[05:02] *** wallacer has quit IRC (Remote host closed the connection)
[05:03] *** wallacer has joined #linux-media
[05:15] *** mvaittin has joined #linux-media
[06:14] *** mvaittin has quit IRC (Ping timeout: 480 seconds)
[06:31] *** shoragan has quit IRC (Read error: Connection reset by peer)
[06:35] *** shoragan has joined #linux-media
[06:52] *** GBenji has joined #linux-media
[06:53] *** ao2 has joined #linux-media
[06:53] <anuw[m]> <KieranBingham[m]> "I think you can ignore teh..." <- yes it is. no more error except s5k5e9.yaml not found. I cant get a real image, horizontal vertical color bars instead.
[06:54] <anuw[m]> <pinchartl> "how much RAM does that device..." <- done now with 64mb, initially default 32mb seemed not enough
[06:54] <KieranBingham[m]> You mean you do get horizontal bars? Grim a test pattern?
[06:55] <KieranBingham[m]> *from a test pattern 
[06:56] <anuw[m]> not sure. if thats the case then is encurraging
[06:56] * anuw[m] uploaded an image: (75KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/LLWzBmULmjXxVdDuFYifzRjj/2024-05-30-14-07-31-167.jpg >
[06:59] <anuw[m]> i will do some test on camss known test pattern first to verify the pipeline 
[07:02] <KieranBingham[m]> I think that looks like 8 bars so it could very well be a test pattern, showing streaming from the sensor but with something misconfigured
[07:05] <KieranBingham[m]> Have you manually enabled the test pattern control yourself? With v4l-ctl at some point perhaps?
[07:05] *** ao2 has quit IRC (Quit: Leaving)
[07:06] *** djrscally has joined #linux-media
[07:12] <anuw[m]> <KieranBingham[m]> "Have you manually enabled the..." <- not manually, i actually do nothing since pmos boot up. I also believe the pattern from sensor. since there are many uncertain through the whole pipeline. and my plan next round will test on camss-csid test generator and this is a well known pattern. if the pipeline works then at least exclude many unsure once something happened later
[07:33] *** tmerciai has joined #linux-media
[07:39] *** mvaittin has joined #linux-media
[08:20] * anuw[m] uploaded an image: (53KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/MdHZHLRIAzcUgOkheXgHpyTH/2024-05-30-16-19-46-338.jpg >
[08:20] <anuw[m]> ah....seemed not so long..
[08:20] <anuw[m]> some glitches but
[08:21] <anuw[m]> i force hack on test pattern addr off on start streaming
[08:21] <anuw[m]> thx guys.
[08:32] *** wallacer has quit IRC (Ping timeout: 480 seconds)
[08:41] *** wallacer has joined #linux-media
[09:53] *** ao2 has joined #linux-media
[11:44] *** lexano has joined #linux-media
[11:46] <jmassot> sailus: Hi I'm waiting for a review on my GMSL v7 patchset do you have some time to have another look at this patchset?
[11:50] <sailus> jmassot: Thanks for the ping. I'll check that again by early next week.
[11:51] <jmassot> Thanks 
[12:25] *** mvaittin has quit IRC (Ping timeout: 480 seconds)
[13:13] *** xroumegue has quit IRC (Ping timeout: 480 seconds)
[13:15] *** BrianG61UK has joined #linux-media
[13:22] *** xroumegue has joined #linux-media
[13:57] *** avolmat has joined #linux-media
[13:57] *** eelstrebor has joined #linux-media
[14:19] <pinchartl> anuw[m]: nice :-)
[14:22] <pinchartl> jmassot: could you CC tomi.valkeinen@ideasonboard.com on the next version ?
[14:36] *** GBenji has quit IRC (Quit: Leaving.)
[14:37] *** GBenji has joined #linux-media
[15:03] *** hansg has joined #linux-media
[15:03] *** hansg has quit IRC ()
[15:58] <anuw[m]> <pinchartl> "anuw: nice :-)" <- pinchartl, i still get weird output on manual frame dump rather than look-normal qcam output. Is there any tricks？
[15:58] *** GBenji has quit IRC (Read error: Connection reset by peer)
[15:58] * anuw[m] uploaded an image: (135KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/qMcbJcUlFkNIjwhSvWbqKoSQ/2024-05-30-23-58-21-647.jpg >
[15:59] <anuw[m]> the manual checkup on some other app like this
[16:05] <pinchartl> is that last image from qcam or from capturing images manually ?
[16:19] <anuw[m]> <pinchartl> "is that last image from qcam..." <- i captured manually
[16:20] <pinchartl> it may be a stride issue. each line can be followed by some padding
[16:20] <pinchartl> what did you capture the frames with ? yavta ? v4l2-ctl ?
[16:21] <anuw[m]> pinchartl: i have that gut felling too. yavta this case
[16:24] <pinchartl> could you share the yavta log ?
[16:24] <pinchartl> along with the command line you used
[16:24] <pinchartl> and the raw2rgbpnm command line
[16:27] <anuw[m]> pinchartl: pardon me, is it enable the debug build of yavta？
[16:27] <pinchartl> no, I mean the messages printed by yavta on the command line
[16:29] <anuw[m]> pinchartl: ah thx for your hints. it shows 3248 as stride of 2592X1944
[16:30] * anuw[m] uploaded an image: (98KiB) < https://matrix.org/_matrix/media/v3/download/matrix.org/RXVMUGYkKrvIkQZWIkXegGhm/2024-05-31-00-29-51-520.jpg >
[16:30] <anuw[m]> pretty much done
[16:31] <pinchartl> that's better :-)
[16:31] <pinchartl> stride-wise at least
[16:34] <anuw[m]> pinchartl: thx alot, i wondering if the capture is a bit low quality, is that a normal case？
[16:38] <pinchartl> the quality of the raw image you get from the sensor isn't expected to be great
[16:38] <pinchartl> that's why an ISP is needed :-)
[16:38] <pinchartl> it's not just about converting from bayer to RGB/YUV
[16:39] <pinchartl> it's also about implementing auto-exposure, auto white balance, denoising, lens shading correction, gamma correction/tone mapping and lots of other processing steps to improve the image quality
[16:40] <pinchartl> Qualcomm SoCs used in phones include an ISP, but there's no mainline driver for them, and no technical information to develop and use drivers
[16:40] <pinchartl> so we're a bit stuck for now
[16:40] <pinchartl> hence the soft ISP project, which aims at implementing a basic ISP in software
[16:41] <pinchartl> at the moment it handles conversion from bayer to RGB, auto-exposure, auto white balance, and gamma correction
[16:41] <pinchartl> it's better than nothing, but won't provide great quality images
[16:42] <pinchartl> future developments may improve that if we get to offload the processing to the GPU. CPU-based processing is too expensive to implement some of the processing steps that would improve the image quality significantly
[16:42] <pinchartl> but the best option, of course, is for Qualcomm to be more open and let us implement support for their chips in libcamera
[16:42] <pinchartl> the ball is in their court, until they make a good move, all Qualcomm users will be stuck with bad image quality
[16:46] <anuw[m]> ah, its a very complicated situation. cheers for what we have.  good job man！
[16:52] *** anuw[m] is now known as geobang[m]
[16:53] <pinchartl> thanks :-)
[18:05] *** cyrinux30 has quit IRC ()
[18:07] *** cyrinux30 has joined #linux-media
[18:28] *** BrianG61UK has quit IRC (Read error: Connection reset by peer)
[18:33] *** BrianG61UK has joined #linux-media
[18:52] *** gouchi has joined #linux-media
[19:35] *** ao2 has quit IRC (Quit: Leaving)
[20:19] *** ten1572377432463050 has quit IRC (Remote host closed the connection)
[20:20] *** ten1572377432463050 has joined #linux-media
[20:46] *** gouchi has quit IRC (Quit: Quitte)
[21:02] *** avolmat has quit IRC (Quit: Leaving)
[22:15] *** ten1572377432463050 has quit IRC (Remote host closed the connection)
[22:15] *** ten1572377432463050 has joined #linux-media
[22:17] *** mrpops2ko has quit IRC (Ping timeout: 480 seconds)
[22:25] *** mrpops2ko has joined #linux-media
[23:21] *** djrscally has quit IRC (Ping timeout: 480 seconds)
[23:25] *** ten1572377432463050 has quit IRC (Remote host closed the connection)
[23:26] *** ten1572377432463050 has joined #linux-media
[23:45] *** lexano has quit IRC (Ping timeout: 480 seconds)