensive amps is better than it is for expensive DACs.  That's why I'd say to spend more on the amp.  That is also assuming you're determined to spend the whole $1000 the op mentioned, not because you need to spend that much on the amp to get good a quality amp.  If someone gave me a kilo-buck I could only spend on what the op mentioned, I'd get a DAC Magic and a LD MKVI+ since that's the cheapest balanced tube set up I'm aware of (Not that I spend a whole lot of time researching things that far out of my price range...) and balanced is the next clearly defined step up in quality, assuming a large stable of unknown 'phones which isn't going to be upgraded or added to.

 

[Shike]

Quote:

I meant the $200 as a general ballpark figure to get something rather good.  Not as some kind of dogmatic statement about the minimum price needed to get good sound.  Sorry if I came off like that.  I was actually going for the opposite, since $200 is cheap compared to some of the DACs that people praise around here.  Not much of my own gear is very expensive, but I'm perfectly happy with it.

As for the amps, let me try to restate my position a little more clearly.  In my experience, the price/performance ratio for expensive amps is better than it is for expensive DACs.  That's why I'd say to spend more on the amp.  That is also assuming you're determined to spend the whole $1000 the op mentioned, not because you need to spend that much on the amp to get good a quality amp.  If someone gave me a kilo-buck I could only spend on what the op mentioned, I'd get a DAC Magic and a LD MKVI+ since that's the cheapest balanced tube set up I'm aware of (Not that I spend a whole lot of time researching things that far out of my price range...) and balanced is the next clearly defined step up in quality, assuming a large stable of unknown 'phones which isn't going to be upgraded or added to.

Err, I think you misunderstood.  The X-head is an amp, not a source.  It has low output impedance, THD+N, IMD, flat frequency response, enough power to prevent clipping . . . hopefully you see where I'm going here.
 
All the amp needs to do is amplify a signal as accurately as possible - a DAC needs to do this while (arguably) keeping jitter relatively low.

 

[maverickronin]

Quote:

Quote:

Err, I think you misunderstood.  The X-head is an amp, not a source.  It has low output impedance, THD+N, IMD, flat frequency response, enough power to prevent clipping . . . hopefully you see where I'm going here.
 
All the amp needs to do is amplify a signal as accurately as possible - a DAC needs to do this while (arguably) keeping jitter relatively low.

Whoops...To many names to keep track of.  Now what the heck was I confusing that with?
 
Your points are correct.  I'd also add that I think the amp would have the more difficult job because it has to output the signal into a reactive load, while the line out on a DAC (or anything else really) doesn't have to.  There's always room for refinement in an analog system like that.  It runs up against the wall of diminishing returns like everything else, but my opinion is that amps hit the wall latter than DACs.
 
When I listened to high end amps at canjam I could tell where that money was going.  The high end DACs didn't sound any different than the same amp and 'phone combo hooked up to a good CDP at someone else's booth.  That's certainly not to say that cheap amps suck either.  It's just from my experience, I can actually tell a cheaper amp from a more expensive amp.  I couldn't do that with the DACs.  Admittedly, there were all kinds of confounding factors, but I wouldn't spend that kind of money for a difference that didn't really jump out and grab me.
 
It would be nice  to see a jitter free DAC, just on principle though.  It shouldn't be that expensive either.  Ethernet jack, takes packet data at faster than playback speed, reassembles, error corrects, and reclocks the data with a chip that doesn't need  to do much but does have to run pretty fast.  Say 4x faster than the smallest perceivable amount of jitter, so even if the length of 2 clock cycles in a row are both off by 100% in opposite directions the maximum jitter can only be 1/2 of what can be perceived.  I don't know what the minimum perceivable jitter is though, or if you'd need some multi Ghz monster chip.  (Warning: this paragraph has contained mostly speculation!)
 
So to reiterate, I'm not saying you need to spend a lot of money to get quite good sound.  If you do want to spend a lot of money to get a little better sound and want to know how to split your budget between DAC and amp, then I would spend more on the amp.
 
I've rambled on a bit, but did that make more sense?  I may be a little new to the scene, so I don't quite have all the products down, but I'd like to think I have a pretty firm grasp of the theory.

 

[Shike]

Quote:

Whoops...To many names to keep track of.  Now what the heck was I confusing that with?

Good question . . . maybe something from Musical Fidelity?
 
 
Quote:

Your points are correct.  I'd also add that I think the amp would have the more difficult job because it has to output the signal into a reactive load, while the line out on a DAC (or anything else really) doesn't have to.  There's always room for refinement in an analog system like that.  It runs up against the wall of diminishing returns like everything else, but my opinion is that amps hit the wall latter than DACs.

 
Maybe - the thing I'd jump to is that amp circuitry has probably already reached its climax, at least in design principles where we still have to cope with newer issues in the digital domain.  Also, only some headphones are really reactive.  Grado, Audio-Technica, and certain AKG headphones (K701) are more resistive than reactive.
 
Furthermore, let's say you do have a reactive load.  As long as it's not clipping (amp has enough power) and a low output impedance (high damping factor) what does it matter?
 
Maybe it's come to the point where entry level DACs and amps are both getting to the level where they're measurably indifferent for a good part against higher priced equipment.  This is worth considering since most good quality SS amps (that aren't driven to clipping when compared) and a universal DVD player vs. a $10K CD player have both been shown indifferent by many in DBT.
 
Quote:

When I listened to high end amps at canjam I could tell where that money was going.  The high end DACs didn't sound any different than the same amp and 'phone combo hooked up to a good CDP at someone else's booth.  That's certainly not to say that cheap amps suck either.  It's just from my experience, I can actually tell a cheaper amp from a more expensive amp.  I couldn't do that with the DACs.  Admittedly, there were all kinds of confounding factors, but I wouldn't spend that kind of money for a difference that didn't really jump out and grab me.

 
There's a problem with this though: did you level match and what type of amps were you listening to?  Also many tube amps have high output impedance that would change the FR substantially of reactive load, whereas more resistive ones would merely be attenuated.
 
I get where you're coming from though in regards to DACs - they may have hit the law of diminishing returns for the most part too if the design is done properly with some decent implementation of reclocking or PLL.
 
Quote:

It would be nice  to see a jitter free DAC, just on principle though.  It shouldn't be that expensive either.  Ethernet jack, takes packet data at faster than playback speed, reassembles, error corrects, and reclocks the data with a chip that doesn't need  to do much but does have to run pretty fast.  Say 4x faster than the smallest perceivable amount of jitter, so even if the length of 2 clock cycles in a row are both off by 100% in opposite directions the maximum jitter can only be 1/2 of what can be perceived.  I don't know what the minimum perceivable jitter is though, or if you'd need some multi Ghz monster chip.  (Warning: this paragraph has contained mostly speculation!)

 
True, but unfortunately DACs really require being the master to be truly jitter free.  The master and slave have their own clocks, which is why slaved DACs usually feature some form of reclocking to help correct the error (via PLL lock or a different clock such as a oscillator).  When a DAC is used as a master clock (common in studio use), the source derives its clock from the DAC so both in much closer in syncing which greatly reduces jitter.
 
Of course some AES papers state that 250ns is required to hear jitter, which would mean even a good chunk of consumer equipment is well suited for audio production.  It's the question of whether you want to try a fly swatter on a fly and risk it getting away or nuke the thing from orbit "just to be sure".
 
Quote:

So to reiterate, I'm not saying you need to spend a lot of money to get quite good sound.  If you do want to spend a lot of money to get a little better sound and want to know how to split your budget between DAC and amp, then I would spend more on the amp.

 
All I can say is it depends, and I still have a hard time on which to prioritize.  I still think it's easy to get an accurate amp though for a low cost , and it's probably the same for DACs anymore.  The Gilmore Lite, while now discontinued at $300 proved this quite well.  Heck, the X-Head Rev. 2 as I mentioned does well (i.e. should be audibly transparent) too.

 

[SP Wild]

amps and dacs I consider a 2nd stage upgrade
cables and jitter in 3rd stage upgrade
headphone is 1st grade.

 

[Uncle Erik]

Of what you listed, the amp makes the most difference.

Overall, I've found it breaks down like this:

1. Quality of recording (*not* necessarily the player/DAC/etc.)
2. Transducer
3. Amp
4. Everything else
5. Cables = jewelry, or a new outfit for your Barbie

Amps do make a difference, but you really just need one capable of fully powering your headphones or speakers. In a lot of cases, a mid-level solid state amp will perform very well. Something in the $300-$1,000 range gets you close to the best.

Tube amps are different. You need heavy iron to really get the power supply right. Beefy power transformers, chokes and large caps are not cheap. But that's what you need to really filter the power well and get the tubes performing at their best. You also, often, need output transformers to get the impedance low enough for a good damping factor. Generally, an OTL only performs well with high impedance headphones. So if you want good power transformers (preferably separate ones for the B+ and filament supplies), a few chokes to smooth things out, then good output transformers, you're going to get $500-$1,000 or more just into the raw cost of the transformers.

Solid state is a little easier in this regard. Most chips run from 16V-24V or so and you can use chips to nicely filter and regulate the power without resorting to heavy - and expensive - iron. When you're running (for example) 350V through a tube, a little chip won't handle it. That's why you can get really good power supply performance from a solid state amp relatively cheaply.

So if you're on a budget, put the money into a mid-level solid state amp. You'll get a good power supply for a lot less than you would with a tube amp. Of course, if you're willing to spend the cash, there are few pleasures greater than tubes.

As for the source stuff... most of it is a leftover from Linn's marketing campaign about 40 years ago. Back then, a Linn turntable really made a difference. They're still great turntables, too. But that was before digital. Today, even a $20 DVD player has vanishingly low distortion and great sound. The biggest difference is in output voltage, which makes players and DACs sound different. Of course, the cheapies might break down after a few months, which is why I recommend commercial/pro gear. You can get a studio DAC reasonably that will hold up for decades. A Denon CD player meant for studios or radio stations is the same way. Go for reliability, not for the tweako stuff with magical parts and glowing reviews in an audiophile rag.

Don't bother with cables. The industry has to step up with hard facts, figures and measurements. Otherwise, you're just buying folklore and collective wisdom. Not particularly different from someone insisting that a giant bird lays an egg each morning that, somehow, hatches and becomes the sun. If you're still curious, check the prices of raw materials versus the MSRP on cables. There's a monstrous disconnect between what they cost to manufacture and what they're sold for. You'll hear some yapping about R&D, etc., but how is that even possible if you can't measure a cable in the first place and comparative listens are invalid? The simple answer is that cables are a massive profit center for those who make them.

 

[maverickronin]

I'll take your word for it.  You seem quite knowledgeable on the subject.  I washed out of the EE program I enrolled in.

 
I'd just call myself an educated layman.
 
As for the theoretical DAC, maybe I need to be clearer about what I meant.  It only uses one clock.  The DACs.  You hit play, and the computer sends all the raw data as fast as it can.  The DAC stores most of it in RAM and at least several seconds worth in on the die cache that the DAC may need.  The DAC gives periodic requests to the computer to keep its RAM topped off.  The DAC chip itself, which is probably just a modified cpu, gets the sample rate information from the computer, clocks itself at an appropriate multiple, and "plays" the samples out of its cache, fully buffered and error corrected with all its ducks in a row.
 
Of course it won't be any good for anything that requires real time playback, like music or video production, or games for that matter.  It should work just fine with normal music though.
 
That's the theory anyway.  Is it sound? (I can't decide if that's intended or not)

 

[SP Wild]

Diy is cheap.

 

[Shike]

Quote:

I'll take your word for it.  You seem quite knowledgeable on the subject.  I washed out of the EE program I enrolled in.

 
I'd just call myself an educated layman.
 
As for the theoretical DAC, maybe I need to be clearer about what I meant.  It only uses one clock.  The DACs.  You hit play, and the computer sends all the raw data as fast as it can.  The DAC stores most of it in RAM and at least several seconds worth in on the die cache that the DAC may need.  The DAC gives periodic requests to the computer to keep its RAM topped off.  The DAC chip itself, which is probably just a modified cpu, gets the sample rate information from the computer, clocks itself at an appropriate multiple, and "plays" the samples out of its cache, fully buffered and error corrected with all its ducks in a row.
 
Of course it won't be any good for anything that requires real time playback, like music or video production, or games for that matter.  It should work just fine with normal music though.
 
That's the theory anyway.  Is it sound? (I can't decide if that's intended or not)

Well, here's the thing - what you did sounds more like a network streamer than anything else.  If you're using a CPU for buffering and whatnot, you're still going to need a different chip for D/A more than likely.  Basically, all you've done is taken PC A and sent data to PC B.  This won't necessarily improve jitter, and would once again depend on internal DAC quality which when combined with switching PSU noise isn't a good idea.
 
For a decent explanation of methodologies of reducing jitter check out this http://www.tnt-audio.com/clinica/diginterf2_e.html - though he's a cable believer so take any opinions with a grain of salt and pay more attention to the actual technology explanations.

 

[maverickronin]

What I'm proposing is basically a very specialized network streamer, but with the DAC chip actually integrated into the cpu.  The point of this design was to eliminate the need for more than one clock, as well as the proprietary protocols that the DAC and the transport use to sync their clocks, that were described in your link.  I've heard those techniques work quite well, but unless someone makes a standard, it would seem lock you in to a brand.  Is that correct?
 
The point of my idea was to create an interface that can give it's best output from any crappy computer, being almost plug and go.  Just a simple driver, if the marketroids don't turn it into bloatware.
 
Obviously, its not for a studio where you'd need to sync multiple clocks, but it should be fine for home use.

 

[Sumflow]

What he say's about wires is backed up by Roger Russell at MacIntosh Labs: The Truth About Speaker Wire
Quote:

  Don't bother with cables. The industry has to step up with hard facts, figures and measurements. Otherwise, you're just buying folklore and collective wisdom.

 

[Shike]

Quote:

What I'm proposing is basically a very specialized network streamer, but with the DAC chip actually integrated into the cpu.  The point of this design was to eliminate the need for more than one clock, as well as the proprietary protocols that the DAC and the transport use to sync their clocks, that were described in your link.  I've heard those techniques work quite well, but unless someone makes a standard, it would seem lock you in to a brand.  Is that correct?

 
Well - in a studio it is a standard.  Some people on the forums have done it, it's not really proprietary as it is just uncommon or expensive.  Integrating the DAC and processor onto a single chip may or may not be a good idea depending on implementation.
 
Quote:

The point of my idea was to create an interface that can give it's best output from any crappy computer, being almost plug and go.  Just a simple driver, if the marketroids don't turn it into bloatware.

 
If you think the studio interface is proprietary or rare - this would only make the situation worse IMO.
 
Quote:

 
Obviously, its not for a studio where you'd need to sync multiple clocks, but it should be fine for home use.

 
While it may work the answer is who would adopt it?  It's not standard for studio use, and many companies are in the process of trying to integrate USB to the best of their abilities still.
 
Those are the major issues I see.

 

[maverickronin]

I suppose you are correct then.  I was under the impression that each brand had their own DAC to transport interface.
 
"Driver packages" are prone to turning into bloatware, but I was figuring that if my idea was viable, the type of company to implement it would be a small boutique company or a start up, which wouldn't be encumbered with the layers of PHBs who probably mandate the stuff.
 
Is that async usb you were talking about?  How does that work, and does it work well?

 

[angle_sh00ter]

This sounds like very valuable advice to me.  I've got a good amp in the burson HA160 and am using a musiland 02 as a dac at the moment,
but was considering using it as just a digital transport and getting a new dac in the range of $500-$700. 
Thing is Im really happy with how my setup is sounding but just reading all the threads about dacs and jitter and all the other buzz words is giving me an itchy trigger finger.
In light of your advice I think I will wait until the Sydney meet in a couple weeks to see how much difference a good DAC would make to my rig.

 
Quote:

Of what you listed, the amp makes the most difference.

Overall, I've found it breaks down like this:

1. Quality of recording (*not* necessarily the player/DAC/etc.)
2. Transducer
3. Amp
4. Everything else
5. Cables = jewelry, or a new outfit for your Barbie

Amps do make a difference, but you really just need one capable of fully powering your headphones or speakers. In a lot of cases, a mid-level solid state amp will perform very well. Something in the $300-$1,000 range gets you close to the best.

Tube amps are different. You need heavy iron to really get the power supply right. Beefy power transformers, chokes and large caps are not cheap. But that's what you need to really filter the power well and get the tubes performing at their best. You also, often, need output transformers to get the impedance low enough for a good damping factor. Generally, an OTL only performs well with high impedance headphones. So if you want good power transformers (preferably separate ones for the B+ and filament supplies), a few chokes to smooth things out, then good output transformers, you're going to get $500-$1,000 or more just into the raw cost of the transformers.

Solid state is a little easier in this regard. Most chips run from 16V-24V or so and you can use chips to nicely filter and regulate the power without resorting to heavy - and expensive - iron. When you're running (for example) 350V through a tube, a little chip won't handle it. That's why you can get really good power supply performance from a solid state amp relatively cheaply.

So if you're on a budget, put the money into a mid-level solid state amp. You'll get a good power supply for a lot less than you would with a tube amp. Of course, if you're willing to spend the cash, there are few pleasures greater than tubes.

As for the source stuff... most of it is a leftover from Linn's marketing campaign about 40 years ago. Back then, a Linn turntable really made a difference. They're still great turntables, too. But that was before digital. Today, even a $20 DVD player has vanishingly low distortion and great sound. The biggest difference is in output voltage, which makes players and DACs sound different. Of course, the cheapies might break down after a few months, which is why I recommend commercial/pro gear. You can get a studio DAC reasonably that will hold up for decades. A Denon CD player meant for studios or radio stations is the same way. Go for reliability, not for the tweako stuff with magical parts and glowing reviews in an audiophile rag.

Don't bother with cables. The industry has to step up with hard facts, figures and measurements. Otherwise, you're just buying folklore and collective wisdom. Not particularly different from someone insisting that a giant bird lays an egg each morning that, somehow, hatches and becomes the sun. If you're still curious, check the prices of raw materials versus the MSRP on cables. There's a monstrous disconnect between what they cost to manufacture and what they're sold for. You'll hear some yapping about R&D, etc., but how is that even possible if you can't measure a cable in the first place and comparative listens are invalid? The simple answer is that cables are a massive profit center for those who make them.

 

[xnor]

Indeed, very nice post Uncle Erik. Time to try out the *thumbs up* feature.