Latest Announcement: On Going Site Upgrades & Data Sharing
View Post
While I was away I said I would take a look at how our scheduling software was working in relation to the pier queue
Myself and Phil have just done out best to go through the queue and check objects,
horizon's and what filters are being used as well as stating if the
object is setting on the west where possible.
Whilst we are only armed with some of the information it gives us a better idea of why its
struggling to collect some of the data,
The moon is above the horizon after dark approximately 25 nights per calender month in one
phase or another and alters position in the sky by about 50 minutes a
day so will travel through quite a large proportion of the
southerly aspect throughout it's different phases. When you take for
example our luminance moon avoidance parameters, which I will add an
example of below this obviously means that objects will only be imaged
on certain days due to moon constraints. Narrowband however has much
lower moon constraints by nature so allows more data collection
throughout the whole month.
Scheduler at present contains:
put in, so if we missed something them my apologies. However, to us its
obvious that the scheduler is not being given enough decent jobs in
multiple filters across varying areas of the sky to keep itself busy
However we would like your opinions after reading the attached image so we can
move forward and get the best out of the pier for us and all the group members
Steve & Phil
Luminance Moon Constraints
HA Moon Constraints
Whilst I think about it, I also think my suggestion while I was away of using filter prioritisation would also help on top of better
object/sky/filter use by members
So suggested filter priorities would be :
0 = Narrowband
1 = RGB
2 = Luminance
Signal strength per filter:
We are all guilty of it including myself at times
CMOS is very very different to CCD so please forgive my very crude
inaccurate non scientific example below but it serves the purpose I am
trying to make
(not taking into account outside factors like light pollution, moon etc)
CCD
The reason we have all been going long exposures for so long is signal to
noise and with CCD the exposure lengths required to overwhelm its noise
floor. So a 600 second exposure is perhaps 500s of noise and 100s of
data
CMOS
A 300s exposure is perhaps 5/10/20s of noise and the rest is data
Myself and Phil have just done out best to go through the queue and check objects,
horizon's and what filters are being used as well as stating if the
object is setting on the west where possible.
Whilst we are only armed with some of the information it gives us a better idea of why its
struggling to collect some of the data,
The moon is above the horizon after dark approximately 25 nights per calender month in one
phase or another and alters position in the sky by about 50 minutes a
day so will travel through quite a large proportion of the
southerly aspect throughout it's different phases. When you take for
example our luminance moon avoidance parameters, which I will add an
example of below this obviously means that objects will only be imaged
on certain days due to moon constraints. Narrowband however has much
lower moon constraints by nature so allows more data collection
throughout the whole month.
Scheduler at present contains:
- Lots of luminance data which is the hardest data to collect[/*]Lots of RGB objects[/*]Few narrowband objects[/*]Many objects that are low on the horizon[/*]Some forward planning but not enough[/*]little narrowband that us up at the moment[/*]quite a few objects on the west and setting[/*]
put in, so if we missed something them my apologies. However, to us its
obvious that the scheduler is not being given enough decent jobs in
multiple filters across varying areas of the sky to keep itself busy
However we would like your opinions after reading the attached image so we can
move forward and get the best out of the pier for us and all the group members
Steve & Phil
Luminance Moon Constraints
HA Moon Constraints
Whilst I think about it, I also think my suggestion while I was away of using filter prioritisation would also help on top of better
object/sky/filter use by members
So suggested filter priorities would be :
0 = Narrowband
1 = RGB
2 = Luminance
Signal strength per filter:
- You also may need to consider the collection of too much data in any one
filter, HA is a good example of this as it needs very little imaging - Broadband on the whole does not need 20hr plus job lengths except for the most special or dim projects as its very
high in signal to noise ratio[/*]
time compared to SII[/*]
We are all guilty of it including myself at times
CMOS is very very different to CCD so please forgive my very crude
inaccurate non scientific example below but it serves the purpose I am
trying to make
(not taking into account outside factors like light pollution, moon etc)
CCD
The reason we have all been going long exposures for so long is signal to
noise and with CCD the exposure lengths required to overwhelm its noise
floor. So a 600 second exposure is perhaps 500s of noise and 100s of
data
CMOS
A 300s exposure is perhaps 5/10/20s of noise and the rest is data
Please ignore my dylexia wherever possible, just be thankful I can control my Tourettes ;)
Things to do, so little time!
Steve
Roboscopes Tea Boy
0
Undo
Votes
Hi gentlemen,
You've done a sterling job analysing the queues and may be forgiven for suggesting that the job 1132 does not rise sufficiently. Not my job by the way.
This entire subject is a can of worms and possibly why there has been no response so far. Too many variables involved not just with the jobs themselves but also participation by members of the forum.
Roboscopes team, you need to accept that keeping a pier occupied is not something as a top priority to some members, I cannot speak for them all. Personally I'll submit targets that are of particular interest to me and welcome all data from targets submitted by others. As a result I've discovered many excellent objects I'd never come across or considered. In addition taking pier 1 for example, looking at the job queue as it stands, there appears to be many outstanding jobs of current objects. Until such time the queue reduces I'm reluctant to add to it, even by including not yet risen items, since they may impact upon these current jobs.
You have stated before that we can submit what we like, so that does somewhat contradict what you are suggesting we do. I however realise you have made this post in response to some negative feedback, and are just being helpful.
I repeat, this is a can of worms!
From the side of the members of the forum it would be very helpful if there was more interest in what was going on overall. It's extremely easy to duplicate a submission due to the different names objects have, however, pointing out a duplication on the forum does not always initiate a response. The question is, why is that? There have been suggestions on limiting the integration times by splitting them, or, waiting for the data to check more is required. It is not possible to be very precise, but submitting too many hours may result in either that job not being completed, or, other much shorter jobs being affected. I'll say it again, a can of worms.
Steve and gang, stop fretting over these piers not being used to their full capacity. To me and the rest of the forum, it's up to us to keep a check on our jobs and submit a ticket to get them to raise its priority, providing the job was originally made while the target was rising, but it's now setting and nothing has yet been captured. Also, we cannot rely solely on notifications, and need to reasonably often check what others have written.
Maybe you could assign priorities depending on the filters, I've no view on that at the moment.
Now to pier 14. Wish I hadn't started this!
There is a risk that some jobs, particularly those for additional data, may not be collected by the time of the visit. So, do we need to list those which fall into that category and others where the current set up is better suited to.
All comments welcome.
Cheers,
Ray
You've done a sterling job analysing the queues and may be forgiven for suggesting that the job 1132 does not rise sufficiently. Not my job by the way.
This entire subject is a can of worms and possibly why there has been no response so far. Too many variables involved not just with the jobs themselves but also participation by members of the forum.
Roboscopes team, you need to accept that keeping a pier occupied is not something as a top priority to some members, I cannot speak for them all. Personally I'll submit targets that are of particular interest to me and welcome all data from targets submitted by others. As a result I've discovered many excellent objects I'd never come across or considered. In addition taking pier 1 for example, looking at the job queue as it stands, there appears to be many outstanding jobs of current objects. Until such time the queue reduces I'm reluctant to add to it, even by including not yet risen items, since they may impact upon these current jobs.
You have stated before that we can submit what we like, so that does somewhat contradict what you are suggesting we do. I however realise you have made this post in response to some negative feedback, and are just being helpful.
I repeat, this is a can of worms!
From the side of the members of the forum it would be very helpful if there was more interest in what was going on overall. It's extremely easy to duplicate a submission due to the different names objects have, however, pointing out a duplication on the forum does not always initiate a response. The question is, why is that? There have been suggestions on limiting the integration times by splitting them, or, waiting for the data to check more is required. It is not possible to be very precise, but submitting too many hours may result in either that job not being completed, or, other much shorter jobs being affected. I'll say it again, a can of worms.
Steve and gang, stop fretting over these piers not being used to their full capacity. To me and the rest of the forum, it's up to us to keep a check on our jobs and submit a ticket to get them to raise its priority, providing the job was originally made while the target was rising, but it's now setting and nothing has yet been captured. Also, we cannot rely solely on notifications, and need to reasonably often check what others have written.
Maybe you could assign priorities depending on the filters, I've no view on that at the moment.
Now to pier 14. Wish I hadn't started this!
There is a risk that some jobs, particularly those for additional data, may not be collected by the time of the visit. So, do we need to list those which fall into that category and others where the current set up is better suited to.
All comments welcome.
Cheers,
Ray
Ray
Roboscopes Guinea Pig
Hi Ray, I am easy on this I promise. It was an exercise in due diligence for us is all so we know What/Why/When/How with our software for future. and my post above was simply our conclusions after an analysis of the data
You lads can do as you wish, I was just covering my arse when the "pier 1 is not doing much" statement happens lol
2 points
1, Prioritisation I think should be used as it will help
2, May I also make another suggestion, Pier 1 has Astrodon filters 3nm, I think we could tweak the moon avoidance a little for the HA filter to say 80º max distance from the moon. Please let me know your thought on this
Again thoughts please folks ?
Steve
You lads can do as you wish, I was just covering my arse when the "pier 1 is not doing much" statement happens lol
2 points
1, Prioritisation I think should be used as it will help
2, May I also make another suggestion, Pier 1 has Astrodon filters 3nm, I think we could tweak the moon avoidance a little for the HA filter to say 80º max distance from the moon. Please let me know your thought on this
Again thoughts please folks ?
Steve
Please ignore my dylexia wherever possible, just be thankful I can control my Tourettes ;)
Things to do, so little time!
Steve
Roboscopes Tea Boy
"Now to pier 14. Wish I hadn't started this!
There is a risk that some jobs, particularly those for additional data,
may not be collected by the time of the visit. So, do we need to list
those which fall into that category and others where the current set up
is better suited to."
I forgot to answer this question last night Ray, P14 swap will not be happening until all the parts have arrived for the new Pier-11 widefield and the E160 Pier-15. if we are very lucky it will be November but if I am being honest that date is sliding as I am awaiting some special adapters that are taking longer than expected so it this rate it may actually be early December (10th December onwards) before we do this trip.
It's a little fluid is all
Steve
There is a risk that some jobs, particularly those for additional data,
may not be collected by the time of the visit. So, do we need to list
those which fall into that category and others where the current set up
is better suited to."
I forgot to answer this question last night Ray, P14 swap will not be happening until all the parts have arrived for the new Pier-11 widefield and the E160 Pier-15. if we are very lucky it will be November but if I am being honest that date is sliding as I am awaiting some special adapters that are taking longer than expected so it this rate it may actually be early December (10th December onwards) before we do this trip.
It's a little fluid is all
Steve
Please ignore my dylexia wherever possible, just be thankful I can control my Tourettes ;)
Things to do, so little time!
Steve
Roboscopes Tea Boy
0
Undo
Votes
It's pita again Steve.
Would you have an explanation for the imaging carried out last night on pier 1. I've only looked at that pier so far by the way.
Below should be the queue for pier 1 for yesterday and today. You will notice that all the time last night was given to a broadband image, submission 1292. This is a very recent submission only made on the 6th and appears to have been given the entire night to the detriment of many prior broadband jobs. If it is not prioritising earlier jobs then that may give rise to a situation where a target sets before being entirely captured, despite having been submitted many weeks earlier.
If indeed that is the case then I feel that sort of makes a nonsense of our trying to submit jobs to keep the pier busy, particularly if submitting them results in that scenario arising.
Cheers,
Ray
Would you have an explanation for the imaging carried out last night on pier 1. I've only looked at that pier so far by the way.
Below should be the queue for pier 1 for yesterday and today. You will notice that all the time last night was given to a broadband image, submission 1292. This is a very recent submission only made on the 6th and appears to have been given the entire night to the detriment of many prior broadband jobs. If it is not prioritising earlier jobs then that may give rise to a situation where a target sets before being entirely captured, despite having been submitted many weeks earlier.
If indeed that is the case then I feel that sort of makes a nonsense of our trying to submit jobs to keep the pier busy, particularly if submitting them results in that scenario arising.
Cheers,
Ray
It's pita again Steve.
Would you have an explanation for the imaging carried out last night on pier 1. I've only looked at that pier so far by the way.
Below should be the queue for pier 1 for yesterday and today. You will notice that all the time last night was given to a broadband image, submission 1292. This is a very recent submission only made on the 6th and appears to have been given the entire night to the detriment of many prior broadband jobs. If it is not prioritising earlier jobs then that may give rise to a situation where a target sets before being entirely captured, despite having been submitted many weeks earlier.
If indeed that is the case then I feel that sort of makes a nonsense of our trying to submit jobs to keep the pier busy, particularly if submitting them results in that scenario arising.
Cheers,
Ray
[attachment]Screenshot_20211108-112607.jpg[/attachment]
[attachment]Screenshot_20211108-114749.jpg[/attachment]
Hi Ray
I will have a proper look tomorrow and come back to you but just a quick thought
All the new objects are now using the new priority system for Lum, colour and Narrowband and this could have had an impact and none of the rest have any priorities in
Also the camera did have some strange USB download issues for a few nights behind the scenes. The group are not privy to this in the morning download section. So it is way down on productivity the last few days because of this as we had a good few hours worth of data fail.
However this appears to have now been solved with a driver roll back
Steve
Please ignore my dylexia wherever possible, just be thankful I can control my Tourettes ;)
Things to do, so little time!
Steve
Roboscopes Tea Boy
It's pita again Steve.
Would you have an explanation for the imaging carried out last night on pier 1. I've only looked at that pier so far by the way.
Below should be the queue for pier 1 for yesterday and today. You will notice that all the time last night was given to a broadband image, submission 1292. This is a very recent submission only made on the 6th and appears to have been given the entire night to the detriment of many prior broadband jobs. If it is not prioritising earlier jobs then that may give rise to a situation where a target sets before being entirely captured, despite having been submitted many weeks earlier.
If indeed that is the case then I feel that sort of makes a nonsense of our trying to submit jobs to keep the pier busy, particularly if submitting them results in that scenario arising.
Cheers,
Ray
[attachment]Screenshot_20211108-112607.jpg[/attachment]
[attachment]Screenshot_20211108-114749.jpg[/attachment]
Ray
I was keeping an eye on things as I said I would and the new priority system is working, however its had the knock on effect off pushing some older jobs down the queue as they have no prioritisation settings. To that end I have just done the super long winded job of altering all the older jobs up to the new prioritisation standards.
However the only thing I have not the time to do with older jobs is alter the HA moon parameters to the new settings as I have to delve to deep in the system and the law of diminishing returns on my time starts to apply lol
I will keep a close eye the next few nights to see what happens, as I am a one change at a time sort of a person rather than big en-mass changes!
Steve
Please ignore my dylexia wherever possible, just be thankful I can control my Tourettes ;)
Things to do, so little time!
Steve
Roboscopes Tea Boy
0
Undo
Votes
Steve,
I really appreciate you looking into this and what you have so far done. Just let it run to its conclusion without any further intervention, since we'll soon reach the point where the new priority system will apply to all the jobs and result in them all having a more equitable prioritisation. You certainly have explained very clearly what had been happening, so thanks.
Cheers,
Ray
I really appreciate you looking into this and what you have so far done.
Just let it run to its conclusion without any further intervention, since we'll soon reach the point where the new priority system will apply to all the jobs and result in them all having a more equitable prioritisation. You certainly have explained very clearly what had been happening, so thanks. Cheers,
Ray
Ray
Roboscopes Guinea Pig
It will be interesting to see how it unfolds over the coming month as we go into the next new moon...
None the less we are still having a few guiding fails on this pier I need to keep an eye on but I do suspect its because we are imaging at peak levels of 0.8/0.9"/PX for our location so as good as the mount is it cannot override local atmospherics. its one of the good reasons for shorter subs and more of them as you don't have to guide. Behind the scenes I am sorting this however and I will let you know how it goes
Steve
None the less we are still having a few guiding fails on this pier I need to keep an eye on but I do suspect its because we are imaging at peak levels of 0.8/0.9"/PX for our location so as good as the mount is it cannot override local atmospherics. its one of the good reasons for shorter subs and more of them as you don't have to guide. Behind the scenes I am sorting this however and I will let you know how it goes
Steve
Please ignore my dylexia wherever possible, just be thankful I can control my Tourettes ;)
Things to do, so little time!
Steve
Roboscopes Tea Boy
0
Undo
Votes
Steve,
You need this as much as a hole in the head right now, however......
In my humble opinion a major omission in the queue priority criteria and therefore not being taken into consideration, is the submission date. This is now starting to show. A secondary one is where a request is made to lower the elevation limit. Such targets at a lower elevation will have shorter periods of time when they can be imaged. A couple of examples on pier 1, the one I see mainly affected due partly to the number of submissions.
First a low elevation example.
Ngc253 Sculptor galaxy was submitted August 13th and I believe could have started to be captured towards dawn at that time. It never rises very much, maximum 26°, and will be setting at the start of next month. So far nothing has been captured and looks likely not to be completed this time around. That's a shame.
The next example is where the submission date is not being used, M81 submitted back on July 31st. At that time this was not visible, however, a recommendation by you that we enter 'future' objects to keep the piers busy was followed by the person. If you look at the job list today on pier 1 the neighbouring object M82 submitted November 25th has started to be captured.
I am not requesting that you do something about those 2 instances or any others , that is up to whoever made each one to do so if they want to.
Those 2 shown were standout examples but I see a few more that have been waiting for a long time. Maybe people ought to check on Telescopius to see if any of their earlier jobs may need prioritising.
I'm assuming that the solution right now is for manual intervention on your part if you receive a notification to that effect for a particular job. In future it needs to be up to each of us to keep an eye on our submissions and notify you accordingly.
Caveat, it takes quite a while to look into things like this, so I may have got something wrong.
Cheers,
Ray
You need this as much as a hole in the head right now, however......
In my humble opinion a major omission in the queue priority criteria and therefore not being taken into consideration, is the submission date. This is now starting to show. A secondary one is where a request is made to lower the elevation limit. Such targets at a lower elevation will have shorter periods of time when they can be imaged. A couple of examples on pier 1, the one I see mainly affected due partly to the number of submissions.
First a low elevation example.
Ngc253 Sculptor galaxy was submitted August 13th and I believe could have started to be captured towards dawn at that time. It never rises very much, maximum 26°, and will be setting at the start of next month. So far nothing has been captured and looks likely not to be completed this time around. That's a shame.
The next example is where the submission date is not being used, M81 submitted back on July 31st. At that time this was not visible, however, a recommendation by you that we enter 'future' objects to keep the piers busy was followed by the person. If you look at the job list today on pier 1 the neighbouring object M82 submitted November 25th has started to be captured.
I am not requesting that you do something about those 2 instances or any others , that is up to whoever made each one to do so if they want to.
Those 2 shown were standout examples but I see a few more that have been waiting for a long time. Maybe people ought to check on Telescopius to see if any of their earlier jobs may need prioritising.
I'm assuming that the solution right now is for manual intervention on your part if you receive a notification to that effect for a particular job. In future it needs to be up to each of us to keep an eye on our submissions and notify you accordingly.
Caveat, it takes quite a while to look into things like this, so I may have got something wrong.
Cheers,
Ray
Ray
Roboscopes Guinea Pig
Hi Ray
I am out and about for a day or two so i will take a look then
It may be because they are so low and if you remember when we put the new system in place with priorities we said in order for it to work it need to be kept to strictly otherwise there was no point in doing a priority based system if we threw it out the window when it suits members. as a month later members moan why normal jobs have not been done and finished LOL
see the link to the sticky at the top of the page below which set out the new system some time ago after much discussion
https://roboscopes.com/index.php/forum/pier-5-objects-to-image-6/865-how-our-scheduling-software-works
However as i said I will take a look, off to Wales now to do deliveries in this nasty storm
You have my mobile if you need me....
Steve
I am out and about for a day or two so i will take a look then
It may be because they are so low and if you remember when we put the new system in place with priorities we said in order for it to work it need to be kept to strictly otherwise there was no point in doing a priority based system if we threw it out the window when it suits members. as a month later members moan why normal jobs have not been done and finished LOL
see the link to the sticky at the top of the page below which set out the new system some time ago after much discussion
https://roboscopes.com/index.php/forum/pier-5-objects-to-image-6/865-how-our-scheduling-software-works
However as i said I will take a look, off to Wales now to do deliveries in this nasty storm
You have my mobile if you need me....
Steve
Please ignore my dylexia wherever possible, just be thankful I can control my Tourettes ;)
Things to do, so little time!
Steve
Roboscopes Tea Boy
- Page :
- 1
There are no replies made for this post yet.
Be one of the first to reply to this post!
Be one of the first to reply to this post!
Proud to use
Resources
Company Details:
Roboscopes
802 Kingsbury Road
Birmingham
B24 9PS
United Kingdom
