I have one of the sd cards reader discussed in this forum and regardless of theory discussed here I tell you just try it. It's not cheap but it cost like a very budget cd player. If you talk of user interface you are right , it need much improvment but sound quality is so wonderful that is difficult to go back and last but not least I can listen music 24h/day for three days without touch anything just the volume.
This is the thinking behind the thread I started http://www.diyaudio.com/forums/digital-source/156633-ipod-i2s-hack.html
An iPod modded & reboxed with an SSD or SD card reader & I2S out might make for a very nice battery powered unit? It solves the UI issues with one of the best UIs around!
An iPod modded & reboxed with an SSD or SD card reader & I2S out might make for a very nice battery powered unit? It solves the UI issues with one of the best UIs around!
Oh but they don't!. With a CDROM drive you can use software that may read a sector 82 times. Tell me. In which way is that the same as an ordinary CD player?
None of the above has anything whatsoever to do with jitter.
Jitter is entirely an artefact of the digital signal processing circuits, and nothing to do with the mechanics of reading the disc. Re-reading data and taking a statistical average is a means to get a more reliable datastream from a mediocre disc drive, but once the bit steam is assembled and verified as correct (note 1) the jitter is still just as likely to be introduced by the subsequent data processing and DA conversion.
Sector re-reading may improve the bit error rate. But CD has a very robust error correction system, and beyond that, error concealment. Random bit errors from a well engineered drive very rarely propogate down to the DAC stage.
Even so, bit errors and jitter are not the same thing
Software is far cheaper than hardware, so it is financially more more rewarding to cut costs on the mechanics, and use statistical correlation models to improve the binary signal-to-noise ratio.
None of the above has anything whatsoever to do with jitter.
Jitter is entirely an artefact of the digital signal processing circuits, and nothing to do with the mechanics of reading the disc. Re-reading data and taking a statistical average is a means to get a more reliable datastream from a mediocre disc drive, but once the bit steam is assembled and verified as correct (note 1) the jitter is still just as likely to be introduced by the subsequent data processing and DA conversion.
Sector re-reading may improve the bit error rate. But CD has a very robust error correction system, and beyond that, error concealment. Random bit errors from a well engineered drive very rarely propogate down to the DAC stage.
Even so, bit errors and jitter are not the same thing
Software is far cheaper than hardware, so it is financially more more rewarding to cut costs on the mechanics, and use statistical correlation models to improve the binary signal-to-noise ratio.
Last edited by a moderator:
Who said anything about jitter?
You tried to argue with the fact that it makes no difference if you copy the data to a storage medium or play it with a CD. And I quote:
Post #2 and Post #15
And I still maintain that it makes no difference, unless the CD-audio player is of extremely low quality, or is grossly mis-aligned. Good players do not suffer from uncorrectable bit errors. Defective discs may be the cause of such, but a CDrom drive will cope no better.
Most CDrom drives REQUIRE multiple sector reading for CD-DA discs, since, unlike data discs, CD-DA sector boundaries are not explictly identified. When a high speed drive tries to read a specific sector, it cannot explicitly determine the sector start and end points, so they use a method of overlapping and sector comparison to assemble the data structure.
An audio player does not need to do this, since it reads the data at 'real time' speed, and plays the sectors in sequence, ignoring the sector boundaries.
I should claify that, before someone assumes I'm saying all digital players sound the same.
There may be an audible difference between one player and another. There are many factors that differ between different models of players and between brands. But the sonic differences are attributable to the characteristics of the digital* circuitry each model uses, and are not specifically due to the fact that one uses solid state memory, and the other an optical disc.
* I say 'digital' since to be meaninful, one has to ignore/eliminate the sonic signature of whatever analogue circuits follow the DAC, which have a much greater impact on the sound.
We seem to be talking about different things. You in essence seem to be claiming that it does not matter that much listening to a good cd player or a solid state system. Which I'll agree with in a jiffy. What I'm trying to say is that any system that does not have the ability to verify the correctness (if that is a word) of the data it is reading, is fundamentally flawed. I.e. CD.
Solid state has no such limitation. Thus it is inherently superior.
But CD-DA DOES verify the data, that's what the interleave, parity data and redundany checking are for. The mathematics of ths process allows the data to be CORRECTED (not concealed) even if there is an optical read error.
Solid state memory is not free of random bit corruption. That why your PC RAM also uses parity checking/correcting, and why Hamming codes are used to validate & correct memory in servers that need extra reliability.
Are you going to prove me wrong? Or simply tell me I'm wrong? 😉 I'd prefer the first.
Actually your post made me read my Digital Audio and Compact Disc Technology book. And there is plenty of error detection and correction. So I'll admit I'm wrong that when a "normal" cd is used...i.e. without fingerprints. The error detection and correction will pick up on any errors and correct them.
From Wikipedia: "The result is a CIRC that can completely correct error bursts up to 4000 bits, or about 2.5 mm on the disc surface. This code is so strong that most CD playback errors are almost certainly caused by tracking errors that cause the laser to jump track, not by uncorrectable error bursts." So here it is Steerpike. You were right.
Last edited:
I would expect nothing less from Dutch engineers! 🙂
(Although perversely, the CIRSC used on Cd was a Sony contribution)
And as one reader of a magazine asked in the early days of CD 🙂
He was concerned about playing less than perfect discs from the local library... "they play OK but what happens when the error correction device is full, does that mean it won't correct any more errors... how do you empty it"
He was concerned about playing less than perfect discs from the local library... "they play OK but what happens when the error correction device is full, does that mean it won't correct any more errors... how do you empty it"
CD transports are mechanical contraptions that attempt to detect reflections caused by microscopic dents, placed on a wobbly (often not perfectly centered)plastic disk using a spindle motor, laser positioning mechanism, dynamically focused laser beam and servo circuits to keep the minute focused laser spot on target.
And o wonder, it actually works, and this ancient technology that reminds me of the flintstones is still in use today, but so are primitive combustion engines...
Fact is, sooner or later these will be replaced by more modern devices as technological developments advance.
Apart from obvious downsides like laser aging, problems with mechanical wear, sensitivity to shock and vibration, and high cost, these mechanical devices seem to do a good job.
Unfortunately, the servo electronics produce electrical and electromagnetic interference when performing required dynamic corrections to keep the laser spot on the moving target. The result of these actions is the pollution of power supplies and connected electronics.
If both, source interference and jitter would be completely inaudible on the receiving side (DAC), then yes, a CD transport would be fine. Unfortunately it's virtually impossible to isolate both, source interference and jitter from the DAC electronics without using extreme measures.
Integrating both CD transport and DAC is also not optimal either, because of this source interference.
What we actually need for ultimate CD playback is a digital audio source that provides 100% bit-perfect playback, zero interference and zero jitter. CD-transports are not able to provide this, not even the very best, highly modded audiophile devices. They can deliver the data, but unfortunately they also deliver some unwanted waste products in the form of electromagnetic and electrical interference.
Computers won't be able to provide this either as both interference and jitter levels are usually much higher than CD transports. Slaving a sound card could reduce jitter levels, but the interference will make it all the way to the connected DAC as it "rides" on the required digital interface signals like USB, Firewire, and SPDIF (Toslink) and ground loops. Even if full galvanic insulation between computer and DAC is provided, the interference still enters with the digital audio interface signal in the form of timing jitter. In other words, all computer interference is neatly preserved in the interface signal (SPDIF, USB, FireWire) and is just waiting to pollute each and every circuit in the connected DAC (power supplies, logic circuits, and so on).
Oh yes, jitter is by far not the only problem with digital audio playback, interference (source, mains, EM, and so on) that will inter-modulate with the audio spectrum may even cause more perceived sound quality degradation than jitter does.
Here is why a memory card player makes a lot of sense, especially in extreme performance audiophile equipment:
- Extreme low jitter levels possible (<100 femto seconds rms).
- Extreme low interference signals possible (<10uV rms on both I2S signals and power supplies).
- Dead quiet, completely silent operation.
- Fast response, no seek time.
- Over 50 CDs (WAV format) on a single 32GB SDHC memory card "CD changer".
- No laser aging.
- Insensitive to schock and mechanical vibration.
- Insensitive to pollution (dust particles on laser lens and CD).
- Insensitive to ambient temperature and humidity.
- Compact size and much lower cost compared to audiophile CD transports (high quality SD-transport could be mass manufactured for less than approx. $10).
I agree it's amazing the things ever work at all, let alone when let loose on the public.
And I agree with why you say it makes sense to have a format like flash memory.
From a commercial view though a "standard" would have to developed. Even memory cards keep changing... SD cards will be historic probably in 10 years... outlasted by wibbly wobbly CD lol
And as a music lover it's nice to have those CD notes etc. Could you imagine a small memory chip in a large "Jewel case".
And I agree with why you say it makes sense to have a format like flash memory.
From a commercial view though a "standard" would have to developed. Even memory cards keep changing... SD cards will be historic probably in 10 years... outlasted by wibbly wobbly CD lol
And as a music lover it's nice to have those CD notes etc. Could you imagine a small memory chip in a large "Jewel case".
Any interference that does not directly cause bit transition errors (transposes a zero to a one or vice versa) is limited in its effect to timebase noise (jitter). If a design suffers from bit transition errors, it is seriously flawed right from the start.
Any jitter that enters the transmission medium prior to the DAC (the I²S / SPDIF, USB, etc. connection) can be entirely removed by timebase correction. CAN BE, doesn't necessarily mean it is - depends on the budget.
Jitter originating beyond that cannot be eliminated, but can be minimised by careful design (regardless of the data source).
A mechanical transport does induces noise into common wiring and power supplies. But so does the microprocessor used in a solid state playback device. In both cases, the designer needs to provide adequate power supply rejection. [/quote]
I agree with all those BUT for the first two. Both of these can be achieved in a conventional CD player if the manufacturer so desires.
The key issue is the last one listed above. Solid state playback is cheap and small. That's why it will succeed, not because it produces inherently better audio.
[OT] Well I do use full galvanic insulation between computer and DAC, so at least I feel better about it. =)
Actually, it cleaned up a lot of trouble I was having. [/OT]
Actually, it cleaned up a lot of trouble I was having. [/OT]
Thanks for all reactions.
The average consensus is that trick is in the playback, which is in line with the many publications on jitter, its origin, its propagation and all that. (Good reading btw is Bob Katz at Audio CD Mastering, Mixing & Replication). Either with disk or SC we not suffer from data transfer problems. Error correction circuits help us there, yes these introduce jitter. And yes drive mechanics introduce jitter when their power supplies interfere the digital decoding circuits etc. And yes SC has an advantage here as it is easier to build smaller etc. BUT, all of this I want to put aside; my original question: “does it make sense to have SC assuming we have a good CD transport and good DAC (I2S connected, buffered, reclocked, etc etc)“
My hypothesis is that, with all respect to beautiful memory card projects I see in this forum, it will not add to the sound quality. Maybe I should rephrase my question: does a card player fundamentally raise the potential of digital playback? I think not at all.
Isn't it better to spend effort at the DAC side, isolating the jittery source and recreating a jitter free digital stream.
The average consensus is that trick is in the playback, which is in line with the many publications on jitter, its origin, its propagation and all that. (Good reading btw is Bob Katz at Audio CD Mastering, Mixing & Replication). Either with disk or SC we not suffer from data transfer problems. Error correction circuits help us there, yes these introduce jitter. And yes drive mechanics introduce jitter when their power supplies interfere the digital decoding circuits etc. And yes SC has an advantage here as it is easier to build smaller etc. BUT, all of this I want to put aside; my original question: “does it make sense to have SC assuming we have a good CD transport and good DAC (I2S connected, buffered, reclocked, etc etc)“
My hypothesis is that, with all respect to beautiful memory card projects I see in this forum, it will not add to the sound quality. Maybe I should rephrase my question: does a card player fundamentally raise the potential of digital playback? I think not at all.
Isn't it better to spend effort at the DAC side, isolating the jittery source and recreating a jitter free digital stream.
- Status
- Not open for further replies.