Nor do i. Threw away all my cds anyway. Their physical presence always annoyed me and unlike LPs, they contibuted negatively towards the acoustic environment. Having 5-6k of cds on a server is a completely different proposition and as far as sound quality goes it turned out a big step up for me. As for convenience, it's getting harder and harder to bother with LPs, better sound or not.
The books inevitably followed the cds btw. Thousands of paperbacks collecting dust on a wall shelf just doesn't seem so cool anymore. All my reading is done on a Kindle these days, couldn't be happier.
I for one am struggling to see how there can be any difference in sound between an SSD, USB stick or normal hard disk.
1) No matter what the medium is, the PC will read the data bit for bit, performing CRC (and in some cases parity) checks along the way.
2) This data is then cached in the memory(note 1,2) of the computer, which is then streamed to the device in question, usually a sound card or network player.
3) The player, sound card or DAC then processes this data and produces music.
Surely, with all things being equal, the only influence on the sound can be jitter.
So it would appear that differences heard between various media suggests that jitter is occurring somewhere else in the chain. If so, how can this be? Surely any device (HDD,USB,SSD) would be affected by jitter if this were the case?
Note 1: Some PCs will have ECC ram where even further checking is performed. However, if there are errors in memory, this usually results in operating system failure.
Note 2: Parity checking is also performed in many PCs, any parity errors usually result in "blue screen" errors or instant system HALT.
Disclaimer: I am not a member of the "bits is bits" brigade. I am merely trying to understand why there should be audible differences between different media.
1) No matter what the medium is, the PC will read the data bit for bit, performing CRC (and in some cases parity) checks along the way.
2) This data is then cached in the memory(note 1,2) of the computer, which is then streamed to the device in question, usually a sound card or network player.
3) The player, sound card or DAC then processes this data and produces music.
Surely, with all things being equal, the only influence on the sound can be jitter.
So it would appear that differences heard between various media suggests that jitter is occurring somewhere else in the chain. If so, how can this be? Surely any device (HDD,USB,SSD) would be affected by jitter if this were the case?
Note 1: Some PCs will have ECC ram where even further checking is performed. However, if there are errors in memory, this usually results in operating system failure.
Note 2: Parity checking is also performed in many PCs, any parity errors usually result in "blue screen" errors or instant system HALT.
Disclaimer: I am not a member of the "bits is bits" brigade. I am merely trying to understand why there should be audible differences between different media.
Not so hard to see imho if you consider shared PS lines. If a clock inside a PC uses the same PS as a HDD some cross modulation (jitter) is quite likely.
Yes, but then that should surely affect any device within the PC, not just the HDD.
Yes, I used to think like you. Then I connected my home-brew DAC to an el-cheapo DVD player (SMPSU, considerable leakage current to mains) and found the sound less than pleasing as compared to my QA550 SD card player (outboard wall-wart transformer, relatively low mains leakage current). After investigation I tracked it down to the RF hash entering the DAC via the SPDIF input. Rewiring the grounding of this input reduced the glare.
Jitter could be an issue too, but RF hash is the main one in my experience.
RF hash would surely introduce jitter, as it will in effect modulate the signal?
I know that somebody on here (John Westlake?) advocates the use of Optical rather than Coax connections, to minimize the effects of ground loops etc.
IMHO a well designed COAX output (with an isolating transformer at each end) would eradicate any earth loop problems and provide better impedance matching anyway.
I tend to look at the SPDIF interface as that of a transmitter / antenna interface. If you don't get the tuning right, you get standing waves. My theory is the same must apply with an incorrectly matched COAX connection, obviously on a much smaller scale though.
I guess it does yeah. So to fix jitter we'd definitely need to fix RF hash first.
Sound advice up to a point - Toslink has low bandwidth and introduces a lot more jitter than coax. But yeah, it definitely fixes the hash problem.
Not sure that two transformers are required - I'd have one at the input only. Transformers introduce jitter by means of them being band-limited. Impedance matching is more difficult to achieve via a transformer.
If the transformer has low coupling capacitance then that will help with RF hash. However even 2pF looks like 800R at 100MHz. Ferrites may well be a solution.
All I can say is:
If reading of CDs was a error prone as some people would have us believe, then DVDs and Blu Ray discs would be unwatchable.
The data contained on these discs is far more densely stored than the data on a typical CD.
Yet even the cheapest of players still manage to render a picture free of artifacts, which suggests that the data is being read in a bit perfect manner, with errors (if any) successfully corrected.
I'm wondering if the AV forums out there are going to be flooded with posts stating how much better a film looked when played off an SSD or HDD as against using a DVD / Blu Ray player.
Because surely, if it really IS possible to hear differences between audio media, then it should be possible to SEE differences when video is played from different media.
If reading of CDs was a error prone as some people would have us believe, then DVDs and Blu Ray discs would be unwatchable.
The data contained on these discs is far more densely stored than the data on a typical CD.
Yet even the cheapest of players still manage to render a picture free of artifacts, which suggests that the data is being read in a bit perfect manner, with errors (if any) successfully corrected.
I'm wondering if the AV forums out there are going to be flooded with posts stating how much better a film looked when played off an SSD or HDD as against using a DVD / Blu Ray player.
Because surely, if it really IS possible to hear differences between audio media, then it should be possible to SEE differences when video is played from different media.
Not sure if i get what you mean. It is the HDD with its high pulsed current consumption that contaminates PS lines, not the other way round.
audio_tony said:
I misunderstood your original statement. I thought you were implying that PSU noise affects the ability of the PC to read the HDD accurately.
What I meant was, that in many cases, there will be a conventional HDD in use alongside the SSD / USB disks that are being used here. At least, there has been no suggestion to the contrary.
In any case, even the act of pushing data around the various buses, reading / writing memory etc. is going to create some small pulses on the PSU lines.
After all, the average CPU is switching thousands (millions??) of transistors on and off millions of times a second. This in itself is going to create an immense amount of noise on any PSU.
Last edited:
Actually no, that doesn't follow. Video is sent via HDMI these days, entirely a digital transmission method. It, like SPDIF and USB, is an error free channel. The contents of the screen are a digital copy of the bits sent on this channel, so no errors occur there either.
In contrast, parts of the audio chain must still remain analogue and thus subject to degradation.
I agree however....
It is largely implied in earlier posts in this thread, that the variations in sound are taking place in the digital domain, and not the analogue domain.
I accept that there can (and always will be) variations in the analogue domain, but according to theory *note there should never be any variation in quality of the signal (or sound) as it passes through the digital domain.
note: I use the term theory with caution lol.
No, not in the case of LCD screens. They are analog. DLP is too - but in another way. But with HDMI the D/A conversion does take place within the display, not the player.
As for noise and jitter, I've seen an M-Audio Audiophile card used as a measurement device and the noise floor was very, very low (as seen by ARTA). However, it may have been modified. I'll have to check.
Does anyone on this forum have access to jitter measurement? If so, can we measure the jitter of a file played from a USB stick and the same file played from a hard drive? It shouldn't be that difficult.
As for noise and jitter, I've seen an M-Audio Audiophile card used as a measurement device and the noise floor was very, very low (as seen by ARTA). However, it may have been modified. I'll have to check.
Does anyone on this forum have access to jitter measurement? If so, can we measure the jitter of a file played from a USB stick and the same file played from a hard drive? It shouldn't be that difficult.
So the D/A is hidden within the LCD screen itself? I have no experience of the innards of LCDs, I just know about the interface to the LCD as an OEM module, and that's definitely digital. Usually 6 bits per pixel too - extra bits of colour gradation can be created by creative use of dither or 'colour shaping' between subsequent frames.
My M-Audio 192 PCI card is stone stock. It resides in a stone stock cheap HP computer. The noise is ridiculously low, something like -120dB, with no hash in the spectrum, just a nice baseline. Maybe I've just got dumb luck...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.