Page 1 of 1

Experimental very fast tape loading

Posted: Wed Aug 03, 2016 8:54 pm
by Symoon
It's all there !
https://youtu.be/ZjN3SdGQVQE

Sadly can't seem to be that fast and work on any machine. As soon as I change, I have to (very painfully) set up again several parameters.

Re: Experimental very fast tape loading

Posted: Wed Aug 03, 2016 9:21 pm
by iss
Congrats! The speed is really impressive.
I think you can extend the idea and create kind of networking trough the tape port ;).

Re: Experimental very fast tape loading

Posted: Wed Aug 03, 2016 9:35 pm
by Symoon
iss wrote:Congrats! The speed is really impressive.
I think you can extend the idea and create kind of networking trough the tape port ;).
Thanks !
Networking PC and Oric through the tape port is an idea I've explored for yeaaars, but needed to improve skills on various sides:
- how to allow the Oric to be interrupted by an incoming signal (which is a real data signal)
- how to decode an Oric signal "live" on a PC/Mac
Quite another story !

About this attempt, I wanted to try to reach the fastest speed. But sadly it's not reliable (that's why TAP2CD doesn't go that fast: reliability!). What you see is a mix between TAP2CD-like accelerated loading, mixed with a little RLE compression (an empty Hires screen will load in 0.8s instead of 3.2 in the video)

Re: Experimental very fast tape loading

Posted: Wed Aug 03, 2016 9:48 pm
by Godzil
I would be curious to see side by side with the "normal" way of working.

For the network part I think that using the audio card is not a good idea, it would be better and more reliable to use a small microcontroller, probably connected to the PC via USB

Re: Experimental very fast tape loading

Posted: Wed Aug 03, 2016 9:57 pm
by Symoon
Godzil wrote:I would be curious to see side by side with the "normal" way of working.
Routines loading excluded:
Very fast: 3.2 seconds
Tap2CD: 6 seconds
Standard: 57 seconds :lol: (ok, with synchro)
Godzil wrote:For the network part I think that using the audio card is not a good idea, it would be better and more reliable to use a small microcontroller, probably connected to the PC via USB
Certainly better indeed, but the basic idea was, for non harware guys like me, to just have to buy a tape cable and be able to access to the PC hard drive ;)

Re: Experimental very fast tape loading

Posted: Thu Oct 20, 2016 4:26 pm
by ThomH
Symoon wrote:Networking PC and Oric through the tape port is an idea I've explored for yeaaars, but needed to improve skills on various sides:
- how to allow the Oric to be interrupted by an incoming signal (which is a real data signal)
...
For what reason did you reject using the CB1 interrupt as a 'start listening' trigger? So the Oric would be off doing its other things, upon getting a CB1 interrupt it would attempt to receive a packet, which either would timeout quickly (if it was noise on the line) or else would be a real piece of content.

Re: Experimental very fast tape loading

Posted: Thu Oct 20, 2016 6:30 pm
by Symoon
ThomH wrote:
Symoon wrote:Networking PC and Oric through the tape port is an idea I've explored for yeaaars, but needed to improve skills on various sides:
- how to allow the Oric to be interrupted by an incoming signal (which is a real data signal)
...
For what reason did you reject using the CB1 interrupt as a 'start listening' trigger? So the Oric would be off doing its other things, upon getting a CB1 interrupt it would attempt to receive a packet, which either would timeout quickly (if it was noise on the line) or else would be a real piece of content.
I didn't reject anything ;) Just never found enough spare time to explore how to program: "how to time out when the Oric has recieved crap, and how to prevent the Oric blocking again and again if it recieves lots of crap".
It's certainly doable and maybe not too complicated, the only problem is... Me, and life ;)

Re: Experimental very fast tape loading

Posted: Wed Feb 22, 2017 10:08 pm
by Symoon
Hello! I need your help to test the reliability of an expected improvement on the very fast tape loading routines. Fabrice gave me (huge) hints on how to measure things in the best possible way (with a 3µs precision).

How to operate, once you downloaded the test kit:
- you need a perfect quality WAV player: quality of playback, and signal fully played (not cut at the beginning or ending)
- move away your cellphone and any other audio interference
- load MD3456-Interrupt.wav on an Atmos (no Oric-1, except if it has a ROM 1.1)
- RUN the program then load MD3456_test_sample_3_--___.wav and write the interval that appears after "Résultat final :"
- RUN the program then load MD3456_test_sample_4_--___.wav and write the interval that appears after "Résultat final :"
- RUN the program then load MD3456_test_sample_5_--___.wav and write the interval that appears after "Résultat final :"
- RUN the program then load MD3456_test_sample_6_--___.wav and write the interval that appears after "Résultat final :"
- finally send me those 4 results!

This allows me to know if all the Atmos will measure the same time interval between various very fast tape signal combinations. I need those results on a maximum of different Atmos machines. :D

Here's an example of results on one of my machines:
Atmos Paris: 3:[B7-AE] 4:[A2-9C] 5:[8D-87] 6:[75-6F]

Thanks in advance! :wink:
MD3456-Interrupt-test-kit.zip
New test kit
(5.86 KiB) Downloaded 41 times