Raspberry pi Camera Module tips needed

[SirNight]

Hi everyone,

I have been trying to get better quality photos with my camera module.
How can I get less blur and noise ? Even sometimes i try with spot lights to have good lightning but can not get good photos.
I do now that i mustn't expect the best photos from camera module but i am also curious if there is some useful tips to snap better photos.. I use "raspistill" command..

Cheers,

[gordon77]

What settings are you using in raspistill ?

For lower noise use low ISO.

Why is it blurred ? Shutter speed or focus ?

[SirNight]

What settings are you using in raspistill ?

For lower noise use low ISO.

Why is it blurred ? Shutter speed or focus ?

Hi Gordon,

Thx for your answer,
I use like this, " raspistill -awb off -awbg 1.9,1.3 -ss 30000 -ISO 100 -n -t 1 -o test.jpg "
Can i do some improvements? So, please tell me what you think..

[gordon77]

You have a low ISO and a fast shutter speed.

So unless you are moving the camera or its out of focus l don't know what to suggest.

Expecting too much from a £20 camera ? I think they are very good for the cost.

[DirkS]

Could it be a focusing problem. Maybe less than minimal focusing distance?

[SirNight]

Yes gordon, Thx for your help..
Dirks , How can adjust to focus then ? any tips ?

[AndrewS]

http://www.raspberrypi.org/macro-photog ... era-board/

[DirkS]

Minimum distance (w/o changes) seems to be about 60cm
http://www.raspberrypi.org/forums/viewtopic.php?t=43985

[RaTTuS]

post an image so we can see what it is you are seeing

[SirNight]

post an image so we can see what it is you are seeing

Hi, There are the sample photos from two different cameras

[6by9]

Something is not right with your setup. The AWB gains are stored in the EXIF header ("strings foo.jpg | grep gain" to read them) and for your two images read
"gain_r=1.664 gain_b=1.000" for 1.jpg and
"gain_r=1.781 gain_b=1.156" for 2.jpg.
Those should both be the same at or around your specified settings.
Likewise the exposure times on both are 15ms when you said you added the -ss 30000 to the command line. Analog gain is at x1.0 (256) which is what you would expect for ISO100, but the rest doesn't add up.

[SirNight]

Something is not right with your setup. The AWB gains are stored in the EXIF header ("strings foo.jpg | grep gain" to read them) and for your two images read
"gain_r=1.664 gain_b=1.000" for 1.jpg and
"gain_r=1.781 gain_b=1.156" for 2.jpg.

Yes definately something is wrong with my setup.. Let me explain what i do.
All Raspberries use the same settings . Actually I push all shot parameters to Raspberries.
And then broadcasting sends signal to execute this code below; (I so often change it and push it with ansible to see differance)

I dont understand what's the matter.

[AndrewS]

You haven't yet mentioned if you've checked that all your Pis are running up-to-date firmware

What do the commands 'vcgencmd version' and 'uname -a' report on each of your Pis?

Using a limited sample-set of just a single Raspberry Pi (with up-to-date firmware) and a single Pi camera, but taking 3 pictures with pointing the camera at very different areas of brightness, running your sample command of

Code: Select all

raspistill -awb off -awbg 1.7,1.5 -ISO 100 -n -t 400 -sh 40 -ss 11000 -o captureN.jpg

and then running the 'grep gains' command from 6by9's post, gives me three different images that all have 'gain_r=1.699 gain_b=1.500'

And it's not clear where your original problem report of 'bluriness' comes into this?

[SirNight]

Hi,

I made a new test.

First , i checked all my raspis version by using ' vcgencmd version ' and ' uname -a ' commands.
They reported; (only 6 of them)

They have the same version. also I remember 6by9 told me "AWB gains were added back in March, so it must be after that date." according to him there is no problem.

And then I used raspi still command that is "raspistill -awb off "
later I read their info by " strings test.jpg | grep gain "
They reported; They all look the same , except " f= " and " tg= ", what do they stand for ?


Then, I was thinking perhaps my trigger script makes problem when it calls "raspistill".
I triggered with my script and took all the photos. Gains are the same, however " tg= " and " f= " changes


The most interesting thing is, They still look different from each other..

I am going to be crazy soon...

Thank you for your help guys, I really appreciate it.

And it's not clear where your original problem report of 'bluriness' comes into this?

The situation changed... I wish, there was only focus problem...

[SirNight]

any comments or suggestion so far?

[RaTTuS]

my comment would be :-
each camera needs to be calibrated with WB to itself - setup a white screen and get the figures for each one ,
then you can use that as the basis on the same camera

[SirNight]

I see,
So we can say Raspberry pi camera modules somehow aren't produced at the same quality.
Perhaps sensor or other problem with the camera module should be issued?

[RaTTuS]

dunno - but there will be differences - the awb option will correct ...
what happens when you have that switched off and you read the vales - are they vastly different from each other ?

also just do a
raspistill -o test.jpg
on each of them and post them to see if they look bad?

[Cancelor]

I see,
So we can say Raspberry pi camera modules somehow aren't produced at the same quality.
Perhaps sensor or other problem with the camera module should be issued?

Quality is not the same as identical, that is why things need calibrating.

[pluggy]

Nobody appears to have suggested the obvious (to me at any rate) step of swapping the cameras between the Pis and seeing what changes......