As you may have read in another thread, the allowed plate-solve time and number of catalogs checked have been reduced with the hope of reducing the number of images that get "dramatically slowed" in the queue. In the past year, literally 10% of the VPhot images have taken more than 100 minutes to get through the queue. Most get through in less than one minute. 40% take less than a few seconds.
What caused these long delays that all ended in plate solve failure? It has become apparent that these "bad images" are mainly a combination of images with clouds and images with very short exposures that have such few stars present that they fail pin-point after serious "trying"! As a result they historically have sat in the queue for over 200 sec each before they fail plate-solving and get forwarded as unusable for photometry.
Alternatively, we determined that most images plate-solve within a few secs (<10 sec) on the first UCAC4 catalog. The subsequent GSC and Tycho catalogs are never needed. So in the interest of removing the annoying (for many) queue delays, we changed the protocol to plate-solving for 10 secs x 2 catalogs (UCAC4 and GSC). We continue to ask/hope that most VPhot users will pre-plate-solve their images before uploading to VPhot. This is not a requirement but a request!
SO, perhaps you have a particular image set for which that new procedure may not be reliable. Can you explain a few things:
1. Do you pre-plate-solve your images before uploading?
2. Do most of your images still upload properly? Is there something different about this target set?
3. Do your problem images exhibit any of the characteristics that I mentioned above?
4. Can you give us an example of the targets/images that have recently failed? Share a few of these previously successful images to MZK and SGEO?
We would certainly like to understand what may be happening and try to fix this issue! Let's see if we can figure this out.
There may be a simple alternative explanation. Perhaps you were trying to upload your images this morning at the same time that George had the system down to fix the version problem? Try again? (Fingers crossed). ;-)
During my runs, all my images are plate solved through SGPro with either PinPoint or Astromery.net (for blind solve), so that I am assured the image is what I want before it is taken through SGPro. since my FOV is 15' x 20'.
Today's image that does not upload is RU Her_I (I filtered image). It is a difficult image - a stack of thirty 0.5 second images. Calibration and stack were performed with MPO Canopus. MPO Canopus has difficulty plate solving this image while VPHOT is able to do so, hence my use of VPHOT for the image's photometry.
Basically, no image I've tried has uploaded today - either from last night (excellent night, no clouds) or previous nights.
I'll send a couple of images as requested through Dropbox. I've erased all images from my in-box so I don't have past images that calibrated on the system at present.
The first 5 images are from last night's run - all processed with darks and flats. They plate solved quickly in MPO Canopus, so I did not upload them through VPHOT till I wanted to check whether VPHOT was uploading any images.
The last image - DY HER - is from about 4 weeks ago. I cannot remember if this particular image is one I had matched in VPHOT in the past. I typically use MPO Canopus for processing. However, I've used VPhot with DY HER images in the past without difficulty, so I thought I would send an example of an image that does not upload now but did in the past.
Mike:
As you may have read in another thread, the allowed plate-solve time and number of catalogs checked have been reduced with the hope of reducing the number of images that get "dramatically slowed" in the queue. In the past year, literally 10% of the VPhot images have taken more than 100 minutes to get through the queue. Most get through in less than one minute. 40% take less than a few seconds.
What caused these long delays that all ended in plate solve failure? It has become apparent that these "bad images" are mainly a combination of images with clouds and images with very short exposures that have such few stars present that they fail pin-point after serious "trying"! As a result they historically have sat in the queue for over 200 sec each before they fail plate-solving and get forwarded as unusable for photometry.
Alternatively, we determined that most images plate-solve within a few secs (<10 sec) on the first UCAC4 catalog. The subsequent GSC and Tycho catalogs are never needed. So in the interest of removing the annoying (for many) queue delays, we changed the protocol to plate-solving for 10 secs x 2 catalogs (UCAC4 and GSC). We continue to ask/hope that most VPhot users will pre-plate-solve their images before uploading to VPhot. This is not a requirement but a request!
SO, perhaps you have a particular image set for which that new procedure may not be reliable. Can you explain a few things:
1. Do you pre-plate-solve your images before uploading?
2. Do most of your images still upload properly? Is there something different about this target set?
3. Do your problem images exhibit any of the characteristics that I mentioned above?
4. Can you give us an example of the targets/images that have recently failed? Share a few of these previously successful images to MZK and SGEO?
We would certainly like to understand what may be happening and try to fix this issue! Let's see if we can figure this out.
Ken
Hi Mike:
There may be a simple alternative explanation. Perhaps you were trying to upload your images this morning at the same time that George had the system down to fix the version problem? Try again? (Fingers crossed). ;-)
Ken
Hello! Thank you for your note.
During my runs, all my images are plate solved through SGPro with either PinPoint or Astromery.net (for blind solve), so that I am assured the image is what I want before it is taken through SGPro. since my FOV is 15' x 20'.
Today's image that does not upload is RU Her_I (I filtered image). It is a difficult image - a stack of thirty 0.5 second images. Calibration and stack were performed with MPO Canopus. MPO Canopus has difficulty plate solving this image while VPHOT is able to do so, hence my use of VPHOT for the image's photometry.
Basically, no image I've tried has uploaded today - either from last night (excellent night, no clouds) or previous nights.
I'll send a couple of images as requested through Dropbox. I've erased all images from my in-box so I don't have past images that calibrated on the system at present.
The first 5 images are from last night's run - all processed with darks and flats. They plate solved quickly in MPO Canopus, so I did not upload them through VPHOT till I wanted to check whether VPHOT was uploading any images.
The last image - DY HER - is from about 4 weeks ago. I cannot remember if this particular image is one I had matched in VPHOT in the past. I typically use MPO Canopus for processing. However, I've used VPhot with DY HER images in the past without difficulty, so I thought I would send an example of an image that does not upload now but did in the past.
Thank you for your help and best regards.
Mike
https://www.dropbox.com/s/l3cunpfre9ftx5t/V386%20SER_300sec_Clear_0001_…
https://www.dropbox.com/s/atjcs0pg3519s87/U%20HER_4sec_I_0001_0C_P.FIT?…
https://www.dropbox.com/s/f0nisltpopl8fjh/T%20SER_10sec_I_0001_0C_P.FIT…
https://www.dropbox.com/s/abvt21bmpsjm2d2/RU%20HER_I.FIT?dl=0
https://www.dropbox.com/s/7p2d6w8pv86jbq1/RU%20HER_60sec_I_0001_0C_P.fi…
https://www.dropbox.com/s/82anma44ftpy3ue/R%20HER_300sec_V_0001_0C_P.fi…
https://www.dropbox.com/s/59o3ybp3yabvfg1/DY%20HER1_120sec_I_0001_0C_P…
VPhot is upload is not working right now.
Working on it.
George
Vphot is now able to upload files again.
The warning message on the website will be taken down in the morning.
Meanwhile, upload at will!
George
Phew! I thought it was me!
I just tried RU HER_I. Still no upload. I'll wait till the announvement that everything is back up and running. Best regards.
Mike
Mike:
Try again and let's talk. Can talk/skype off-line if needed. Email at kenmenstar@gmail.com
Ken