Bob McC
Living the life of Riley
Yep, talk about a little knowledge being a dangerous thing!
A challenge to ItemAudio
- Thread starter Crash!
- Start date
Yep, talk about a little knowledge being a dangerous thing!
Steven Toy
Accuphase newbie
Perhaps I just prefer the expectation of the 'tank' to the writing of the Yank.
Steven Toy
Accuphase newbie
I was fully aware of why it is often spelt with a k. It is still wrong.
If non-Americans are going to use Americanisms they should do so knowingly.
Steven Toy
Accuphase newbie
That is pretty much what we're planning on but this remains inexplicably unacceptable to some.
Peter Stockwell
Back for more ...
Aren't you being Frenchist, there?
Steven Toy
Accuphase newbie
Ça se peut, oui.
sonddek
Trade: SUPATRAC
No, the French prefer to be different to us, too.
sonddek
Trade: SUPATRAC
If you read the excerpt I posted a little more carefully you will notice that the word first appeared in English as skeptic, before the European settlement of the Americas. It is the only spelling in Johnson's dictionary and is in the original Fowler, and as I've stated, the writings of the blessed John Locke, who had a few interesting things to say about skepticism. Skeptic can therefore hardly be described as a non-English spelling, or an American spelling. It's the first English spelling and the more popular spelling in the ex-colonies. It is not "wrong".
Interesting that post-war Australian slang had Yanks as "septic tanks".
Ciunas Audio
Trade: Ciunas Audio
No, it's not just down to the quality of the local clock used to reclock - it's down to the nature of how reclocking works - it's not a "magic" solution as you are constantly pushing. There is a mathematical process involved in converting the digital samples from one clock domain to another. This process necessarily involves rounding of values in the calculation i.e. rounding errors. It's these rounding errors that result in noise being embedded in the new digital stream that results from this process.
As I said, for high/med jitter sources the balance of this is probably beneficial; for low jitter sources, it definitely is not beneficial.
Sonddek, time to stop spreading mis-information & "magic" thinking & learn something!
I agree "Yep, talk about a little knowledge being a dangerous thing!"
Steven Toy
Accuphase newbie
Indeed. Their attempts at legislating to protect their language are quite comical.
Personally, I would rather put less cultural distance between ourselves and the French and more between ourselves and the bible-thumping, gun-totin', war-mongering, obese loud-speaking Americans.
Yellow school buses and whining sirens accompanying Blues and Twos [shudder!]
God bless America!
sonddek
Trade: SUPATRAC
I've read your description several times, but I still don't understand it. Can you explain it in clearer terms?
In particular, why can't the data from the bus pile up in a stack, and be processed with perfect regularity using the convertor clock? With typical transports a buffer of a fraction of a second ought to provide several minutes worth of clock drift at least. Where do the rounding errors come from?
sonddek
Trade: SUPATRAC
This kind of comment from a Brit always reminds me of the story about France's exit from NATO and de Gaulle's demand that all US troops be evacuated. The US secretary of state asked if the demand included the ones buried there.
You can sacrifice young men's lives to help to save Europe from the Kaiser, Hitler and Stalin, but don't expect Europeans to have any gratitude, or even understand their own history.
Could you provide references for this, and put them in the thread I started for this purpose?
I think it as well to have the technical docs accessible to clarify issues such as these.
Peter Stockwell
Back for more ...
Thanks Tony, there's interesting and useful stuff in here, sorry you had to come in like the school teacher.
Peter Stockwell
Back for more ...
sonddek
Trade: SUPATRAC
OK - I'll try to explain my confusion again. I don't see how there can be "rounding errors" in taking a series of, let's say 48K samples, which admittedly have the studio ADC's jitter embedded in the samples, which can't be fixed now, and reclocking them to the DAC clock's best effort at 48K. It seems to me that rounding errors only creep in if you upsample at too low a bit depth, but that's not what we're talking about here. Here we're talking about why a buffered DAC need take any notice of transport jitter.
Why can't the following scheme work?:
- the sequenced samples sit in a stack on the buffer, and the DAC slides samples off the bottom as needed according to its own clock. Then it does its upsampling/filtering algorithms on the re-clocked data coming off the bottom of the buffer. Assuming an adequate buffer, i.e. no under/overflow, How is the DAC seeing transport jitter in this scheme?
No it's already digitised. It's "there" by the way. Rounding errors occurs when digitising and when changing the sampling frequency. If you upsample 16/44.1 data to a multiple no problem, but if you don't there has to be a rounding decision. (or something like that). if you sample at the correct depth and frequency you end up with the rounding errors being too small to matter.
Reclocking has to be a bit crap if it can't reproduce the original data. That's the whole point, isn't it?