Re: mt9m111/mt9m131: kernel 3.8 issues.

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Fri, 8 Mar 2013, Guennadi Liakhovetski wrote:

> On Fri, 8 Mar 2013, Benoît Thébaudeau wrote:
> 
> > Hi Javier,
> > 
> > On Friday, March 8, 2013 1:37:38 PM, Javier Martin wrote:
> > > Hi Benoît,
> > > 
> > > On 8 March 2013 12:53, Benoît Thébaudeau <benoit.thebaudeau@xxxxxxxxxxxx>
> > > wrote:
> > > >
> > > >> Regarding 3, you say it works nicely for you in kernel 3.4.5. I've
> > > >> migrated my code to that version but I still get colours that lack
> > > >> enough intensity.
> > > >> This is a snapshot "a" taken with my mobile which is much more similar
> > > >> to what I can really see with my eyes:
> > > >> http://img96.imageshack.us/img96/1451/20130307171334.jpg
> > > >>
> > > >> This is a similar snapshot "b" taken with mt9m131 in my board. It
> > > >> shows that colours tend to be dull and darker, specially green:
> > > >> http://img703.imageshack.us/img703/6025/testgo.jpg
> > > >>
> > > >> Are the snapshots you take with your HW  more similar to "a" or to
> > > >> "b"? Perhaps I am being too picky with the image quality and this is
> > > >> all what mt9m131 can do?
> > > >
> > > > I fear that my captures are closer to "b". Your description of "3" was
> > > > giving
> > > > the impression of flashy colors. But the impression that this sensor gives
> > > > me is
> > > > rather a superimposed gray film. This effect is more or less visible
> > > > depending
> > > > on the lighting conditions, but it never seems to produce high quality
> > > > colors.
> > > 
> > > Yes, yours is probably is the best description for the image quality
> > > this sensor provides with the current settings.
> > > Well, this is a bit disappointing. Let's see if some other user has a
> > > similar experience with it or comes up with a way to improve it.
> > > 
> > > I've tested several things such as disabling auto white balance and
> > > auto exposure and try to manually change colour gains but it seems the
> > > sensor simply ignores the latter.
> > > 
> > > There is an evaluation board  from Aptina
> > > (http://www.digikey.com/product-detail/es/MT9M131C12STCH%20ES/557-1251-ND/1643271)
> > > but, unfortunately, I don't have one of these available. It could be
> > > very useful to test the sensor with this board with the configuration
> > > Aptina recommends and see whether the "grey layer effect" still
> > > persists.
> > 
> > It is perhaps also possible to find their recommended register settings
> > somewhere without having this evaluation board.
> 
> I just tested my mt9m131 camera on a i.MX31 board, if not this your email 
> I don't think I'd be alarmed by the image quality it's producing, maybe 
> I'm just less picky:-) And yes, in general I agree, I think, this level of 
> image quality tuning is difficult to achieve on modern cameras with 100s 
> of fine-tuning knobs. I'll try to re-test this camera in day light 
> conditions and post my best shot :)

http://download.open-technology.de/mt9m131/

.ppm is taken with mt9m131. Note, that I shot 10 frames and took the 3rd 
one - the first two are much darker, #3 is where autoexposure has kicked 
in, I suppose. Also note, that the comparison shot has been fired from a 
much smaller distance to cover a similar area due to obviously different 
lenses. I'll leave any colour-quality judgement to the reader(s) :-)

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer
http://www.open-technology.de/
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html


[Index of Archives]     [Linux Input]     [Video for Linux]     [Gstreamer Embedded]     [Mplayer Users]     [Linux USB Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [Yosemite Backpacking]
  Powered by Linux