iraf.net

Dear Sir/Madam,I always encounter something problem while running IRAF on Ubuntu system. This time I install IRAF V2.14 on Ubuntu 7.10. It sounds great that I didn't encounter previous posting problems. However, this time after I run [b:21ca970759]imalign[/b:21ca970759] task and displayed the image to DS9 window, I found that all the processed images had regular horizontal strips. Just like a clear image with a regular horizontal strips mask on it. It has two types. One is the black strips and the other is the white ones.I feel very confused because the posted problems didn't occur in another notebook installed with the same version of Ubuntu and IRAF. Is this kind of problem a systematic one? Could somebody tell me why? and how to fix it?Thansks for your kindly helps!!
hymong

The first thing to do is check whether it is a display problem (i.e. does displaying [b:752e83ec57]any[/b:752e83ec57] image show the same patter) or an effect of running IMALIGN. On the processed image, you can use a task like CONTOUR or HAFTON to see whether the banding is real in the image, you might even get some hints from IMSTAT. If so, then you'll need to check your IMALIGN params or post them. If it appears to be a display problem, are you setting the 'fill' param? Is the image larger than your 'stdimage' size?Otherwise, this isn't a familiar problem.Cheers,
-Mike

Dear Mike,I think it should not be a display problem because my posted problems only occur after running IMALIGN.Here I post the results after running IMSTAT.Before running IMALIGN
[code:1:e7991688ee]plot> imstat @A12i.list[/code:1:e7991688ee]
# IMAGE NPIX MEAN STDDEV MIN MAX
ngc7000_A12_HaE_-001.fit 3214848 53.5 295.3 -11734. 55037.
ngc7000_A12_HaE_-002.fit 3214848 53.31 296.5 -11788. 55160.
ngc7000_A12_HaE_-003.fit 3214848 53.49 295.1 -11676. 55132.After running IMALIGN
[code:1:e7991688ee]plot> imstat @A12o.list[/code:1:e7991688ee]
# IMAGE NPIX MEAN STDDEV MIN MAX
ngc7000_A12_HaE_1.fit 3203889 46.88 255. -4289. 54904.
ngc7000_A12_HaE_2.fit 3203889 46.74 258.6 -5969. 54266.
ngc7000_A12_HaE_3.fit 3203889 46.92 263.2 -7897. 54448.Ooops, I don't know why the MIN pixel value is negative in all processed images (even those pre-processed images only). Are my data processes right?And the IMALIGN, DISP params list below.
[code:1:e7991688ee]plot> lpar imalign[/code:1:e7991688ee]
input = "@A12i.list" Input images
reference = "ngc7000_A12_HaE_-002.fit" Reference image
coords = "cood" Reference coordinates file
output = "@A12o.list" Output images
(shifts = "") Initial shifts file
(boxsize = 7) Size of the small centering box
(bigbox = 11) Size of the big centering box
(negative = no) Are the features negative ?
(background = INDEF) Reference background level
(lower = INDEF) Lower threshold for data
(upper = INDEF) Upper threshold for data
(niterate = 3) Maximum number of iterations
(tolerance = 0) Tolerance for convergence
(maxshift = INDEF) Maximum acceptable pixel shift
(shiftimages = yes) Shift the images ?
(interp_type = "linear") Interpolant
(boundary_typ = "nearest") Boundary type
(constant = 0.) Constant for constant boundary extension
(trimimages = yes) Trim the shifted images ?
(verbose = yes) Print the centers, shifts, and trim section ?
(list = "")
(mode = "ql")[code:1:e7991688ee]plot> lpar disp[/code:1:e7991688ee]
image = "ngc7000_A12_HaE_1.fit" image to be displayed
frame = 1 frame to be written into
(bpmask = "BPM") bad pixel mask
(bpdisplay = "none") bad pixel display (none|overlay|interpolate)
(bpcolors = "red") bad pixel colors
(overlay = "") overlay mask
(ocolors = "green") overlay colors
(erase = yes) erase frame
(border_erase = no) erase unfilled area of window
(select_frame = yes) display frame being loaded
(repeat = no) repeat previous display parameters
(fill = yes) scale image to fit display window
(zscale = yes) display range of greylevels near median
(contrast = 0.25) contrast adjustment for zscale algorithm
(zrange = yes) display full image intensity range
(zmask = "") sample mask
(nsample = 1000) maximum number of sample pixels to use
(xcenter = 0.5) display window horizontal center
(ycenter = 0.5) display window vertical center
(xsize = 1.) display window horizontal size
(ysize = 1.) display window vertical size
(xmag = 1.) display window horizontal magnification
(ymag = 1.) display window vertical magnification
(order = 0) spatial interpolator order (0=replicate, 1=line
(z1 = ) minimum greylevel to be displayed
(z2 = ) maximum greylevel to be displayed
(ztrans = "linear") greylevel transformation (linear|log|none|user)
(lutfile = "") file containing user defined look up table
(mode = "ql")Hope my posting information is useful....
Thanks....hymong

Did you find a solution to your problem ?I have the same thing happening when I try to register images with 'xregister', horizontal black strips appears on the images. The original ones have no problem. I try to shift images with 'imshift' and no black lines appears. (you tell that you have sometimes white lines but I always have black ones)I'm also on Ubuntu 7.10 with IRAF 2.14If anyone have a solution I'm very interrestedThank you very much
Pierre

Could somebody please post an image (or url to one) of the problem image? I doubt it is specific to Ubuntu but would need an input image, parameters and task call to reproduce so I can either explain what's happening or track a bug.Thanks,
-Mike

Dear Pierre,I feel very sorry to reply late. Actually, I didn't have any idea and still struggle to figure the problem out. Sorry I didn't response promptly... I am glad to hear that you or somebody have encountered the same problem as me. I wondered that it would be my personal specific problem but not general one. Hi dear Mike,Thanks for your kindly reply. I was eager to post my image before and after IRAF processing. But I still didn't know how to do it. Have you read my previous post on another subject yet? I did try to upload my images to the ftp site. But I encountered some error message. Could you please help me to figure it out? And I have noticed that in the forum you have provided the function to insert images Insert image: [code:1:855bd74bac][img]http://image_url[/img] (alt+p)[/code:1:855bd74bac]
But I still cannot get the idea. Sorry!! Could you please give me an example or previous posting how to post an image on the forum? That would be helpful to me. Sorry for my poor comprehension.Hope my image posting could clearly point it out my previous questions. And give you the clues to find out the problem. Thanks for your hard work very much!! Thanks!!hymong

HelloThank you very much for your answer.Here is a link to the bad image [i:2f70c0aad1][removed link][/i:2f70c0aad1].I get this image with the following lines in IRAF. It uses the dev$pix image installed with IRAF :
[code:1:2f70c0aad1]
imshift dev$pix image 12 14
xregister image dev$pix [*,*] shift.db xwindow=101 ywindow=101 correlation=fourier output=out
[/code:1:2f70c0aad1]
This code shifts dev$pix to a new image called 'image' and try to register the two images in a new image 'out'. This is this latter which presents black lines. The shift found is correct.I have tried the same code on another PC with IRAF 2.12.2 (I don't know which distribution of linux is installed) and all works fine.I think the two functions 'xregister' and 'imalign' use the same function or algorithm.Thank you very much for your help
Pierre

Hi!
I have also encoutered the black horizontal stripes after running imalign. I use the latest available iraf on a Mac pro (Tiger). The black stripes seems to have pixelvalue zero when I look at them with ds9.Would very much appreciate help or ideas!

We are three to have black stripes I've made some tests : the same problem happens with 'imshift' when the xshift or the yshift is [b:b1c063cd55]not an integer[/b:b1c063cd55]. 'imalign' uses 'imshift' to shift the images so it is probably the point which poses a problem. Maybe xregister also uses imshift ???[code:1:b1c063cd55]imshift dev$pix image 12 14[/code:1:b1c063cd55] works fine but
[code:1:b1c063cd55]imshift dev$pix image 12.1 14[/code:1:b1c063cd55] makes black linesIs it a problem of utilisation (bad parameters) or a bug in the function ?
I hope it will be helpfull
Pierre

Hi again!
Yes, i also tried imshift and still got the stripes. When I tried imalign on my old unix-machine it worked fine. Maybe it is a bug in the newest iraf-version? Are you also using a new version?Elisabet

I tried xregister also, with even worse result than for imalign... Elisabet

Thanks to all who reported the bug. This is an artifact of a system bug that was just uncovered as a result of multiple problems in any task that accesses an image section (either explicitly or during internal processing such as looping over rows). This resulted in the critical update just released and the suggestion that all users reinstall their IB and NB binaries from the patched version now in the archives.Unfortunately there is no workaround since the bug is in a library used by all tasks. I've verified that the change fixes the behavior shown by IMSHIFT/XREGISTER and so upgrading to the new binaries should (he said with crossed fingers) fix the problems reported here.-Mike

Thank you very much, it works Pierre