Good analysis....

This ties in quite nicely with the guy who used to post on the Digital Asylum a lot about how the sound of a CD player changed depending on how you navigated to the track. He had a whole complex set of rules about how to get the best sound from the player.

It is certainly within the realm of possibility that there could be something to this (pre-loading of data) with a modern sigma-delta DAC chip (although it's hard to imagine how the effect could last more than a few milliseconds!). And I don't really see how it could have any effect with the "older" R-2R "ladder" DAC chips such as the PCM1704.

But it certainly couldn't hurt to do a bit more listening and see what can be learned. It doesn't hurt to keep an open mind. We can always laugh later if we decide the entire theory is rubbish...

RE: Interesting!, posted on November 4, 2009 at 11:01:27
audioengr Manufacturer Posts: 4088 Location: Oregon Joined: April 12, 2001	If you want some test tracks, email me. I can put them in "Dropbox" for you. Steve N.

"audio files with the same md5 checksum that sound different", posted on November 3, 2009 at 00:11:46

Gevo
Audiophile

Posts: 51
Location: California &
Joined: June 30, 2009

This statement is flawed and confusing. If there are *any* differences in the header/padding between files, the MD5 will be different. If you actually take out just the audio data and create new files, sure the MD5 can be the same for these new files. But to make a proper WAV file, you need header info, and if its different between files (quite possible if you use different rippers), then MD5 will be different.

Not., posted on November 2, 2009 at 18:14:28

Tony Lauck
Audiophile

Posts: 3048
Location: Vermont
Joined: November 12, 2007
Contributor
Since: February 24, 2009

Seems like pretty poor analysis to me. For starters, MD5 checksums are calculated on complete files, and will be different unless the two files contain the same identical bits and have identical length, unless world-class cryptographers have been hacking the files. Based on the quality of the one post, one has grounds to question the entire thread.

If something happening at the start of a track affects the sound beyond a fraction of a second then something is seriously broken in the playback process, probably some kind of software or firmware bug. This seems unlikely. I think this is more a case of delusion coupled with group-think.

The people proclaiming that these offsets make a difference have some serious work to do, if they expect to be taken seriously by anyone but fools. Either that or they need to get new equipment.

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

I think it is., posted on November 2, 2009 at 19:28:32

alan m. kafton
Dealer

Posts: 3941
Joined: April 7, 2000

While I don't have the textbook credentials you have, I found the information put forth to have merit, and logic, in addition to the very idea that someone is looking *elsewhere* for a reason for these perceived/heard differences in software.

On the other hand, if the process doesn't meet your criteria (whatever that may be), you always presume that "something's broken". You did it again....and that solves nothing.

The poster is on to something, whether *you* agree with it or not. No man is the sole repository of knowledge, (which is) in contrast with the positions your posts always seem to put forth. If you're so sure, conduct similar experiments and prove him wrong.

RE: I think it is., posted on November 2, 2009 at 20:50:11

Tony Lauck
Audiophile

Posts: 3048
Location: Vermont
Joined: November 12, 2007
Contributor
Since: February 24, 2009

Actually, my contribution to this effort was to move the focus away from where it didn't belong, the ripper. That was several months ago. That seems to be about the only progress that's been made so far.

Those people who are claiming there are differences are not doing the necessary work, or at least they are moving at a glacial pace. Until they have made significant progress they will have little credibility with me.

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

RE: Not., posted on November 2, 2009 at 19:13:39

audioengr
Manufacturer

Posts: 4088
Location: Oregon
Joined: April 12, 2001

Tony, please try to keep an open mind. This is not mass-hysteria. There are quite a few folks with the test files that you made that are listening to them now. Virtually all listeners heard these differences in my room at RMAF.

I believe we may be onto something. I appreciate the fact that you generated these files and all of the help I'm getting with this. I'm too busy to do much analysis myself.

The next step is to create files with increasing offset to identify the point where the benefit stops.

People use to believe the world was flat you know. Lots of science has been turned on its ear in the last century. Newtons laws and the theory of relativity for instance.

Steve N.

LOL (nt), posted on November 3, 2009 at 02:16:29
Scrith Audiophile Posts: 851 Location: Los Angeles Joined: July 19, 2005	Congratulations...in my opinion the farce that is Empirical Audio has reached a new high today. Did you really just compare the people who question your ridiculous opinions to flat-earthers?

RE: Flat Earth beliefs., posted on November 2, 2009 at 20:26:30

rick_m
Audiophile

Posts: 2529
Location: Oregon
Joined: August 11, 2005

I think people figured out the round Earth thing well over a century ago. The Earthy bit that has happened in my life is plate tectonics. Solid Earth used to be a solid belief. But every school child that has ever done jigsaw puzzles, including yours truly, took one look at the globe and said that Europe and America fitted together to which the august teachers snidely told us that just demonstrated how appearances can be deceiving to the uneducated mind. But we were right..

Rick

RE: Not., posted on November 2, 2009 at 20:09:00

Tony Lauck
Audiophile

Posts: 3048
Location: Vermont
Joined: November 12, 2007
Contributor
Since: February 24, 2009

Look. Some people claimed to have heard differences. Others also heard the files on various equipment and claimed not to have heard differences. This doesn't surprise me. People hear false differences all the time, and equipment malfunctions all the time, and software and/or firmware has bugs all the time.

What really bothers me about this whole discussion is that almost nobody seems to appreciate that if these differences actually exist they must be the fault of the player and DAC and not the ripping software. That people still talk in terms of one ripping being more accurate than another because of the presence or absence of minute offset errors is complete nonsense (except in the very special case where we are discussing gapless playback). To ascribe magical properties to one rip vs. another without regard to the process by which the original audio samples were created to go on the disk is simply complete nonsense. But I've seen equally absurd nonsense in past decades. The prize was the claim that playing digitally mastered LPs would damage one's turntable.

In all of your reports, no mention was made for how long into a track the effect continued. This alone renders all of these observations useless, to my way of thinking.

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

DACs that 'benefit' from this so-called 'reset', posted on November 3, 2009 at 07:39:00

cfmsp
Audiophile

Posts: 353
Joined: October 21, 2006

Tony says:
"What really bothers me about this whole discussion is that almost nobody seems to appreciate that if these differences actually exist they must be the fault of the player and DAC and not the ripping software."

If indeed there is a real difference in the sound of the various offsets here, perhaps the names of the DACs benefiting from such a 'reset' (as described in the referenced article) should be shared in the public interest.

I know I wouldn't want one. :)

clay

RE: DACs that 'benefit' from this so-called 'reset', posted on November 3, 2009 at 11:08:59
audioengr Manufacturer Posts: 4088 Location: Oregon Joined: April 12, 2001	Here are a few: VALAB, Benchmark DAC-1, Northstar 192 Most DACs exhibit the effect.

RE: DACs that 'benefit' from this so-called 'reset', posted on November 3, 2009 at 12:30:10
cfmsp Audiophile Posts: 353 Joined: October 21, 2006	Steve, What about your own DACs? Do they 'benefit'? As a DAC manufacturer, what do you think DACs could be doing differently based on this info, and/or what might be causing this so-called 'reset'? very curious, clay

RE: DACs that 'benefit' from this so-called 'reset', posted on November 4, 2009 at 11:05:06
audioengr Manufacturer Posts: 4088 Location: Oregon Joined: April 12, 2001	A host of different DAC chips seem to exhibit the effect, including the ones that I use in my products. The Sigma-Delta thing is a good explanation, except for the VALAB seems to exhibit it too. It uses the Philips TDA1543. I really have no explanation yet as to why this is happening. Steve N.

RE: DACs that 'benefit' from this so-called 'reset', posted on November 4, 2009 at 14:32:30
idiot_savant	Hi, Have you tried my little thought experiment at the bottom of the thread? I'd be interested to know what I've missed? your friendly neighbourhood idiot

DACs are garbage collectors – right?, posted on November 2, 2009 at 19:09:19

Audio Bling
Audiophile

Posts: 194
Location: Australia
Joined: October 9, 2007

The proposition being floated (in that thread) is that a “burst of zeros” is necessary to “reset the D/A chip”. Why the D/A chip would need resetting is perhaps the next question.. To clean out the garbage? Does this mean that the sound progressively deteriorates throughout a CD (or track) unless a sufficiently large enough group of zeros is encountered in which case the recording comes right? Hmm..

RE: DACs are garbage collectors – right?, posted on November 2, 2009 at 21:44:05
fmak Audiophile Posts: 4099 Joined: June 1, 2002	I remember that they used to do annual 'spring cleaning ' of mainframes!

RE: DACs are garbage collectors – right?, posted on November 2, 2009 at 22:42:26
Audio Bling Audiophile Posts: 194 Location: Australia Joined: October 9, 2007	fmak, I can see where your thinking is headed: a self-vacuuming D/A chip! As it happens, I have one of those new bag-less cleaners that can lift bowling balls and family cars. I reckon I could solder the cleaner nozzle directly to the D/A chip.. I’ll report back after suitable burn-in and DBT.

RE: DACs are garbage collectors – right?, posted on November 3, 2009 at 00:45:46

Werner
Audiophile

Posts: 1829
Joined: September 30, 1999

I've decided to create an audio company that will make and sell
cleansamplers that insert 7 zeroes between every two input samples.

That is the only way to guarantee a constant level of quality.

The first device will be launched soon and will cost $2500.

This new and massively innovative technology is expected to
trickle down into a 3-zero $900 hardware implementation, and
a 15-zero $500 software plugin.

In all cases will the zeroes be calculated with 64-bit accuracy.

bring back dynamic range

RE: DACs are garbage collectors – right?, posted on November 3, 2009 at 07:38:49

Tony Lauck
Audiophile

Posts: 3048
Location: Vermont
Joined: November 12, 2007
Contributor
Since: February 24, 2009

"I've decided to create an audio company that will make and sell
cleansamplers that insert 7 zeroes between every two input samples."

I believe that the problem is symmetric with 1's having the same cleaning effect as 0's. Therefore, so long as there are sufficient 1's in the signal being sent to the DAC, there is no need to worry. Note that USB uses bit stuffing to insert extra ones on occasion, as it is based on the IBM SDLC format. This explains quite well why USB DACs are superior to SPDIF DACs.

Be sure and cite the IBM work on bit-stuffing as prior art when you file your patent application.

Tony Lauck

"Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

RE: DACs are garbage collectors – right?, posted on November 2, 2009 at 20:24:31
Tony Lauck Audiophile Posts: 3048 Location: Vermont Joined: November 12, 2007 Contributor Since: February 24, 2009	I think that problem only happens with rock music. With classical music, such as at least one of John Cage's works, there would be no such problem. :-) Thanks for getting to the heart of this nonsense. Tony Lauck "Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

RE: DACs are garbage collectors – right?, posted on November 3, 2009 at 03:26:07
lord addleford Audiophile Posts: 336 Location: new england Joined: July 5, 2005 Contributor Since: April 11, 2006	once again, we return to the central piece of busines: the ongoing ideological struggle, or is a " war of attrition"? between the adherents (of varying degrees of persuasion),of the "objective" and "subjective" analysis. to look on the bright side, some sense of humor and mutual respect remains in attendance....for the moment. congratulations. now, back to battle.

RE: DACs are garbage collectors – right?, posted on November 3, 2009 at 13:40:08

idiot_savant

Hi,

I'd just like to conduct a short "thought experiment" here, to see if the act of some extra zeroes at the start of a WAV file can do anything good, or indeed bad. I'd just like people to work through, and see what conclusions they can bring.
Firstly, just about every digital audio interface known to man, and definitely all the good ones share a characteristic that the "audio stream" continues, irrespective of audio. So, if you have a DAC locked to the SPDIF output from say a PC, that audio stream is running all the time ( or it wouldn't be locked ). So, before you press "play" on Foobar, Winamp, iTunes, etc, the DAC is receiving a lot of... zeroes. The DAC can't tell the difference between these "player stopped" zeroes and the zeroes from the start of the file.
Secondly, sigma delta systems are, as pointed out, heavily recursive/fed back. Unfortunately, the act of putting in a few zeroes doesn't reset it - if it did, there would be horrific pops and clicks between tracks ( another source of zeroes ). In fact, a zero input does nothing - the sigma-delta sees no noise to feed back, and so the modulator will just churn round continuously.
As to whether the software is faulty, and behaves differently between tracks with zeroes prepended, it would seem unlikely that they all do?

just my thoughts,

your friendly neighbourhood idiot

RE: DACs are garbage collectors – right?, posted on November 5, 2009 at 13:47:16
audioengr Manufacturer Posts: 4088 Location: Oregon Joined: April 12, 2001	Actually, if you look at an idle signal from a digital source and then start a track on a scope, there is a difference. I believe the control frames are usually not there in the idle signal. The average frequency changes radically, almost double when the track starts. Am I off-base?

RE: DACs are garbage collectors – right?, posted on November 7, 2009 at 03:38:41

idiot_savant

Are you looking at SPDIF?
If so, SPDIF represents a '0' as no transition in a bit period, and a '1' as a transition during a bit period. So, a silent track ( all zeroes ) will look like a slower signal than one with data ( as the track playing will have lots of '1's in it). In both cases the preambles should be constant.
Now, there may be interesting things happening when you suddenly go from all 0 to lots of data, but why would the precise moment this happens make a difference to the whole track?

your friendly neighbourhood idiot

RE: DACs are garbage collectors – right?, posted on November 7, 2009 at 11:02:55
audioengr Manufacturer Posts: 4088 Location: Oregon Joined: April 12, 2001	There is a considerable amount of "control" info in the actual data frames, as opposed to "idling" I think. This control info includes sample-rate, bit depth, length and other info. If it is just idling, then the source does not know the nature of what is coming next. DVD players dont even idle. They have no output until the CD is read.

RE: DACs are garbage collectors – right?, posted on November 7, 2009 at 14:41:19
Tony Lauck Audiophile Posts: 3048 Location: Vermont Joined: November 12, 2007 Contributor Since: February 24, 2009	One could easily mod the player software so that it output zeros all day long as if it were reading them from disk, waiting for the mouse click telling it to start playing the zeros at the start of the WAV file. Come to think of it, why couldn't the sound card driver do this? Why isn't it? Tony Lauck "Diversity is the law of nature; no two entities in this universe are uniform." - P.R. Sarkar

RE: DACs are garbage collectors – right?, posted on November 7, 2009 at 17:03:53

idiot_savant

My contention is this:
The control bits as referred to earlier I am presuming are the status bits in an SPDIF stream. These bits, which I can go into in more detail if required, should stay the same on say the output of a CD player ( i.e. 44.1k, 16 bit, no de-emphasis ).
I would say that they are always output between tracks constantly, and my denon CD player seems to do this, as does a few PC cards I've tried. I would add that most DACs "like" seeing this constant stream, as a few of them mute for a while if it goes away.
So, assuming my facts are true ( that most devices output a valid stream of audio which is '0' ), how do extra '0's prepended to a stream affect the output?

your friendly neighbourhood idiot

Not so idiotic....quite interesting. [nt], posted on November 3, 2009 at 18:43:36
alan m. kafton Dealer Posts: 3941 Joined: April 7, 2000	nt

Not the DACs are garbage collectors, posted on November 3, 2009 at 13:55:17
Roseval Audiophile Posts: 506 Joined: March 31, 2008	Don’t think a couple more of these heuristics experiments will harm the forum The Well Tempered Computer

Page processed in 0.097 seconds.

Computer Audio Asylum

Good analysis....

24.255.40.175