PINE64
Camera support progress - Printable Version

+- PINE64 (https://forum.pine64.org)
+-- Forum: PinePhone (https://forum.pine64.org/forumdisplay.php?fid=120)
+--- Forum: PinePhone Software (https://forum.pine64.org/forumdisplay.php?fid=121)
+---- Forum: Debian on PinePhone (https://forum.pine64.org/forumdisplay.php?fid=139)
+---- Thread: Camera support progress (/showthread.php?tid=10578)

Pages: 1 2 3 4 5 6 7 8


RE: Camera support progress - aancel - 07-14-2020

Thanks a lot for the answer, that is at least reassuring to know that it might not be a hardware issue !

I once had some some hope for it to work as it worked once today, but I got an image that was slowly degrading to be green only ...
Since then, I couldn't make it work again. I'll try to do some debug if I get some time !


RE: Camera support progress - a-wai - 07-14-2020

Starting pinhole from a terminal with `G_MESSAGES_DEBUG=all pinhole` might give you some clue, fwiw


RE: Camera support progress - aancel - 07-15-2020

I've tested the command you told me to launch and here is the output:

Code:
mobian@mobian:~$ G_MESSAGES_DEBUG=all pinhole
(pinhole:7081): GLib-GIO-DEBUG: 18:24:19.375: _g_io_module_get_default: Found default implementation dconf (DConfSettingsBackend) for ?gsettings-backend?
(pinhole:7081): dconf-DEBUG: 18:24:19.376: watch_fast: "/org/gnome/desktop/interface/" (establishing: 0, active: 0)
(pinhole:7081): dconf-DEBUG: 18:24:19.377: watch_fast: "/org/gnome/settings-daemon/peripherals/mouse/" (establishing: 0, active: 0)
(pinhole:7081): dconf-DEBUG: 18:24:19.377: watch_fast: "/org/gnome/desktop/sound/" (establishing: 0, active: 0)
(pinhole:7081): dconf-DEBUG: 18:24:19.378: watch_fast: "/org/gnome/desktop/privacy/" (establishing: 0, active: 0)
(pinhole:7081): dconf-DEBUG: 18:24:19.379: watch_fast: "/org/gnome/desktop/wm/preferences/" (establishing: 0, active: 0)
(pinhole:7081): dconf-DEBUG: 18:24:19.379: watch_fast: "/org/gnome/settings-daemon/plugins/xsettings/" (establishing: 0, active: 0)
(pinhole:7081): dconf-DEBUG: 18:24:19.380: watch_fast: "/org/gnome/desktop/a11y/" (establishing: 0, active: 0)
(pinhole:7081): dconf-DEBUG: 18:24:19.385: watch_established: "/org/gnome/desktop/interface/" (establishing: 1)
(pinhole:7081): dconf-DEBUG: 18:24:19.386: watch_established: "/org/gnome/settings-daemon/peripherals/mouse/" (establishing: 1)
(pinhole:7081): dconf-DEBUG: 18:24:19.388: watch_established: "/org/gnome/desktop/sound/" (establishing: 1)
(pinhole:7081): dconf-DEBUG: 18:24:19.388: watch_established: "/org/gnome/desktop/privacy/" (establishing: 1)
(pinhole:7081): dconf-DEBUG: 18:24:19.389: watch_established: "/org/gnome/desktop/wm/preferences/" (establishing: 1)
(pinhole:7081): dconf-DEBUG: 18:24:19.389: watch_established: "/org/gnome/settings-daemon/plugins/xsettings/" (establishing: 1)
(pinhole:7081): dconf-DEBUG: 18:24:19.389: watch_established: "/org/gnome/desktop/a11y/" (establishing: 1)
(pinhole:7081): Gdk-DEBUG: 18:24:19.515: Compositor prefers decoration mode 'client'
(pinhole:7081): GLib-GIO-DEBUG: 18:24:19.541: _g_io_module_get_default: Found default implementation gvfs (GDaemonVfs) for ?gio-vfs?
(pinhole:7081): GLib-DEBUG: 18:24:19.708: unsetenv() is not thread-safe and should not be used after threads are created
(pinhole:7081): Gtk-DEBUG: 18:24:19.708: Connecting to session manager
(pinhole:7081): dconf-DEBUG: 18:24:21.519: watch_fast: "/net/flyingpimonster/Camera/" (establishing: 0, active: 0)
(pinhole:7081): dconf-DEBUG: 18:24:21.521: watch_established: "/net/flyingpimonster/Camera/" (establishing: 1)
(pinhole:7081): dconf-DEBUG: 18:24:21.527: change_fast
(pinhole:7081): dconf-DEBUG: 18:24:21.527: change_notify: /net/flyingpimonster/Camera/capture-mode
(pinhole:7081): dconf-DEBUG: 18:24:21.528: change_fast
(pinhole:7081): dconf-DEBUG: 18:24:21.529: change_notify: /net/flyingpimonster/Camera/capture-mode



I'm not used to Glib logging, but I hope this can help in identifying the issue !

For reference here is the log I get when switching to the front camera (which works):
Code:
mobian@mobian:~$ /usr/bin/pinephone-camera-setup.sh front
mobian@mobian:~$ G_MESSAGES_DEBUG=all pinhole
(pinhole:7333): GLib-GIO-DEBUG: 18:25:16.478: _g_io_module_get_default: Found default implementation dconf (DConfSettingsBackend) for ?gsettings-backend?
(pinhole:7333): dconf-DEBUG: 18:25:16.478: watch_fast: "/org/gnome/desktop/interface/" (establishing: 0, active: 0)
(pinhole:7333): dconf-DEBUG: 18:25:16.480: watch_fast: "/org/gnome/settings-daemon/peripherals/mouse/" (establishing: 0, active: 0)
(pinhole:7333): dconf-DEBUG: 18:25:16.480: watch_fast: "/org/gnome/desktop/sound/" (establishing: 0, active: 0)
(pinhole:7333): dconf-DEBUG: 18:25:16.480: watch_fast: "/org/gnome/desktop/privacy/" (establishing: 0, active: 0)
(pinhole:7333): dconf-DEBUG: 18:25:16.481: watch_fast: "/org/gnome/desktop/wm/preferences/" (establishing: 0, active: 0)
(pinhole:7333): dconf-DEBUG: 18:25:16.481: watch_fast: "/org/gnome/settings-daemon/plugins/xsettings/" (establishing: 0, active: 0)
(pinhole:7333): dconf-DEBUG: 18:25:16.482: watch_fast: "/org/gnome/desktop/a11y/" (establishing: 0, active: 0)
(pinhole:7333): dconf-DEBUG: 18:25:16.491: watch_established: "/org/gnome/desktop/interface/" (establishing: 1)
(pinhole:7333): dconf-DEBUG: 18:25:16.502: watch_established: "/org/gnome/settings-daemon/peripherals/mouse/" (establishing: 1)
(pinhole:7333): dconf-DEBUG: 18:25:16.502: watch_established: "/org/gnome/desktop/sound/" (establishing: 1)
(pinhole:7333): dconf-DEBUG: 18:25:16.502: watch_established: "/org/gnome/desktop/privacy/" (establishing: 1)
(pinhole:7333): dconf-DEBUG: 18:25:16.506: watch_established: "/org/gnome/desktop/wm/preferences/" (establishing: 1)
(pinhole:7333): dconf-DEBUG: 18:25:16.506: watch_established: "/org/gnome/settings-daemon/plugins/xsettings/" (establishing: 1)
(pinhole:7333): dconf-DEBUG: 18:25:16.506: watch_established: "/org/gnome/desktop/a11y/" (establishing: 1)
(pinhole:7333): Gdk-DEBUG: 18:25:16.632: Compositor prefers decoration mode 'client'
(pinhole:7333): GLib-GIO-DEBUG: 18:25:16.669: _g_io_module_get_default: Found default implementation gvfs (GDaemonVfs) for ?gio-vfs?
(pinhole:7333): GLib-DEBUG: 18:25:16.863: unsetenv() is not thread-safe and should not be used after threads are created
(pinhole:7333): Gtk-DEBUG: 18:25:16.864: Connecting to session manager
(pinhole:7333): dconf-DEBUG: 18:25:17.856: watch_fast: "/net/flyingpimonster/Camera/" (establishing: 0, active: 0)
(pinhole:7333): dconf-DEBUG: 18:25:17.859: watch_established: "/net/flyingpimonster/Camera/" (establishing: 1)
(pinhole:7333): dconf-DEBUG: 18:25:17.864: change_fast
(pinhole:7333): dconf-DEBUG: 18:25:17.865: change_notify: /net/flyingpimonster/Camera/capture-mode
(pinhole:7333): dconf-DEBUG: 18:25:17.866: change_fast
(pinhole:7333): dconf-DEBUG: 18:25:17.866: change_notify: /net/flyingpimonster/Camera/capture-mode



RE: Camera support progress - a-wai - 07-17-2020

Hmmm, nothing there could explain why it doesn't work for you, that's disappointing...
I'm afraid I can't help you more there atm, sorry


RE: Camera support progress - aancel - 07-17-2020

Thanks for all the help !

I managed to make some progress with additional tests. Initially I wanted to unscrew the inside plastic cover to ensure that the camera module was correctly plugged to the board, but I ended up having trouble with one of the screw (the closest to the camera) and being stuck on a hidden screw ... I could however barely see that the module is apparently correctly plugged
In the end, I rebooted to mobian with the back cover of the phone not properly closed (one screw closest to the camera module removed), and to my surprise the camera was working !

After more testing I think that there might be a connection issue with the camera module.
Because of the missing screw, I have a small gap between the inside plastic cover and the part with the screen:
- If I leave the device as is with the small gap, I either have an image with garbage green and magenta pixels or a black image
- If I apply some pressure for the inside plastic cover to be in contact with the part with the screen, I get a clear image of the back camera

But what seems strange is that sometimes the reverse behavior happens:
- With the gap, I get a clear image
- With the contact, I get either no image of an image with garbage pixels

Does anyone has an idea from where this kind of issue from ?
Should I request assistance directly from Pine64 ? (As this does not seem to be a software issue in the end)


RE: Camera support progress - puffchumpy - 07-18-2020

I'm having trouble with the camera as well. Using a just flashed mobian-pinephone-20200716 image on an SD card. Pinhole runs but shows the default test image. Attempting to take a picture the message "Could not save Image! Unknown error: The viewfinder is not in the Ready state." is returned. This thread has the first check as the dmesg log, it show a "failed with -6". Details pasted below. Is there a easy fix?

I might try examining the hardware like anncel. My BraveHeart phone also has issues with the proximity sensor, see Proximity sensor misbehaving.

Code:
mobian@mobian:~$ dmesg | grep -E "(csi|ov5640)"
[    1.429506] dovdd-csi: Bringing 3300000uV into 1800000-1800000uV
[    1.432047] avdd-csi: Bringing 2900000uV into 2800000-2800000uV
[    1.433955] dvdd-csi: Bringing 700000uV into 1800000-1800000uV
[    4.253957] i2c-gpio i2c-csi: using lines 141 (SDA) and 140 (SCL)
[    9.778135] sun6i-csi 1cb0000.csi: creating ov5640 3-004c:0 -> sun6i-csi:0 link
[    9.793349] sun6i-csi 1cb0000.csi: creating gc2145 3-003c:0 -> sun6i-csi:0 link
[  10.348554] ov5640 3-004c: ov5640_init_slave_id: failed with -6
[  11.578215] ov5640 3-004c: ov5640_init_slave_id: failed with -6
[  219.185022] ov5640 3-004c: ov5640_init_slave_id: failed with -6
[  220.259847] ov5640 3-004c: ov5640_init_slave_id: failed with -6
[  536.724229] ov5640 3-004c: ov5640_init_slave_id: failed with -6
[  721.236655] ov5640 3-004c: ov5640_init_slave_id: failed with -6
[  920.136994] ov5640 3-004c: ov5640_init_slave_id: failed with -6
[ 1047.866158] ov5640 3-004c: ov5640_init_slave_id: failed with -6



RE: Camera support progress - ivanB1975 - 07-18-2020

(07-17-2020, 05:56 PM)aancel Wrote: Thanks for all the help !

I managed to make some progress with additional tests. Initially I wanted to unscrew the inside plastic cover to ensure that the camera module was correctly plugged to the board, but I ended up having trouble with one of the screw (the closest to the camera) and being stuck on a hidden screw ... I could however barely see that the module is apparently correctly plugged
In the end, I rebooted to mobian with the back cover of the phone not properly closed (one screw closest to the camera module removed), and to my surprise the camera was working !

After more testing I think that there might be a connection issue with the camera module.
Because of the missing screw, I have a small gap between the inside plastic cover and the part with the screen:
- If I leave the device as is with the small gap, I either have an image with garbage green and magenta pixels or a black image
- If I apply some pressure for the inside plastic cover to be in contact with the part with the screen, I get a clear image of the back camera

But what seems strange is that sometimes the reverse behavior happens:
- With the gap, I get a clear image
- With the contact, I get either no image of an image with garbage pixels

Does anyone has an idea from where this kind of issue from ?
Should I request assistance directly from Pine64 ? (As this does not seem to be a software issue in the end)

I have exactly the same issue and I have a BH version. It looks like an hardware problem. If I try to start the app I hear just the clicks and everything remain black. If I press with two fingers the part of the phone where the camera is located the black image become green or shows green lines. I don't want to unmount the phone camera since it was supposed to be hardware ok and identical to the final version, but it is not and quite buggy. At this point I will write to Pine64 as well and ask about this problem.

NOTE: after pressing hard with two fingers on the camera (without umounting the phone) the camera started to work. I am pretty disappointed about this. Not a good impression about the quality of this component.

(07-17-2020, 05:56 PM)aancel Wrote: Thanks for all the help !

I managed to make some progress with additional tests. Initially I wanted to unscrew the inside plastic cover to ensure that the camera module was correctly plugged to the board, but I ended up having trouble with one of the screw (the closest to the camera) and being stuck on a hidden screw ... I could however barely see that the module is apparently correctly plugged
In the end, I rebooted to mobian with the back cover of the phone not properly closed (one screw closest to the camera module removed), and to my surprise the camera was working !

After more testing I think that there might be a connection issue with the camera module.
Because of the missing screw, I have a small gap between the inside plastic cover and the part with the screen:
- If I leave the device as is with the small gap, I either have an image with garbage green and magenta pixels or a black image
- If I apply some pressure for the inside plastic cover to be in contact with the part with the screen, I get a clear image of the back camera

But what seems strange is that sometimes the reverse behavior happens:
- With the gap, I get a clear image
- With the contact, I get either no image of an image with garbage pixels

Does anyone has an idea from where this kind of issue from ?
Should I request assistance directly from Pine64 ? (As this does not seem to be a software issue in the end)

Could you post a picture of the module?


RE: Camera support progress - aancel - 07-22-2020

(07-18-2020, 03:50 PM)ivanB1975 Wrote: I have exactly the same issue and I have a BH version. It looks like an hardware problem. If I try to start the app I hear just the clicks and everything remain black. If I press with two fingers the part of the phone where the camera is located the black image become green or shows green lines. I don't want to unmount the phone camera since it was supposed to be hardware ok and identical to the final version, but it is not and quite buggy. At this point I will write to Pine64 as well and ask about this problem.

NOTE: after pressing hard with two fingers on the camera (without umounting the phone) the camera started to work. I am pretty disappointed about this. Not a good impression about the quality of this component.


Could you post a picture of the module?

Ok, that is good to know that I am not the only one encountering this kind of issue, and it seems to be a hardware issue indeed.
Regarding your request for a picture, I cannot unfortunately get the camera to work again ... I've tried again pushing a bit around the module but without success.

I had the exact problem that you have, my guess is that the green lines appear when the camera is starting to loose the connection with the system board and garbage pixels are output to the viewfinder.
I will try to contact pine64 to see get their opinion on this issue and if I can do something about it.


RE: Camera support progress - chris_02 - 07-22-2020

@a-wai today pinhole is down with error:

the viewfinder's not working.


RE: Camera support progress - arturo2bodegas - 07-22-2020

just tried on my CE ubports phone, and it crashed first try, then worked second...took the shot and saved it to photos...do you have braveheart, or CE edition phone? seems like many people are struggling with braveheart phones...maybe hardware?