From: Hársfalvi Levente (hlpublic_at_freestart.hu)
Date: 2005-08-16 20:59:28
Hi!,
Marko Mäkelä wrote:
> On Tue, Aug 16, 2005 at 07:07:38PM +0200, Laze Ristoski wrote:
>
>>Loaders usually (probably always) work with equal phases.
>
> The built-in loader of the Oric-1 works with pulses, not with cycles.
> That loader is immune to the polarity of the signal. For some reason,
> in my experience some Commodore machines (maybe even the 64) are picky
> about signal polarity. I didn't figure out why; I just wanted to make
> the C2N232 work.
Most non-Commodore computers that used regular tape recorders are like
this (for those this is neccessary, as the signal polarity is unknown).
But neither those formats are inherently polarity-free; the loaders just
detect the polarity from the header signals.
The problem with polarity is, that bits are usually encoded by pulse
lenghts. Those are measured by the time spent between just one type of
signal edges (in c64 this is the falling edge). When signal polarity is
swapped, the falling edges will correspond to the half of two
consecutive bits pulses, not one full encoded bit.
> Also, some fastloaders on the plus/4 work with pulses. What you say
> may be true on the Commodore 64, which cannot detect a rising edge
> of the CASS READ signal.
Well, some really just use "inverted polarity", but most of them are C64
turbo derivatives otherwise (ie. they still use symmetric pulses).
> Some audio editors can show the signal in the frequency domain. I
> wonder if the quantization errors are significant when sampling at
> 48 kHz or more, 16 bits per sample.
They aren't.
Best regards!,
Levente
Message was sent through the cbm-hackers mailing list
Archive generated by hypermail pre-2.1.8.