OSSC v1.xx series beta firmware

[marqs]

The code is now on github (v1-series branch) and SD card version is also available for download (link added on opening post).

[uchristo]

Hi marqs!
Are there any differences to the former uploaded beta build?

BTW: I’d like to say “thank you!” for all the work your doing to keep this masterpiece of hardware up to date and support all those creepy non-standard input sources out there!

[marqs]

There are some fixes compared to the beta build but nothing major.

[BuckoA51]

I hooked up my Amiga (CD32 with Terriblefire) and did a round of testing with some of the Amgia’s more unusual screen modes. Some of which are documented here:-

https://amiga.lychesis.net/knowledge/ScreenModes.html

The only one that caused any issue was NTSC super high res. This failed to sync once, however try as I might I couldn’t get it to fail again subsequently. Obviously I’m running the OSSCs output through a splitter to several displays so perhaps it was just something else in the video chain playing up. The OSSC itself locked onto the signal and seemed perfectly happy with it.

Otherwise, every screen mode worked and produced a stable picture across OLED TV, computer monitor and capture card. I changed between modes multiple times and in different orders trying to trip things up but everything was fine. As with the pictures previously, my capture card doesn’t like the bob deinterlace modes, I need to connect up my old f/w OSSC and see if that’s the case there too.

Even tried the old “Productivity” mode which is reported as 384p and works just fine. 29.07kHz 57.34hz according to the LCD on the OSSC

Obviously these modes didn’t all look great as I expect sampling phase etc is well out from optimal on some.

Overall though a resounding success when paired with original Amiga hardware.

[uchristo]

There seems to be an overflow-issue with h.backporch. In UI, there‘s the ability to adjust more than 256. but internally there seems to be uint_8 in use because when increasing to a value over 254, the image jumps to the right and will vontinue to move to the left (until we reach 512 and the next overflow appears resulting in another jump to the right)

Edit: using the latest 1.01 SD-Image

• This reply was modified 1 year, 6 months ago by uchristo.

[marqs]

The issue with H.backporch range is now fixed and will be included in the next fw.

[xtrasmiley]

I’m happy with my current OSSC but what specific arcade PCBs will this improve compatibility with? I’m willing to test, but don’t want to do the mod unless there is something it (and I) can help with.

[marqs]

I can’t say for sure which specific arcade boards are more compatible with the new firmware. From the ones listed in wiki and which I’ve tested over the years, I suppose at least some Taito boards using constant even field indicator (e.g. G. Darius) should work more reliably. With the new firmware sync details should get always detected correctly and shown on the LCD, but that doesn’t yet guarantee line locking if there is major deviations in hsync period like with some Toaplan V2 boards.

[BuckoA51]

This won’t help with boards like Mortal Kombat that use weird refresh rates because OSSC cannot change refresh rate, otherwise I’d say it was worth a try.

[xtrasmiley]

Well PCBs like Mortal Kombat won’t sync to the TV, but the OSSC should still lock them in, right? I have a few PCBs that won’t even sync with the OSSC, not the TV itself.

So I’m wondering if this mod would help with that.

[BuckoA51]

Yes that’s correct, this mod could well help with that and we’d definitely appreciate your feedback.

[Aluminite]

I just installed the firmware and it’s working fine, except for one weird bug:

I created two profiles: one with the horizontal resolution set to 1920 for 240p, 288p, 480i, 480p, 576i, and 576p, and the other with a few settings modified. I use this first profile with a HDMI to VGA adapter on a CRT monitor to get better quality.

Whenever I turn on the OSSC when it’s set to the first profile, the test pattern does not show and I get a flashing “Signal Out Of Range” message on the monitor. When I turn the OSSC off, it stops flashing and goes back to being solid. Turning it on with the second profile does not cause this problem, and the only difference is the sampling settings. I also tried it on my regular gaming monitor and it does the same thing, but instead of showing a message I just get a black screen. Also, if I load the first profile after starting with the second profile, the screen stays on, and when I do the reverse the screen stays off. This only seems to affect the test pattern, if I set it to an input it works fine regardless of the profile. I had the exact same profiles on 0.90 and it didn’t cause this problem.

I exported the profiles if you want to try testing it: https://drive.google.com/file/d/1Ut7luFKIl8Ey-2QqN-gzLxuP1jtwIYQI/view?usp=sharing

 

EDIT: I tried reloading the profiles from the card and now none of them do anything. The “Adv. timing” option shows “———-” as the current resolution and going into it shows everything set to 65535. Something is broken.

• This reply was modified 1 year, 5 months ago by Aluminite.
• This reply was modified 1 year, 5 months ago by Aluminite.

[Aluminite]

Seems like something is broken with the profile saving/loading and export/import functions. I can get the bugginess to spread to other profiles and make them broken too.

Also, would downgrading to 0.90 work fine after doing the mod?

[BuckoA51]

No sadly you can’t downgrade without reversing the hardware mod. I’m sure Markus will look into this though.

[marqs]

It is possible to downgrade without reversing the mod. The red LED just doesn’t indicate anything then.

Parameters for test pattern mode now get read from 480p sampling preset which explain why your first profile breaks it. I’ll probably need to change it to get them from default 480p preset instead. There seems to be something wrong with importing profile from SD, I’ll need to take a closer look into that.

[Author]

Posts