Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Audio-enthusiast insanity (Read 22098 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Audio-enthusiast insanity

Reply #50 –
Quote
well, try to rip badly scratched CD with EAC. quite often you'll be able to get it right, but much slower than real-time. so, the quality of writting does matter, even if you can get your bits back.
(also, most players don't do such accurate error correction as EAC - they just play errors. this means that ripping with EAC and making a good burn will in fact increase "the quality" in cd player)

Stop it...  all cd players must use C1 error detection and correction, otherwise it can't be called a cd player !

Your example with badly scratched cd's falls within Exception 1 of my latest post.  It also falls (mostly) outside the scope of Gary's PDF paper.

For a player to play lower quality sound, there MUST be uncorrectable read errors on the first read. According to the specifications, a player is allowed to "give up" error correction ONLY if BLER raises above 220. This will NOT happen on good condition cd's, and most cd drives will tolerate much more than 220.

Believe me, EAC (or the drive, for that matter) won't slow down when there are only correctable C1 errors. If the cd is badly scratched, EAC will slow down because of the C2 correction calculations, and of course the re-reads.

Again, if there's no unrecoverable C1 error there's no problem. Only the rest of your audio chain (including cd player's DAC converters) can then make a difference.

Audio-enthusiast insanity

Reply #51 –
Quote
(source: http://www.cdfreaks.com/document.php3?Doc=51 )
If the CD were being formulated anew today, as better technology now exists, one might see CDs with a frequency range of 20,000 Hz but with a slow roll-off to 100,000 Hz and with sampling rates of five samples per second at 20,000 Hz.

uh.. wha..
.
.

.
.
.
*faints*
A riddle is a short sword attached to the next 2000 years.

Audio-enthusiast insanity

Reply #52 –
This New Scientist "Feedback" article is amusing for the same reasons, though it's nothing about audio, it's about non-ABXable new age beliefs in crystals and homeopathy and how somebody challenged them.

The article will probably be online for about one week only.

Audio-enthusiast insanity

Reply #53 –
Quote
The Last word on this is about half way down this: http://www.cdfreaks.com/document.php3?Doc=51

Quote
Thanks for the excellent article!


I'd like to state that the quotes from this page are not from CD Freaks, but from Kyrowire and Yeme3 users, and that the "exellent article" is not the paper we are speaking of, but an old CDR paper by CD-RW.org.

Audio-enthusiast insanity

Reply #54 –
Who here has seen the movie Pooti-Tang???

Well, this is on the level of hilarious as his smash hit entitled "............."  Where the whole song is absolute silence!

Remember???

Audio-enthusiast insanity

Reply #55 –
Quote
...By the way, any so-called "jitter" issues would be included in exception 3, because the player is responsible for re-clocking the data without altering ANY bits.


I never said bad players alter bits. I also know that error correction can correct extreme numbers of errors. But I'm not talking about the numbers, for indeed, any "half decent" player should get the correct numbers to the DAC. I'm talking about timing.

Quote
Bottom line: Either a given cd player reads a given cd correctly, or it doesn't.  Of course, with an analog output it might be difficult to make sure.


Analog stages aren't the first step where things can get messed up. I like to see the DAC and the digital lines going to it as part of the analog section too.

In the perfect digital world, FIFOs like DAC chips have them, are capable of totally reclocking the data. Also, power supply lines don't carry noise, and keep constant voltages no matter what happens on the board. Not so in reality...

Many circuit designs unfortunately do carry a substantial part of the input noise over to the FIFO's output. Also, there is such a thing as power supply noise. On the really cheap devices, voltages may even vary with the amount of C1/C2 correction being done, as the chips draw current to work on it. This is even worse, since it is noise, somewhat correlated to the signal.
Anyway, I'm not going to type all out here, but as you see there are many reasons why its infact difficult to get a truly constant (jitterless) D/A conversion rate.

Besides this, it seems people are far more sensitive for timing problems than they are for noise levels. On recordings with a high noise floor people can well ignore the hiss. The distortions caused by bad timings in the signal seem to be harder to ignore. Thus, timing must be very precise.

Audio-enthusiast insanity

Reply #56 –
Quote
This New Scientist "Feedback" article is amusing for the same reasons, though it's nothing about audio, it's about non-ABXable new age beliefs in crystals and homeopathy and how somebody challenged them.

The article will probably be online for about one week only.

That was very interesting... I actually followed the whole thing, including reading part of the paganism message board specified on the "crystals" website.

Turns out that the discussion started pretty calm, with sane rebuttals and people (rightly, imho) asserting that the "science vs. religion" argument is pointless and that science can become a form of religion (scientism).

It seems it was only later in the discussion that the flamers and "true believers" came in and started arguing that crystals work, the science guy was just plain wrong, etc. etc. 

It seems people are people, and there are 'bigots' and 'flamers' both on the science and the 'mystical' side of things -- those who cling to their beliefs, whether they be in the assertion that "nothing unproven exists" or "unproven things exist." 

In the end, I didn't learn anything from the 'crystal homeopathy' thing that I didn't already know: Many people are insecure and cling to beliefs (including scientists), all people want to feel good, and all people want to feel validated by others and not belittled.  Doh!

Edit -- sorry for going (way) off-topic, and I'm certainly not saying that there's any lack of false claims floating around... this topic just put me in a philosophical mood I guess .

Audio-enthusiast insanity

Reply #57 –
Quote
Anyway, I'm not going to type all out here, but as you see there are many reasons why its infact difficult to get a truly constant (jitterless) D/A conversion rate.

Yes, this is explained by Bob Katz in digido.com.
But he says also that such a thing (power supply noise creatng audible jitter) is an hypothesis, and that he has not got the gear needed to verify it.
Also, most engineers say that the jitter at the DAC input is too low to have an audible effect. Bob Katz says that great CD drives have an RMS jitter between 10 and 100 ps.

Let's compute the effect of a 100 ps jitter. 100 ps is 0.000005 times 1/44100 s. When a sample suffers from jitter, the level played at the right time is no more the level recorded, but something between it and the next one. If a sample gets moved by 1/44100 s, the worst that can happen is that the value switches from -32k to +32k, that is 65k. Now, the variation is 0.000005 times bigger, thus the error, under catastrophic condition, may reach as high as 0.3 steps.
Thus such a jitter, in the worst case, may be equivalent to a distortion inferior to the quantization noise.

If this is audible, after all I've not got the peak jitter, nor the value for average players, nor even the PLL efficiency, I'd like to know how much really is the jitter in an average DAC.
I've never seen it reported anywhere. So how can we say that it causes differences in the sound (never ABXed, by the way), if we don't even know its value ?

Audio-enthusiast insanity

Reply #58 –
It's a good thing that you point this out Pio. Rereading my writing, I can see it might be confusing.

The original point was this: if you have the same track on two cdr's, it could sound different. That's all I'm trying to say, and I shouldn't have mentioned power supply noise since its beside the argument.

I believe that any recent cd-player will indeed not have problems with reading a good cd, and reading a good copy will produce significantly the same analog output. But having bad cdr's, where the jitter is in fact present on the disc (and if your mouse movement influences sound, like is said here, then I assume this is the case), you may have too much jitter from the transport into the rest of the player to correct it.

To finish, I shall stress here that I haven't seen these effects in proper measurement either. Perhaps I should do some proper measurement, then... Already got a plan, but I'll need a better recording interface. Don't expect measurement shortly.

edit: oh and I hardly ever read it since it's so expensive, but I believe Stereophile magazine reports jitter measurements. Will look if I can find an old issue in my mess.

Audio-enthusiast insanity

Reply #59 –
Quote
But having bad cdr's, where the jitter is in fact present on the disc (and if your mouse movement influences sound, like is said here, then I assume this is the case), you may have too much jitter from the transport into the rest of the player to correct it.

If there are too serious timing errors, this will cause error at the extracted data values. The disc is not longer well recorded, and a cd rip will show this.

About less serious timing errors, in any decently engineered cd-player this should not be an issue. In *any* cd player the data gets reclocked before sent to the DAC. I guess that only in poorly engineered players the servo movements due to physical jitter at the media can affect the power supply so much that the DAC clock is severely affected. Even if there's some small jitter in the output, it can be due to the own DAC clock, and not related to media jitter. Anyway, I've not seen any seriout verification of this media jitter influence in a regular cd player, in form of either measurements or blind tests.

Audio-enthusiast insanity

Reply #60 –
Quote
The original point was this: if you have the same track on two cdr's, it could sound different. That's all I'm trying to say, and I shouldn't have mentioned power supply noise since its beside the argument.

That's the original point indeed.

Quote
I believe that any recent cd-player will indeed not have problems with reading a good cd, and reading a good copy will produce significantly the same analog output.

Exactly.  Similarly, all decent cd writers will produce burns with negligible jitter. It is then the job of the cd player to annihilate any jitter by buffering a few audio bits, and reclocking all of them.

Quote
But having bad cdr's, where the jitter is in fact present on the disc (and if your mouse movement influences sound, like is said here, then I assume this is the case), you may have too much jitter from the transport into the rest of the player to correct it.

- No no no, NO. A bad cdr media will introduce C1 errors, but NOT jitter !
- Reality check: most modern, very cheap cd players will handle ANY reasonable jitter without any problems. Storing a few bits in advance is cheap, you know.
- USB mouse:  do you have any idea how a cd burner works ?  it's an independant embedded device, which receives audio data at one end of a variable-length (max ~2-8MB) buffer, and takes out the data to burn, from the other end. The bits are handled using an internal clock. No USB mouse will have ANY influence on the burner's INTERNAL clock !  The only USB issue that can happen, is that of (obviously) data corruption - just like with any other pc component.

Audio-enthusiast insanity

Reply #61 –
Quote
all decent cd writers will produce burns with negligible jitter.

CDRinfo has made jitter measurments (3T pit and 3Tland jitter) on various media/burners/speed/anti-jitter technologies.

The highest value they got was 60 ns. They don't seem to state it clearly, but they always refer to 35 ns as a limit not to overcome. So it might not be negligible at all.

Audio-enthusiast insanity

Reply #62 –
To get an idea, if I take 44100*4*8/8*17 bits per second, I get 3 Mbps. ECMA 130 specifications say 4.3218 Mbps.

So the value of T (one bit) would be 231 or 333 ns, to be compared to the 35ns "limit" and the maximum 60 ns recorded for the error on these numbers (of twice the error if the jitter is given peak to peak)

Audio-enthusiast insanity

Reply #63 –
Quote
do you have any idea how a cd burner works ?

Calm down. I do.

And to go with your reality check: I've seen enough pc's with devices that send so much shabba back into their power lines, that you could hear it on the soundcard outs. That is reality: there's a lot of bad engineering in the world. Don't think there are no writers out there that have their clock crystal supplied with insufficiently buffered power.

And to be clear about this: like I stated in my earlier post, I too have severe doubts this guy is really hearing all that. But technically it could be.

Being someone who has an idea how cd writers work  , I ofcourse have bought myself a proper writer. So, I'm still wondering how to get it to write an ugly cd for my testing plans. Any ideas?

Audio-enthusiast insanity

Reply #64 –
Burning an ugly CD from the error rate point of view is easy : burn a correct one, and let it in sunlight until Nero CDSpeed shows some errors in it.

From the jitter point of view, it is quite a challenge ! Maybe have a look at CDRinfo tests, and try to reproduce the conditions that lead to the ir worst results... but you'll never know if your CD has jitter without having it analyzed.

Audio-enthusiast insanity

Reply #65 –
Quote
Burning an ugly CD from the error rate point of view is easy : burn a correct one, and let it in sunlight until Nero CDSpeed shows some errors in it.

From the jitter point of view, it is quite a challenge ! Maybe have a look at CDRinfo tests, and try to reproduce the conditions that lead to the ir worst results... but you'll never know if your CD has jitter without having it analyzed.

I know sunlight is bad for a lot of things, but can it do this? The point of focus of a writeing laser is raising the recording layer to thousands of degrees for very small moments as it passes, so can the sun, no matter how long it shines, damage this? I think it would have to be focussed.

And thanks for the good stuff in this thread.  pj pointed it out to me, and it's been funny stuff. NumLOCK, you're a class act. First post, I thought someone had poisoned a babel fish.

Audio-enthusiast insanity

Reply #66 –
You're all wrong. Let me explain why  :

CDs with different colors sound better or worse...: don't you see that they "color" the sound?. You ignorants!!!! 

Audio-enthusiast insanity

Reply #67 –
Quote
I know sunlight is bad for a lot of things, but can it do this?

Certainly... AFAIR, a CDR can resist from two or three days to two or three monthes outside in sunlight.

Edit : the UV are attacking it, rather than the heat.

Audio-enthusiast insanity

Reply #68 –
Focus people, we need to FOCUS!! Enough of the serious - go back to the real point, which is making fun of the horrible, misleading article.

Audio-enthusiast insanity

Reply #70 –
Quote
Calm down. I do.

And to go with your reality check: I've seen enough pc's with devices that send so much shabba back into their power lines, that you could hear it on the soundcard outs. That is reality: there's a lot of bad engineering in the world.

Of course, but that's why there are norms. I don't argue that most PC's are pleagued by noise, which begs for significant soundcard shielding - but besides, if your cdplayer has a digital out, you can easily verify whether it was decently engineered or not.

Quote
Don't think there are no writers out there that have their clock crystal supplied with insufficiently buffered power.

A clock crystal is not like a regular oscillator: its frequency & stability are independant of the input power supply.

Quote
And to be clear about this: like I stated in my earlier post, I too have severe doubts this guy is really hearing all that. But technically it could be.

Nope 

Quote
Being someone who has an idea how cd writers work  , I ofcourse have bought myself a proper writer.

Everyone who owns a cd writer, has a more or less fuzzy idea about how it works. That's no substitute for a real *understanding* though.

Quote
So, I'm still wondering how to get it to write an ugly cd for my testing plans. Any ideas?

For the jitter ?  I don't know, I suppose the most practical way would be to replace the internal clock with a jittery oscillator - or better, one could replace the oscillator output with the output of some soundcard     

This way, you could reproduce the soundcard's jitter (while playing a near-square wave) onto a cd-r..

Audio-enthusiast insanity

Reply #71 –
Quote
Quote
Don't think there are no writers out there that have their clock crystal supplied with insufficiently buffered power.

A clock crystal is not like a regular oscillator: its frequency & stability are independant of the input power supply.


What about the circuitry that divides the frequency ?

Quote
Quote
And to be clear about this: like I stated in my earlier post, I too have severe doubts this guy is really hearing all that. But technically it could be.

Nope 


As I understand it, it is a valid qualitative argument, but we have nothing quantitative : among those who say that jitter is audible, and those who say it isn't, so far I only saw the ones saying it is giving numbers and analysis (digido.com) : 10 to 100 ps RMS on a high end audiophile drive, 1000 to 3000 ps on a DAT deck, and the fact that the jitter can be periodic, random, or signal correlated, and that its spectrum distribution will affect its audibility, because the RMS values themselves seem too low to have an audible effect as a random constant jitter.
I'd like to see the arguments of those who say it isn't audible : maximum peak jitter ? Effect on the analog side ? In which way is it inferior to the threshold of hearing ?

Audio-enthusiast insanity

Reply #72 –
Ridiculing this article is hardly productive. While I do not defend it, I find is most interesting:

That a USB burner can be affected by a USB mouse while using Windows and CD Creator does not surprise me at all. Badly written drivers/software and flawed hardware can easily produce strange errors. I have experienced this myself too many times. It is possible that his computer is fucked. If this is the case, the whole article falls.

The reference CD player, the Sony SACD-1, is mightly expensive - $6,000 if I remember correctly. The audio equipment behind Arnie on the picture looks mighty expensive too.  I will assume that both guys have high-end equipment and a decent listening room. They are clearly very interested in audio too - otherwise they would never have spent their time listening to all the different brands of cd-rs or writing the article. It is possible that both guys inherited a lot of money and never did anything intelligent in their lives. It is also possible they were high on drugs when listening or just plain insane. I will however not ignore the possibility that they could hear a difference. Whether or not the differene sounds good or bad is actually irrelevant at this point. 

When I think back on my first and second cd-players, I remember the first one (cheap model) being able to play all my cds, while the second (better model) could not play a certain cd properly. It stalled at a specific position. The cd was only lightly scratched. When I got my third cd-played (lower high-end), I had to rip and burn 9 of my 250 cds. The cd-player simply refused to play the originals. Yes they were lightly scrathed, but could easily be ripped with EAC/secure mode. A very funny thing: The cd-player can easily play three of my cds that cannot be ripped properly with EAC/secure mode.

This third/current cd-player is from Sony, so back to the reference cd-players used in the articles: Sony SACD-1 @ $6,000. It is cabable of playing SACDs, so it use a different laser than normal cd-players. I believe it was one of the first SACD capable players Sony made. Would I be surprised if it has a flawed CDDA implementation? Not really.

Audio-enthusiast insanity

Reply #73 –
I must add that in Winamp forums, back in 2000, we had a lot of people getting skips in MP3 playback moving their mouse.

Audio-enthusiast insanity

Reply #74 –
Quote
A clock crystal is not like a regular oscillator: its frequency & stability are independant of the input power supply.

Hmmm, I guess you're right there. Must confess, I didn't finish my course on solid state physics yet. But I was kind of assuming that sinusoidal noise going in to a crystal could spawn other modes...? Well, anyway, even if that isn't so, having a stabile clock is one thing, preserving it across your circuitry is another...

edit: cut out some unnecessary qouteing here

Quote
Everyone who owns a cd writer, has a more or less fuzzy idea about how it works. That's no substitute for a real *understanding* though.

Please tell me, what do you want me to do? Prove that I am smart enough to be in this discussion? I've been too lazy to put up my curriculum vitae in my forum profile, but if you're suggesting it will gain me some credibility, I might edit it...

Quote
I suppose the most practical way would be to replace the internal clock with a jittery oscillator - or better, one could replace the oscillator output with the output of some soundcard     

Actually I was thinking of something like that. Or perhaps I'll just get a very old, and reportedly "bad sounding" cd-writer somewhere.