On 2009-05-11, at 07:43, <ruud.baltissen@apg.nl>  
<ruud.baltissen@apg.nl> wrote:

> Yesterday I wrote Joe Foster a PM asking if he could tell me how he  
> read
> the data from disk. During writing this email I got this idea:
> - use one turn of the floppy to find out in which order the sectors  
> are
> stored.
> - then start reading/writing the sectors with an interleave.
>
> This interleave depends on how long it takes it to perform three
> actions. The timing is measured in sectors:
> 1) reading a sector - 1 sector.
> 2) en/decoding the sector - I have no idea.
> 3) reading/writing the data from/to the harddisk + initialising the HD
> for the next setor - IMHO < 1 sector.
>
> If 2) would take one sector ie. interleave = 3, then it would take 4
> revolutions to read a track or 24 seconds to store a disk. Much faster
> then the 3 minutes I need now.

If you recall my disk imager for IDE64: it reads, decodes, and stores  
image on the harddrive - I managed to get below 22s but I do not use  
interleave reading there. I buffer the whole track (in the 64) before  
decoding and writing. We would have to count the cycles on the GCR  
decoding to get exact time but this was the most time critical place  
where I did the optimising and I feel like there is still some room  
for improvement. I guesstimate that you should be able to put all the  
stuff in three sectors reading time. I do all the processing and  
writing in the 64, which is slower than the 1541 and I still have the  
image ready and written below 22s. I believe this gives you some order  
of magnitude information.

> I know there is a routine the 1541 uses to find the right header to
> cover 1). 3) isn't a problem either. Regarding 2) I have two qestions:
> - I know the routines exists and should be able to find out how to use
> it but if somebody happens to have some experience on this matter,  
> info
> is welcome.

As above - the routines are available. AFAIR I used the 1571 version  
as the base for optimising and partially rewrote them but there is ROM  
version ready.

> - I have no idea how much time it takes to en/decode a sector. Anyone?

We have to count the cycles on that to get the exact numbers.


       Message was sent through the cbm-hackers mailing list