User avatar
neodev
Posts: 452
Joined: 09 Apr 2018, 14:47

Firmware 1.2 WIP status

26 Apr 2018, 13:14

Hello

We have been working on a new firmware release to address the bugs people have found. At the same time we keep investigating the audio noise issue.

These are the fixes and features that have been included so far:

- [FIX] World Heroes 2 music now restarts each round (probably other games with same issue are fixed too)
- [FIX] Tengai Makyou - Ziria Hibaihin is now playable and doesn't show corrupted tiles
- [FIX] Adjusted the CD read speed for Gulliver Boy so video doesn't show glitches now
- [FIX] Improved Backup ram write/ SD read contention so it doesn't cause crashes.
- [FIX] Selecting the CDROM tab entering and exiting options now keeps the CDROM tab selected.
- [FIX] Solved random noise when switching tabs if user has a games.db and certain directory structure.
- [FIX] Fatal Fury 2 locking up randomly during fights
- [FIX] Sherlock Holmes video glitches (and probably some other games too)
- [FIX] Improve in game trigger in some games that were not recognizing the button presses
- [FIX] Fix cue sheet INDEX 00 entries processing
- [FEATURE] A previous backup file (.bup.bak) is created when saving new .bup file. On startup, if the .bup file doesn't exist or is 0 bytes, the .bup.bak file will be read.
- [FEATURE] Ability to adjust the CD block volume. Adjusting the volume will reduce the audio distortion in games with high volume level.
- [FEATURE] HuCard dumper and checker.

Now working on:
- Improve CUE parsing for some Trurip dumps.
- Magical Fantasy Adventure - Popful Mail broken cutscenes.
- Sherlock holmes audio/video desync (not too high prio)
- Glitches on arcade card that a couple of users have reported

I'll keep editing this post with progress.
Also at some time, we'll ask for the users that reported or tested the issues for their cart serials to send the firmware in advance and ensure it's all fixed before releasing it to the public.
Last edited by neodev on 23 May 2018, 11:46, edited 6 times in total.

User avatar
Todd
Admin
Posts: 1141
Joined: 08 Apr 2018, 22:25
Location: Cincinnati, OH
Contact: Twitter YouTube

Re: Firmware 1.2 WIP status

26 Apr 2018, 14:44

Excellent news!

:D
My PMs are disabled. Please send support requests to support(at)terraonion.com.

For real time assistance, please consider joining our very active Discord server: https://discord.gg/JexuJVW

User avatar
RevQuixo
Posts: 176
Joined: 09 Apr 2018, 15:31

Re: Firmware 1.2 WIP status

26 Apr 2018, 15:39

Great news. I'd be glad to test beta firmware as it becomes available for testing fixes. Also I know it is listed as low priority, but if you can fix the Sherlock Holmes games that would be great as it was one of the games highlighted by the My Life in Gaming video as being buggy. Would be good to squash it.

User avatar
neodev
Posts: 452
Joined: 09 Apr 2018, 14:47

Re: Firmware 1.2 WIP status

26 Apr 2018, 16:39

RevQuixo wrote:
26 Apr 2018, 15:39
Great news. I'd be glad to test beta firmware as it becomes available for testing fixes. Also I know it is listed as low priority, but if you can fix the Sherlock Holmes games that would be great as it was one of the games highlighted by the My Life in Gaming video as being buggy. Would be good to squash it.
Yes, I *think* I found the cause of the glitches in Sherlock and other HuVideo games. It seems to be related to the same problem than FF2. If the games constantly read some the CD block registers, they lock the ARM processing data read from the SD card (interrupt) because it has to answer the the status to the fpga and that has high priority, as the PCengine needs the value asap, if that happens while it's reading from the SD card fifo, it might become full and data was lost, as the hardware continuously streams data from the SD and expects the ARM to read it fast enough. That could be solved by using DMA in these reads to, so the ARM doesn't need to transfer the data itself, like it's done for audio reads, but this is going to need a lot of testing, as it's changing one of the very low level libraries of the firmware.

User avatar
GadgetUK164
Posts: 36
Joined: 14 Apr 2018, 11:19
Location: United Kingdom
Contact: Twitter YouTube

Re: Firmware 1.2 WIP status

26 Apr 2018, 17:53

Great news!!!! =D

SavagePencil
Posts: 83
Joined: 13 Apr 2018, 17:02

Re: Firmware 1.2 WIP status

26 Apr 2018, 18:11

I would also be happy to put it through its paces to help test!

User avatar
Mobiusstriptech
Posts: 340
Joined: 13 Apr 2018, 15:41

Re: Firmware 1.2 WIP status

26 Apr 2018, 18:26

Obviously I would be happy to test the new firmware and provide constructive feedback.

User avatar
AlexUnderBase
Posts: 26
Joined: 14 Apr 2018, 01:51
Location: Europe
Contact: Website Facebook Twitter YouTube

Re: Firmware 1.2 WIP status

27 Apr 2018, 02:17

:o ... Love this!
SSS3 - MEGASD - SD2SNES rev F - EVERDRIVE-64 v3 - EVERDRIVE N8 Fam - PSIO

atman
Posts: 10
Joined: 13 Apr 2018, 16:02

Re: Firmware 1.2 WIP status

27 Apr 2018, 08:46

Sounds great!

Tho, when I'm trying to register for downloads on the neosdstore.com it says the serial does not exist, tripplechecked the serialno on my SSDS3.

How can I proceed?, I have a neosd aswell.

User avatar
neodev
Posts: 452
Joined: 09 Apr 2018, 14:47

Re: Firmware 1.2 WIP status

27 Apr 2018, 14:52

I think the SSDS3 serials aren't set on the BD yet, as there is no firmware available for them (yet).

A bit of progress. Fatal Fury 2 now works great, I've been it playing for several hours and it's not freezing, it froze after 3 or 4 fights before. The CD sector read now goes through DMA so it's not affected by the game doing many CD status requests in a row. This also seemed to fix Gulliver boy without needing to reduce the read speed, and probably other HuVideo games. I need to check them.

For sherlock, I think I've found the source of the glitches. I have a fix, but it's not proper. It works, but it's a hack. There seems to be a wrong state handling in the ssds3 SCSI state machine that causes the game to try to read data before it has actually been transferred from the SD, and so, causes glitches. I need to trace the code to see what flags is it actually checking and what is wrong with them. This game uses its own cd access routines, so it might be doing something different than the bios routines and we are not handling it properly

Magical Fantasy Adventure still puzzles me. It's not a issue of random values read, as the glitches are always the same. Either some sector read is being missed or there is some data in the read request that is not being properly handled.

Return to “Super SD System 3 Discussion”