| Author |
Super camera driver for k750/w800/w810 |
lpachuong
Joined: Jun 06, 2007
Posts: > 500
From: Vietnam
PM |
A4A - 11.00 Final
http://www.mediafire.com/?4wdxvugynmk
There are two version, 1 for w800i/k750i and 1 for w810i with different compression.
Yellowish viewfinder is fixed
Fixed macro
No more test on compression => stable I gave up with this now.
Functions changed for w810i users:
EV_M_20= CYAN;
EV_M_17= RED;
EV_M_13= BLUE;
EV_M_10= YELLOW;
EV_M_07= ZOOM;
EV_M_03= ISO;
EV_0_00= FOCUS;
EV_P_03= EXPOSURE;
EV_P_07= SPOT;
EV_P_10= SHUTTER_SPEED;
EV_P_13= SATURATION;
EV_P_17= SCENE;
EV_P_20= SET_PROFILE; Raiderski's profile
NEW SHARPNESS + NOISE REDUCTION (Haven't tested on w810i at high compression)
Thanks to terminator, you have waken me up 
I will stop for few days to carry on with my new idea.
Looking forward to your comments.
Enjoy.
[addsig] |
|
|
tommymar
Joined: Aug 10, 2007
Posts: 76
PM |
On 2007-12-18 18:08:53, metallica1 wrote:
On 2007-12-18 15:34:31, phoenix1985 wrote:
@metallica1:
So...your w800 never got crash with strong 5.6?
can you upload strong 5.6 camdriver that you used ?
Cheers...!
phoenix1985
My phone is a W810i White, and the phone never crash with Driver Strong 5.6, just in function of supermod "Profile" is selected, but I never use this function.
The driver 5.6 from my phone that I had already put here at the forum:
http://rapidshare.com/files/7[....]5.6_from_W810i_21-11-2007_.rar
this driver have diferent function of 5.6 original, instruction together the link.
good advantage!
Your phone is w810i and that is why it is stable with any high compression.
Besides, Strong's driver has no compression, which means its compression is at default value (75%). (Search for Fine-Quality in prop_5_1 and you will see)
---------------------------
@lpachuong: Sound good. I will give it a try. Thank you. |
lpachuong
Joined: Jun 06, 2007
Posts: > 500
From: Vietnam
PM |
@Raiderski.
If it is possible, we can force the events to work first, then use if/loop to regenerate the properties. That will be awesome.
Besides, as far as I know, we can force the phone to work on i2c address (in our case, it is 1F) to halt a step of a command. Hope this can help.
[addsig] |
terminator17
Joined: Jul 04, 2007
Posts: 155
PM, WWW
|
On 2007-12-19 17:16:32, lpachuong wrote:
A4A - 11.00 Final
http://www.mediafire.com/?4wdxvugynmk
There are two version, 1 for w800i/k750i and 1 for w810i with different compression.
Yellowish viewfinder is fixed
Fixed macro
No more test on compression => stable I gave up with this now.
Functions changed for w810i users:
EV_M_20= CYAN;
EV_M_17= RED;
EV_M_13= BLUE;
EV_M_10= YELLOW;
EV_M_07= ZOOM;
EV_M_03= ISO;
EV_0_00= FOCUS;
EV_P_03= EXPOSURE;
EV_P_07= SPOT;
EV_P_10= SHUTTER_SPEED;
EV_P_13= SATURATION;
EV_P_17= SCENE;
EV_P_20= SET_PROFILE; Raiderski's profile
NEW SHARPNESS + NOISE REDUCTION (Haven't tested on w810i at high compression)
Thanks to terminator, you have waken me up
I will stop for few days to carry on with my new idea.
Looking forward to your comments.
Enjoy.
thanx a lot mate...just superb driver...u simply rock!!
thanx again...ur driver is just perfect for w810....
|
SPU
Joined: Feb 07, 2007
Posts: 131
From: Russia, SPb
PM |
Raiderski
Events seems more tricky than we think about them...
Actually we don't need to "wait" the end of execution of previous command, but we need to remove generated event from the events queue to unlock further execution.
Very pity that we don't have some property called when frame changes...
Btw, you can check logging possibility in the next way:
insert somewhere code like this: $"text string %0" 06 02 00 01 04 61 #CMD_END;
if "log" command can't be executed, then interpreter will skip everything till the end of code part.
But... even if logging works... may be logs are available only in special "debugging" version of firmware or we need to modify some internal settings =/ Note, that log size is not limited (even if contains only one "session"), but logging lines are not removed from drivers.
W800i CID36 R1BD001 |
chengrong
Joined: Sep 01, 2006
Posts: 81
PM |
On 2007-12-19 18:24:28, SPU wrote:
Raiderski
Events seems more tricky than we think about them...
Actually we don't need to "wait" the end of execution of previous command, but we need to remove generated event from the events queue to unlock further execution.
Very pity that we don't have some property called when frame changes...
Btw, you can check logging possibility in the next way:
insert somewhere code like this: $"text string %0" 06 02 00 01 04 61 #CMD_END;
if "log" command can't be executed, then interpreter will skip everything till the end of code part.
But... even if logging works... may be logs are available only in special "debugging" version of firmware or we need to modify some internal settings =/ Note, that log size is not limited (even if contains only one "session"), but logging lines are not removed from drivers.
I think the log system is connected to the computer via USB.
http://chengrongblog.blogspot.com/
|
Raiderski
Joined: Jul 03, 2006
Posts: > 500
From: Poland, Hell, Mountains
PM, WWW
|
1. events
that was easier than i expected, with simple test code i had results after 15 minutes. here they are:
51 = scene change started / change scene
EB = scene changed
small loop waiting for EB event resolved my biggest problem, finally i can release 6.7
2. log
i must fully agree with all comments. code interpterer does not remove line $"test" other code but $"" probably doesn't relate with txt file in filesystem. this doesn't make sense. the best way (as always) is to observe debugging results in realtime
- phone is connected to PC (with active supervising software) in some kind of service mode
-special firmware logging all operations with possibility to read them on demand
3. loop with [] brackets
doesn't work on our phones at all which is obvious when we see macros like WAIT_FOR_ 
|
Raiderski
Joined: Jul 03, 2006
Posts: > 500
From: Poland, Hell, Mountains
PM, WWW
|
MALICIOUS ISSUE!
all drivers are affected! (where zoom keys can control anything else than zoom)
modders and users should pay attention on this!
simplest path to reproduce in supermode drivers:
* = function example
** = value example
- go to EV value corresponding to manual focus control*
- set value to x1.6**
- go to EV value corresponding to contrast*
and from this point you have two ways:
1. use key 7 or camera menu to switch between normal/night mode
2. make photo
result: contrast* value x1.6** will be applied...
1. after normal/night switch
2. after back from still photo view to realtime viewfinder
path to reproduce in drivers with manual focus in macro mode:
- enable macro
- set value x1.6**
- turn off macro
result: zoom x1.6** will be applied
solution: block code execution in seq_set_prop_6 after every normal/night mode transition (key2, night mode in camera menu, $u07)
i checked also other transitions for example macro mode on/off (key 4 or camera menu) and they also generate issue
[ This Message was edited by: Raiderski on 2007-12-21 11:43 ] |
leetiran
Joined: Aug 15, 2007
Posts: 134
From: China
PM |
There is WB bug in it because some cods are from raider6.6.1. replace the cods"#SSP_UNSET" after "seq_set_prop_7_0" and "seq_set_prop_7_1" with "$b6 00 4 1 #CMD(04,00)" seems can solve it. |
Randolph
Joined: May 21, 2007
Posts: 115
PM |
hello again... today I noticed a strange behavior in my pictures: When I open the with exif reader they all show prgCXC_125952_EU_1_CL 6.6
I have used since a long time the patch "move driver to another location"
This ...6.6 stays the same and does not respond what driver I use.
What stands behind this 6.6?
regards,
Randolph |
Raiderski
Joined: Jul 03, 2006
Posts: > 500
From: Poland, Hell, Mountains
PM, WWW
|
nope, this will help only if you don't use manual shutter speed. if you use MSS then ISO100 scene will be activated and you will not retrieve automatic scene - this is very serious bug
problems with WB in 6.6.1 were due to scene changes in MSS which also affected noMSS mode. i spent one week trying to fix this and today i got full success in this matter
[ This Message was edited by: Raiderski on 2007-12-20 14:21 ] |
Raiderski
Joined: Jul 03, 2006
Posts: > 500
From: Poland, Hell, Mountains
PM, WWW
|
Randolph
last post by lpachuong on this page:
http://www.esato.com/board/viewtopic.php?topic=135737&start=6675
|
leetiran
Joined: Aug 15, 2007
Posts: 134
From: China
PM |
On 2007-12-20 14:18:53, Raiderski wrote:
nope, this will help only if you don't use manual shutter speed. if you use MSS then ISO100 scene will be activated and you will not retrieve automatic scene - this is very serious bug
problems with WB in 6.6.1 were due to scene changes in MSS which also affected noMSS mode. i spent one week trying to fix this and today i got full success in this matter
[ This Message was edited by: Raiderski on 2007-12-20 14:21 ]
Yes, you are absolutely right. It helps only on the other funtions except manual shutter speed. long for the raider6.7.
[ This Message was edited by: leetiran on 2007-12-20 15:15 ] |
Randolph
Joined: May 21, 2007
Posts: 115
PM |
On 2007-12-18 17:55:25, Raiderski wrote:
guys, if you want to fight with noise then enable pixel interpolation (filtering, reg5B to 00) and use low sharpness value (reg56 to 00 - 04) but this is something for something, in this case sharpness vs. noise
Hello Raiderski, thanks for the hint to the comment on exif reader...
In you configtool I can en/disable filtering: Is this what you ment "enable pixel interpolation"= "bilinear filtering" on?
kind regards,
Randolph |
Raiderski
Joined: Jul 03, 2006
Posts: > 500
From: Poland, Hell, Mountains
PM, WWW
|
Randolph
yes, exactly this what you wrote
|
|
|
Posted: 2007-12-19 17:16
