Computer based Hifi

[lazycatken]

Originally Posted by Idiocratease

Hi Lazycatken,
But I would say lack of depth in the mid to upper frequencies was noticeable. I listened to bit of KD Lang's Ingenue album, and the depth and articulation of her voice seemed to suffer a little. Not a huge amount, but enough for me to be sure there was a difference...

Thanks, this information is good enough for me...

In the post, ZeroGain Forum - View Single Post - Rob Watts on the DAC64, Michaelab mentioned coaxial cable was
Just a touch less detail though, things ever so slightly less well defined as with the TOSlink
Supposedly, higher jitter(ie, Toslink) could result in less detail/definition, but the result was not!

And he mentioned Toslink was
everything sounds terribly "thin"
coaxial was
Lots more drive and punch in the bass aswell and none of that horrid treble harshness.
I got similar experience, using oversampling, from 44.1KHz to 88.2KHz(you know oversampling results in less jitter component), I heard much "thin" sound and less bass/punch. But I would say, it's much close to the real recording space/room!
Less bass/punch may be due to more clear and defined. So as "thin" & treble.
Perhaps, Michaelab's system tended to be too much treble... And he may prefer sound with much bass...

Per the difference you heard, is it possible due to electrical noise carried by coaxial cable?
You know, sometimes, better doesn't mean more precise. Replace a specific component of a system with an ideal one, may result in a not-ideal system, depends on other components of the system.

[ChrisA]

Originally Posted by lazycatken

Supposedly, higher jitter(ie, Toslink) could result in less detail/definition, but the result was not!

I don't think jitter could be an issue with S/PDIF. The data are packetized and needs to be re-clocked at the receiving end. S/PDIF is sent in blocks, each block has some control words at the front of the block that define what follows. It is not un-like the way CD-ROM data must be buffered before being clocked into a DAC.

Have we ruled out any bias with the listener? Did he know which cables were in use while listening?

[Idiocratease]

Originally Posted by lazycatken

T
Per the difference you heard, is it possible due to electrical noise carried by coaxial cable?
You know, sometimes, better doesn't mean more precise

I got similar experience, using oversampling, from 44.1KHz to 88.2KHz(you know oversampling results in less jitter component), I heard much "thin" sound and less bass/punch. But I would say, it's much close to the real recording space/room!
Less bass/punch may be due to more clear and defined. So as "thin" & treble.
Perhaps, Michaelab's system tended to be too much treble... And he may prefer sound with much bass...

I certainly take your point about more precise (=transparent) not necessarily being better... I have for many years used an Audio Note DAC 2.1 Signature, which I think is not at all hi-spec, but does sound wonderful. I am prepared to concede this this could be what's going on, but judgement/feel tells me that it's not the issue.

When you mentioned upsampling, I'm interested to know what platform you upsampled on. Was this a PC, or some other device ?

[Idiocratease]

Originally Posted by ChrisA

I don't think jitter could be an issue with S/PDIF. The data are packetized and needs to be re-clocked at the receiving end. S/PDIF is sent in blocks, each block has some control words at the front of the block that define what follows. It is not un-like the way CD-ROM data must be buffered before being clocked into a DAC.

Have we ruled out any bias with the listener? Did he know which cables were in use while listening?

I think that my description and my mention of the placebo effect should have addressed listener bias questions. To be honest, I think it's a pretty cheap shot, when people are trying to discuss interesting observation.

If you had read my original post, you would have seen the Coax cable specified, and the optical was some inexpensive monster item.

[sandyK]

Ideocratease
Sorry for delay. I didn't get an update email for this thread.
Have a look at the attached link. In the U.S.A. a while back, the Editor of Computer Audiophile, organised a Symposium which included DAC designers, software experts, Recording engineers, Chesky etc.
Check out especially,the reply "Hi Paul, I definitely hear ..." from Chris Connaker, the Editor of C.A.

SandyK

Does SSD sound better than Hard Disk? | Computer Audiophile

[phofman]

Originally Posted by ChrisA

I don't think jitter could be an issue with S/PDIF. The data are packetized and needs to be re-clocked at the receiving end. S/PDIF is sent in blocks, each block has some control words at the front of the block that define what follows. It is not un-like the way CD-ROM data must be buffered before being clocked into a DAC.

SPDIF is carrying the master clock, which is recovered via synchronous PLL, rather sensitive to the incoming jitter. It is the opposite of CD-ROM/HDD data processing which involves feedback control of the reading speed (unfortunately SPDIF offers no feedback).

Unless asynchronous reclocking is used which is a completely different story.

[phofman]

Originally Posted by sandyK

Does SSD sound better than Hard Disk? | Computer Audiophile

Perfectly possible since SSD puts way less noise on power supply lines. Certainly not a function of timing unless it affects the level of noise in the case - the sound card has its clock and the rest is slaved to it via software/interrupts.

[lazycatken]

Originally Posted by Idiocratease

IWhen you mentioned upsampling, I'm interested to know what platform you upsampled on. Was this a PC, or some other device ?

My source is MacBook Pro. DAC is an AKM AK4380 evaluation board, powered by a car battery. Fiber cable in between Mac & DAC.
Mac optical output can be set as 44.1/48/88.2/96KHz.
I tried to play music by iTune. 88.2 is better than 44.1.
And set output as 88.2. Then played music by Audacity 1.3.9. Made comparison between fast & high-quality sinc sampling rate conversion setting. High-quality sinc was indeed better.

The improvement is thoroughly, clear/transparent/precise/detail/emotional...
Although, not much, but audible. Even a non-audiophile can easily recognize the improvement.
Once you heard 88.2, you'll never back to 44.1.

Perhaps, CS8414, the S/PDIF receiver, can't generate good enough clock... Thus, I'm thinking about building a buffer board... It might be based on FPGA or micro-controller.

I also tried this on an Accuphase DP-500 CD player, which can work as a DAC. Got the same result. You know this player is not cheap...

[ChrisA]

Originally Posted by Idiocratease

I think that my description and my mention of the placebo effect should have addressed listener bias questions. To be honest, I think it's a pretty cheap shot, when people are trying to discuss interesting observation.

If you had read my original post, you would have seen the Coax cable specified, and the optical was some inexpensive monster item.

Not a cheap shot. It was asked as a question because few listeners take the time to collect statistically valid sample sizes.

But I'm certain you can eliminate the possibility of jitter causing the problem. The S/PDIF data protocol introduces HUGE jitter. For example the left and right (and any other) channels are sent alternately.

If there is a different then it must be something else. Likely, almost certainly the same DAC is used for the RCA and optical jacks. I would expect only one set of buffer memory and clock too.

S/PDIF is self clocking and the data is the same even if the bit rate doubles. the speed of the its can vary over a huge range with no effect.

Hard to think what could case the effect. Possible the RCA cable could have some kind of ground loop effect. The optical cable would not be conductive.

The described effect is "thinness" and not "drop-outs" or "static" so It's not a failure of a digital link but analog effects. Possably the RCA or optical jacks load the power supply differently. Could the device sense which cable is in place and switch off the other jack? The receivers on the jacks the DAC chip might share a power supply. I'd be inclided to beleive their is a small design problem (or shortcut) with the equipment.

If the problem is really in the data transition then you'd have to have a theory that could explain it. When the data are packeted and channel multiplexed it is hard to think it hard to explain how data errors can cause the observed effect

[phofman]

Originally Posted by ChrisA

The S/PDIF data protocol introduces HUGE jitter. For example the left and right (and any other) channels are sent alternately.

I2S multiplexes the channels too. Multiplexing does not cause jitter.

Originally Posted by ChrisA

Likely, almost certainly the same DAC is used for the RCA and optical jacks. I would expect only one set of buffer memory and clock too.

SPDIF receiver does not use a stable clock and buffer, but the clock is recovered from the incoming SPDIF stream via PLL.

Possible the RCA cable could have some kind of ground loop effect. The optical cable would not be conductive.

Most decent sound cards use an isolation transformer for SPDIF coax output. I would assume the same holds for CD players.