Skip to main content

Images shown as just white in VPHOT?

11 posts / 0 new
Last post
LMN
Images shown as just white in VPHOT?

Hi!

I'm brand new to VPHOT, so probably just missing out on something simple. What I've done: 

I used my own telescope and CCD to captura a number of images, in this case of SS Cyg. Then calibrated them with darks, flats and bias in Pixinsight. THen saved the calibrated images as fits (xisf by default in PI). 

So: then I uploaded to VPHOT. After I while, I figured out how to do that. I uploaded both uncalibrated a fits image and a calibrated version of the same. Now the trouble: the calibrated image just shows as white in the image view - meaning I cannot select stars. The un-calibrated image shows nicely. 

So: what is wrong here? You can find the fits files in a dropbox folder - the calibrated one is with the _c appended to the name:

https://www.dropbox.com/sh/fs744ujp8oed79d/AACfnztnt-E98N8km8Tv2dUBa?dl=0

 

 Secondly, the images are tagged as not calibrated. How do I make VPHOT understand that they ARE calibrated?

Grateful for any advice. 

Magnus

 

spp
spp's picture
White images

Magnus,

Something has gone wrong in your calibration process.  The pixels in the uncalibrated image have normal pixel values (a few hundred ADU's in the background, up to  a few thousand ADU's in the stars).  In the calibrated image the pixels have values of a very small fraction of an ADU.

To make the Cal box turn green you can edit the FITS header, adding keyword  CALSTAT and the value BDF.  (...for calibration status and Bias Dark Flat.)  This is not really necesary, and it makes no difference in the way VPhot processes the image.

Phil

LMN
Hi!

Hi!

I'm not sure what you mean with fractions of ADU. Did you look at the fits files? If I measure the stars in the calibrated fits file, the brightest star is around 65000 ADU. So how to you see fractions of ADU's there? I've utilized the standard calibration process in PI... or do I measure something wrong here?

Magnus

spp
spp's picture
ADU values in calibrated image

Magnus,

I have attached four screen shots which show the central regions of your raw and calibrated images.  In the upper right you'll see a zoomed in box for each which shows the pixel for which the information is displayed in the table.  The "Value" for the pixel in the table is given in ADU's. 

I suspect the fractional pixel values for the calibrated images are an artifact from Pixel Insight.  I think this is why your calibrated images will not display properly in VPhot. 

Phil

 

 

 

LMN
Hi!

Hi!

Yes, I get it. There is something strang in these files. However, they show me very different numbers when measuring in PI - background around 200 ADU, for instance, in the calibrated - far from any fractions.

I belive now (working hypothesis) that is something to do with either: 

1. PI does something to files when saving in nateive xisf-format, that translates into the fits when re-savting as fits. 

2. Something about the fits format that I re-saved in. These are actually in "32-bits floating point" and I think that what the camera (Atik 383L+) delivers is 16 bit integer. But so far I have not found a fits format from PI that works. Any suggestions or details about the format might be helpful, actually. 

So, if I do not calibrate in PI - what program to calibrate would you recommend? Preferably free.... I would only use it for this calibration. 

Best, 

Magnus

 

 

JJLB
free calibration

I use AstroImageJ for calibration and fits header editing of my aperture photometry images prior to uploading to VPhot...this is the software of choice for the ExoPlanet folks and is FREE and available for PC and Mac platforms...just google "astroimagej".

My original images are saved as Fits out of the capture software for my camera (Atik 314L+)...I use Atik's ArtemisCapture for acquisition...if that helps.

Jamey

spp
spp's picture
Your hypothesis

Magnus,

I think your hypothesis #1 is probably correct.

I agree with Jamey.  Use AstroimageJ for calibration, then resubmit the images to VPhot.

Phil

MZK
MZK's picture
PixInsight

Magnus:

I believe PI has tired to "improve" their software for "pretty pictures". IMO, it has caused many problems for plain photometry. I think they normalize some values that really mess with the screen stretch and cause many problems for VPhot. Try other software for calibration?

VPhot likes to find a CALSTAT=BDF header. Note that although VPhot may not otherwise acknowledge that they are calibrated and show a green CAL button, it is calibrated and you can ignore this. There are utilities that can modify fits headers if you really want to.

Ken

LMN
Hi!

Hi!

I'm now using AstroImageJ, and that works nicely. I am still very curious, though, about what it is that PI does. I take it to be a very serious program, although focused on pretty pictures, and I really don't understand how the file conversion between xisf and fits can cause problems such as these. Interestingly, AstroImageJ have no problems in opening my fits files from PI and showing correct values... I'd much like to understand it better, so far, however, I have not gotten any response on the PI forum. Will return if something turns up. 

Best, 

Magnus

spp
spp's picture
Calibration problem

Magnus,

In general I think it is better to shy away from software designed for astro-photography.  ...better to use software which is geared more to science:  AIP, MaximDL, CCDSoft, AIJ, and others.

If you ever find out what happened with that image in PI, yes, please post it here so we can all learn from it.  There are plenty of AIJ users around here.  Since AIJ opens that image perhaps someone in this forum can tell us.

Phil

Eric Dose
Eric Dose's picture
Sorry to come late to this...

Sorry to come late to this...

You can't use PixInsight for steps in a photometry workflow with other software. You just can't.

PixInsight has its own flux scaling, incompatible with and uninterpretable by other software including MaxIm and VPhot. Been there. Most PixInsight functions do not guarantee linearity, that is, they may apply a gamma or other non-linear flux transform to make the images look better, but which (irreversibly) ruin the linearity needed for photometry. I can't speak for other purely photoimaging software, but PixInsight is a dead end for any photometry I would do.

Log in to post comments
AAVSO 49 Bay State Rd. Cambridge, MA 02138 aavso@aavso.org 617-354-0484