BEST DAC

[ivo]

Peter,

What can you say about this one: http://www.lampizator.eu/szop/L-Szop/BEST_DAC.html ?

I do not know why, but so far I trust this person. It is based on what he has done. I think it is valuable work and can be taken serious in choosing the *best* DAC.

Ivo

[pedal]

Thanks for the link, IVO. It looks interesting. Lampizator doesn't mention anything about oversampling or (absens of) lowpass filter, though...

PETER: While we are on the subject:
Is your analog section based on tubes, opamps or discrete?
Does it offer XLR balanced outputs?

All the best,

[ivo]

Well, he does not, but he offers at least 3 finalized DACs. What we do not know is what exactly is in those DACs ,e.g., what chips he utilizes. To my mind there are OS DAC chips surrounded by very delicate electronics inside and this is the answer I guess. I myself have a DAC with OS chip (PCM1798) and in company with XX it works fantastic. The question is how much more fantastic can I get with the *best* DAC???

Ivo

[PeterSt]

Peter,

What can you say about this one: http://www.lampizator.eu/szop/L-Szop/BEST_DAC.html ?

I do not know why, but so far I trust this person. It is based on what he has done. I think it is valuable work and can be taken serious in choosing the *best* DAC.

Ivo

Hey, you are both at an equal number of posts I see. So, both deserve an answer, but this first is first ...

Yes, I followed him for quite some years, and I would say he is similar(ly) stupid as I am. Very involved, very "ignorant" hence very much pioneering for the best cause. If anything, I would buy from him ... (and I mean it).

But the approach is different. Lampizator (and excuse me for judging at this far distance) is merely empirically finding the best possible sound with the means available. And mind you, for sure a "best sound" (or DAC ) can come from that; find the best combinations of everything (at the expense of a lot of time) and you will get to something.
But this is not what I did;

I will try to keep the story short, but (as you most probably know) I have a strategy right from the beginning (of XXHighEnd) which is based on 1:1. This sounds far too simple than it is, just because 1:1 is to be avoided for 16/44.1, but *still* it can be the strategy. So, in the very end all is derived from that (and I mean *all*), and where it failed for good sound, I did not give up but looked elsewhere for the flaws. Well, read the Phasure NOS1 topic, and you know where I come from, and what it took. It includes stupidities like sarting off with a filterless DAC, knowing that a filter would be needed *and* knowing that all existing filters (for mathematics) would not do the job (or otherwise I could just as well have started with a filterED DAC). So, I knew that along the way I had to "invent" something which hadn't been invented before, no matter all the Philipses, Sonys and the whole world. So, quite a challenge. It also includes the knowledge of the necessity of 352.8 (384) from a computer interface which doesn't exist (dig deep and you will see that I knew that too from the beginning, although I had no clue how - which is why the DAC topic is safely called "First 24/192 etc.").

So you see, because of my stupidity in following a strategy as a dream, I *had* to come up with a couple of things which hadn't been done before. And thus I did, this is just me ...
Whether I succeeded in it all is up to you in the end.

Continued in the next post ...

[Marcin_gps]

Lampizator aka Łukasz Fikus is Polish guy (same as me). 'The best DAC' of his is highly tuned, lampized Behringer DAC, which is available for everyone. I'm going to try his DAC soon, so I'll let you know how it performs. That is if you trust my judgement haha :D

[PeterSt]

Thanks for the link, IVO. It looks interesting. Lampizator doesn't mention anything about oversampling or (absens of) lowpass filter, though...

PETER: While we are on the subject:
Is your analog section based on tubes, opamps or discrete?
Does it offer XLR balanced outputs?

All the best,

Now it's your turn ...

Let's start with the last one - the easy one;
Both SE (RCA) and Balanced (XLR). In 2 channels, 4 channels or 8 channels (but the 8 channel to choose in advance for SE/Bal).

Now the hard part ...

As said, all is based on 1:1 as much as possible, which includes the culprit of 16/44.1 not *allowed* to be that. But this was solved by Arc Prediction.
As how I started off (see first picture in the Phasure NOS1 topic), it all was as discrete as possible (look at the pile of parts). Now, that it was unbuildable commercially (what about a couple of 1000 parts to solder) is one thing (as it turned out later, because I originally just bought that for myself), but that it didn't sound 100% was another. Not that I knew that back then (November 2008), but I found that during the process at not being able to improve where it should. It was too slow ...


So yes, te discrete solution was too slow to "follow" the 1:1 strategy.
From then on, I started to trust measurements, and while dozens solutions exist for the analogue stage, they all measured just plain wrong. From there I just refused to believe that wrong measurement could sound good, which btw is related to the DAC (board) itself, which just output the THD of the DAC chips itself (after a third version of it). So, how could I destroy that to begin with. This immediately created all the problems, because no single solution would measure good (enough), no matter it sounded good to my ears. And yes, 1 year ago or so I tried an opamp solution as well, that being the only solution which *did* measure right, but sounded like sh*t. Well, don't we think this all ...
(except a few like I think you Pedal ... I think we had a small discussion over it).

:wackco:

Well ... I had an extensive further explanation here, but I didn't like it much. Too many unveiled. So I quit here.
OpAmp.
Without telling how or what exactly. Measures the best, sounds the best.
Now.

[ivo]

Peter,

Some time ago in Phasure DAC section you mentioned that analog stage must be passive one and that op-amps color the sound. So, now you have changed opinion? Or am I mistaken?

Ivo

[ivo]

'The best DAC' of his is highly tuned, lampized Behringer DAC

From this page:http://www.lampizator.eu/szop/L-Szop/BEST_DAC.html it follows that actually it is not really Behringer:

"Putting all my reputation on the table, I do believe that there is no other DAC box sounding even close on the market today. It is already in another class compared even to lampized Sabre Buffalo, Behringer, Satch or Lampucera."

Ivo

[Marcin_gps]

Maybe so or maybe not, check the bottom of this page - http://lampizator.eu/LAMPIZATOR/TRANSPORT/behringer/Behringer.html Maybe blue PCB on his Lampizator DAC is just a coincidence. Also, put attention to his entries, each one of them has in title adjectives like: best, ultimate, ultra etc. It's to enthusiastic imho, but maybe I'm wrong.

[PeterSt]

Peter,

Some time ago in Phasure DAC section you mentioned that analog stage must be passive one and that op-amps color the sound. So, now you have changed opinion? Or am I mistaken?

Ivo

Hi Ivo,

It depends a bit how you look at it. So, to be honest, I did NOT change by opinion, but I changed the environment of the opamp, so the environment now meets the specs of the opamp. In very brief, think like this (and I sure did not know this one year ago) :
When we talk "NOS", this is about the highest transients possible. Like in digital from 0V to 1V in one sample step. If you feed this without active power to analogue, you will get a smoothening implied by the physics of electrics. But if you do this by active means, the "force" of it will create distortion. And, the distortion upon the distortion will create sibilance as how *I* always perceived sound from an opamp; This is (as it appeared to me) not the sound from te opamp itself, but from all which can't cope behind it (including sufficient current supply).

If you again read my posts about the "gaining stage" in the DAC topic, you will see there how it evolved and how I got there. Yes, you already read that, but you didn't know I was talking about an opamp all the time. A year ago already;
All the time it wouldn't work because of other things witholding it from proper "operation" ... which is what I know now. I came to it by accident, at noticing that the (then) new step from 176.4 to 352.8 suddenly didn't sound good anymore. Although this is totally unrelated to the gain stuff, the reason why it didn't work anymore was the same : things couldn't cope. And once I passed that stage, I again tried the opamp, and now it sounded 100% the same as a passive solution (for sound character), although it is also clear that other things improve because of a sheer lack of capacity otherwise (the high frequencies eating the current for the low (!!) frequencies).

Peter

[Telstar]

I think this is the right thread to ask you this: does this opamp output a transient perfect square wave?
What worries me the most about IC is the multi-stage global NFB. afaik all opamps are multistage, but i could be wrong.

Anyway, speed is king.

[pedal]

Congratulations Peter, reaching this conclusion makes you a Mythbuster! The very first generations of opamps didn’t sound good. That is a fact. And many manufactures still use the cheapest ones they can find. (Image the hundreds of cheap Philips opamps inside most mixing consoles in recording studioes, degrading the sound quality of our beloved music, recorded during the 70s and 80s).  Because of this, many audiophiles keeps the myth alive about opamps not sounding good in year 2010.  But the very best opamps of today, for instance the new top of the line from National Semiconductor, measures better and probably sound BETTER than any passive circuit. (The cost and space saved can be used to power supply and other important matters).

My old Dynamic Precision DAC had removable opamps, so I was able to upgrade 2 times. Each time the sound become less grey, more full bodied and with blacker background. This happened a few years ago.

Last year I upgraded the opamps inside my power and preamp, to the very best ones from National Semiconductor (I don’t remember the name. I think there are several ones of same quality, but suitable for different purposes). Holy smoke! They sound more “tubey” than tubes. Extremely detailed in a natural way, with so much bloom and nuances in the treble. Never heard cymbals sounding so good before.

Opamps applied in a DAC can sometimes be a disaster, if it is fed the raw unfiltered signal from the DAC chip. All the HF (noise) will stress the opamp making it sound a little coarse. According to an insider I know, this error has been done with quite a many DACs… But within the right environment, like Peter says, opamps are the best and certainly the most convenient solution.

In case of even better opamps in the future: If the opamp is not soldered to the circuit board, but attached mechanically to a universal adapter, then it can be easily upgraded by the owner without need for shipping back to the manufacturer. (hint, hint). 

[PeterSt]

(I typed the most of the below before Pedal's post, and watched the match before I finished it)

The one I use is more than 3000 times over-rated for 16/44.1.
Before the "data" is through it is corrected already (say 1500 times). At least that was my idea about it (I drew an engineer to tears because of thinking it would work, but as known now, it actually is about the outside). I mean it ... when you are working very explicitly on this matter, you immediately hear it when it does not change the sound. And notice I had the reference already as long (1 year) ... with the passive setup, and too few gain for many of you (as I estimate).

For all the time I had been trying. I always was sure that *if* I could ever find something which was acceptable, I had to make it an option. So, those needing the gain could have it in and would have "acceptable" sound to my own standards, and those who wouldn't need it, well, wouldn't have it. They would be off the best. But it didn't turn out like that, because after all the "tweaks" done, it turned out to be just *better* than without (like Pedal just said). Notice the importance of the setup without preamp, and the necessary drive. It worked very okay without, but it works just better with. This is not to forget, because it is my intention not to use a preamp if you don't want, so it all has to comply to at least that.

Pedal, making it removable ... I don't know. All is within half of a square cm, and it has a reason. But, the whole board is removeable and replaceable, so I did think about that. But also keep in mind my strategy : If it measures good (which means better than the DAC itself) it doesn't make sound. So, it should be more or less true that a better future version does not exist.

For Telstar, see below. 2V in 2 or 3 ns or so. Notice that one channel 192KHz (32 bits) is something like 160ns.

[AHe059]

Opamps applied in a DAC can sometimes be a disaster, if it is fed the raw unfiltered signal from the DAC chip. All the HF (noise) will stress the opamp making it sound a little coarse. According to an insider I know, this error has been done with quite a many DACs… But within the right environment, like Peter says, opamps are the best and certainly the most convenient solution.

Thats right! I just building a new DAC with the TDA1541 and use first a Passive filter and then an opamp with some gain (32dB). Opamps don't like HF signals comming from the clocking and other noise, blocking the noise to the opamps is an essential part of the design.
Suggested opamps: LME49710, AD797, LT1115, OPA637

In the the digital part I use an upsampler AD1896 controlled by a tentlabs clocking so the TDA is running at 192KHz, in my opinion the best solution today.

[Calibrator]

For those looking for a little light reading on OPAmps, the following link may be useful:

http://www.sg-acoustics.ch/analogue_audio/ic_opamps/index.html

At the top of that page is a link ( http://www.sg-acoustics.ch/analogue_audio/ic_opamps/pdf/opamp_distortion.pdf ) to a series of tests performed on commonly available devices. The pdf is ~38MB

Enjoy the read!

Cheers,

Russ