Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[kstuart]

I can see the point when it comes to downsampling, but I canīt see the point doing this in real time. Why not just use a stand-alone resampler and resample the audio files or convert them if needed?
When it comes to upsampling I donīt see the point at all. Could anyone explain whatīs the point with upsampling? (Upsampling in the DAC is another matter)

In some DACs, upsampling with SOX in MC22 is better quality resampling than the hardware chip upsampling in the DAC.

This is highly dependent on the design of the DACs.

Some do not upsample internally, and some have patented upsampling algorithms.

[Asawendo]

In some DACs, up sampling with SOX in MC22 is better quality resampling than the hardware chip upsampling in the DAC.

This is highly dependent on the design of the DACs.

Some do not upsample internally, and some have patented upsampling algorithms.

Noted Kstuart. I'm using iFi iDSD Nano and Chord Hugo for my Laptop. Both of them seems to prefer SOX enabled in MC22. The difference is not night and day but on the specific song it become cleaner and clearer.

[Magellan]

Because then I would have to store multiple copies of the same files to accommodate different needs at different times.

OK. That make sense.

[Magellan]

In some DACs, upsampling with SOX in MC22 is better quality resampling than the hardware chip upsampling in the DAC.

This is highly dependent on the design of the DACs.

Some do not upsample internally, and some have patented upsampling algorithms.

But in real life, this seems to be a very theoretical issue. Considering the impact of speakers, listening room, amplifiers, source material (i.e. recording quality) and the design of the analogue parts of the DAC itself, it seems to me the upsampling algorithm in the DAC has to be very poor if external upsampling makes a significant difference.

If the DAC do not upsample internally, then I guess it probably wont play the upsampled audio either, i.e. the only reason I can see for say a 24/192 DAC is the need for internal upsampling.

[kstuart]

But in real life, this seems to be a very theoretical issue. Considering the impact of speakers, listening room, amplifiers, source material (i.e. recording quality) and the design of the analogue parts of the DAC itself, it seems to me the upsampling algorithm in the DAC has to be very poor if external upsampling makes a significant difference.

That sort of analysis almost never works for sound quality issues.

If you have an imperfection early in the reproduction chain, then later stages - analog output stage*, amplifiers, speakers, and listening room only magnify the imperfection.

And thus, source material is the most important.  (This is why I have no interest in Tidal, because for pre-1980 recordings, as a general rule - with some exceptions - the earlier the mastering, the better the sound quality.)

----

* The DAC that I use comes in two flavors - one with a delta-sigma DAC chip and an elaborate analog output stage, and one with a multibit DAC chip, patented internal upsampling and (due to board space contstraints) a more simple analog output stage based on an IC-chip provided by the DAC chip maker.  The latter sounds significantly better, especially on 16/44.1 material.

[pschelbert]

Hi

yes, if there is a good DAC (all good impelemtations of oen of the four leading DAC Chip producers), you will not hear a difference.
I made some blind-tests with several "HighEnd listeners" and even with a mediocre DAC people could not tell a difference between the original and the DA-AD converted file.

I do not believe in multibit DAC as better, they are clearly lower quality today. Sigma-Delta is the leading edge, no doubt. Superb performance if hardware is done as the eval board.
DAC are so high quality at a relative low price its not worth to think to much about. May be the most important thing is to have good sound and no opertating system issues is using ASIO (in PC). If the DAC does have ASIO drivers, its okay.

Oversampling, downsampling is nice for some trials or if the source material does not match the DAC. In modern DAC you can't expect a sound improvement by upsampling.
I would like to try just what a difference minimumphase filters versus linearphase filter make. However it can be done in a different way (offline with SOX), just a bit more circumstances.

Peter

[blgentry]

I do not believe in multibit DAC as better, they are clearly lower quality today. Sigma-Delta is the leading edge, no doubt.

That's a bold statement.  Schiit DACs don't use old multi-bit chips.  They use modern technology that's not Delta Sigma.

Brian.

[kstuart]

That's a bold statement.  Schiit DACs don't use old multi-bit chips.  They use modern technology that's not Delta Sigma.

Brian.

Brian, the current state of the world is that people with very limited understanding and information, believe they know something.  The Internet gives them a whole group of people with similar misunderstanding to say "we are right".

One of the Schiit designers uses the saying "In my system.  With my recordings.  And I have not heard everything out there".

In contrast, the people whose ego leads them to believe their reading of articles with simplifications of digital technology means that they understand, are the ones saying "this sort of thing cannot sound better than that sort of thing".

[BillT]

Of course, there's another view on this. The ear has very limited abilities, the brain has an amazing susceptibility to suggestion.

It is very difficult to run valid comparison test, it's very easy to run invalid ones.

Occam's razor would suggest that electronics have developed well past the ability of the human ear to detect differences but suggestibility is still alive and well.

[pschelbert]

That's a bold statement.  Schiit DACs don't use old multi-bit chips.  They use modern technology that's not Delta Sigma.

Brian.

Hi

Schiit specifies in which product they use what chip.
Each type of DAC uses a different chip, some modern, some older designs.
The akm-chip they use is most probably a Sigma-Delta (32Bit). The others are Ladder (14, 16, 20Bit), some EOL (not for new designs according to chip manufacturer).
See on their website for details.

Peter

[pschelbert]

Of course, there's another view on this. The ear has very limited abilities, the brain has an amazing susceptibility to suggestion.

It is very difficult to run valid comparison test, it's very easy to run invalid ones.

Occam's razor would suggest that electronics have developed well past the ability of the human ear to detect differences but suggestibility is still alive and well.

Hi

yes the point is, and up to now nobody could show otherwise: what you can hear, you can easily measure.
Double blind tests are very seldom done. Mostly professionals do that, like Harman Kardon and others. See publications for example in the AES-journals.

Bias is so strong if you do not double-blind tests that the result is completely useless.

Peter

[blgentry]

Schiit specifies in which product they use what chip.
Each type of DAC uses a different chip, some modern, some older designs.
The akm-chip they use is most probably a Sigma-Delta (32Bit). The others are Ladder (14, 16, 20Bit), some EOL (not for new designs according to chip manufacturer).
See on their website for details.

I'm quite familiar with Schiit's DAC lineup.  All of their multi-bit DACs use chips that have never been used for audio before.  These Analog Devices chips are intended for medical imaging and military targeting applications.  But Schiit has adapted them for audio use, since they are the only modern multi-bit chips that Schiit could find that are being manufactured currently.  You're claiming that some of these are End Of Life by the DAC chip manufacturer?  Which ones?

Brian.

[pschelbert]

I'm quite familiar with Schiit's DAC lineup.  All of their multi-bit DACs use chips that have never been used for audio before.  These Analog Devices chips are intended for medical imaging and military targeting applications.  But Schiit has adapted them for audio use, since they are the only modern multi-bit chips that Schiit could find that are being manufactured currently.  You're claiming that some of these are End Of Life by the DAC chip manufacturer?  Which ones?

Brian.

I read the 14 and 16 bit ones, but I am not sure.
Nothing against using DACS from medical, machinery for audio. However they are lower quality in terms of audioperformance (precision 20Hz-20kHz).
These Ladder DAC shine if it goes to DC-performance (which Sigma-Delta cannot do).
Overal I see no benefit, not pricewise nor technical.

Modern sigma-Delta have oversampling and filtering very much optimized, even selectable filters. I doubt a separate solution could do better.
To test, you can DA and then AD-record and put in on a drive for others to hear.

There are others which use such Ladder DACS like Audionote and Lampizator (some say NOS, non oversampling).

Peter

[kstuart]

Of course, there's another view on this. The ear has very limited abilities, the brain has an amazing susceptibility to suggestion.

It is very difficult to run valid comparison test, it's very easy to run invalid ones.

Occam's razor would suggest that electronics have developed well past the ability of the human ear to detect differences but suggestibility is still alive and well.

The phrases "very limited", "amazing", "very difficult", and "very easy" have zero scientific content.

You are convincing yourself that your opinion is "scientific" when it is just what you want to believe.

The quality of consumer audio electronics is circular.  Certain things are measured, and then devices are designed to perform well in those same measurements.   The other aspects of sound quality that are not covered by the measurements are ignored - because most people cannot tell the difference simply because they are not interested.

[kstuart]

To test, you can DA and then AD-record and put in on a drive for others to hear.

What?  How is adding a sound-deteriorating AD step make something a test??