Neat. Proof of concept.
I'm surprised that the PS1 actually can notice the ATIP wobble. I've always had my doubt about this. Normally read-only drives don't notice it (the signal from the pregroove is at least an order of magnitude weaker than the main signal)… in addition to the small amplitude.
This makes the task significantly hard than I thought; and I already thought it was VERY hard.
I've not really understood what has been modified on the writer in order to write
Feeling dumb after reading this.
I'm surprised that the PS1 actually can notice the ATIP wobble. I've always had my doubt about this. Normally read-only drives don't notice it (the signal from the pregroove is at least an order of magnitude weaker than the main signal)… in addition to the small amplitude.
This makes the task significantly hard than I thought; and I already thought it was VERY hard.
I've not really understood what has been modified on the writer in order to write
a) bigger (like Plextor GigaRec in reverse mode)
b) strange patterns simulating the wobble
My own thoughts went more into the direction of abusing the tracking coils while writing a CD with the help of an external signal. No idea if this is even possible at the insane speed needed (22.05KHz multiplied with the current write speed).Feeling dumb after reading this.
There is a 22khz filter specifically for testing the wobble on the PSX (among other hardware I believe). From my understanding, what he is doing is burning abnormally long pits and lands to the point that it is affecting the tracking, generating the exact same signal as i.e. a modchip. He is sort of writing over the actual pits and lands by making them so large (which effectively removes the 22khz CD-R signal since this causes the CD-R to spin faster and the PSX 'misses' the telltale sign that this is a CD-R and not a licensed CD-ROM). He does this in the lead-in (which is normally zero data) right next to (and sometimes over) the TOC since this is very early days and the method hasn't been perfected.
All of the issues this has currently (besides booting reliability) can be worked around by making a self-booting Tonyhax Boot CD which then just plays whatever normal CD-R backups or import discs. I really can't wait for more info to be released.
Blathering ahead!
I've a lot to write.
To the person who made it this far:
=========
I still want to know if the tracking coils are able to wobble fast enough to write a wobbled track. I have to emphasize that neither the PS1 nor a PC drive actively follows the ATIP wobble. The lens normally does not move that fast (a service manual mentioned 5KHz as cutoff frequency where the pickup stops trying to follow tracking error). The deliberate ATIP wobble has such a low amplitude that it doesn't cause any problems not actively following it. It can be noticed by different light intensity on each light detector while staying in center of the track.
Some explanation without warranty for correctness – to the best of my knowledge – might be wrong, or incomplete:
==============================================================================================
The sled can not remotely be positioned as accurate as it would be needed. In normal operation the tracking hardware gradually moves the lens from one side to the other following the spiral track – roughly moving the laser 1.6µm per rotation. At some point the sled has to move a bit and tracking coils move the lens back to the other side… gradually following the track until the sled moves a bit…
This is very, very far from 22.05KHz movement frequency. Let alone faster! If we burn at 4x the movement frequency required would be 88.2KHz so that the resulting signal meets the expectations of the PS1.
Yesterday – lucky coincidence – I bought a used and not really working all-in-one HiFi thing (3CD changer with MP3 support, 2 music cassette, radio, external input). Gave the lens a cleaning and now it is running smoothly. What has this to do with the topic? I could see the servo in extreme action on this thing! Not the tracking, but the focus servo. The CD drive is of exceptionally low quality with a HUGE imbalance. Not defective or worn, just really bad. Even when spinning the CD while the machine is off shows to the naked eye the disc goes up and down… and so does the focus servo with the lens to keep in focus (completely successful!)
That is the speed in which the focus (and tracking) usually work. As we can easily see it ourselves by just looking, it should be pretty clear that this is not in kilohertz range.
Granted, the wobble movement required for introducing a slight tracking error is much smaller – a tiny fraction of the maximum movement sideways. But the required frequency is insane!
A comment on the larger pit size speeding up the CD in the read process:
===========================================================
In certain boundaries this is within the specification. A 63min CD-R rotates roughly 17% faster compared to common CDs. Going to 150% speed like mentioned in the thread is pretty extreme, but can be done with Plextors out of the box (and usually does not cause trouble reading). The question is: Is it necessary to go that far? How much faster does it have to spin for the detector not noticing ATIP anymore? The other mentioned problem – reducing the pit size back to normal after the lead-in area – would be easier if the speedup was less extreme.
Despite the CD standard demanding the pit size to not have more than 0.1% deviation across the whole disc(!), drives are very lenient and can synchronize with changing data density (SecuROM new for example has sudden changes of 3% in data density up/down and for the most causes no problems despite violating the standard heavily). I fully agree with the suggested idea: Gradually reduce size back to normal in small steps. Should not lose sync this way or cause read errors and burn full commercial games without torturing the aging mechanical parts by permanent higher speed than intended.
What I can add for the other direction – slowdown: PS1 does not perform well with smaller pits. I could get GigaRec 1.2 (83% of normal pit size) to be detected as audio CD. No chance with level 1.3 (77% of normal pit size). Generally no drive is obliged to read discs with smaller than normal pit size.
A burner giving accurate control over the pit size (with ability to change on the fly) would be able to fully beat SecuROM (maybe also the DVD version).
I've a lot to write.
With this new development squash some of my assumptions:
ATIP will not to be detected as wobble due to very tiny amplitude.- This is why I've never considered speeding up or slowing down – this idea suggests itself – to prevent the 22KHz filter noticing it. This makes the PS1 protection even more impressive than it already is.
A burner with CFW can't do anything to create wobble.- I still believe the firmware has no direct control over the tracking hardware and can't introduce deliberate tracking error
- Not writing wobble but something else to simulate it fooling the detector is what I've not considered
To the person who made it this far:
=========
I still want to know if the tracking coils are able to wobble fast enough to write a wobbled track. I have to emphasize that neither the PS1 nor a PC drive actively follows the ATIP wobble. The lens normally does not move that fast (a service manual mentioned 5KHz as cutoff frequency where the pickup stops trying to follow tracking error). The deliberate ATIP wobble has such a low amplitude that it doesn't cause any problems not actively following it. It can be noticed by different light intensity on each light detector while staying in center of the track.
Some explanation without warranty for correctness – to the best of my knowledge – might be wrong, or incomplete:
==============================================================================================
The sled can not remotely be positioned as accurate as it would be needed. In normal operation the tracking hardware gradually moves the lens from one side to the other following the spiral track – roughly moving the laser 1.6µm per rotation. At some point the sled has to move a bit and tracking coils move the lens back to the other side… gradually following the track until the sled moves a bit…
This is very, very far from 22.05KHz movement frequency. Let alone faster! If we burn at 4x the movement frequency required would be 88.2KHz so that the resulting signal meets the expectations of the PS1.
Yesterday – lucky coincidence – I bought a used and not really working all-in-one HiFi thing (3CD changer with MP3 support, 2 music cassette, radio, external input). Gave the lens a cleaning and now it is running smoothly. What has this to do with the topic? I could see the servo in extreme action on this thing! Not the tracking, but the focus servo. The CD drive is of exceptionally low quality with a HUGE imbalance. Not defective or worn, just really bad. Even when spinning the CD while the machine is off shows to the naked eye the disc goes up and down… and so does the focus servo with the lens to keep in focus (completely successful!)
That is the speed in which the focus (and tracking) usually work. As we can easily see it ourselves by just looking, it should be pretty clear that this is not in kilohertz range.
Granted, the wobble movement required for introducing a slight tracking error is much smaller – a tiny fraction of the maximum movement sideways. But the required frequency is insane!
A comment on the larger pit size speeding up the CD in the read process:
===========================================================
In certain boundaries this is within the specification. A 63min CD-R rotates roughly 17% faster compared to common CDs. Going to 150% speed like mentioned in the thread is pretty extreme, but can be done with Plextors out of the box (and usually does not cause trouble reading). The question is: Is it necessary to go that far? How much faster does it have to spin for the detector not noticing ATIP anymore? The other mentioned problem – reducing the pit size back to normal after the lead-in area – would be easier if the speedup was less extreme.
Despite the CD standard demanding the pit size to not have more than 0.1% deviation across the whole disc(!), drives are very lenient and can synchronize with changing data density (SecuROM new for example has sudden changes of 3% in data density up/down and for the most causes no problems despite violating the standard heavily). I fully agree with the suggested idea: Gradually reduce size back to normal in small steps. Should not lose sync this way or cause read errors and burn full commercial games without torturing the aging mechanical parts by permanent higher speed than intended.
What I can add for the other direction – slowdown: PS1 does not perform well with smaller pits. I could get GigaRec 1.2 (83% of normal pit size) to be detected as audio CD. No chance with level 1.3 (77% of normal pit size). Generally no drive is obliged to read discs with smaller than normal pit size.
A burner giving accurate control over the pit size (with ability to change on the fly) would be able to fully beat SecuROM (maybe also the DVD version).
Never thought I'd see another Utopia Boot Disc style hack, interesting times ahead.
Similar threads
-
-
-
-
-
-
-
-
-
-
@
TwoSpikedHands:
yall im torn... ive been hacking away at tales of phantasia GBA (the USA version) and have so many documents of reverse engineering i've done -
@
TwoSpikedHands:
I just found out that the EU version is better in literally every way, better sound quality, better lighting, and there's even a patch someone made to make the text look nicer
-
@
TwoSpikedHands:
Do I restart now using what i've learned on the EU version since it's a better overall experience? or do I continue with the US version since that is what ive been using, and if someone decides to play my hack, it would most likely be that version?
-
@
Sicklyboy:
@TwoSpikedHands, I'll preface this with the fact that I know nothing about the game, but, I think it depends on what your goals are. Are you trying to make a definitive version of the game? You may want to refocus your efforts on the EU version then. Or, are you trying to make a better US version? In which case, the only way to make a better US version is to keep on plugging away at that one
-
@
Sicklyboy:
I'm not familiar with the technicalities of the differences between the two versions, but I'm wondering if at least some of those differences are things that you could port over to the US version in your patch without having to include copyrighted assets from the EU version
-
@
TwoSpikedHands:
@Sicklyboy I am wanting to fully change the game and bend it to my will lol. I would like to eventually have the ability to add more characters, enemies, even have a completely different story if i wanted. I already have the ability to change the tilemaps in the US version, so I can basically make my own map and warp to it in game - so I'm pretty far into it!
-
@
TwoSpikedHands:
I really would like to make a hack that I would enjoy playing, and maybe other people would too. swapping to the EU version would also mean my US friends could not legally play it
-
@
TwoSpikedHands:
I am definitely considering porting over some of the EU features without using the actual ROM itself, tbh that would probably be the best way to go about it... but i'm sad that the voice acting is so.... not good on the US version. May not be a way around that though
-
-
@
The Real Jdbye:
@TwoSpikedHands just switch, all the knowledge you learned still applies and most of the code and assets should be the same anyway -
@
The Real Jdbye:
and realistically they wouldn't
be able to play it legally anyway since they need a ROM and they probably don't have the means to dump it themselves -
-
-
-
-
@
Karma177:
do y'all think having an sd card that has a write speed of 700kb/s is a bad idea?
trying to restore emunand rn but it's taking ages... (also when I finished the first time hekate decided to delete all my fucking files
)
