Jitter a myth too?

[richidoo]

http://www.novaphysicsgroup.com/NonTechRURguidePAGE8.html

[Carlman]

Thanks for the fun read.  Weird company but some interesting info.

Jitter's not a myth, just an easy way to describe a 'digital sound'... or 'digititis' or whatever makes it sound non-analog. 

Their site explains that playing from a hard drive sounds better, which isn't really news. Then they're selling something like an Olive or one of the memory players out there... but I don't see prices or how to buy it.

On a side note, if you ever wanted to know what today's date was, just scroll through the article.  You will be reminded frequently and somewhat randomly.

[miklorsmith]

Yeah, weird.  I think we should string Reed and Soloman up by their toes.

It looks like they think streaming cures the problems though I imagine any streamer is going to have some kind of clock, does this re-create the problems they're trying to avoid?  Once a disc is ripped, however it's ripped, are the error correction codes copied to the drive and kept with the audio files?  If so, are they copied/synthesized upon burning a copy?  Or are the error codes embedded with the audio file and proprietary Nova Physics hardware and software are needed to extricate the demons?

I think the Memory Player is around $10k, if memory serves.

[_Scotty_]

It seems to me that the error codes are embedded in the music data and if we can burn a CD from the image file on the hard drive we must also be transferring the EEC data as well so that the copy can be played back on a CD player.The other alternative is that the software used to burn the CD copy supplies the code. It would be nice to hear from someone who knows what is going on. I have just taken possession of a highly modified Cambridge 740 C and all I have to add to playback files from my laptops hard drive is a USB to SPDIF converter to interface with the DAC section in the 740C. I haven't had a problem with digititis for several years. In fact,the modified DVD player I have been using has shown the need for redesigning my phono stage,it's always something.
Scotty

[richidoo]

http://www.novaphysicsgroup.com/RetailPricingPage6.html

Starting $12,500 for transport only.

[richidoo]

Doesn't all this hinge on the optical disc media? The original digital recording doesn't have error codes in it before it is mastered to disk. Once we start downloading music directly from the publisher the difference should be obvious.

I have some original digital recordings. It would be interesting to burn them to CD, then rip back to WAV and compare files bit for bit. Doesn't a CD-RW drive write the error codes also? Nova implies that ripping and streaming doesn't remove the ECC, don't they? Their machine takes 10 minutes to rip a CD!

SB buffers the stream in memory, what's different about that? No DAC has a provision for removing embedded ECC errors, it just takes PCM and DACs it.  I think the optical reader portion of the chain removes the error codes.  I think they have found something interesting, but the guy doing the explaining doesn't fully understand what is going on technically - my .02  The case they make is not linear and logical.

[_Scotty_]

If there is another way around error correction, which is what they have identified as the problem,they certainly aren't going to tell us we can use the data stream from our computer and bypass their $12,000
solution to the "problem".
It looks like we still need a thorough explanation from someone without a vested interest in the contents of our wallets.
Scotty

[mgalusha]

From what I have read the ECC is not stored in the audio but is part of the physical sector stored on the disc, so the raw WAV file won't have any of the ECC as part of the file. Programs such as ECC and dbPoweramp reread the sectors to make sure the bits extracted match and don't rely on the redbook ecc. At least I'm believe this is correct.

I'm sure the nova physics player works well but once the data is extracted to a WAV (or whatever) and written to a hard disk I don't see how the bit order is going to change. I could be wrong but unless something corrupts the file the computer is going to read it from start to finish in order. It treats a WAV just like any other data file.

[richidoo]

Recent TAS magazine has a broiling letter to editors section (as usual) where Harley defends the concept of jitter hurting sound quality, that jitter as low as 5ps is audible. Seems reversing bits (likely not just one at a time) would be even more audible and become a known problem after 30 years. I have never heard it mentioned.  Of course, error correction prevents it in optical media. Other media like hard drive are more stable, I guess.

One feature I noticed on the linked page was that when the stream is interrupted due to the error correction being ignored during a scratched section of the disc, the player just plays silence instead. Well silence could be a gigantic transient if it happens at high volume at the peak of a highvoltage signal. If a scratch causes multiple short patches of "silence" to beat the hell out of my amp and speakers I wouldn't be too happy. The blurb also proudly states that the silent patches only happen 2% of the time. I think I would rather have "faked" music instead of 2% of the music turned to silence. That's every 50th note blanked out. The guy writing the hype is not as technically sophisticated as the theory they are trying to sell.

Scotty's right, though, I hope to learn more about the transport process and this Nova innovation.

[mgalusha]

I too think jitter has an impact. When I was replacing the input/src and clock in the DCX's I was able to A/B the stock clock vs the new clock and the improvement was easily noticed with the music seeming to have a better flow and having a more analog type sound with less of the irritation that digital can bring. In the DCX the new clock allows everything (PLL, SRC, DSP and DACS) to run synchronously at 96Khz from that clock as the SRC is changed at the same time. So I'm sure that had something to do with it as well. The clock is supposed to have 2.6ps of jitter from 10Hz to 20Mhz.

Since the incoming signal is being re-clocked I've heard it said the jitter on the S/PDIF input is supposed to matter less but I have no way of verifying this. I can say that different digital cables affect the sound, so I'm guessing it's not immune.

mike

[richidoo]

That's interesting, that the internal clock is not the whole issue. Altmann uses a scheme to filter out jitter but rollo reported fancy digi cables were a further improvement.   A digital cable that doesn't roll off the edges of the square wave data signal is supposed to sound better too. I forget why, but digital signal transmission is a big factor.

[_Scotty_]

The only way the bit transition from one to zero can be reliably recognized by the input receiver is if the squarewave has no distortion or rounding of the edges of the wave form. If a one is interpreted as a zero because the edge is rounded off then you have a timing error at a minimum.
Scotty

[rollo]

If there is another way around error correction, which is what they have identified as the problem,they certainly aren't going to tell us we can use the data stream from our computer and bypass their $12,000
solution to the "problem".
It looks like we still need a thorough explanation from someone without a vested interest in the contents of our wallets.
Scotty

Scotty could not agree more. Long live the LP.


charles