PS3 [Attempt] Restoring XMB UI of Firmware 1.00 - 2.70 to 4.84+

[Danxx444]

By the way ... did you check the images? ... noticed that there is something "pulsing" that comes out from inside the clock (I was trying to send a video but was giving error) ... since there are no images per frame in watch..so what would this be?

 

[Danxx444]

Ok, anyway... is the same problem i mentioned, it seems the explore_plugin_full.rco used by HEN have some custom texts
And the explore_plugin_full.rco i made doesnt have them because i used the official file as the base for the mod

At this point is easy to fix it though, you can take the explore_plugin_full.rco from HEN and...
1) extract contents with rcomageGUI (and dont enable the checkboxes for file conversions)
2) replace the "tex_playing.gim" and "text_playing_shadow.gim" with the gim files i made
3) compile (and enable the checkbox for zlib header compression)

Hmmm..okay then...thanks

 

[pinky]

yes, it's easier to commit out some code by disabling it, than rewriting. I'm sure microsoft and nintendo probably do the same thing. I've looked at some of the 360s flash files. it's kinda safe with xell being there, in case you brick. xell is only 50 blocks, and can be used to install a firmware update if need be. I've only updated a couple times that way. the most delicate of the three seems to be nintendo. the 3ds and switch have some recovery menus (never seen either), but I don't know about the wii u.

 

[Danxx444]

@sandungas..we may not even be able to use the old loading animation, but I'm pretty sure that we can make visible only this pulsating loading animation ... maybe editing the tex_busy, tex_busy_l and tex_busy_s images to invisible.

 

[pinky]

something like that happened with one of my mods. I changed the colors, and I wanted to show a friend. iirc, he was on 4.50, and I was one or two firmwares higher. anyway, instead of it showing different colors, the clock was gone completely.

 

[Danxx444]

something like that happened with one of my mods. I changed the colors, and I wanted to show a friend. iirc, he was on 4.50, and I was one or two firmwares higher. anyway, instead of it showing different colors, the clock was gone completely.

oh ... really? ... the things that seem to be the simplest ... are the ones that make the most trouble.

 

[pinky]

yeah, speaking of modding other systems. I've seen some people change the power off screen on the 3ds to various colors, including pink. I think you can save yourself with godmode and a NAND backup if something goes wrong. I've heard of bad things happening like frankenstein firmware when something gets corrupted. I've also heard that the default apps or even game apps may appear as a black square if something is very wrong with them. it seems to be a lot more complicated than it is on the ps3.

on the 360, you can make custom themes, but they have to be skinned to a specific version of the hacked dash or parts of it will be missing. that's happened to me before, so now I just use the default theme.

 

[pinky]

I've corrupted a game icon before on the wii u by not using the right save settings. I figured out those settings later on. if an icon is corrupted, it will show a question mark over the game, and the name of the game will be ???. I know the meta.xml is what displays the name for a game. I can't remember if the icon is in the same location. @sandungas , I think, mentioned there being some ps3 games that won't run if the icon0.png is missing or corrupted. never encountered one, but it might be some sort of verification process. I think that's what happens with games like uncharted and infinite loading (missing or possibly corrupted files).

 

[Danxx444]

yeah, speaking of modding other systems. I've seen some people change the power off screen on the 3ds to various colors, including pink. I think you can save yourself with godmode and a NAND backup if something goes wrong. I've heard of bad things happening like frankenstein firmware when something gets corrupted. I've also heard that the default apps or even game apps may appear as a black square if something is very wrong with them. it seems to be a lot more complicated than it is on the ps3.

on the 360, you can make custom themes, but they have to be skinned to a specific version of the hacked dash or parts of it will be missing. that's happened to me before, so now I just use the default theme.

Cool ... but particularly I just like to modify the system to try to restore some of its previous firmware appearance or functions ... I can modify certain things from smarphones (custom rom's) is "easier" than modifying a console .

 

[sandungas]

Before moving on to other stuff i would like to mention a couple of things of what we was doing, just for curiosity sake, and to clarify some details and the doubts you have @Danxx444

I just tested the files ...

the first image is from tex_busy2.17 for tex_playing4.84

and the second image is from tex_busy2.51 for tex_playing4.84

Overall that images are displayed fine, this is how it was looking in old firmwares (but located in the top-right corner of the screen)
I see 2 problems, one made by sony and another made by me

The problem made by sony is the dark pixels of the shadows are too much intense, the shadow needs to be dark because otherway if you display a shadow in white on top of a white background you are not going to see anything (white pixels on top of white = nada)
i guess this can be solved by making the pixels of the "playing_shadow" more darker... it looks in the shadow image the black color represents transparency

The problem made by me is what i explained in one of the first post when we was talking about this
https://www.psx-place.com/threads/a...re-1-00-2-70-to-4-84.26076/page-5#post-204040

The purpose of this images is the same, we can replace them, but is needed to rebuild them because have different size
The good news is both are composed by 30 "frames", the bad news is the frame size is different
To do it well (without losing any quality) is needed to crop all the frames, resize all them, then join all frames together again, etc...
It requires some work (i guess half an hour or so), so before doing it i would like to know if this replacement is what you want to do

Also, before doing it i would like you to test a "low quality" version of the rco (a bit pixelized) to see how it works

The problem is the size of the frames, for 4.84 the frames are a perfect square, but in older firmwares are not a perfect square
What i did in all this tests is to scale the image (with the 30 frames) entirelly, and this "streches" the frames... so what originally was a circle could become an oval
Im not sure how much the frames was deformed, is needed to do some maths to see it, basically is like this (copyed from one of the readme.txt i uploaded)

-tex_busy217.gim             MD5:AC292D7AF3BA478A87400AC40FDA7574 (29x750 pixels -ps3rgba8888)
-tex_busy_shadow217.gim      MD5:862F858E85BEA609FF8E9C78EE73101A (32x960 pixels -ps3rgba8888)

-tex_busy251.gim             MD5:AC24824CF4BD2DB82AE2FE512802968A (32x960 pixels -ps3dxt5)
-tex_busy_shadow251.gim      MD5:3D45ED809E9A81774D3CF5B3481105ED (32x960 pixels -ps3dxt5)

-tex_playing484.gim          MD5:85DEB25B347E39BAA66FD9A940F7C625 (30x900 pixels -ps3rgba888)
-tex_playing_shadow484.gim   MD5:04FDF3DC5D06D36F2643210A8E8C5019 (30x900 pixels -ps3rgba888)

tex_busy217.gim <--------------------- frame size 29x25
tex_busy_shadow217.gim <------------ frame size 32x32

tex_busy251.gim <--------------------- frame size 32x32
tex_busy_shadow251.gim <------------ frame size 32x32

tex_playing484.gim <------------ frame size 30x30
tex_playing_shadow484.gim <------------ frame size 30x30

So... the frames needs to be increased in size up to 30x30 individually (but not resizing them, it needs to be made by changing the canvas size in photoshop, this preserves the pixel data and the original dimmensions of the circles inside the frames)

I was aware that this "problem" of deformation of the circle could happen, i was not sure because i didnt made the math calculations or tests in photoshop to see how much it was going to be deformed, in your photos cant be seen well but you mentioned it so i guess is notable

 

[Danxx444]

Before moving on to other stuff i would like to mention a couple of things of what we was doing, just for curiosity sake, and to clarify some details and the doubts you have @Danxx444


Overall that images are displayed fine, this is how it was looking in old firmwares (but located in the top-right corner of the screen)
I see 2 problems, one made by sony and another made by me

The problem made by sony is the dark pixels of the shadows are too much intense, the shadow needs to be dark because otherway if you display a shadow in white on top of a white background you are not going to see anything (white pixels on top of white = nada)
i guess this can be solved by making the pixels of the "playing_shadow" more darker... it looks in the shadow image the black color represents transparency

The problem made by me is what i explained in one of the first post when we was talking about this
https://www.psx-place.com/threads/a...re-1-00-2-70-to-4-84.26076/page-5#post-204040

The problem is the size of the frames, for 4.84 the frames are a perfect square, but in older firmwares are not a perfect square
What i did in all this tests is to scale the image (with the 30 frames) entirelly, and this "streches" the frames... so what originally was a circle could become an oval
Im not sure how much the frames was deformed, is needed to do some maths to see it, basically is like this (copyed from one of the readme.txt i uploaded)

-tex_busy217.gim             MD5:AC292D7AF3BA478A87400AC40FDA7574 (29x750 pixels -ps3rgba8888)
-tex_busy_shadow217.gim      MD5:862F858E85BEA609FF8E9C78EE73101A (32x960 pixels -ps3rgba8888)

-tex_busy251.gim             MD5:AC24824CF4BD2DB82AE2FE512802968A (32x960 pixels -ps3dxt5)
-tex_busy_shadow251.gim      MD5:3D45ED809E9A81774D3CF5B3481105ED (32x960 pixels -ps3dxt5)

-tex_playing484.gim          MD5:85DEB25B347E39BAA66FD9A940F7C625 (30x900 pixels -ps3rgba888)
-tex_playing_shadow484.gim   MD5:04FDF3DC5D06D36F2643210A8E8C5019 (30x900 pixels -ps3rgba888)

tex_busy217.gim <--------------------- frame size 29x25
tex_busy_shadow217.gim <------------ frame size 32x32

tex_busy251.gim <--------------------- frame size 32x32
tex_busy_shadow251.gim <------------ frame size 32x32
now i get it .... but..you will use math to figure out how many frames have been deformed? ... wow ... but that's ok..no more need to modify this icon to make it work in explore_plugin_full.rco. .my intention was different..unless you want to solve your personal challenges.
tex_playing484.gim <------------ frame size 30x30
tex_playing_shadow484.gim <------------ frame size 30x30

So... the frames needs to be increased in size up to 30x30 individually (but not resizing them, it needs to be made by changing the canvas size in photoshop, this preserves the pixel data and the original dimmensions of the circles inside the frames)

I was aware that this "problem" of deformation of the circle could happen, i was not sure because i didnt made the math calculations or tests in photoshop to see how much it was going to be deformed, in your photos cant be seen well but you mentioned it so i guess is notable

Now i get it .... but..you will use math to figure out how many frames have been deformed? ... wow ... but that's ok..no more need to modify this icon to make it work in explore_plugin_full.rco. .my intention was different..unless you want to solve your personal challenges.

 

[sandungas]

Now i get it .... but..you will use math to figure out how many frames have been deformed? ... wow ... but that's ok..no more need to modify this icon to make it work in explore_plugin_full.rco. .my intention was different..unless you want to solve your personal challenges.

Yep, at this point is a bit pointless to continue with what we was doing, it requires some additional work to do all that cropping of the frames to preserve the circle geometry, also what i mentioned about the intensity of the dark pixels in the shadow is the kind of thing that needs several tests to do a very good adjustment
I know there is an small room for improvement in what i did, but i think im not going to do it, the files i uploaded are a nice example incase someone wants to do it though

Now im more interested in what you have been mentioning in your last posts... when i sent one of my posts where i said the actual clock doesnt have an animation at the back i reminded i saw something like "expanding" in my XMB at some point behind the clock
Yeah, it seems to have something, is needed to see what is it, by now i dont know

Have you found some trick to trigger that animation for long in XMB ? (this way we can see it better sticking the nose in the TV)

 

[DeViL303]

this way we can see it better sticking the nose in the TV

This might help, not sure, have not been following this thread really.

2.17

2.51

 

[sandungas]

The animation for 2.51 sucks btw, look at this, i took the images used in the rco files i uploaded before, are rotated 90º to fit better in screen, scaled up to 160% zoom, every image have his shadow at back, i modifyed the shadows by "invert colors" + "replace color" (white by a middle gray tone)
This allows to see better how they looks together with his shadows, his geometry and the intensity of the shadows, from top to bottom:
tex_busy217
tex_busy251
tex_playing484

The "tex_busy_shadow" of 2.17 and 2.51 looks exactly the same visually, in other words... is the same source image but when they converted them to GIM became different because uses different GIM settings

-tex_busy_shadow217.gim      MD5:862F858E85BEA609FF8E9C78EE73101A (32x960 pixels -ps3rgba8888)
-tex_busy_shadow251.gim      MD5:3D45ED809E9A81774D3CF5B3481105ED (32x960 pixels -ps3dxt5)

tex_busy_shadow217.gim (made with rgba8888) is lossless
tex_busy_shadow251.gim (made with dxt5) is lossy, it have lower quality than tex_busy_shadow217.gim

 

[sandungas]

This might help, not sure, have not been following this thread really.

2.17

2.51

Ohhhh, now i got the design concept, i always thought that it was like a waterdrop filmed from a perfect vertical (in that case the circle should be a perfect circle). But is an oval because is a waterdrop filmed from a perspective view !!!
What a bad idea, lol... if i never realized about what they was trying to represent im sure many more people could not get it either
The perspective angle is pretty much the same than this image, and yeah is an oval:

So... in the rcos i made with the images from 2.17 have the biggest deformation (because was the smallest), they looks like a circle because a coincidence, originally was ovals and when i stretched i made them to look like a circle

And the rcos i made with the images from 2.51 are ovals, probably i deformed them a bit, but overall was intended to be displayed like that

 

[sandungas]

I realized a cool visual trick they did in tex_busy_shadow251
As i mentioned before tex_busy_shadow217.gim and tex_busy_shadow251.gim are exactly the same image
This shadow images was generated in a image editor app by using tex_busy_217.gim or older
In 2.17 the shadow matches perfectly with the image that goes on top

But in 2.51 they changed a bit the image that goes on top of the shadow, and also they displaced some frames from his positions for 3 reasons:
-because the first frame of this new animation is better, it starts with an small "hole" in the center (instead of a "dot" in the center)
-the circles increases his diameter along time, like in real world if you throw an stone to a lake
-because this way the shadow is disaligned with the image that goes on top, lets say... the dark pixels "follows" the white pixels from behind, this increases the effect that is "swirling"

Is hard to explain, not sure if that disalignment of the shadow was made on purpose or if it was a mistake, but i like the idea, lol

Take a look at the original images, and keep in mind the frame most at top is frame=1, and the frame most at botom is frame=30. And the animation plays at 30 frames per second, so 1 loop = 1 second

 

[sandungas]

At this point im pretty sure how they did it, there is only one way because is very smooth
They used a 3D program with fluid dynamics, like 3Dmax, the fluid container was a cylinder closed at the base, and the cylinder is animated to move up and down
This is going to create the circles, by adjusting the distance and speed of the cylinder movements they made the fluid to enter in "reverberation" with an intervale of 1 second... in other words, the fluid dynamics of the circle gets balanced to make a perfect loop of 1 second

But additionally it have another object (probably an sphere) rotating in circles under the water, this is what creates the "swirling" effects at the most external circles

Then they recorded it, exported it in 2D and cropped the 30 frames

But... think in it... if you fill a can with water and move it up and down the circles that appears in the surface are going to decrease in diameter, because the walls of the can "pushes" them to the center making them smaller, this is how is made the animation in 2.17 with the circles decreasing his diameter

In 2.51 they decided that it was better to invert the animation, so... (i think) they kept the same frames that they had already recorded and changed his positions
And... they forgot to change the positions of the frames in the shadows to make them match (or they made it on purpose... or both first it was a mistake and when they saw it they thought "lulz, that mistake looks nice")

------------------------
I have the feeling this was only 2 chapters of a long story when they was doing some experiments with that animation... i wonder if there are better versions of it in other firmwares

Btw @Danxx444 you uploaded a system_plugin.rco and you said it was from 2.51 but in wiki it looks like the only 2.51 firmware it was specific for kiosk/shop PS3 models
https://www.psdevwiki.com/ps3/System_Firmware
Thi is where you took the rco file, from a 2.51 kiosk firmware ?... and if this is right, why you choosed that version ?... is because it was the last version before they disabled this animation of the waterdrop ?

Do you know which firmware version is the last with this waterdrop animation ? (i guess it needs to be the version inmediatly before they implemented the new clock)

 

[sandungas]

Definitivelly, the main image tex_busy251.gim is damaged and have lot of pixelation, probably it happened because they converted it to a GIM -ps3dxt5, this image should not be used for a mod, is bad quality, look at this comparison:

I rotated them just to have a better preview of it, the frame most at right (located most at top in the original, not rotated) is frame 1... so in this preview the animation plays "from right to left"
This images doesnt includes the shadows, im ignoring the shadows for this test
Are the original pixels, not escaled, not resized, but the ones from 2.17 have more zoom because are smaller
From top to bottom:
tex_busy217.png <---- original
tex_busy217.png <---- replaced black pixels by alpha (in gimp), then added a 100% black layer at bottom (in photoshop)
tex_busy251.png <---- added a 100% black layer at bottom (in photoshop)
tex_busy251.png <---- added a 100% black layer at bottom, merged layers, replaced black pixels by alpha (in gimp), then added a 100% black layer at bottom (in photoshop)

When i do the "replace black pixels by alpha" in gimp im keeping only white pixels in a layer specific for white pixel tones, then i add a black layer at bottom, so i have very good control of the white and black pixels separated each in a layer
The result is the same, so what im doing is working, is mostly because the images was made in a similar way

The geometry is the original, it can be seen the tiny circles looks like dots, and the bigger ones are ovals, is because the camera is located in a perspective view

It can be seen how the frames of the animation has been inverted for 2.51
The frame most at right in 2.17 (first frame) has been moved to the most left in 2.51 (last frame) and so on...

It can be seen how much damage caused the dxt5 format in tex_busy251.gim (the 2 images most at bottom), if you compare them with the images most at top from tex_busy217.gim
And is awesome because tex_busy217.gim is 29x750 and tex_busy251.gim is 32x960... so initially we could think the biggest one it was going to look better, but thats not the case... the dxt5 format damaged it completly

And now im completly convinced the problem i mentioned... when they forgot to invert the shadows in 2.51 it was a mistake, and it looks bad for sure, is not an smart trick to displace the shadows as i thought
Because the shadow image is used like a "alpha mask" and tells wich pixels are visibles, and doesnt makes sense because is "inverted" instead of "displaced" as i mentioned before

As example... in 2.51 frame 3 is using the shadow that was originally intended to be used by frame 28 (simetrical positions)
1 uses the shadow of 30
5 uses the shadow of 26
...and so on
This is why it looks that have a lot of black pixels when is displayed in the PS3, because doesnt matches well

Long story short... this images from 2.51 should not be used for a mod, the main image with the most white pixels is very pixelated, and for the shadow they forgot to invert it
I bet they updated them in the next firmwares (2.52 2.53 or whatever) because this looks too much broken, is a regression

The idea of inverting the frames was nice to make the animation to increase the size of the circles, but the way how they did it for 2.51 was bad, gg though

 

[Danxx444]

Have you found some trick to trigger that animation for long in XMB ? (this way we can see it better sticking the nose in the TV)

The trick is simple ... but I think not so simple if your internet is absurdly fast ... however..for you to make it spin for a long time..just access heavy sites in the PS3 browser, as his browser It is not so good to access certain websites. To take the test. Try to get into wallpaper sites, these are pretty heavy.

 

[DeViL303]

There is another trick, if you load some offlline JS from an xml, it will load the page fine, but the clock will spin forever.

BTW are we meant to be changing the clock activity animation or the activity animation that shows when tracks are playing or files are downloading, or both?