SACD 2 channel vs Redbook 2 Channel

Ritteri writes:

Its a well known fact that the 99% of the SACD's on the market have been "remastered" to give the illusion of "better sound".

What about brand new DSD recordings such as Telarc have produced. These are not remastered.

Sorry, but I have yet to hear ANY SACD player put out a better musical signal than a competently built "redbook" player.

What players have you heard?

Regards,

Ritteri writes:

I dont think there are even close to 500 out.

I have 600 SACDs. If I was more of a classical music buff I'd probably have a lot more.

Regards,

Ritteri writes:

Still some that arent even in English, alot are not even music CD's(or 2 channel)native to this country and alot of these arent even true SACD's with the higher upsampling.

Many SACDs are not native to the United States of America, although that's hardly surprising given that when SACDs were first released only three plants existed - one in Japan, one in the USA, and one in Europe. Sonopress in Germany was the first plant to produce hybrids, so many SACDs came from there. That matters not one iota. It's a global village.

By "true SACD" I'm guessing you are referring to recordings that were made with DSD right through the chain. There have been some, but it is only in recent times that expanded mixers that operate in DSD have become available. We are sure to see many more completely DSD SACDs in the future.

I personally don't think this matters much. I have excellent sounding SACDs made from analogue recordings and various resolution PCM recordings.

Regards,

Ritteri writes:

I believe it can reproduce a perfect signal up to 22khz.

That's incorrect.

What follows is a simplification, but you'll get the idea. Members, feel free to correct me as I'm only an enthusiastic amateur and am keen to learn.

In order to avoid "aliases" (byproduct of the sampling) when converting the original analogue signal to digital, no signal at half the sampling frequency must be present. Since the Redbook sampling frequency is 44.1kHz, this means that no signal must be present at 22.05kHz.

Let's say there was a signal at 24kHz. Sampling would produce an "alias" - an artifact - at 12kHz, which you can hear. Clearly we don't want this to happen. So the signal must be way down in level at 22.05kHz.

Yet, to have accurate reproduction to 20kHz (the nominal limit of human hearing), we want normal signal strength (whatever there is in the performance) at 20kHz.

So the signal must be passed through a very steep filter which is not affecting the signal at 20kHz, and is 90dB down at 22.05kHz. The famous "brick wall".

Oversampling attempts to overcome this problem. If the sampling frequency is (say) 88.2kHz, then we have to pass the signal through a filter that is flat at 20kHz and 90dB down at 44.1kHz. Still pretty steep and difficult to make without nonlinearities, but doable. Now we need a mathematical algorithm to choose (essentially) every second sample point and save the amplitude, thus making the digital recording.

Let's say the studio is recording digitally at 96kHz and 24-bit words. They make the recordings, mix it in the digital domain, and now they have to prepare it for the Redbook format. Lots of very funky mathematics to convert down to 16/44.1.

Consider now a studio recording in DSD. They make the recording, mix it (DSD mixers are more available now) and put that on the disc.

Similarly with DVD-Audio. The studio could record stereo at 192kHz, probably mix digitally at that resolution, and save this on the DVD using lossless compression.

In principle, both are superior to Redbook.

Regards,