How to change the color of the download progress bar ?

[jeka26]

We need to know 2 values to do a math calculation with them
The color of the image (from one of the 5 i uploaded)
And the color in the screenshot

The formula is
color in the screenshot - original color = this is what the sprx does





--------------------------------
If the original color is black, then in the formula we are substracting zero so...
color in the screenshot = this is what the sprx does

But this needs to be tested 5 times (with all the images i made) because in the tests made by @LuanTeles it did look the sprx was doing something with blue channel too
Also, because maybe the value of the channel "loops" when the sprx is increasing it... lets say... imagine we have the value 0xFE and the sprx adds a + 0x2... this could result in 0x00 incase is looping

 

[Louay]

@sandungas Color in screenshot by @jeka26 using remote play and tex_default_progress_slider is 0000ff and the one appread in the screenshot is 07083b ,i do a screenshot with my capture card now its totaly diffrent

• Original image 8f8f8f gives this result :

so result is 78950e-8f8f8f = -16FA81 (what sprx does)

• Black image 000000 gives this result :

so result will be : 000002 - 000000 = 2 (what sprx does)

• Blue image 0000FF gives this result :

so result will be : 000047 - 0000FF = -B8 (what sprx does)

• Green image 00FF00 gives this result :

so result will be : 00FF00 - 00FF00 = 0 (sprx does nothing)

• Red image FF0000 gives this result :

so result will be : EA0000 - FF0000 = -150000 (what sprx does)

• White image FFFFFF gives this result :

so result will be : EA0000 - FFFFFF = -1C00D0 (what sprx does)

Strange why there is big diffrence of the sprx work for colors

 

[LuanTeles]

@sandungas Color in screenshot by @jeka26 using remote play and tex_default_progress_slider is 0000ff and the one appread in the screenshot is 07083b ,i do a screenshot with my capture card now its totaly diffrent

• Original image 8f8f8f gives this result :

• Black image 000000 gives this result :

• Blue image 0000FF gives this result :

• Green image 00FF00 gives this result :

• Red image FF0000 gives this result :

• White image FFFFFF gives this result :

Strange why there is big diffrence of the sprx work for colors

WOWWWW, seems like you got better results than me, dunno why i always have different kind of red color,

are you trying to archieve White as well?

 

[sandungas]

Great @Louay thats what i meant
But the results are a bit weird, you used the images i uploaded in the other thread or you made your own ?
Im asking because the images i made are fully opaque at the center (to prevent the color to merge with the other colors at his back)

Edit:
Also, there is something that doesnt makes sense
When @LuanTeles made his first test i bet he used a white image and he got a red color as result
But when you use a white image you have a yellow (and i know you use some XMB mods to have your XMB in gold color tones)

Edit:
Hmmm... so it seems the color of that image is affected "on real time" by the monthly color filters

 

[sandungas]

I think i got how is working, you made some mistakes in the calculations @Louay btw

The first time i was aware this image of the progress bar was "modifyed" in real time by a sprx i thought the sprx was increasing the values of the color channels, i imagined this tests making screenshots to try to find what he sprx does, and i thought the test with an image using full opaque black was going to be the most successful (because you know... if the black is 0x000000 and you add something to it we are going to see the exact value, not much maths needed)

But is the other way around, the sprx is doing a substraction... now while thinking in it it has lot of sense because this is intended to work with images that are originally white/bright... or have a degrade of whites
If you do this "recoloration" with code in a normal image... lets say a photo from a landscape the result is weird
So... this sprx doing color adjustments "on the fly" is intended to work with a base image in white

And this is the reason why the test with 0xFFFFFF is the best

The new formula is... original color - what the sprx substracts = color in the screenshot
Or reordered... original color - color in the screenshot = what the sprx substracts

In other words... for this test you did it seems the sprx substracted 0x1c to the red channel and 0xd0 to the blue channel
0xffffff - 0xe3ff2f = 0x1c00d0


-----------------------------------------------
And this test with an original black image made with 0x000000 it can be seen the sprx doesnt makes a "loop" when it reaches the minimal values
Probably is trying to do the same substraction than the other tests (0x1c00d0 for all ?) but it cant go lower than 0

*The 02 that appears in photoshop in the "color picker" is just garbage added either by the software algorithms, or by the capture card, or the screenshot software or whatever, it seems this trick using screenshots have an error rate or 02 or so, not bad


-------------------------------------------
And btw, as mentioned before im not really sure if the amount substracted by the sprx changes "by month of the year" or "by hour of day", so it seems the results of this tests could vary a lot if you do them in a non-controlled enviroment
But im not sure, i think both you @LuanTeles and @Louay are using a custom lines.qrc
Is something like that what is causing you different results

 

[Louay]

I think i got how is working, you made some mistakes in the calculations @Louay btw

The first time i was aware this image of the progress bar was "modifyed" in real time by a sprx i thought the sprx was increasing the values of the color channels, i imagined this tests making screenshots to try to find what he sprx does, and i thought the test with an image using full opaque black was going to be the most successful (because you know... if the black is 0x000000 and you add something to it we are going to see the exact value, not much maths needed)

But is the other way around, the sprx is doing a substraction... now while thinking in it it has lot of sense because this is intended to work with images that are originally white/bright... or have a degrade of whites
If you do this "recoloration" with code in a normal image... lets say a photo from a landscape the result is weird
So... this sprx doing color adjustments "on the fly" is intended to work with a base image in white

And this is the reason why the test with 0xFFFFFF is the best

The new formula is... original color - what the sprx substracts = color in the screenshot
Or reordered... original color - color in the screenshot = what the sprx substracts

In other words... for this test you did it seems the sprx substracted 0x1c to the red channel and 0xd0 to the blue channel
0xffffff - 0xe3ff2f = 0x1c00d0


-----------------------------------------------
And this test with an original black image made with 0x000000 it can be seen the sprx doesnt makes a "loop" when it reaches the minimal values
Probably is trying to do the same substraction than the other tests (0x1c00d0 for all ?) but it cant go lower than 0

*The 02 that appears in photoshop in the "color picker" is just garbage added either by the software algorithms, or by the capture card, or the screenshot software or whatever, it seems this trick using screenshots have an error rate or 02 or so, not bad


-------------------------------------------
And btw, as mentioned before im not really sure if the amount substracted by the sprx changes "by month of the year" or "by hour of day", so it seems the results of this tests could vary a lot if you do them in a non-controlled enviroment
But im not sure, i think both you @LuanTeles and @Louay are using a custom lines.qrc
Is something like that what is causing you different results

Yes i have custom lines.qrc hhhh thats why i have black background i will recopy original lines.qrc and then retest

 

[Louay]

I think i got how is working, you made some mistakes in the calculations @Louay btw

The first time i was aware this image of the progress bar was "modifyed" in real time by a sprx i thought the sprx was increasing the values of the color channels, i imagined this tests making screenshots to try to find what he sprx does, and i thought the test with an image using full opaque black was going to be the most successful (because you know... if the black is 0x000000 and you add something to it we are going to see the exact value, not much maths needed)

But is the other way around, the sprx is doing a substraction... now while thinking in it it has lot of sense because this is intended to work with images that are originally white/bright... or have a degrade of whites
If you do this "recoloration" with code in a normal image... lets say a photo from a landscape the result is weird
So... this sprx doing color adjustments "on the fly" is intended to work with a base image in white

And this is the reason why the test with 0xFFFFFF is the best

The new formula is... original color - what the sprx substracts = color in the screenshot
Or reordered... original color - color in the screenshot = what the sprx substracts

In other words... for this test you did it seems the sprx substracted 0x1c to the red channel and 0xd0 to the blue channel
0xffffff - 0xe3ff2f = 0x1c00d0


-----------------------------------------------
And this test with an original black image made with 0x000000 it can be seen the sprx doesnt makes a "loop" when it reaches the minimal values
Probably is trying to do the same substraction than the other tests (0x1c00d0 for all ?) but it cant go lower than 0

*The 02 that appears in photoshop in the "color picker" is just garbage added either by the software algorithms, or by the capture card, or the screenshot software or whatever, it seems this trick using screenshots have an error rate or 02 or so, not bad


-------------------------------------------
And btw, as mentioned before im not really sure if the amount substracted by the sprx changes "by month of the year" or "by hour of day", so it seems the results of this tests could vary a lot if you do them in a non-controlled enviroment
But im not sure, i think both you @LuanTeles and @Louay are using a custom lines.qrc
Is something like that what is causing you different results

@sandungas
i restore lines.qrc from ofw and result is the same :/ and when i make transparent image no progress bar i there percentage run to 100% without bar héhéhé

 

[sandungas]

@sandungas
i restore lines.qrc from ofw and result is the same :/ and when i make transparent image no progress bar i there percentage run to 100% without bar héhéhé

Ok, so doesnt seems to be lines.qrc but you have something diferent than @LuanTeles

When you use a white image you have yellow
When he uses a white image he have red
You should talk with each other to figure why you are having different results

WOWWWW, seems like you got better results than me, dunno why i always have different kind of red color,

are you trying to archieve White as well?

 

[Louay]

Ok, so doesnt seems to be lines.qrc but you have something diferent than @LuanTeles

When you use a white image you have yellow
When he uses a white image he have red
You should talk with each other to figure why you are having different results

Nope @LuanTeles recolor image to pink thats why he get reds as i get red

 

[sandungas]

Nope @LuanTeles recolor image to pink thats why he get reds as i get red

View attachment 20550

Ok, it seems you are right, he was using pink colors in the first tests
https://www.psx-place.com/threads/ps3-pro-mod-release-information-thread.14641/page-35#post-205991
https://www.psx-place.com/threads/ps3-pro-mod-release-information-thread.14641/page-35#post-206095

 

[Louay]

Ok, it seems you are right, he was using pink colors in the first tests
https://www.psx-place.com/threads/ps3-pro-mod-release-information-thread.14641/page-35#post-205991
https://www.psx-place.com/threads/ps3-pro-mod-release-information-thread.14641/page-35#post-206095

So how can we achieve white color ? i get black color can we reverse that so from 000000 to FFFFFF

 

[sandungas]

In the test you made with 0xffffff the result was the sprx is substracting this amount: 0x1c00d0

In other words... it substracts:
-a bit from the red channel
-nothing from the green channel
-a lot from the blue channel

Is imposible to achieve a white because the sprx always is going to substract something :/

 

[Louay]

In the test you made with 0xffffff the result was the sprx is substracting this amount: 0x1c00d0

In other words... it substracts:
-a bit from the red channel
-nothing from the green channel
-a lot from the blue channel

Is imposible to achieve a white because the sprx always is going to substract something :/

File being used when i click on a pkg to install are :
game_ext_plugin sprx/rco
nas_plugin sprx/rco
game_plugin rco

 

[sandungas]

See this example (for guinea pigging), im going to try to do the calcualtions to ballance all channels and considering he sprx is going to substract this: 0x1c00d0

The blue channel is the most problematic, lets say... if we start with 0xff in the blue channel the result is going to be... 0x2f
So by now our guinea pig color is 0x????ff

Green channel is not modifyed by the sprx, i can use 0x2f for it and the sprx will respect it... so our guinea pig color is 0x??2fff

Now for the red channel... i want to achieve another 0x2f but the sprx substracts 0x1c (so 0x2f + 0x1c = 0x4b)

So our final guinea pig color is 0x4b2fff




---------------------
Try to make a custom image using color 0x4b2fff then take an screenshot of it, open it in photoshop and use the "color picker"
The result shoud be 0x2f2f2f (a very dark gray tone)

Im just saying this as an example about how to control the colors, probably is not much precise, it seems the sprx uses some special algorithm for recoloring it that varies the amount substracted
By looking at your results it looks most of them matches almost perfectly with that i was explaining before.... but some doesnt matches so well

 

[LuanTeles]

One thing i realized recoloring the slider in the system_plugin.rco

also affects video/music category even they having the same image in their rco

 

[sandungas]

I guess that progress bar appears in different places with different colors, and this is why that "recoloring" is made with a sprx

 

[Louay]

See this example (for guinea pigging), im going to try to do the calcualtions to ballance all channels and considering he sprx is going to substract this: 0x1c00d0

The blue channel is the most problematic, lets say... if we start with 0xff in the blue channel the result is going to be... 0x2f
So by now our guinea pig color is 0x????ff

Green channel is not modifyed by the sprx, i can use 0x2f for it and the sprx will respect it... so our guinea pig color is 0x??2fff

Now for the red channel... i want to achieve another 0x2f but the sprx substracts 0x1c (so 0x2f + 0x1c = 0x4b)

So our final guinea pig color is 0x4b2fff




---------------------
Try to make a custom image using color 0x4b2fff then take an screenshot of it, open it in photoshop and use the "color picker"
The result shoud be 0x2f2f2f (a very dark gray tone)

Im just saying this as an example about how to control the colors, probably is not much precise, it seems the sprx uses some special algorithm for recoloring it that varies the amount substracted
By looking at your results it looks most of them matches almost perfectly with that i was explaining before.... but some doesnt matches so well

Result is diffrent from 2f2f2f its 372539

 

[sandungas]

Result is diffrent from 2f2f2f its 372539

View attachment 20552

Is really close to what i calculated, it looks like a proof that what i said is right
The small imprecissions could be caused by different reasons... but are very small

See the difference in between 2f2f2f and 372539
In between 2f and 37 (for red) there is only 8 units
In between 2f and 25 (for green) there is only 10 units
In between 2f and 39 (for blue) there is only 10 units

 

[Louay]

Is really close to what i calculated, it looks like a proof that what i said is right
The small imprecissions could be caused by different reasons... but are very small

See the difference in between 2f2f2f and 372539
In between 2f and 37 (for red) there is only 8 units
In between 2f and 25 (for green) there is only 10 units
In between 2f and 39 (for blue) there is only 10 units

i change the display option

Display Settings before :

Display Settings After :

Bar Screenshot :

 

[sandungas]

Hmmm, interesting, yep that video settings seems to be one of the things that could affect the calculations we are trying to do for this tests
Right now i dont know which video settings are the best for this kind of experiments though



Edit:
But btw... note the new value is pretty close to the previous one you had... so we are not much far away
Overall... the calculations i did was not much precise but are not bad

Most probably we are dragging an small error because i did all calculations thinking the sprx is substracting 0x1c00d0 (this was taken from your test with 0xffffff resulting in a yellow)

But dont get crazy with this trying to find the exact values... if you want to do it for fun or for curiosity sake go for it... but i have the feeling is not going to be so easy