NEOSD MVS Support thread !

[Sceptre_JLRB]

yes, zintrkcd does not have the issue because I made that conversion... Proper like.

Upsss, haven't noticed, sorry!

Razoola's conversions r00l!

 

[greatfunky]

Do the wrong sprites flicker up on the red lady though when it shows a close up of her?

I didn't notice specially sprites flickering around the red lady but clearly during all the intro .
I think there's also a difference depending of the motherboard , i noticed different glitch between my mv1fz and mv1fzs boards , less or more glitch but glitch anyway....
You can see the difference in my video.

 

[GadgetUK]

This is the glitch I get (private not a public link) - just so you can see what I mean. https://www.youtube.com/watch?v=zaiN8lw9S44 It will get deleted later probably.

 

[greatfunky]

No, not like that ,i haven't seen a flickering like this one before , for me it's always the same kind of little horizontal glitch as in my video.

 

[neosd]

This is the glitch I get (private not a public link) - just so you can see what I mean. https://www.youtube.com/watch?v=zaiN8lw9S44 It will get deleted later probably.

Hello,

This video is a different issue, it looks tiles are bad
Can you check the .neo rom you are using with the validator ? It may be wrong
If the rom passes the check, can you flash the game again ?
Thanks

 

[Razoola]

Hello,

This video is a different issue, it looks tiles are bad
Can you check the .neo rom you are using with the validator ? It may be wrong
If the rom passes the check, can you flash the game again ?
Thanks

I don't think that is the case because the tiles are being drawn but in the wrong place. You can't rule out a vram issue or a P ROM issue (corrupt or banks) but its not bad Sprite tiles I feel.

 

[GadgetUK]

I agree its definitely not the ROM. I've validated it OK, and I also tested the AES rom (completely different ROM with different files) and the same problem happens in the same spot on screen. What does change between power cycles is the exact corruption in those positions on screen, it can vary a little each time you boot it, but always in that part of the intro with the red lady.

 

[neosd]

I agree its definitely not the ROM. I've validated it OK, and I also tested the AES rom (completely different ROM with different files) and the same problem happens in the same spot on screen. What does change between power cycles is the exact corruption in those positions on screen, it can vary a little each time you boot it, but always in that part of the intro with the red lady.

Thanks,

We are going to investigate it tomorrow aswel.
Can anyone with an MV1FZ confirm this is happening there aswel ?

Also, GadgetUK have you got another MVS to check if this is happening into another board ?

Thanks

 

[Razoola]

I agree its definitely not the ROM. I've validated it OK, and I also tested the AES rom (completely different ROM with different files) and the same problem happens in the same spot on screen. What does change between power cycles is the exact corruption in those positions on screen, it can vary a little each time you boot it, but always in that part of the intro with the red lady.

You could quickly try the unibios crc32 check with kof2003 a few times and confirm the P ROM banks pass every time. Given how games pull up gfx from the banks it might be an outside chance of a problem there.

 

[GadgetUK]

You could quickly try the unibios crc32 check with kof2003 a few times and confirm the P ROM banks pass every time. Given how games pull up gfx from the banks it might be an outside chance of a problem there.

I think its possibly a cross talk problem. Thanks for that suggestion Raz, ROM Bank 2 and 4 give occasional NG results. For example, on first CRC check, all banks good except for bank 2. I then repeat the test and now bank 2 is OK and bank 4 is NG. Repeating this several times it seems to always show either bank 2, bank 4, or both as NG. Every other game I've tested work fine and all games show good CRCs. It does seem like a cross talk / noise issue. If Neo SD team can identify the problem I can easily add a cap or resistor if it is a cross talk issue or something. I also suspect if I put some grounded copper tape in the right place it might fix the problem too, but I am not touching the board until we've tried some firmware updates. Pretty cheesed off tbh =/

See how Neo SD team get on tomorrow.

 

[neosd]

You could quickly try the unibios crc32 check with kof2003 a few times and confirm the P ROM banks pass every time. Given how games pull up gfx from the banks it might be an outside chance of a problem there.

Hello

We have just tried kof2003 into two different MV1FZ boards, and we cant get the glitches on any of those.

Could you check the voltage on the jamma connector to see if you have 5v there and not less ?

Also, could you try to pass unibios crc checksum on kof2003 ?

Thanks

 

[GadgetUK]

4.97v! Good clean PSU! See my comments just posted - CRC changes each time you run the CRC check (Not Good reported on bank 2 or bank 4, or both - they change every time) - ONLY on KoF2003!

EDIT: Are you using PLL somewhere because it crossed my mind that the way the behaviour changes with the lines (as shown in greatjunky's video), and the way I can stop those lines by keep switching on and off until they disappear makes me think a clock is not syncronised precisely.

It's absolutely not the MVS because I've got a few different boots of KoF2003 and they work fine, and if I try the boot version on the Neo SD the same exact graphical corruption appears. So the fact that the same bootleg version on a boot cart works fine, vs that exact same bootleg ROM running on the Neo SD not working shows its not my MVS.

 

[Razoola]

I feel the failing banks is the cause of the hair glitch issue you have in kof2003. The information you gave on that will help neosd. You could try the crc32 check with svc and mslug5 also just to see if you get the same issue there with banks failing sometimes.

 

[neosd]

4.97v! Good clean PSU! See my comments just posted - CRC changes each time you run it - ONLY on KoF2003!

Hello,

We are going to look at it tomorrow, but we wanted to reproduce the intro issue here right now.

We will contact you tomorrow as soon as we know more about the issue.

Thanks for the info

 

[GadgetUK]

I feel the failing banks is the cause of the hair glitch issue you have in kof2003. The information you gave on that will help neosd. You could try the crc32 check with svc and mslug5 also just to see if you get the same issue there with banks failing sometimes.

I will do that now! Thanks, I was going to ask you which games made similar use of the program banks.

 

[Razoola]

Of course the other issue you have is something different, maybe a slight voltage increase will help there, who whows. A little worrying neosd does not yet reproduce the glitching in the other video on their mv-1fz. Will be much easier if they can reproduce it.

 

[GadgetUK]

Of course the other issue you have is something different, maybe a slight voltage increase will help there, who whows. A little worrying neosd does not yet reproduce the glitching in the other video on their mv-1fz. Will be much easier if they can reproduce it.

Yes, that worries me too! I didn't see the 'lines' issue on their video either - yet two people now have that problem. If it is a crosstalk / impedance problem the temperature and humidity is factor. I am testing this in a really cold area at the moment. The power is definitely not a problem, I've tried a few really good 5v supplies I have and there's no ripple and remains at a 5v or just above.

Just tested SVC and all checksums OK and the game is OK (except for the 1st issue reported of the short lines jittering over sprites).

Testing MS5 now.

 

[Rot]

I LOVE nerd talk...

It reminds me that I have no idea what you're talking about... but the fact that I will read anything on the forums...

Have fun playing lads... hope its an easy fix...

xROTx

 

[neosd]

Yes, that worries me too! I didn't see the 'lines' issue on their video either - yet two people now have that problem. If it is a crosstalk / impedance problem the temperature and humidity is factor. I am testing this in a really cold area at the moment. The power is definitely not a problem, I've tried a few really good 5v supplies I have and there's no ripple and remains at a 5v or just above.

Just tested SVC and all checksums OK and the game is OK (except for the 1st issue reported of the short lines jittering over sprites).

Testing MS5 now.

Hello,

I think i have myself similar temperature and humidity here as you have there.
The rest of the team are in different parts of Spain, so we will know more for sure tomorrow.

What puzzles me is that i have tried it in two different MV1FZ boards and it dosent happens here (both videos are from different boards)

We will know more tomorrow for sure.

We really apreciate the time you are spending on helping us to figure the issue.

Thanks

 

[GadgetUK]

Checked MS5 - CRC problems with that reported too. Bank 0 checksum changes each time you run the CRC test. My original and boot MS5's both work fine, so definitely something related to the prog bus.

In summary there are two issues:-

1) Short lines of pixels flickering over sprites - often powering off and on a few times can make this go away.

2) Graphical issues that I see in KoF2003 intro, and as we've seen this is detected as different (changing) checksum failures in bank 2 and bank 4 on KoF2003. SVC was OK in all respects, and MS5 reported different checksums each time on bank 0, yet more original and boot versions of both KoF2003 and MS5 work fine and always report their correct checksums.

 

[GadgetUK]

Hello,

I think i have myself similar temperature and humidity here as you have there.
The rest of the team are in different parts of Spain, so we will know more for sure tomorrow.

What puzzles me is that i have tried it in two different MV1FZ boards and it dosent happens here (both videos are from different boards)

We will know more tomorrow for sure.

We really apreciate the time you are spending on helping us to figure the issue.

Thanks

Thanks, and no problem! I just don't want you to think this is just my system - the lines issue has been reported by another person earlier (so I am pleased theres at least two of us with that problem), but this CRC / corrupt sprites thing is my main concern now as it seems hard for others to recreate. I guess as more people receive them theres more chance of others reporting the same problem.

I also think its not a hardware fault because other games that utilise similar (or more) flash ROM aren't reporting bad CRCs, nor are they giving any corruptions (except for that line flicking thing).

 

[neosd]

Thanks, and no problem! I just don't want you to think this is just my system - the lines issue has been reported by another person earlier (so I am pleased theres at least two of us with that problem), but this CRC / corrupt sprites thing is my main concern now as it seems hard for others to recreate. I guess as more people receive them theres more chance of others reporting the same problem.

I also think its not a hardware fault because other games that utilise similar (or more) flash ROM aren't reporting bad CRCs, nor are they giving any corruptions (except for that line flicking thing).

Hello,

The lines issue is not your issue for sure, cause as i told you by pm, we had those during development and it was a timming issue. We thought this was fixed cause we stopped to see it after we changed some things, but it seems its not ...

The seccond issue is what puzzles me, cause i cant get it reproduced.
Not saying its your fault, but i had a faulty MV1FZ board that was reseting itself sometimes and we spent a lot of hours thinking it was an issue and it was the board itself. Those boards are pretty old and some may have faults. Thats why i asked so many things ...

Again, i know you are not the usual user and you know what you are talking about, thats good cause it will help us to locate and fix the issue.
We have to reproduce the issue to see whats happening, but with the info you and Raz have posted, we know where to look at it tomorrow.
We will prepare you a firmware changing timmings so we can know if that improves it or not ...

Thanks

 

[Razoola]

Thanks, and no problem! I just don't want you to think this is just my system - the lines issue has been reported by another person earlier (so I am pleased theres at least two of us with that problem), but this CRC / corrupt sprites thing is my main concern now as it seems hard for others to recreate. I guess as more people receive them theres more chance of others reporting the same problem.

I also think its not a hardware fault because other games that utilise similar (or more) flash ROM aren't reporting bad CRCs, nor are they giving any corruptions (except for that line flicking thing).

If you go into the UI and then the options screen it may be helpfull if you report the version numbers reported for all areas (don't give your neoSD serial publically though).

 

[X68]

Thanks,

We are going to investigate it tomorrow aswel.
Can anyone with an MV1FZ confirm this is happening there aswel ?

Also, GadgetUK have you got another MVS to check if this is happening into another board ?

Thanks

Works OK for me no glitching on MV-1FZ, here's the vid
https://youtu.be/8jWz2_QXSOw

EDIT: using the roms from // [THE FUCK IS THIS?]

 

[Rot]

You know what this reminds me of lads?

...An episode of Poirot... where there's lots of Red Herrings and the murderer is the one you least expect...

xROTx