Hi everyone, This post is for the members to discuss and come back to me with a decision :)
As we have discussed many times before the reducer on Pier-14 is struggling with the 6200's 3.76um pixels when combined with the full frame 35mm sensor. This combined with the tilt inherent in all larger ZWO cameras has led to funky corner stars and obvious colour fringing.
On the next visit we intend to deal with this:
first off we have a new jig for ZWO cameras to improve sensor tilt.
Secondly this leaves us with two options which I will outline below with merits etc.
Option-1
Option-2
I have attached an image below showing fields of view with both options, you will see that option 2 has a slightly smaller FOV than option 1 but it is minimal. I will not lie and say I don't already personally believe option 2 is far better with only a minimal compromise on FOV whilst maintaining good star shapes and a fast F/3.6 telescope :)
If anybody needs it, I can share a data file from Pier 6 which has the same sensor albeit OSC fitted as option-2. This sensor has not been jig flattened as yet as we had not built the jig when it was installed. You can then make an informed decision with what I hope is the data evidence, FOV image and the other points I have made here.
I have added a poll below to make life easier for me tracking what you all want.
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
Technically I believe we have the same image quality between these two. F5 vs F3.6 and 62MP vs 26MP, when resize/resample and print at the same size, the image quality/noise level will be essentially the same at the same imaging parameters.
I have no particular preference, but the F3.6 image quality is really bad on the 6200 even when I heavily cropped so I am hesitant to keep that piece
I have voted option 1, but I think it is finely balanced.
Both are back-illuminated chips with excellent performance. The 6200 is a mighty beast, but I have been managing OK with the large files. It has not caused me too much pain. Pier 1 has smaller files with the 2600, but is binned 2x2. Native it should have file sizes about 32MB. Still much smaller than the 6200 at 120MB, but substantial nevertheless.
Our original thought was to stick with the 6200 and remove the reducer, but if anyone thinks that is really not a good idea, I am happy to try the 2600.
Using a terrestrial photography calculator I find a 200 sec exposure at f3.5 (nearest I could get to 3.6) is equivalent to a 400 sec exposure at f5. So a substantial difference in acquisition time.
I have voted option 1, but I think it is finely balanced.
Both are back-illuminated chips with excellent performance. The 6200 is a mighty beast, but I have been managing OK with the large files. It has not caused me too much pain. Pier 1 has smaller files with the 2600, but is binned 2x2. Native it should have file sizes about 32MB. Still much smaller than the 6200 at 120MB, but substantial nevertheless.
Our original thought was to stick with the 6200 and remove the reducer, but if anyone thinks that is really not a good idea, I am happy to try the 2600.
Using a terrestrial photography calculator I find a 200 sec exposure at f3.5 (nearest I could get to 3.6) is equivalent to a 400 sec exposure at f5. So a substantial difference in acquisition time.
Pier 1 has a much smaller 1600 4/3rds sensor, the 2600 is fitted to piers-3,4,5,6,9
HTH
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
I have voted option 1, but I think it is finely balanced.
Both are back-illuminated chips with excellent performance. The 6200 is a mighty beast, but I have been managing OK with the large files. It has not caused me too much pain. Pier 1 has smaller files with the 2600, but is binned 2x2. Native it should have file sizes about 32MB. Still much smaller than the 6200 at 120MB, but substantial nevertheless.
Our original thought was to stick with the 6200 and remove the reducer, but if anyone thinks that is really not a good idea, I am happy to try the 2600.
Using a terrestrial photography calculator I find a 200 sec exposure at f3.5 (nearest I could get to 3.6) is equivalent to a 400 sec exposure at f5. So a substantial difference in acquisition time.
However, 6200 has 2.25 times more pixels than 2600 with the same pixel quality. If you consider that, at the same print size, these two are almost equivalent in nose level with the same integration time at F5 and F3.6, respectively.
first off, whatever the users decide as a group is how we will go, so if option 1 is the most popular then that's what Roboscopes will do :)
However if I explain my own personal and I hope logical thought process on the matter and why I personally have a preference as an individual user rather than as "Roboscopes"
I suppose I am trying to say is at this stage my personal vote went on Option-2 as I am more of a bird in the hand rather than 2 in the bush kinda guy. So something I can see that works (2600) is far better than something that is unproven (6200)
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
Good afternoon,
The original discussion of removing the reducer was in part due to the galaxy season when the vast majority of targets are fairly small. I remember seeing requests from Steve to keep submitting jobs to keep the scope imaging. From what I recall most of what came back was rather disappointing with little detail, something important when imaging galaxies. That is why the idea of removing the reducer first arose. By doing so there would be a gain, admittedly not enormous, and more importantly after galaxy season the absence of weird star shapes. The downside would be a reduced fov something that both presented options will still result in.
The increase in resolution from option 1 would effectively result in almost a doubling, in area, of a target compared to now.
My preference therefore remains with option 1 despite all my complaining over file sizes. :) Sorry Steve.
Anyway to show no bias, I am attaching a couple of screenshots now to show how good the setup on pier 6 using the aps size sensor is. This wes captured within the last week
It does currently have a small problem to.the left hand side, however, Steve has mentioned about slight sensor tilts which they aim to fix. You can see what I mean from one of the screenshots. The main image and the images of the corners, each of which are 768 x 768 pixels, are taken from a full image. That is 6248 x 4176 resampled at 18 percent for the forum. Not a single pixel was harmed in creating this image. :)
Maybe members will be satisfied with these examples and not require Steve to supply anything more.
Cheers,
Ray
Good afternoon,
The original discussion of removing the reducer was in part due to the galaxy season when the vast majority of targets are fairly small. I remember seeing requests from Steve to keep submitting jobs to keep the scope imaging. From what I recall most of what came back was rather disappointing with little detail, something important when imaging galaxies. That is why the idea of removing the reducer first arose. By doing so there would be a gain, admittedly not enormous, and more importantly after galaxy season the absence of weird star shapes. The downside would be a reduced fov something that both presented options will still result in.
The increase in resolution from option 1 would effectively result in almost a doubling, in area, of a target compared to now.
My preference therefore remains with option 1 despite all my complaining over file sizes. :) Sorry Steve.
Anyway to show no bias, I am attaching a couple of screenshots now to show how good the setup on pier 6 using the aps size sensor is. This wes captured within the last week
It does currently have a small problem to.the left hand side, however, Steve has mentioned about slight sensor tilts which they aim to fix. You can see what I mean from one of the screenshots. The main image and the images of the corners, each of which are 768 x 768 pixels, are taken from a full image. That is 6248 x 4176 resampled at 18 percent for the forum. Not a single pixel was harmed in creating this image. :)
Maybe members will be satisfied with these examples and not require Steve to supply anything more.
Cheers,
Ray
No worries Ray, that's the whole point of this thread so we move forward and debate it and vote :)
I agree the first option put forward was removal of the reducer, I offered option 2 after some careful behind the scenes consideration of what the situation at hand was and what the groups possible goals moving forward could be. As I have been told many times before, it's always good to have options :)
PS thanks for sharing a sub from Pier-6 and I suspect that will save me messing around from my end
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
Good afternoon,
The original discussion of removing the reducer was in part due to the galaxy season when the vast majority of targets are fairly small. I remember seeing requests from Steve to keep submitting jobs to keep the scope imaging. From what I recall most of what came back was rather disappointing with little detail, something important when imaging galaxies. That is why the idea of removing the reducer first arose. By doing so there would be a gain, admittedly not enormous, and more importantly after galaxy season the absence of weird star shapes. The downside would be a reduced fov something that both presented options will still result in.
The increase in resolution from option 1 would effectively result in almost a doubling, in area, of a target compared to now.
My preference therefore remains with option 1 despite all my complaining over file sizes. :) Sorry Steve.
Anyway to show no bias, I am attaching a couple of screenshots now to show how good the setup on pier 6 using the aps size sensor is. This wes captured within the last week
It does currently have a small problem to.the left hand side, however, Steve has mentioned about slight sensor tilts which they aim to fix. You can see what I mean from one of the screenshots. The main image and the images of the corners, each of which are 768 x 768 pixels, are taken from a full image. That is 6248 x 4176 resampled at 18 percent for the forum. Not a single pixel was harmed in creating this image. :)
Maybe members will be satisfied with these examples and not require Steve to supply anything more.
Cheers,
Ray
No worries Ray, that's the whole point of this thread so we move forward and debate it and vote :)
I agree the first option put forward was removal of the reducer, I offered option 2 after some careful behind the scenes consideration of what the situation at hand was and what the groups possible goals moving forward could be. As I have been told many times before, it's always good to have options :)
PS thanks for sharing a sub from Pier-6 and I suspect that will save me messing around from my end
Steve
"The increase in resolution from option 1 would effectively result in almost a doubling, in area, of a target compared to now."
Ray I am not entirely sure what you mean by this ? Its Sunday and I am tired so please be gentle :)
EDIT I must be asleep Ray, I think you are saying the object of interest will under either of the proposed systems be almost twice the size it is in the FOV than it is now in the existing FOV ?
Steve
I have put all three true FOV in an attached image below including showing the existing FOV as well as the text to the side showing it in degrees for the smarter amongst us :)
Hi Steve,
Okay to answer your question regarding area.
Take an average size of the full moon at around 1800 arc seconds, 30 arc minutes 0.5 degrees.
Cropping that to the edges and you'll have 1230x1230 pixels option 1, and, 890x890 pixels option 2.Using a print resolution of 240 dpi (dots per inch), these would print as 13.0 x 13.0 cms [5.1 inches] and 9.4 x 9.4 cms [3.7 inches] respectively.
That is 169 square cms and 88 square cms. That's almost double the area, and double the ink when you print it.
Anyway the main reason for this reply is to attach a decent size image showing the corners. I believe I may have reduced it too much before, so hopefully this will be satisfactory.
Cheers,
Ray
Hi Steve,
Okay to answer your question regarding area.
Take an average size of the full moon at around 1800 arc seconds, 30 arc minutes 0.5 degrees.
Cropping that to the edges and you'll have 1230x1230 pixels option 1, and, 890x890 pixels option 2.Using a print resolution of 240 dpi (dots per inch), these would print as 13.0 x 13.0 cms [5.1 inches] and 9.4 x 9.4 cms [3.7 inches] respectively.
That is 169 square cms and 88 square cms. That's almost double the area, and double the ink when you print it.
Anyway the main reason for this reply is to attach a decent size image showing the corners. I believe I may have reduced it too much before, so hopefully this will be satisfactory.
Cheers,
Ray
Thanks for that Ray :)
Ok so we are talking DPI to use a printing vernacular not FOV. I get this, Option-1 due to it having a higher "/PX resolution over option-2 will always have a higher DPI when printing, that to me this was always a given as the higher resolution you go the higher your DPI is and the larger you can print :)
I was taking FOV as the proposed FOV is not so dissimilar with either Option-1 or 2 (Option-2 is a little smaller than option-1)
Whilst the resolution for option-2 would give you the same resolution or DPI as well as speed you are currently enjoying with P-14 presently
So to summarise
Option-1
Option-2
It really depends on where you all want it to go I suppose but I would trade a small amount of FOV anytime over loosing a system that collects data twice as fast as the other.
Any way I have had the stage long enough and I hope made my case for Option-2 so I will leave you all to muse before I get murdered by Ray :)
Cheers
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
OK - having slept on it and thought about everyone's arguments, I would like to change my vote to Option 2.
Why? It's about acquisition time.
I see Pier 14 as a widefield set-up. So I want to use it for larger objects and mosaics. M42, Elephant's Trunk, Rosette, Heart etc all fit nicely into both the native FSQ106 with 6200 and reduced FSQ 106 with the 2600.
If I look at the photo printing size at 300 dpi for the 2600 the full-frame is roughly 21 x 14 inches (53 x 36 cm). I am unlikely to want to print larger than that. So for my preferred targets, both the arcsec/pixel and pixel count are fine.
What you lose out on is what Ray calls 3 for 1 imaging. With the 6200 you can crop down to a small segment of the larger image. So I have the whole Veil complex, the Eastern Veil and the Witches Broom all from one imaging run. Nice, but I have only done it a couple of times since joining.
You will do a better job of smaller targets with the native FSQ 106 and 6200 because you are imaging at ~1.5 arcsec/pixel instead of ~2.4 arcsec/pixel with reduced plus 2600, but I have never been happy with that approach, because to me you are wasting the widefield properties that are the purpose of Pier 14. I would not image M51, M81, or the Pacman nebula on Pier 14, I would go to Pier 1.
So in terms of FOV it is a wash for me for the targets I want to image and the size I am likely to print.
Then we come to the 1-stop difference in acquisition time. That is big in terms of total imaging time, number of targets you can image for the syndicate, and signal to noise ratio vs time imaging.
For my interests, I think Option 2 is better.
old_eyes
OK - having slept on it and thought about everyone's arguments, I would like to change my vote to Option 2.
Why? It's about acquisition time.
I see Pier 14 as a widefield set-up. So I want to use it for larger objects and mosaics. M42, Elephant's Trunk, Rosette, Heart etc all fit nicely into both the native FSQ106 with 6200 and reduced FSQ 106 with the 2600.
If I look at the photo printing size at 300 dpi for the 2600 the full-frame is roughly 21 x 14 inches (53 x 36 cm). I am unlikely to want to print larger than that. So for my preferred targets, both the arcsec/pixel and pixel count are fine.
What you lose out on is what Ray calls 3 for 1 imaging. With the 6200 you can crop down to a small segment of the larger image. So I have the whole Veil complex, the Eastern Veil and the Witches Broom all from one imaging run. Nice, but I have only done it a couple of times since joining.
You will do a better job of smaller targets with the native FSQ 106 and 6200 because you are imaging at ~1.5 arcsec/pixel instead of ~2.4 arcsec/pixel with reduced plus 2600, but I have never been happy with that approach, because to me you are wasting the widefield properties that are the purpose of Pier 14. I would not image M51, M81, or the Pacman nebula on Pier 14, I would go to Pier 1.
So in terms of FOV it is a wash for me for the targets I want to image and the size I am likely to print.
Then we come to the 1-stop difference in acquisition time. That is big in terms of total imaging time, number of targets you can image for the syndicate, and signal to noise ratio vs time imaging.
For my interests, I think Option 2 is better.
old_eyes
No worries, when you get a moment just go in and edit your vote response once you are certain please as it will keep the Poll up-to date
Cheers
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
I fully agree with this: "Then we come to the 1-stop difference in acquisition time. That is big in terms of total imaging time, number of targets you can image for the syndicate, and signal to noise ratio vs time imaging."
So, as old_eyes, I am in favour for option 2.
CS
Manuel
Manuel
Roboscopes General Technical
hi all
i vote for the 1 option, i allways prefer to use a scope at is nominal focal lenght. But i understand that if result without reducer is not optimal , then use a less demanding cam .
but i wondering , that can be use with a 6200MM if a taka is not good :)
hi all
i vote for the 1 option, i allways prefer to use a scope at is nominal focal lenght. But i understand that if result without reducer is not optimal , then use a less demanding cam .
but i wondering , that can be use with a 6200MM if a taka is not good :)
No worries and noted :)
FYI
The FSQ with or without reducer was designed 10+ years ago for 6-9um pixel CCD cameras like the Kodak 11002, with these type cameras its epic. However, Multi 62MP cameras were not around back then.This information we has discovered the hard way using pier-14.
Some of the new Tak scopes like the Eplison 160 etc have new style correctors with much smaller spot sizes over a larger field that will work better with modern large format CMOS MP cameras
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