| Message ID | 20260105-softisp-agc-v2-1-9013e693ffd5@mainlining.org |
|---|---|
| State | Superseded |
| Headers | show |
| Series |
|
| Related | show |
Hi, Thank you for the new version. On 5-Jan-26 13:41, Vasiliy Doylov wrote: > On most imx sensors the formula seems to be: > > again_float = 512.0 / (512.0 - again_ctrl_value) > > So the default of 0 makes sense and actually translates to a gain of 1.0. > And the max gain of 400 leads to 512.0 / 112.0 = 4.57 which is about typical for > a max sensor gain. s/default/minimum again/ s/max sensor gain/maximum again/ And can you add some extra text explaining the change a bit more here, e.g. : Since a minimum again value of 0 is actually normal, the special handling of againMin == 0 is undesirable and this is actually causing problems on these IMX sensors. again10 correctly gets set to 0 which is less then the adjusted againMin which causes the AGC code to not work properly. Fix this by dropping the special handling of againMin == 0. > > Signed-off-by: Vasiliy Doylov <nekocwd@mainlining.org> > --- > Changes in v2: > - Drop old solution > - Fix minimal analog gain initialisation > - Link to v1: https://lists.libcamera.org/pipermail/libcamera-devel/2026-January/056045.html > --- > src/ipa/simple/soft_simple.cpp | 19 +------------------ > 1 file changed, 1 insertion(+), 18 deletions(-) > > diff --git a/src/ipa/simple/soft_simple.cpp b/src/ipa/simple/soft_simple.cpp > index b147aca2..dde11666 100644 > --- a/src/ipa/simple/soft_simple.cpp > +++ b/src/ipa/simple/soft_simple.cpp > @@ -237,26 +237,9 @@ int IPASoftSimple::configure(const IPAConfigInfo &configInfo) > camHelper_->blackLevel().value() / 256; > } > } else { > - /* > - * The camera sensor gain (g) is usually not equal to the value written > - * into the gain register (x). But the way how the AGC algorithm changes > - * the gain value to make the total exposure closer to the optimum > - * assumes that g(x) is not too far from linear function. If the minimal > - * gain is 0, the g(x) is likely to be far from the linear, like > - * g(x) = a / (b * x + c). To avoid unexpected changes to the gain by > - * the AGC algorithm (abrupt near one edge, and very small near the > - * other) we limit the range of the gain values used. > - */ > context_.configuration.agc.againMax = againMax; > context_.configuration.agc.again10 = againDef; > - if (againMin) { > - context_.configuration.agc.againMin = againMin; > - } else { > - LOG(IPASoft, Warning) > - << "Minimum gain is zero, that can't be linear"; > - context_.configuration.agc.againMin = > - std::min(100, againMin / 2 + againMax / 2); > - } > + context_.configuration.agc.againMin = againMin; Please move this to above: context_.configuration.agc.againMax = againMax; So that the values are initialized in the same order as in the if (true) code block above the else. Regards, Hans
Hi, thanks for the patch! On 05.01.26 14:23, johannes.goede@oss.qualcomm.com wrote: >> context_.configuration.agc.againMax = againMax; >> context_.configuration.agc.again10 = againDef; >> - if (againMin) { >> - context_.configuration.agc.againMin = againMin; >> - } else { >> - LOG(IPASoft, Warning) >> - << "Minimum gain is zero, that can't be linear"; >> - context_.configuration.agc.againMin = >> - std::min(100, againMin / 2 + againMax / 2); >> - } >> + context_.configuration.agc.againMin = againMin; > Please move this to above: > > context_.configuration.agc.againMax = againMax; > > So that the values are initialized in the same order as in > the if (true) code block above the else. > > Regards, > > Hans I'm afraid this part was forgotten in v3
Quoting Robert Mader (2026-01-05 14:35:11) > Hi, thanks for the patch! > > On 05.01.26 14:23, johannes.goede@oss.qualcomm.com wrote: > >> context_.configuration.agc.againMax = againMax; > >> context_.configuration.agc.again10 = againDef; > >> - if (againMin) { > >> - context_.configuration.agc.againMin = againMin; > >> - } else { > >> - LOG(IPASoft, Warning) > >> - << "Minimum gain is zero, that can't be linear"; > >> - context_.configuration.agc.againMin = > >> - std::min(100, againMin / 2 + againMax / 2); > >> - } > >> + context_.configuration.agc.againMin = againMin; > > Please move this to above: > > > > context_.configuration.agc.againMax = againMax; > > > > So that the values are initialized in the same order as in > > the if (true) code block above the else. > > > > Regards, > > > > Hans > I'm afraid this part was forgotten in v3 Argh, sorry - I just merged v3 as I thought it was all handled. I think the ordering is trivial, but happy to merge a patch fixing the order too! -- Kieran > > -- > Robert Mader > Consultant Software Developer > > Collabora Ltd. > Platinum Building, St John's Innovation Park, Cambridge CB4 0DS, UK > Registered in England & Wales, no. 5513718
diff --git a/src/ipa/simple/soft_simple.cpp b/src/ipa/simple/soft_simple.cpp index b147aca2..dde11666 100644 --- a/src/ipa/simple/soft_simple.cpp +++ b/src/ipa/simple/soft_simple.cpp @@ -237,26 +237,9 @@ int IPASoftSimple::configure(const IPAConfigInfo &configInfo) camHelper_->blackLevel().value() / 256; } } else { - /* - * The camera sensor gain (g) is usually not equal to the value written - * into the gain register (x). But the way how the AGC algorithm changes - * the gain value to make the total exposure closer to the optimum - * assumes that g(x) is not too far from linear function. If the minimal - * gain is 0, the g(x) is likely to be far from the linear, like - * g(x) = a / (b * x + c). To avoid unexpected changes to the gain by - * the AGC algorithm (abrupt near one edge, and very small near the - * other) we limit the range of the gain values used. - */ context_.configuration.agc.againMax = againMax; context_.configuration.agc.again10 = againDef; - if (againMin) { - context_.configuration.agc.againMin = againMin; - } else { - LOG(IPASoft, Warning) - << "Minimum gain is zero, that can't be linear"; - context_.configuration.agc.againMin = - std::min(100, againMin / 2 + againMax / 2); - } + context_.configuration.agc.againMin = againMin; context_.configuration.agc.againMinStep = 1.0; }
On most imx sensors the formula seems to be: again_float = 512.0 / (512.0 - again_ctrl_value) So the default of 0 makes sense and actually translates to a gain of 1.0. And the max gain of 400 leads to 512.0 / 112.0 = 4.57 which is about typical for a max sensor gain. Signed-off-by: Vasiliy Doylov <nekocwd@mainlining.org> --- Changes in v2: - Drop old solution - Fix minimal analog gain initialisation - Link to v1: https://lists.libcamera.org/pipermail/libcamera-devel/2026-January/056045.html --- src/ipa/simple/soft_simple.cpp | 19 +------------------ 1 file changed, 1 insertion(+), 18 deletions(-) --- base-commit: 2861817f09c96a0660b88783eff159631e42030f change-id: 20260105-softisp-agc-16d01cb3da49 Best regards,