Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Digital Source Digital Players and Recorders: CD , SACD , Tape, Memory Card, etc.

MicroSD Memory Card Transport Project
MicroSD Memory Card Transport Project
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 6th May 2016, 07:58 AM   #851
marce is offline marce  United Kingdom
diyAudio Member
 
Join Date: Jun 2007
Location: Blackburn, Lancs
MicroSD Memory Card Transport Project
Quote:
Originally Posted by Greg Stewart View Post



I offer the example of John Swenson. His day job is designing the power networks inside of the chips used in high-end servers, such as run our internet and the Cloud. BTW, you don't do that at a senior level like John without some serious electical engineering knowledge and experience... he IS the real deal!



Two great examples of his examination of situations where many have reported differences and the first kind of engineer would say "this are bit-perfect systems and there are no audible differences" can be found in these two postings of John's. The first is a discussion of why isolation and reclocking does not make downstream DACs immune from the effects of upstream issues and can be found here:

Uptone Audio Regen - Page 7


.
Should have asked for advice, for some designs we have created isolation barriers that work up to 18GHz and solve all the issues he is going on about, posted a few hints and ideas on how to do this on the Jitterbug thread on here. This is often a requirement for many designs as noise getting in (or out) could be catastrophic... I think Mr Swenson's comments are tailored towards the audience and also he has to promote the things he works on and Superdad has to sell...
  Reply With Quote
Old 6th May 2016, 08:23 AM   #852
Julf is offline Julf  Europe
diyAudio Member
 
Join Date: Oct 2011
Location: Amsterdam, The Netherlands
Quote:
Originally Posted by rickmcinnis View Post
There are plenty of folks who report a copy of a CD sounds better than the original.
Real-time replay of a CD is a completely different issue from what we are discussing here. Old-fashioned and primitive CD players are not always able to correct read errors "on the fly" (unlike computers and modern CD players that do pre-buffered reads and, if needed, can re-try reading the blocks, just like a computer).

Quote:
All I am saying, and as you well know I have said it before, there are many who think that all spurious noises can affect the sound of digital.
Yes, spurious noises can cause read errors - but it is very rare these days, and that is why decent ripper programs check the rip both by doing multiple reads and comparing the results (thus completely eliminating the effect of any spurious noise) and compare the whole track against a checksum in the accurip database, thus ensuring a perfect read.

From that point on, the music is stored as symbols that are immune to noise (just like the meaning of the numbers on your pay check don't really change depending on the font style or quality of paper), until they hit the DAC, where analog noise again becomes an issue.

Quote:
How about those early MERDIAN CD players that kept count of errors the number of errors would rise with each playing of the disk. I am not saying this is the same phenomenon but it is an unexplained phenomenon just like what I have found with the cd ripper.
The CD player read errors are not "unexplained" in any way, and as described above, are irrelevant once you have verified your rip.

Quote:
All I can figure is that creating as little noise within the environment that the CD is ripped within could be the reason? I do not know the cause. I can only tell you what hear. And since these CDs were then placed on SD cards and the fact that I had retained the earlier rips on other cards it was simple to go back and forth to hear the difference.
And i have suggested a very, very simple test that would take maybe 30 seconds using a bitwise compare program.

Quote:
It is not a tonal sound quality improvement - it allows a large reduction in digital noise. The most annoying thing about digital music . The noise is pervasive - not like pops and clicks from the LP which are easy to learn to ignore. this noise is within everything you are hearing. It is what makes most LP partisans unable to enjoy digital music for extended periods. The sound with the noise is initially more exciting, but like all additions of "excitement" the ear figures it out and asks that you try something else.

It is immediately apparent on my setup.
As you know, the digital audio consists of 16 or 24-bit words that represent a sample of the audio waveform for each sample point in time, with the sample intervals being a constant 1/441000 s, 1/96000 s etc.

So here is one sample from the left channel from the middle of an 96k/24 audio file:

0010 1011 0101 1111 1001 1000

So where, in that sequence of symbols, would the noise show up if none of the "0"'s and "1"s change?

Quote:
I could email you the OS and you can install it on a machine.. it is not simple to do this but it is far from difficult. More time consuming than anything else. Would take about 30 minutes the first time unless you have used SNAPSHOT to place an OS on a drive it had not resided on previously and you could move quickly then.
Unfortunately I don't have a spare computer for that, and I am sure a windows virtual machine running under linux wouldn't fulfil your requirements. But instead I suggest two simple tests you can do very easily:

First test:

1) rip a CD using your optimized setup
2) rip the same CD using the simplest, cheapest computer setup you can find, but using a ripper program that verifies the result using the accurip database
3) verify, using a bit compare program, that the result is the same
4) label the tracks of the first rip "A1", "A2", "A3" etc, and the tracks of the second rip "B1", B2", "B3" etc.
5) while you are out of the room, have a friend go through the tracks, and make a copy of each track, randomly selecting either an "A" track or a "B" track and naming the copies "X1", "X2", X3" etc, and write down which copy came from which source.
6) listen as many times, and as long as you want, to the tracks, and write down if you think each "X" track sounds like the "A" track or the "B" track. Share the tracks with friends and ask them to do the same.
7) when you (and your friends) have written down their opinions, compare them to the list of sources your first friend wrote down.

Second test:

Take an audio file. Make a script that copies the file first to another hard disk, then from there to an USB stick, then from there over a network to a file server, then from there to some old, crappy magnetic media such as floppy disk. Have the script do this 10.000 times. Then compare the result to the original.

Quote:
I do not think you would get the same result on a system for gaming. The simpler the better. Then you could know for yourself and, of course, I would be interested in hearing your appraisal.
I can tell what my own experiences have been. I haver ripped something like 1200 CDs. I am using a general purpose desktop computer running linux, that also does a lot of other stuff at the same time it is doing the ripping. I do use a ripper program (morituri) that both does multiple reads and does a compare with the accurip database.

During the last 4 years or so I haven't had a single CD that I haven't been able to rip totally error-free (the only problems have been copy-protected CD's and CD's with extra non-audio material, and with those, the rip fails completely).

The only CDs I haven't been able to rip error free have been old, scratched CD's, and in those cases the ripping program has reported the issues very clearly.

Quote:
Just my experience. If it is taken that I am encouraging people to try this, well, there is an aspect of that but the reason would be for collaboration and verification if there is something to it.
Then I am sure you are more than happy to do the experiments I described above.
  Reply With Quote
Old 6th May 2016, 01:00 PM   #853
rickmcinnis is offline rickmcinnis  United States
diyAudio Member
 
Join Date: Jun 2006
Julf,

If you are happy with your setup then what is there to discuss?

I do not understand why we went through this again.

I have offered the thing to you to try. You obviously have no curiosity to try it.

I have found this to work better. You know I have no fancy instruments to prove my point. I doubt there is a measuring scheme that would reveal what is happening but that is beside the point.

I am using my ears for comparison. AS I said numerous times I have the same recordings ripped on the old way similar to yours though I would not run other programs while ripping and with the minimized OS and the difference is immediately audible. I would think this is about as conclusive as one can get. Check sums are ONE aspect not the whole thing.

I think you should find something else to go on about.

Last edited by rickmcinnis; 6th May 2016 at 01:03 PM.
  Reply With Quote
Old 6th May 2016, 01:19 PM   #854
Julf is offline Julf  Europe
diyAudio Member
 
Join Date: Oct 2011
Location: Amsterdam, The Netherlands
Quote:
Originally Posted by rickmcinnis View Post
If you are happy with your setup then what is there to discuss?
Because I am curious about the effect you describe, and suspect the explanation you suggest is not correct - but I am of course fully aware of the fact that I might be wrong. Thus I want to understand exactly what it is you are describing, and on what points we actually disagree.

Quote:
I have offered the thing to you to try. You obviously have no curiosity to try it.
And I have explained why that is difficult for me to do, while suggesting some much easier experiments you could do.

Quote:
You know I have no fancy instruments to prove my point.
And as I have stated before, no fancy instruments (or any instruments at all) are needed.

Quote:
I am using my ears for comparison. AS I said numerous times I have the same recordings ripped on the old way similar to yours though I would not run other programs while ripping and with the minimized OS and the difference is immediately audible. I would think this is about as conclusive as one can get. Check sums are ONE aspect not the whole thing.
So you are not even curious enough to verify the checksums?

Just to be sure I understand your viewpoint, I hope you could clarify a couple of things.

If we look at the chain from CD to speaker, we can, for the sake simplicity split it up into 3 components - ripping (reading the data from the CD into a file), storage/transmission (where the ripped file is moved to disk, over the network, onto a SD card etc.) and playing (where the file is converted into an analog waveform, amplified and fed to a speaker).

Let's start with the simplest of those, storage and transmission. Do you agree that once the audio data has been ripped in an error-free way, there is no change to the audio data, no matter how much it is copied or sent over the network? Do you also agree that two files with the same checksum are extremely unlikely to contain different data?

Next, playing - do you agree that two files containing absolutely identical data will sound the same, independent of how the data happened to be transferred to the files?

As to ripping, do you agree that if the ripping operation transfers the actual data off the CD without errors, and the audio content of the resulting file contains exactly the same data as was written to the CD, the way the ripping was done has no impact on sound quality, as there is nowhere for the noise or distortion to "hide" in the file?
  Reply With Quote
Old 6th May 2016, 01:53 PM   #855
rickmcinnis is offline rickmcinnis  United States
diyAudio Member
 
Join Date: Jun 2006
Julf,

I obviously do not agree with your assertions.

My experience, and that is all I have, tells me there is more at work than simple addition.

I share a preference for logical explanations but sometimes things happen that require additional information to be explained logically and until those tools are available we have to grope through the "unknown".

SO, I have tried ripping CDs both ways - I, too, have used morituri with UBUNTU and dBpoweramp with full installs of XP, 8.1 and 10 tech preview. I was satisfied with these rips. randytsuch made a comment about how lax we are with ripping (he, too, has recently got a SDTrans) and I told him I had had the same suspicions so that got me to work on the thing based upon nothing more than my previous experiences.

It would seem simple for you to try the thing yourself than ask me all of these questions. I have performed the most important experiment, the only one that matters for me, the CDs ripped with this thing sound better than the ones I ripped nonchalantly. I have both versions available and I have compared them numerous times.

If you would like to objectify what is happening I would be thrilled. That is your place in the world since you are trained to do such things. I am one who enjoys listening to music and this motivates me to find things to improve the experience. I make no claims to being anything more than a tinkerer. My instruments are my ears and brain.

I suspect there is nothing statistically surprising about the check sums of the early rips and the new ones so what does that tell us? Nothing more than that.

Tell me again how to find these check sums - if it is easy I am glad to do it but I assert that is not what matters. How much difference in the check sums would be required for there to be sonic consequences? I thought the error correction schemes would take care of that, anyway.

There is just as much likelihood there is something else going on here than not. Is it less need for error correction? I have no idea.
  Reply With Quote
Old 6th May 2016, 02:05 PM   #856
Julf is offline Julf  Europe
diyAudio Member
 
Join Date: Oct 2011
Location: Amsterdam, The Netherlands
Quote:
Originally Posted by rickmcinnis View Post
I obviously do not agree with your assertions.
None of them?

Quote:
I share a preference for logical explanations but sometimes things happen that require additional information to be explained logically and until those tools are available we have to grope through the "unknown".
And I have explained some very simple experiments you could perform. Why are you so unwilling to do them? To me it seems like someone who sees the Loch Ness monster, but can't bother to take a picture.

Quote:
I have performed the most important experiment, the only one that matters for me, the CDs ripped with this thing sound better than the ones I ripped nonchalantly.
And I assume you don't consider confirmation bias / placebo effect a possible explanation - or even one worth ruling out?

Quote:
Tell me again how to find these check sums - if it is easy I am glad to do it but I assert that is not what matters.
Probably the easiest way is using Exact Audio Copy (EAC) that has a "compare just the audio contents of two files" function.

Quote:
How much difference in the check sums would be required for there to be sonic consequences? I thought the error correction schemes would take care of that, anyway.
So you agree that if the rip doesn't produce errors, the resulting file is a bit-perfect copy of the CD?
  Reply With Quote
Old 6th May 2016, 04:58 PM   #857
formatcd3 is offline formatcd3  Netherlands
diyAudio Member
 
Join Date: Jun 2015
Quote:
Originally Posted by marce View Post
It has no bearing on what I have said wots so ever, I am a professional PCB design engineer and have many years of experience in doing both sensitive analogue and digital layout... If you don't want to listen to my advice that is your prerogative but what I listen to and on what has NO bearing on the advice I give.
My points of view are from years of learning and study, a never ending task with electronics, I have over 5G of data, white papers design guides etc covering all aspects of layout and techniques for signal and power integrity, I use these add-ons as well...
CADSTAR PCB Analysis and Verification | Zuken
The projects I work on are often quite complex life and mission critical so a failure would be critical, every design I have done is tested and quantified and has been for 30+ years and over that time I have picked up a lot of knowledge and experience....


As to music preference, I have none I just love music be it the latest pop song or some caterwauling Puccini opera.....
If you search in the "curvy chang" thread and the pictures of "OB speakers" you will find pictures of my various systems well speakers.... again not that it makes any difference to my advice...
Hi, marce. Actually it does matter as DIY high end audio is not a mission critical project (sorry for comparison, I am not beeing sarcastic) but rather something that might require compromises (from common practice point of view) to achieve the best SQ. Sometimes, for example, at expense of circuit stability and often greatly sacrificing conveniency of operation. And you are right when you say (or mean) that certain solutions are not acceptable. You are right if you see this in the context of professional or commercial product that must comply with regulations, certifications, specs, etc. In DIY we do not have these limitations. We are not selling anything to anybody. Just share what works and what does not in our systems.

That's why the most important aspect of this discussion, from my point of view, is the mind set. And I asked you the question about your system and music preferences to see where you are and if you are one of these guys that tweak their systems. You already know the answer.

I should admit, that I actually like you questioning certain things as it puts it in different light and make me question what I believe in and double check it. Why I like it - because its the way to make things better.
  Reply With Quote
Old 6th May 2016, 06:31 PM   #858
rickmcinnis is offline rickmcinnis  United States
diyAudio Member
 
Join Date: Jun 2006
formatcd3,

I think you put the "disagreement" succinctly.

What we do to our audio gear is not something we would do to gear we make our livings with. Reliability is among a professional designer's most important goals. Most of us have some kind of back up (as if professional setups don't have back ups, also) but the time it takes to change stuff out in a studio/giant PA/etc is critical and usually unacceptable where for us audio kooks it is not a big deal.

None of us are suggesting these ideas for the professional world. It is simply to try and evaluate, can you hear a difference at all?, it is fun in a serious kind of way.

When I was eighteen I would do similar things to my car - I would never do that now. I need the thing to work whenever I ask it to.

So there is no lack of respect or appreciation for the work you fellows do and there is no question you know stuff we will never know. If it wan't for folks like you we would have nothing to play with. No chance I would ever come up with the SOEKRIS DAC. But that does not mean it is perfectly realized as it comes from the factory. I think the bristling comes about when there is the impression that whatever the tinkerer comes up with or talks about is ALWAYS nonsense.

The manufacturer has to find the right spot where the sales will justify the production expense. The manufacturer knows they cannot please everyone and certainly would go out of business trying to please people like me. And if they did make something that pleased me I probably could not afford it!

Many of us get into the evolution approach. You buy something that has the basics right and then as time goes by you can gild it. Let's face it, if something sounds better to you even if it doesn't really, what's wrong with that? The problem is time tells the tale. I try to listen to something for a week or two before getting too excited about it and say something.

Which brings me to julf and his tests. I will eventually give those things you suggest a try. Since UBUNTU is what my home machine runs I would prefer a "checksummer" that worked with LINUX if you could suggest one.

But if I find that the sums are the same for both approaches I don't think that is going to make my new rips sound the same as the old ones. There is more to it that the checksum.

If I was twenty years younger I would be unsure of my abilities to listen without bias but at this point I can handle failure. I know I am hearing a much cleaner presentation - like I keep saying - this is not so much about tonality (though there is more "there" in the midbass) but a vastly cleaner window. It is like a sonic version of the TV fuzz HBO uses for their logo during the beginning of their own shows (only one I watch is GAME OF THROWNS (spelled wrong on purpose)) has been removed.

I thought I had become good at hearing digital noise but must admit I did not detect this being there until it was gone - it is like a fuzzy scrim a foot or two in front of the loudspeakers when I go back to the old rips for comparison. Maybe not everyone is sensitive to this stuff and I do not think many people are until they sensitize themselves to it. Is that crazy? Well, it might be considered slightly masochistic but then it is akin to the folks who divine all kinds of flavors out of wine and scotch, etc. I drink to get a buzz not a tasting experience so we all have our hobbies that require attaining a sensitivity in order to experience something that is important to us as fully as possible.

Good thing DIYAudio doesn't charge me by the word.
  Reply With Quote
Old 6th May 2016, 06:48 PM   #859
Julf is offline Julf  Europe
diyAudio Member
 
Join Date: Oct 2011
Location: Amsterdam, The Netherlands
Quote:
Originally Posted by rickmcinnis View Post
Which brings me to julf and his tests. I will eventually give those things you suggest a try. Since UBUNTU is what my home machine runs I would prefer a "checksummer" that worked with LINUX if you could suggest one.
Glad to hear, as I am mostly Ubuntu-based too. What I use is "flac" to convert the files to pure PCM WAV files, and then good old "cmp" to compare the files.

Quote:
But if I find that the sums are the same for both approaches I don't think that is going to make my new rips sound the same as the old ones. There is more to it that the checksum.
Glad you have an open mind.

Please explain why you don't think the checksum is sufficient...

Quote:
If I was twenty years younger I would be unsure of my abilities to listen without bias but at this point I can handle failure. I know I am hearing a much cleaner presentation - like I keep saying - this is not so much about tonality (though there is more "there" in the midbass) but a vastly cleaner window. It is like a sonic version of the TV fuzz HBO uses for their logo during the beginning of their own shows (only one I watch is GAME OF THROWNS (spelled wrong on purpose)) has been removed.

I thought I had become good at hearing digital noise but must admit I did not detect this being there until it was gone - it is like a fuzzy scrim a foot or two in front of the loudspeakers when I go back to the old rips for comparison. Maybe not everyone is sensitive to this stuff and I do not think many people are until they sensitize themselves to it. Is that crazy? Well, it might be considered slightly masochistic but then it is akin to the folks who divine all kinds of flavors out of wine and scotch, etc. I drink to get a buzz not a tasting experience so we all have our hobbies that require attaining a sensitivity in order to experience something that is important to us as fully as possible.
And have you tested your abilities with double-blind listening tests?
  Reply With Quote
Old 6th May 2016, 06:53 PM   #860
DPH is offline DPH  United States
diyAudio Member
 
Join Date: Dec 2008
Location: Portland, OR
Rick, the legit question here is that you're making claims that by most Bayesian analyses would be considered extremely improbable just off priors. Or there's something else going on that isn't getting described. Digital works so well because it IS so noise resistant. A copy of sedra/Smith elucidates this well in a readily digestible fashion (input output curves).

I'd hope you could understand how others would be unwilling to take your impressions at your word, given the unreliability of unblinded human perception and that the mods you're describing go against a very large body of digital best practices. Best practices that were very hard learned and, in such a mainstream practice as this, sounds not be ignored.

Battery power can do some awesome things wrt low level analog, but more in EUVL'S sen/cen application where you need that floating potential.

I appreciate you're frustrated with many of the late posts, but it's best to step back and ask if you're arguing with their content or their effect on your beliefs. Kinda goes to the saying "in God we trust, everyone else being data". :-)
__________________
Happy DIYing, Daniel
  Reply With Quote

Reply


MicroSD Memory Card Transport ProjectHide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Usb Memory Stick/flash Card sdman Digital Line Level 5 5th September 2009 01:37 PM
FS: AMD 1 GHz CPU, motherboard, 256MB RAM, video card, power supply, network card Jimmy154 Swap Meet 0 14th June 2006 08:15 PM
My CD-Transport project YENFU Digital Source 17 24th September 2005 05:27 PM


New To Site? Need Help?

All times are GMT. The time now is 10:24 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki