|
Post by John on Jan 9, 2016 16:06:40 GMT
I get a daily email around digital media and thought I share this one as as it challenges thinking around how data is stored and how to get the best out of it
" In yesterday’s post I suggested that to answer to the age old question, are bits just bits, requires us to first change the question.
The question we should be asking is, how can identical bits sound different when played back on different hardware. Asking the opposite–how can identical bits sound different–is a circular question with only one correct answer: they cannot.
For clarity sake, let’s all agree that stored data is the same regardless of its storage medium: solid state hard drive, mechanical hard drive, USB stick, optical disc. Armed with these sets of facts, let’s dive in and see how things might differ.
Take for example the fact that nearly every CD transport made sounds different, then add to that disc treatments, like cleaning and optical enhancers, seem to further improve performance, and a pattern begins to emerge. Identical data sounds different depending on the mechanism delivering that data.
Optical data has no timing information included on the disc itself. Same with hard drives. Data, is data. Timing information associated with retrieved data is added later. In the case of a transport, a variable clock is used to send bits out to the DAC. That clock within the transport becomes the master timing element for the DAC – and it must be variable. Why? Because the output speed of the optical reading mechanism varies according to the quality of your disc, and where it is reading–beginning, middle or end. These timing changes are what make the audible differences, not the bits themselves. "
This backs up my own experience storing music on solid state and playing back on solid state has a positive affect on the quality of playback. I prefer SSD to SATA thankfully SSD has come down a lot in price
I also suggest that data is sometimes manipulated via algorithms This can be at source something like a SOX batch or when the data is being transferred to waveform as in the Caiman DAC
I am interested in learning more so be interested to hear other peoples thoughts on this
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 9, 2016 16:22:00 GMT
Some people (not me) don't agree with you that bits are just bits and spend significant amounts on USB cables. My personal view is that it's only really worth spending on the analogue elements of a system, and even then not on wires.
|
|
|
Post by John on Jan 9, 2016 16:30:30 GMT
Its suggesting to get the best out of the bits requires a bit of thought and I as I said I see how data can be manipulated. I have not spent huge amounts on USB cables myself life is to short for me for that. But I am wondering if there is more to it You be surprised how good a top of the range CD player is these days I have no idea what better I just think it comes down to personal preferences and of course budgets I would suggest it can cost a lot less to get file based playing sounding pretty good a turntable can require a lot more spent to get similar quality
|
|
|
Post by brian2957 on Jan 9, 2016 17:44:45 GMT
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 9, 2016 17:47:53 GMT
How do you know that wasn't expectation bias? What was the format of the bake off please?
|
|
|
Post by pre65 on Jan 9, 2016 17:48:31 GMT
But, is it the cables themselves, the actual connectors, or the way the connectors are terminated.
Three variables.
|
|
|
Post by brian2957 on Jan 9, 2016 17:55:56 GMT
How do you know that wasn't expectation bias? What was the format of the bake off please? What all three guys who have been into this hobby for years . I don't think so . I'm afraid we don't do blindfolded testing . It was sighted all the way .
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 9, 2016 17:58:16 GMT
Digital cable discussions inevitably end in tears (or blindfolds) but to add my tuppence worth I did notice a distinct improvement in SQ when I swapped the 'freebie' USB printer cable I was using between Mac and DAC for a more robustly constructed cable (Atlas to be specific).
Personally I couldn't give a hoot whether it was expectation bias or a hallucination based on too much cheese for lunch - the improvement meant that the sound was more than tolerable rather than a tad harsh and mushy so the cable is staying.
|
|
|
Post by brian2957 on Jan 9, 2016 17:59:12 GMT
But, is it the cables themselves, the actual connectors, or the way the connectors are terminated.
Three variables. Agreed Philip , I can't prove if it was any of the three variables , or any one for that matter . I simply stated which cables sounded better to all our ears . Two of the guys ( one was me ) subsequently bought the Furutech cables .
|
|
|
Post by John on Jan 9, 2016 18:05:21 GMT
Lets keep it to understanding bits first otherwise the debate becomes circular So a few questions Can bits be manipulated? If they can be manipulated Should we consider this as worthwhile to get the best out of the data Does digital sound get corrupted and if so how can we get over this I do not know the answers but hoping to gain a deeper insight
|
|
|
Post by brian2957 on Jan 9, 2016 18:09:20 GMT
Apologies John , I'm not very technical I'm afraid .
|
|
|
Post by John on Jan 9, 2016 19:27:57 GMT
No worries neither am I Its about enjoying the music
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 9, 2016 20:17:06 GMT
Have been messing with digital files (including transcribing from analogue) for about ten years or so but still operating from a position of knowing a little more than nothing I would say the important features of a digital system are origination, conversion, file format, routing and presentation.
Bits may be bits but all that says is that the information is being stored as a code that a computer based system can understand - it says nothing about the processes that occurred to get it into its encoded state (for example degrees of compression or lack of it) or the quality of the original analogue sound signal to start with. You may be able to brush up and polish a sow's ear but what you get will never be a silk purse - and certainly not if the sound came from the other end of the sow.
Assuming a good job was done to present the music for A/D conversion and this was then stored as a lossless file, the next hurdle is transporting the file to a point where it can go through the reverse D/A process to enable the amplification process - and here is where the rub comes. I don't know of a system where you can simply plug a drive straight into a DAC and from there output to your amplifier. At some point you will need to pass the file through software that will enable you to see your files and choose which one(s) you wish to play ... and this is one of the many points which I have found to be the weakest in the chain, mainly because this is where a greater or lesser degree of manipulation by sneaky little algorithms can come to pass that will either allow you music to shine like a diamond - or drown in a pool of nasty rusty metal filings.
Enough with the lumpy metaphors..... just some ideas to be going on with
|
|
|
Post by aurender on Jan 9, 2016 21:31:39 GMT
Repeating information I posted elsewhere on this forum.
My experience showed that in digital cabling, the key parameter is that the equipment sees the expected (designed for) cable impedance. If there is an impedance mismatch there will be reflections in the cable which may or may not affect sound quality.
In a connection designed for 75ohms, the.cables tried included Stereovox, Siltech, Chord, Vertex and several others. On measurement, it was clear that very few of the cables actually met the required 75 ohm impedance, some being up to 20% out of spec. The cables were used for both digital audio and for digital clocking. Audible differences were clear.
Another set of tests evaluated different clocking techniques. With a dcs system there is the option of using the dac as master clock or of using an external dcs master clock (both containing decent OCXO clocks)
It is possible to feed a higher accuracy external clock into the dcs master clock in order to get even more accurate clocking. Again comparing the system clocked by internal OCXO with external rubidium clock and an external Oscilloquartz bva clock produced audible differences. The bra clock has around two orders of magnitude better short term timing accuracy than the best OCXO clocks commonly used in audio equipment.
So whilst bits are bits, many factors affect how the bits sound!
|
|
|
Post by John on Jan 9, 2016 22:00:00 GMT
On the whole I tend to agree but then I think algorithms are now being used to manipulate the file with certain software. Is this a good or bad thing For me depends on the result
|
|
|
Post by MartinT on Jan 10, 2016 0:41:12 GMT
So whilst bits are bits, many factors affect how the bits sound! Completely agree. Without the timing info, bits may describe the music data but timing describes the conversion back into an analogue waveform. Therein starts the problems.
|
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Jan 10, 2016 1:26:38 GMT
On the whole I tend to agree but then I think algorithms are now being used to manipulate the file with certain software. Is this a good or bad thing For me depends on the result By my understanding algorithmic processing has always been used by programmes such as iTunes (is there any other way?) As this is one of the dominant audio file processing programmes in use across the world on desktop and mobile systems it occurred to me that this was one of the first things to address to improve the SQ of files delivered to my DAC. For some time I had ignored some of the most manipulative aspects - the inbuilt equaliser and software volume control - but the quality it was outputting still left a lot to be desired in my view. After trying a number of alternatives I settled on Audirvarna which is relatively cheap and, in my experience, a lot less intrusive than the iTunes codec - or if it does 'enhance' playback does it in a way that appeals to me a lot more than the rather harsh and mushy soup that iTunes delivers. The second area was the controversial area of cabling protocols and design ... but I'll leave it there for the moment while I fashion a hard hat out of a saucepan....
|
|
|
Post by John on Jan 10, 2016 6:43:08 GMT
I do not like using Itunes to store my files I dbpoweramp A lot of people swear by exact copy I can think of several ways we changing the data at source for example up-sampling or file conversations Then you something like the caiman that has a chip that helps to improve the waveform maybe Stan can clarify if this happens during the process of converting to analogue which I imagine or after It seems one of the critical areas is the timing of the code
|
|
|
Post by zippy on Jan 10, 2016 10:07:21 GMT
As a long-time computer developer/consultant and equally ancient hifi fan, the main points I'd make are:
Processor-based file storage (PC/laptop/NAS) was developed for computer use. As such (as mentioned above) the aim is to get the bits from storage to the software with 100% accuracy BUT the timing doesn't matter to the system. On some systems I worked on, it was quite normal for the data packets to arrive at the destination completely out of order, and the receiving software had to reorder them. There would also be occasional bad data packets and the receiver had to ask for them to be resent. Neither of these were a problem because it was not time critical. Playing stored music files IS time critical so its easy to see where errors can happen.
Even if the bits arrive at their destination (e.g. a streamer) with 100% accuracy they then have to be decoded and subject to all sorts of other processes including from digital to analogue. This is why I always use RCA connections rather than optical - think about it - to get an optical signal from one place to another requires the original electrical signal to be converted to light, and at the other end converted back to an electrical signal. Is it any wonder that 'errors' can creep in here. It seems perfectly possible that cables can in some way affect those additional processes (e.g. by interference) and hence change the sound.
The big question for me is not whether mains/USB/RCA cables affect the sound (I assume they do) but whether that effect is audible and whether the scale of the effect is such that it's worth spending lots of money on trying to eliminate it (mostly I think not).
Sorry for the rant..
|
|
|
Post by John on Jan 10, 2016 10:41:34 GMT
Here is a series of videos that look at timing issues
|
|