Gah- I like to keep my OJ for breakfast and the vodka for other times. OTOH, my brother makes a nice Raspberry Wheat beer (but flavored with real Raspberry concentrate), which is affectionately known as "liquid panty remover". Though its best served a bit warmer than cryo temps, unlike Budwiser and the rest of that ilk.
Gregm
Even if thats so, then why would the alleged bit flipping happen to the digitized audio in such an organized way as to improve the sound instead of introducing noise?
I think you should run the test as a previous poster suggested, take an arbitrary test disc, copy off its image, freeze the disc & confirm that it sounds better, then read the image again and run a md5 against both images. If the checksums are the same, you are imagining things. If they are not, then a per-track investiagtion would be the next step to find out where the errors occurred and how many occurred. Its just a matter of repetition to track down which samples have flipped bits & evaluate what the impact on the reproduced sound is.
The image reading and comparing is easier in Linux/bsd/Solaris and probably MacOS, but I imagine Windows can do it if you track down the software.
Gregm
Vodka ice cubes. You put them *in* the martini.
Jim
I had a quick look and my comment was not entirely correct. On a data disc you get a 512 byte sector followed by about 150 bytes IIRC of ECC data, if the ECC can't correct it you get that nasty situation were the data is corrupted and possibly irrecoverable. The audio CD differs in the data layout but still has means of some ECC although the error handling strategies still play the CD trying to minimise the impact of errors on the reproduced audio. From this I could see that some method of uniformly tweeking the CD might have an effect on the audio output. I agree though that if you run a MD5 on the disc image and they are the same then I can't see how the freezing has effected anything.
A puzzler, but I can think of a few possibilities. Unlike data, the error correction for music is CDs is not absolute - if there is an error the circuitry will try and play around it - and the method used to approximate that might well involve smoothing the sound out over a certain interval of time with interpolated values, causing better bass and worse treble response; or if not actually improving the bass and/or overall quality, causing a relatively better bass/treble quality ratio.
In the studio I believe it is also an occasional practice to add noise at innaudible level to ameliorate distortion caused by the notes being exact fractions of the sampling ratio, and some non-linearities in the overall response, causing and a slight increase in both actual and perceived quality.
Note that these are pure speculation, and probably wrong. Any effect, if there is an actual effect, may well be caused by other processes. They are just to show that it might be possible.
Like CA Decker's Jitter story, his sort of thing happens in musical reproduction technology - for instance when "they" said "44.1 kHz is enough, you won't get better fidelity with a higher sampling rate" and justified it as mathematically proven under Nyquist theory, they were subtly wrong - it's fine and true for the reproduction of constant tones, but the theory falls down when reproducing changing notes, which after all are what music is composed of, and when some implicit assumptions about your reproduction equipment are untrue in practice.
I'm not buying silver speaker cables though!
More, a general improvement in faithfulness to the original causing "better bass" is much easier to imagine possible reasons for.
Of course not. You'll want the Phacetious Audio quicksilver liquid speaker connectors, where mercury is pumped through hoses to carry your speaker signal. The sound is "liquid" and "flowing" in subtle audiophile ways that mere engineers can't quantify. Pumping the mercury, rather than letting it sit still in the hose adds subtle "motion" to the sound. As if that wasn't reason enough to spend $50,000 on a set of ten foot speaker cables, you'll have a hazardous waste site when your dog chews on the hoses. What could be better? Sign up now. Get your deposits in and be among the phirst to be phleeced...
(somebody wrote, but this guy deleted the attributes for while top-posting,)
What, that a digital disk is digital?
Yes. And if the bit is above the upper limit, it's a one. If it's below the lower limit, it's a zero. If it's in the middle, it's an _error_, and the error correction bits tell your reader how to correct it.
Right, so far.
Of course it is. Scratches, bad pits, and so on. That's why the correction bits are there.
If you're taking signal and turning it into something else, it's called _noise_.
A lot of really stupid ideas have been ridiculed as well. I think this is one thing that has always turned me off from the "high end audiophile" people - they can't quantify their theories, and try to justify them as good because some other theory turned out to be good when it was initially ridiculed.
For every person thought to be a fool who turns out to be a genius, there are 999 who really, really were just fools.
Yes, MacOS is a FreeBSD system, so all of the tools you'd expect on any other Unix box are there, and all of whatever you'd want to build builds just fine.
That's because a better Martini is made with gin.
-G
I've had the occasional problem compiling things on OS X, different ranlib defaults for instance, but its loads easier to manage than Windows.
Gregm
The errored blocks of data are then missing and interpolated, which inevitably introduces noise- sort of like a fleck of dust on the LP groove. Its not just bass frequencies that are magically affected, the lost samples literally stop the speaker cone from reproducing the waveform for the duration of the error. At that point you're presumably relying on the ballistic performance of the cones to "smooth over" the lost samples and damp the sharp transitions from signal to error to signal, which doesn't sound like a good way to improve "quality ratios", whatever they are. Its effectively a decrease in the signal/noise ratio.
The ECC codes are essentially data checksums, computed such that some amount of missing or corrupt data can be recovered given the surrounding data and the ECC code itself. Once the algorithm's capacity is exceeded (because of too many errored bits in the associated data block), then all you'll get out is data corrupted to some unknown extent. So from a reproduction standpoint, you get perfect data from the CD as the number of mis-read bits coming off the CD read head increases up to the point where the bit-error-rate exceeds the capacity of the ECC algorithm, then corrupt data starts coming out. By "perfect", I mean the bits read off the CD are the same as written. So, the CD format provides a limited degree of error recovery built into the format of the audio samples- it alway reads the same bit-stream until too many errors show up for the ECC to deal with.
I look forward to your detailed proposition documenting how bit flips that get past the ECC algorithms improve sound, with emphasis on how CD audio formats make it possible. I'm also interested in how bit flips on dvd's and digital audio tapes improve the reproduction quality but do not corrupt things like software images or compressed data.
Gregm
Sorry let me try that again.
I came across this site which give some background to the technology beh>Peter Fairbrother writes:
Actually, consider it the other way around:
1) The standard for CDs was for Audio *first*. 2) The audio CDs have a sector size larger than the 2048 byte size used for data CDS, and follow that with an additional chunk for The ECC (Error Checking and Correction code). 3) When used for *data* purposes, the difference between the audio and data sector sizes are used for an *additional* layer of ECC, so the data off the CD is *doubly* checked.So -- from your point of view, there is less error checking for audio purposes, but there still is quite a bit. even with audio. If it can't actually *fix* an error, it will replace the bad sample with an average of the adjacent samples, which helps hide the error, at least.
So -- in any case, greater errors on the raw CD result in less perfect audio reproduction, but not the glaring errors to be found with a flipped bit (which would probably come out as a loud "click" on playback without that substitution.)
Enjoy, DoN.
The misconception mentioned earlier being that the change from "one" to "zero" is instantaneous; it takes time to make that transition, and that's an analogue thing.
Like people have pointed out; loss of data is not a problem, but since in SP/DIF (the protocol behind the CD) the clock, which governs the data rate, is embedded into the data itself, the rate at which those transitions occur is extremely important, and any ambivalence as to where a transition takes place potentially forces the system to second-guess itself.
This is called jitter, is usually measured in nanoseconds, and it sounds like, well, an old cheap CD player.
I suppose that getting those pesky little holes in the CD surface to have better defined edges might make CDs sound better. However, I seriously doubt that freezing them will do that ( although I'm open to scientific explanation)
Hans
I don't understand what you mean. If you mean the cone stops, it doesn't. If you mean the cone stops getting a signal, it doesn't. What does happen is the signal sent to the amp may (may, not will) be different to the intended digital signal.
No, you are relying on the interpolation circuitry. This "guesses" a value for the signal and inserts into the stream going to the d/a converter.
The aural effect depends to a great extent on the actual circuitry of the player, and this is done in many different ways.
A simple interpolator might half the signal between the before and the after samples, or add the difference between s-2 and s-1 to s-1, but a good predictive interpolator can involve really very complex DSP circuitry (or software).
A decrease in treble s/n without a decrease in bass s/n might well be perceptually seen as an improvement in bass response. This could come about because eg the interpolation circuitry was better at interpolating bass frequencies.
No, what you get is a "data corrupted" signal, and the active interpolation circuitry takes over and provides the next sample to the d/a converter.
To wander slightly for a moment, jitter is not always relevant nowadays - most modern players will store several seconds of data in ram in case of temporary loss of datastream due to mechanical shock. The feed to the d/a converter comes from the ram store, not the reading in process.
Some of the better players will try to reread any corrupted blocks - computer "ripper" software especially does this, that's why it can take a good ripper a long time to read a badly-scratched disk - but the ECC on music cd's is such that unless both the CD and the player are top quality (and usually not even then) the MD5's of the data stream going to the d/a converter in two seperate playings of the same CD on the same player will not match.
You simply do not get 100% accurate bit-for-bit reproduction or copying of music CD's, at least not without multiple reading. That's why people use (or used to use) expensive CD players for audio, and CD caddies, and so on.
You can get that kind of accurate copying, but you have to record the music as data files, with their extra ECC.
They don't exist - and iirc all single bit flips are detectable and correctable anyway. The ECC is done in such a way that a "data corrupt, interpolate instead" signal replaces any putative bit flips - the ECC detects that the signal is corrupt, and that it doesn't have the ability to correct it. The interpolator then comes into action. If the signal corruption is in the upper half, it is quite likely that the interpolator will come up with a signal identical to the original signal.
It's a bit late now, I may write more tomorrow.
Right, but let's keep this in mind - the OP has agreed that an MD5 checksum of the frozen, and the unfrozen CD, were the same. Theoretical this, that, and the other aside, all of that analog and interpretation layer isn't relevant, when the digital result of all those spinning, reflecting, jittering, and whatever else-ing end up giving you exactly the same data. If you've got the same data, it's going to sound _exactly the same_, because it is.
I never saw the original post on this one, just replies about CDs.
Brass instruments do not respond to Cryogenic treatment... rather, they do respond but somewhat like a "Baptist Fish"
The changes do take place but disappear almost as soon as the instrument thaws out.
Save your money and take some extra lessons instead.
LB
If I recall, it was one of these so-
Nut-case audiophiles don't measure anything, They listen, fully aware of which sample is more expensive, modified with green ink, or is frozen, then they make slanted conclusions. =20
Try burning three CD's, make a small mark on the bottom of one, then play them all after a thorough shuffling. If you can pick out the frozen one 100% of the time, then there's an audible difference.=20
To reply, please remove one letter from each side of "@" Spammers are VERMIN. Please kill them all.
Hmm, I wonder if they'll work as well as some of these items :-)
To reply, please remove one letter from each side of "@" Spammers are VERMIN. Please kill them all.
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.