Forums / The Science / Photometry / V filter and DSLR

V filter and DSLR

Citizen Sky is now officially permanent part of the AAVSO. In the coming weeks we will be moving additional content to the AAVSO site and freezing this site as an archive of the 1st three years of the project. Please visit the new landing page for future updates.
  • strict warning: Non-static method view::load() should not be called statically in /media/disk/www/citizensky.org/drupal/sites/all/modules/views/views.module on line 879.
  • strict warning: Declaration of views_handler_filter::options_validate() should be compatible with views_handler::options_validate($form, &$form_state) in /media/disk/www/citizensky.org/drupal/sites/all/modules/views/handlers/views_handler_filter.inc on line 0.
  • strict warning: Declaration of views_handler_filter::options_submit() should be compatible with views_handler::options_submit($form, &$form_state) in /media/disk/www/citizensky.org/drupal/sites/all/modules/views/handlers/views_handler_filter.inc on line 0.
  • strict warning: Declaration of views_plugin_row::options_validate() should be compatible with views_plugin::options_validate(&$form, &$form_state) in /media/disk/www/citizensky.org/drupal/sites/all/modules/views/plugins/views_plugin_row.inc on line 0.
  • strict warning: Declaration of views_plugin_row::options_submit() should be compatible with views_plugin::options_submit(&$form, &$form_state) in /media/disk/www/citizensky.org/drupal/sites/all/modules/views/plugins/views_plugin_row.inc on line 0.
  • strict warning: Non-static method view::load() should not be called statically in /media/disk/www/citizensky.org/drupal/sites/all/modules/views/views.module on line 879.
  • strict warning: Declaration of views_handler_field_user::init() should be compatible with views_handler_field::init(&$view, $options) in /media/disk/www/citizensky.org/drupal/sites/all/modules/views/modules/user/views_handler_field_user.inc on line 0.
  • strict warning: Declaration of views_handler_argument::init() should be compatible with views_handler::init(&$view, $options) in /media/disk/www/citizensky.org/drupal/sites/all/modules/views/handlers/views_handler_argument.inc on line 0.
  • strict warning: Declaration of views_handler_filter_boolean_operator::value_validate() should be compatible with views_handler_filter::value_validate($form, &$form_state) in /media/disk/www/citizensky.org/drupal/sites/all/modules/views/handlers/views_handler_filter_boolean_operator.inc on line 0.

jmarc's picture
jmarc
User offline. Last seen 2 years 51 weeks ago. Offline
Joined: 07/02/2011
Posts: 2
Teams: None

Hi,
If I use a photometric V filter with a DSLR, in front of a camera lens with an adapter, do I get true "V" data ?
If  yes, shoul I work on the green chanel only or not?   

Roger Pieri's picture
Roger Pieri
User offline. Last seen 49 weeks 4 days ago. Offline
Joined: 08/29/2009
Posts: 136

HI, If you put a V filter in front of a DSLR what you will get is a combination of the response of the V filter and the DSLR filters, in fact the (math) product of the transmission coefficients of both at each wavelenght. Nothing to do with a V filter. To transform the G filters (plus the IR cut that exists in a DSLR and affect strongly R and some of G)to a V filter, you would need a 'blue cut' for a first approach (low pass filter). The reason is the roll-off of the G filter of the DSLR is smoother than the steep roll-off of the V filter. Due to this there is a leak of blue in the G. An interferential blue cut could be adjusted for "aligning the blue side of the G" on the V. But if you look at the Gresponse you will see there is also some missing red transmission on the red side roll-off of the G and this can't be corrected by a filter or at a strong loss of global transmission: unacceptable. The "defiltered" DSLR (replacement or elimination of the IR filter) could beenough forimproving the red response.Equipped with the proper "blue cut" such DSLR could be able toachieve a Johnson V response. There is an existingblue cut used by the exoplanet guys, I have not checked its response as it istoo muchexpensive. In case of DSLR from which the IR filter has been eliminated an interferentialIR cut shall be added in any case. The best is to use the G channel of the DSLR and then apply a color transform to the results. To see the spreadsheet from the tutorials: > Tutorials and Training > Observing > DSLR > Overview of Reduction Packages > Finishing Analysis... Yours truly, Roger

jmarc's picture
jmarc
User offline. Last seen 2 years 51 weeks ago. Offline
Joined: 07/02/2011
Posts: 2
Teams: None

thank you for your very clear answer, Roger.

Dennis's picture
Dennis
User offline. Last seen 2 years 4 days ago. Offline
Joined: 11/02/2011
Posts: 2
Teams: None

 

 Hi everybody.

My name is Dennis and I am very interested in learning more about photometry.
I living in Brasil. Forgive my English :/

 I recently purchased a Canon T2i. I know that it has  a IR block filter. How will this affect my data? I wouldn't want to remove this filter because I didn't know where to start.

I accept suggestions on how to work with this camera.

Best regards

Dennis
Fortaleza city- Ceará - Brasil



Roger Pieri's picture
Roger Pieri
User offline. Last seen 49 weeks 4 days ago. Offline
Joined: 08/29/2009
Posts: 136

Hi Dennis,

No problem, just use your DSLR like it is ! 

I am not familliar with the "T2i" type, it's probably special to Americas ? But it looks similar to the 550 D known in Europe. For photometry you need in any case an IR filter otherwise all RGB pixels are gone to be polluted by IR. Dye RGB filters do not reject IR. The build-in IR filter is just fine as long as you do not consider to make spectro or imaging at Halpha wavelength.

Typical settings:

RAW, ISO 100, 10 sec, for a 50 mm aperture of a telelens is perfect for epsilon and zeta AUR (mag 3~4) You should defocus to about a surface of 150 pixels to increase the electron capacity, avoid saturation and undersampling (issue of the Bayer's pixel structure) If you use the 18-55 mm zoom at 55 mm F/5.6 the aperture is only 10 mm getting 25 x less photons. Then push to ISO 400, use less defocus to about a surface of 30~40 pixels, you could also push a bit the exposure. In other cases just adjust the settings depending the surface of the aperture (and the target magnitude...)

Defocus is easy to adjust on a bright star (eg. Capella) using the "live view" magnifier to 10x.

Just use the green channel under the RAW mode as described in the CS tutorial: 
 http://www.citizensky.org/content/dslr-documentation-and-reduction.

Only one point: if you use IRIS: there is AN ERROR at "step 5" as the tutorial describes a process with stacking. Then the result is only one stacked image. By the way use the command: "RGB separation" (not the sequence one) to get the single green image you will use in step 6.

There is a corrected and improved Tutorial from the DSLR Team but for unknown reason it has not been put on line by the CS staff. If needed contact me directly through the "contact" flap under my profile. 

Clear Skies !

Roger

Wagner Lamonica's picture
Wagner Lamonica
User offline. Last seen 1 year 43 weeks ago. Offline
Joined: 05/19/2011
Posts: 2
Teams: None

Hi people

I am from Brasil, i speak portuguese, so sorry about my poor english.
I wonder if anybody could help me.
I am involved with a project of DSLR photometry of delta scorpii in last july, august, setember and november, and i have a problem.
In one regular nigth of observation i  take usualy about 15 images of the star with 5 seconds exposure each and about 10 seconds interregno between then. I found diferences in diferencial photometry, after correction with darks and flats, about 0,2 vmag ( ex. 1,6 and 1,8 vmag)between two images with 10 seconds of interregno, same settings!
Thats is more than standart deviation of the background noise, so i think that camera is doing some "correction" i dont know.

I use a Canon 450d 55mm/f=5.6/5 seconds, RAW format, slightly defocused, 400 ISO. The reference star is beta Scorpii (V=2,5 mag) 

What would be the normal error expected?

Best regards to all

Wagner Lamonica
 

Roger Pieri's picture
Roger Pieri
User offline. Last seen 49 weeks 4 days ago. Offline
Joined: 08/29/2009
Posts: 136

Hi Wagner,

No problem, I am French and my English is not very good too !

With images exposed for only 5 seconds it seems to me normal to see a +/- 0.1 scatter between images ( the time between images is not involved) This is due to the scintillation of the stars. Under a good sky, the reduction of that effect to 0.01~0.03 mag needs about 60 seconds of integration. Sometimes more... 

With such bright stars and a 10 mm aperture (55mm F/5.6) it would be better to use ISO  200 and expose longueur, like15~20 sec (Check the saturation). With a 55 mm you would get some trail but it's no issue for photometry. If you see a saturation increase the defocus. All this will provide more integration and increase the overall electron storage capacity. That means an increase of the mag range and - more important - getting more electrons and a better SNR (signal to noise ratio)  

In RAW mode the 450D doesn't apply any brightness control, no issue. I use it for two years and I can say it works very well.

You could make three, or better five, series of 5 images of 15~20 seconds, then average the results of each series and calculate the SD between the five series. You should reach SD < 0.01

Clear Skies,

Roger

Wagner Lamonica's picture
Wagner Lamonica
User offline. Last seen 1 year 43 weeks ago. Offline
Joined: 05/19/2011
Posts: 2
Teams: None

Hi Roger

Thank you very much, for the interesting information. I will try.

Reading yor post to my compatriote Dennis I am wondering if a relation between mag-defocus (number of pixels involved) can be established for a minimum necessary amount of electrons in each pixel of a star (for an determined ISO). The relation electron/ADU of the 450D, must be known, rigth?

Mercy encore
ciel clair,

Wagner Lamonica
 

Roger Pieri's picture
Roger Pieri
User offline. Last seen 49 weeks 4 days ago. Offline
Joined: 08/29/2009
Posts: 136

Hi Wagner,

Good French too !  But the sky here is very bad that November (usual) should be better in Brazil !

Yes, it's known for the 450D:  at ISO 100 this is 2.3 e/ADU, then it's proportional to k = ISO 100 / ISO xxxx (the ISO setting is only the gain of the amplifier that normalise the signal to the range of the Analog to Digital Converter) At ISO 400 we have less than one electron per ADU, by the way getting to heigher ISO doesn't improve the quantization, the minimum signal coding step ! But this reduces the dynamic range, the possible mag range.

For stars bright enough and if the aperture of the lens permit, we could reach saturation even at ISO 100, in this case we could increase the defocusing to eliminate the saturation, get more pixel involved, a larger electron count and a better SNR.
It's possible to get far this way, I often use 100~300 pixels (but with the large aperture of about 2000 mm² of a tele-lens,  instead the 80 mm² of the 55 mm zoom)
The limit for bright stars is the blending with nearby background stars. In case of fainter stars that would not saturate, some defocus shall be keept to eliminate the risk of undersampling. This is the two limits of the defocusing range.
 I should also note that we have problems with some softwares that do not recognize defocused stars like DSS. Others like the IRIS ">Automatic Photometry.."  could have both poor location accuracy and photometry depending the defocus shape. 

Clear Skies !

Roger

Dennis's picture
Dennis
User offline. Last seen 2 years 4 days ago. Offline
Joined: 11/02/2011
Posts: 2
Teams: None

 Hi everybody.

 I was working with CCD ST8 XME when I read about undersampling. I solved that problem using binning 2x2. But how can I determine the best measure between linearity, defocusing and ADU? How can I measure this in my 550D ?

 Clear skies!

 Dennis 

Roger Pieri's picture
Roger Pieri
User offline. Last seen 49 weeks 4 days ago. Offline
Joined: 08/29/2009
Posts: 136

Hi Dennis,

The 550D is a recent CMOS camera, it should have no linearity problem in all the ISO 100 range up to the saturation  point (usually somewhat below the end of the range of the ADC eg. between 14000 and 15000) Then the following ISOs use only the bottom of the response curve (1/2, 1/4...) of the sensor and there should never be issue (have a look at my post on the subject: Forum / Photometry / DSLR linearity)

Measuring the linearity is not easy if you do not have access to a lab with a good equipment. One way usually suggested to the amateurs is to make a series of shots of a white target at various speed and measure the ADUs with any of our software. The "stats" function of IRIS ( draw a box on a uniform area then click right... ) is fine (subtract the system offset eg. 1024) The issue here is that the speed displayed by the camera is usually not the exact one ! In fact the speeds used by the camera's electronics follow a geometric progression only approaching the traditional photo speeds that are displayed... It's possible to estimate it with a good accuracy based on a fit over a large range of speeds.
 
For the 450D   k=1.2587498  and  T(s)= 0.008*k^n  starting from 1/125

Normally there is no large difference of sensitivity between sensors, the intrinsic Qe can't be different by a large factor. There could be some difference of transmission of filters. Overall it should not be very different (a question of 10% ~20% ?)  Next is the size of the pixel, the 550D has 1.5 times such of the 450D, this should means 1.5 less electrons per pixel and a calibration factor of about 1.5 e/ADU at ISO 100. To be checked, ISOs are not very well defined (probably there are "commercial" ISOs ! )

Mesuring it is possible, the procedure exists on the C.Buil site http://astrosurf.com/buil/5d/test.htm. It is based on the fact the "shot noise" is just the square-root of the number of electrons being collected. The "read noise" shall be subtracted. If you have a problem let me know I will describe my own procedures for both noises and the e/ADU formula.

Clear Skies !

Roger

Powered by Drupal