The above video goes away if you are a member and logged in, so log in now!
Oh my god! U guys are about to break the Scenix code protection? If you guys manage to make the scrambled data to the real 3.1 hex, please release it public..
I cant wait to see if you guys make it! REAL Magic 3.1 Homebrew....... I just cant wait...
Keep up the good work!
no, that is NOT true.
because the REV16 doesnt have code included for JAPANESE consoles ie, and the Final Magic 3 version have it.
japanese code is in an extra magic 3 version. look at www.china-magic.com and you will find the list.
the final magic 3 is not working in japanese consoles.
magic 3.0 final and beta r16 are nearly the same and and both don't support the japanese ps2.
hmm, what about the comapare of magic 3.1 to 3.0 ?
i see with luck, we can get this based on my idea of comparing the garbled data and find the possible instructions for each garbled output.
nope, i think it's just a typo - those files match exactly- consider this...
In order to add extra code for the japanese version, there would have to be an extra routine (or at least a handful of new instructions). This means we would have less blank space in one file - which we don't. You can see the 000F's start and end in exactly the same place.
What is different is the way we return a byte from our data tables. The way to do this is add W (our offset) to our current address (in file register <02>). We see this instruction is
And from the scenix data sheet, this Opcode should be
01E2..... Which XOR's to the correct value, 000D - we match exactly!
What went wrong? Excel tried to help us out - it interpreted 01E2 as 1*10E2 = 100, and replaced our data 1E2 with the number 100.
Remember to IMPORT AS TEXT if using excel, or it will help you by formatting all your data into useless garbage. I won't make that mistake again.
Sooo... cjack's file is the same as an r16, which probably means a lot of dealers are LYING and saying their chip is 'final' when it is exactly the same as the old r16 code. This is why we need to compare ALL versions to find the truth. Money makes many people do funny things - it is only the people interested in the knowledge that learn the truth!!
We still need a garbled hex file from magic 3.1 - if anyone needs to know how to connect your red/blue PCB to a programmer, send me a PM or email!
ok, to find out, if somebody sells bad stuff, just compare the garbled data. if it matches to r16 or 3.0 code. then they are lying.
i know professionell clones, claimed as 3.1 chips, but it is 3.0 code in it. they are all lying.
i have genuine magic 3.1 here and give the hex file. but it take some time for me.....
have you heard about magic 4? what is this?
Another good reason for people to post their garbled hex readouts from ANY Magic chip - we can make a table of which dealers are selling fake versions..
I see an ad for a 'magic 4' here
Seems fake - the chip is not a SX (looks like FLASH memory) so I think it's a DMS clone or a messiah clone - they're the only chips using a seperate flash memory chip to store programs (yet).
Your scrambled Magic 3.1 hex is going to tell us a lot, though - this is turning into a very good attack!
ok folks, i destroyed my 50 ? = 50 $ expensive magic 3.1 original blue chip.
here is the garbled hex file.
veryfied 5 times.
now give me the good code....
winrar says the file is broken - can you upload the hex without compression?
ooops, this magic 4 is a nh3 and contains magic 3.0 code, but in an asic, not an sx chip. runs in v7 pal, but gives probs on v5 pal.
i have one nh3 here.
ok, here is also garbled output of a green magic 3.0 clone.
i also destroyed this. puh, lots of money, again a 50 ?= $.
i put in this zip file both, magic 3.0 clone and magic 3.1 garbled hex output.
in this zip file contains:
magic 3.0 garbled hex (green 3.0 clone chip)
magic 3.1 garbled hex (original 3.1 blue chip)
modblak 1.0 original hex (he he, that code works perfect)
modblak 1.0 garbled hex
magic 3 beta r16 original hex (he he, that code works)
magic 3 beta r16 garbled hex
all files are tested and 5 times read out and compared, so that no mistake is in it.
it cost a 100 ? = $ to destroy my good chips.
so, homebrew people, help, that i spend my money not for nothing.
so, guichi can verify, if he has original code in his red magic 3.1, he he...
Frankieboy we all thank you very much for your sacrifice! Please know that it is people like you who make this homebrew work possible!
Now on to the fun. I will post some files here soon, but when you look at it with a quick glance - it looks very hard. I don't want people to get discouraged, so I will try to do the easy fixes first to make it look better.
Here is what we're dealing with.
1) Subroutines that have been moved because other routines get a little longer. Shifing the Magic3.1 code down shows us our old matching routines again, no big deal.
2) When things move down a few bytes, like in #1, that makes our JMP and CALL's go to new places - this looks like a long string of matching bytes, but all the JMP and CALL don't match. Again, not a big deal. (this even helps by giving us a hint to the start location so we can align #1 better!)
3) Using different registers for the same function - this shows up as matching before and after, but a block of non-matching instructions. You have to start thinking about what the program needs to do to patch the BIOS, etc. Now we get into programming, but this is simple re-writing of how things work.
4) The hardest - new routines. We know by the advertisements that there will be the new japanese version patching in this one. A good guess is that it's next to all the other country patching we know from r16, but all we have is a big blank spot. This stuff needs to be written from scratch, using the garble as a checksum and the other routines as a template. (Programmers copy and paste lots of code - similar routines probably recycle the setup)
848/2048 differences between MAGIC3.0 and 3.1. That's not exact because some matches are just coincidence - but OVER HALF of the chip is exactly the same, and in the same place.
I like the sound of that