PCI soundcards

[manisandher]

Still 

Hey Peter, I thought the BD-Design CrazyAs that you're using were chip amps, based on the LM3886. If they aren't, then I apologise profusely...

In any event, I've been meaning to do a write up of my experience coming over to your place. Unfortunately, work is just too busy right now (I'm currently in Barcelona, preparing to speak to 600 top guys from Philips on Tuesday!). But I will say that I loved the sound of your setup...

Mani.

[PeterSt]

Hi Mani,

Of course those are chip amps, but I don't call that opamps.
And you know what ? these chip amps measure near 1% harmonic distortion (max) a bit depending on how they are used (gain). Still they sound neutral to me, but they are FAST (which is the consistent objective in my current setup) and they have a nice 200KHz bandwidth. Also, they are completely without audible noise (although it sure can be measured) and they use 15W or so, idle. The latter was the real objective, and after trying them all in that leage (except for your Hypex) they just were the best. Not by a little, but by far, because each of them (think class D) has a very nasty behaviour some way or the other. We (Bert, me, Gerner) really tried and compared over many sessions, but they do something to the sound, hum, again annoy, may blast your fuses, shut off your TV and whatever you can think of with this (very) high frequency operations. With this (GainClone) just nothing seems to be wrong. Also take into account that this search was against my existing SS and Bert's tube amp of which he was sure it couldn't be replaced with any saticfactory other type. The point is (I think), once you are, say, rather far in realistic music reproduction, you are over a hurdle which allows you to judge other things. There's no subjectivity anymore, and -like I said before- there is no way two people, no matter their subjective tastes, won't agree on what you perceive and you will *know* automatically what is better and what not. With *is* I mean *is* because it has become easy to compare with normal live instruments. They are not subjective, they *are*.

While I am blabbering around anyway, here is another one I found only recently (last few days), and I didn't know of the phenomenon before :

I realized an improvement which is theoretically jitter-related (in the DAC of course). Btw, I tried that before, but didn't notice anything special at the time. The XXHighEnd version was older though plus this was Vista. Now it is W7 (Engine#3). Now :

I played a jazzy album, and for a first time I noticed how people talked through their horn instruments. While at first I thought someone was talking to me, I realized that the musician was manipulating his sax so it had a kind of voice. You probably can imagine this better with a trumpet (Jazz at the Lincoln may be an example). However, since then this came over me dozens of times, and it is just the way the musician puts his feelings in the instrument. It is really funny sounding, and I never heard it through loudspeakers.
So, this is another proof of the so many steps of improvement potentially there. It will never end I guess (and hope !).

Peter

[pedal]

An OpAmp sounds "mathematical". It has sibilance. It annoys. But this is nothing much new.

Today this is a myth. Opamps used to sound bad, but that is history. I have the very recent opamp models from National in both my preamp and poweramps. The treble is silky and sweet, with body and details like a (quality) tube amp! 

Interesting read by John Atkinson in Stereophile: http://stereophile.com/news/072307national/#

[PeterSt]

Uhm ... I am not sure what the message is, apart from you liking opamps ? Especially the reference to that article makes me wonder whether you have read what I said about it. No matter the ultra low distortion, no matter the ultra high bandwidth, no matter the slew rate being far over suffcient ... they-do-not-sound (but read : we do not know how to measure it --> please dig into Arc Prediction and measuring, which tells us that we measure the exact same in two situations, which are in the base TOTALLY different. And hear hear, why does A.P. sound better ? because I anticipated on this (false means of measuring)).

But let's finalize this by stating that it is of not much use to *talk* about sound. The only thing I can say is that I hope to know what I'm doing which is based on theories and science first, but on empirical findings next which sadly overrules science. That this needs some recreation of science I might not be able to, is another matter. Haha.

But I have a tip for you : Why do you think the internet is overloaded with means of I/V stages (for DACs that is) ? because a 2c opamp can do the job as well ?
Or am I talking non-sense now ? Yes, your old Philips had a 627, but this was then. You won't find anything like that in a current self-respecting DAC, no matter specs have gone sky-high.

Oh, we were talking amps.
Do you think it makes a difference ? or is it perhaps so that all is much more important at the uV level because everything which is amplified after that goes extra-wrong once it *is* wrong.
I explicitly (and ignorantly !) tried, but it can't work.

Peter

[pedal]

But I have a tip for you : Why do you think the internet is overloaded with means of I/V stages (for DACs that is) ? because a 2c opamp can do the job as well ?
Or am I talking non-sense now ?

And I have a tip for you too:
It’s true that many DAC’s implement opamps in a bad way. The explanation is that (from the beginning) many DACs were build by amplifier companies without much DAC knowledge. The fashion was to buy a semi-finished solution. (Chip suppliers like Crystal, sold integrated chips with DAC and opamp). The only thing missing, was the filter, which the end manufacturer could add by own choice.

The problem then, is that all the HF distortion which a DAC chip outputs, enters unfiltered into the opamp. It’s a severe punishment for any opamp, and they start to sound unnatural.
The solution is to change the order: first DAC, then filter, then opamp. This takes more engineering work, but let’s the opamp perform optimal.

SOURCE: An amplifier manufacturer, producing a DAC in the 90s, who actually knew what he was doing.

-------------

To repeat myself: The best opamps sound wonderful when implemented correctly!

[PeterSt]

Ok, so I'll give you the benefit of the doubt. I'm not sure why, but why not ? haha

But maybe you are around some day. Bring some ties to eat. I'll have a few ready for myself. One can never be sure !

Thanks pedal, always nice to talk to you !
Peter (with greetings from 2010)

[pedal]

But maybe you are around some day. Bring some ties to eat. I'll have a few ready for myself. One can never be sure !

Thanks pedal, always nice to talk to you !Peter (with greetings from 2010)

BTW: I'm not an opamp fundamentalist, but last year I upgraded all the vital opamps inside my pre- and poweramps. I went from the previous "best" models to the very last new generation opamps from National Semiconductors (se link above). The SQ improvment was quite big. The change reminded me somewhat like the sound of a very, very expensive tube preamp. The previous sound was in comparison more "ordinary", the new SQ was slightly "darker" but more organic, more resolved, more body. Fex. cymbals sounds so much more "rich" AND detailed.

I've been a little offline last year, due to kids and work, so havnt been so active here. But it's nice to see that XX gain more and more satiesfied users.

At the moment I'm enjoying the Sabre DAC, but consider me a potentional customer when your filterless DAC is ready for shipping!
You are located in Holland? I have some business there. Not impossible I'll pay you a visit later on.

[Leo]

A quick feedback note on the Cantatis card; ...............
I have only used the card with y5-02 and Engine #3 so far. To put it simply, it sounds very, very good. A different league than the Onkyo and Audiotraks actually. Detailed, snappy, rhythmic sound; tuneful bass and great voices. Films sound much better as well, words are easier to pick up. This card looks like a real killer. And, oh, stereo separation seems much better than the Onkyo, everything is more "there". Big ups to Cantatis.

Hello Fidelio,

I cannot get the Cantatis card going with upsampling, it seems that the Cantatis drivers will only work with fixed sample rates and 88.2 and 176.4 are not in the list I get. Any suggestions (again) ?

thanks,

Leo

ps
in 44.1 it is allready very impressive SQ wise

[Fidelio]

A quick feedback note on the Cantatis card; ...............
I have only used the card with y5-02 and Engine #3 so far. To put it simply, it sounds very, very good. A different league than the Onkyo and Audiotraks actually. Detailed, snappy, rhythmic sound; tuneful bass and great voices. Films sound much better as well, words are easier to pick up. This card looks like a real killer. And, oh, stereo separation seems much better than the Onkyo, everything is more "there". Big ups to Cantatis.

Hello Fidelio,

I cannot get the Cantatis card going with upsampling, it seems that the Cantatis drivers will only work with fixed sample rates and 88.2 and 176.4 are not in the list I get. Any suggestions (again) ?

thanks,

Leo

ps
in 44.1 it is allready very impressive SQ wise

Yes, there is a driver issue, at least under Win 7. It cannot do auto switching, and thus no upsampling in XXHE (except e.g. 96khz for 48khz material of course).

Try getting the standart VIA control panel running and see if it works then:)

[Leo]

Well the helpfull folks at Cantatis prepared a special version of the software and drivers  for Windows 7 which allows for 88.2 and 176.4. Which is impressive treatment of me as customer I think. The only thing is that it is 16 bit as they state that Windows 7 does not allow for 24 bit with these frequencies.

quote:As in Andrew's previous email, Windows 7 only offers us 88.2kHz and 186.4kHz in 16-bit format.

If you want to get the highest quality out of your system then you can set the Overture card to 192kHz, 24-bit (and make sure you do the same in the Windows 7 control panel).

Windows 7 always performs sample rate conversion so your application should still work.

I don't think that is the way Peter sees it but I don't think I could find the correct wording to convince the very nice people of Cantatis (but still ignorant with respect to XX Highend of course )

So any suggestions for the reply to get them on the right track to give me and other Cantatis+Windows7+XX Highend users 176.4 kHz + 24 bits ?

thanks,

Leo

[PeterSt]

I'm not sure what to say.

Of course W7 has no problem with 24 bits (or 32) at any rate, and of course W7 doesn't resample as long as you don't allow the W7 audio engine to jump in (meaning : address the device exclusively and it won't).

Windows 7 always performs sample rate conversion so your application should still work.

Indeed. Also when (in Shared Mode) the output is set to 24/192 and it is fed with 24/192.

As in Andrew's previous email, Windows 7 only offers us 88.2kHz and 186.4kHz in 16-bit format.

It really beats me where this is coming from.

The only thing Cantatis must do is allowing for automatic sample rate setting, and don't think about it. All other applied "thinking" will (obviously) be in the wrong direction. Vista/W2008/W7 just digs everything, as long as Exclusive Mode is used. Do NOT look at those possible settings, which are output settings where the OS resamples to, at your command. and only in Shared Mode.

I hope this helps !
Peter

[Fidelio]

Well the helpfull folks at Cantatis prepared a special version of the software and drivers  for Windows 7 which allows for 88.2 and 176.4.

Could you upload this software so I could try it m8? Also, what do you think of the Cantatis' analogue outs?

[hubsand]

Those Burson OpAmps are seriously well above the calibre of every other op-amp you can shake a stick at. I auditioned dozens in an offboard DAC recently, and the Bursons were night and day better than, for instance, dual OPA627s. As far as I'm concerned, everywhere you can put a Burson OA, put one in: they're just amazingly transformative. If all OAs sounded that good, they wouldn't have the iffy reputation!

I find it hard to believe that the Burson-ised card would just thrash similarly specified competition . . . looks like we need a UK 'bake-off'

As much of an issue as the parts spec is the power supply, though: the same people that stock the modded HD2 cards now do a linear regulated PSU for use with a Mini ITX board: http://www.itemaudio.co.uk/media_server_pc.html

[Leo]

Well the helpfull folks at Cantatis prepared a special version of the software and drivers  for Windows 7 which allows for 88.2 and 176.4.

Could you upload this software so I could try it m8? Also, what do you think of the Cantatis' analogue outs?

Hi Fidelio,

it is in the attachment !

Leo

[Fidelio]

Thx m8!