The Caslink3 project has been released!

Pagina 2/4
1 | | 3 | 4

Van Louthrax

Prophet (2093)

afbeelding van Louthrax

13-05-2016, 17:52

Yes, openMSX has no low-filtering on the signal, with the drawback of generating errors on .WAV files with parasites. Adding a low-pass fitler is on study, so that 25200bps might not work anymore !

Van Manuel

Ascended (15825)

afbeelding van Manuel

13-05-2016, 22:19

Louthrax wrote:

Has anybody tried to load >3600 bauds tapes on turbo / 7Mhz MSXs ? Maybe it could work (I've also never been able to transfer anything over 3692 bauds on any MSX machine).

I've been doing some experiments in openMSX with cas2wav conversion and found this (quoting CasImage.cc):

Quote:
// a higher baudrate doesn't work anymore, but it is unclear why, because 4600
// baud should work (known from Speedsave 4000 and Turbo 5000 programs).
// 3765 still works on a Toshiba HX-10 and Philips NMS 8250, but not on a
// Panasonic FS-A1WSX, on which 3763 is the max. National CF-2000 has 3762 as
// the max. Let's take 3760 then as a safe value.
// UPDATE: that seems to break RUN"CAS:" type of programs. 3744 seems to work
// for those as well (we don't understand why yet)
static const unsigned BAUDRATE = 3744;

I also have the idea that this is purely limited by the exact implementation on the BIOS routines that read out the bits. Exactly as you put it Smile

I guess the SpeedSave4000 and Turbo5000 programs use some custom loader after all? Or they don't actually work on such high baudrates after all?

Van NYYRIKKI

Enlighted (5399)

afbeelding van NYYRIKKI

13-05-2016, 23:00

Louthrax wrote:

Well, I stand corrected on this, there is indeed a filter applied on real MSXs. I was investigating why openMSX was not able to load the GP.WAV file (when a real MSX can), and discovered that applying a simple 14000Hz low pass filter fixed the issue (12000Hz is too low and does not work, values above 16000Hz might generate loading errors).

I actually now looked few schematics (SVI-738, Phillips VG-8000) but could not find anything to backup this theory... (Please note how ever that I'm pretty terrible in analog electronics.) Next natural question is also why would have they done that? I think that C-cassette it self acts pretty well as working low pass filter. I also noticed that in OTLA homepage they said that MSX was able to load with 48Khz & 2.25 samples / bit rate that means way higher than 14000Hz frequencies. Too bad they did not mention any model name where the test was done. :/

Van Louthrax

Prophet (2093)

afbeelding van Louthrax

13-05-2016, 23:17

I'll start coding a little tool to check this low-pass hypothesis : just IN A,(CASPORT) & OUT (BKGND_COLOR),a (to make it simple). Then send increasing frequencies to the CAS port. The stripes on the background should went from very thin to very thick/unique color at some frequency. Nice thing with this little tool is that we could check that openMSX and real MSXs have (almost) the same behavior with the fix. Wondering if there will be differences between different MSX machines too...

I'm no signal / electronic expert but I also have the feeling that this could be caused by the internal circuitry (maybe even the length of the cable ?? Though I have a pretty long (2m) cable to access all my MSX machines, and still can transfer at 3690 bauds to them...)

Van Manuel

Ascended (15825)

afbeelding van Manuel

13-05-2016, 23:27

This is also what is puzzling me: why would there be such a low pass filter be needed? Yes, the cassettes itself also act as a low pass filter, but that will already be directly ending up in the recorded WAV file from the tape, no need to filter it again. Also, this filter is apparently not in the hardware itself, as is shown with the MicroWaver and OTLA projects: the hardware is capable of much higher frequencies than normally used.

So Question ....

Van Louthrax

Prophet (2093)

afbeelding van Louthrax

14-05-2016, 00:17

Microwaver and OTLA should be using a different enconding algorithm as described by Nyyrikki (and still cope with the same frequeny limits) ??

On openMSX the low pass filter seems required to remove small parasites on the .WAV file, maybe on MSX it's the same but for the hardware signal?

Van NYYRIKKI

Enlighted (5399)

afbeelding van NYYRIKKI

14-05-2016, 03:01

Louthrax wrote:

Microwaver and OTLA should be using a different enconding algorithm as described by Nyyrikki (and still cope with the same frequeny limits) ??

On openMSX the low pass filter seems required to remove small parasites on the .WAV file, maybe on MSX it's the same but for the hardware signal?

Ah, the Microwaver!!! Now I only realized that we are talking Wav-file created by PC program, not recorded from MSX tape!

Here is what I think we are talking about:

Microwaver is designed to create WAV-file that will be recorded on tape and loaded on MSX. If it is used like this, it works. Now what happens is that someone has taken that WAV-file without recording it to tape first and tries to load it to openMSX. Because of this, it can't work because the magnetic tape is not there to lose the information. ( -> User error. openMSX is meant to load MSX tapes, not to filter WAV-files created by PC programs. )

C-cassette it self acts as a low pass filter that starts losing information quite a much even before 12KHz. The metallic Dolby-C tapes are a bit better and may on good day hold information quite well up to 15+KHz. How ever PC's, phones and other digital devices do better... If the information has been created on PC it can play even 24KHz tones and that might be problematic if only storage ability of ~12KHz was expected. The program can for example on purpose try to boost the high frequencies volume to cope with the data loss that is going to happen when the WAV is recorded to tape. When it is handled digitally, this expected data loss does not happen.

In Louthrax case the long cable it self may also act as a low pass filter... especially if the extra wire is stored on a roll.

Van Manuel

Ascended (15825)

afbeelding van Manuel

14-05-2016, 09:27

Afaik microwaver was meant to be used without tape, but directly from digital player to MSX.

Van cax

Prophet (3722)

afbeelding van cax

14-05-2016, 10:26

Manuel, you are right about MicroWaver. MTK chipset based DVD players + MP3 proved to be the best. All kinds of MP3 players and PC could not reach the high speeds and generated errors. I am also interested in understanding why, and I guess the reasons vary from device to device.

Van Manuel

Ascended (15825)

afbeelding van Manuel

14-05-2016, 10:56

Also, think about CD Sequential... for some reason it doesn't work properly in openMSX.... but why? On a real MSX it directly connects CD audio out to the MSX cassetteport.

Pagina 2/4
1 | | 3 | 4