I would like to use SCpro to record a TNO occultation of a star (https://www.iota-es.de/2002tc302_2021.html) and there are a two problems I would like to solve.
SC has the sensor analysis and the smart histogram and I tried it, but there is no obvious way for me to calculate the photometric SNR of a star for my setup of camera, telescope and sky brightness. SC already knows about these parameters but I don't see the tree in the forest. There are SNR calculators available (https://www.mirametrics.com/sn_calculator_mvn.php) and something like that should be possible in SC. The goal is to find the exposure time to get a snr of eg 10 for the selected star to have the best time resolution for the event.
SC does record a timestamp in the picture and the SER file. From reading the manual and other literature, I believe it is the time SC actually grabs the frame after the recording for SER but for FITS it is start of exposure. So we have two unknowns. The accuracy of the PC clock and its readout and the time offset between taking the exposure and registering it in SC. The Journal for Occultaton Astronomy 2021-1(https://iota-es.de/JOA/JOA2021_1.pdf) published a test of a device called Shelyak Timebox to precisely set time on a Windows PC. There is also this article about SC timestamps (https://iota-es.de/SC-Timestamps_01.pdf). The offset seems to be constant for SC and a specific camera and somewhere between 6ms - 16ms. Is there more knowledge available or some way of measuring this offset without using an expensive or hard to build device. I have a Raspberry PI based Stratum 1 NTP server (50€) and according to the Meinberg ntp client and stats I can keep the pc clock at <2ms for most of the time. If SC could support a measurement or provide measurements for popular cameras, that would be great.
Markus
occultation recording with SharpCap
Forum rules
If you have a problem or question, please check the FAQ to see if it already has an answer : https://www.sharpcap.co.uk/sharpcap-faqs
If you have a problem or question, please check the FAQ to see if it already has an answer : https://www.sharpcap.co.uk/sharpcap-faqs
Re: occultation recording with SharpCap
Hello Markus,
I would like to add to your request for an on-the-fly Signal/Noise ration (SNR) indicator in Sharpcap, as this would be helpful for timing events like this. From experience a SNR of 4 to 5 would be sufficient in stable seeing conditions. The exposure needed will depend on the telescope aperture and camera off course.
I am one of the 100+ observers hoping to record this TNO occultation, and i expect you could get a lot of responses on your question of timing. I use the QHY174m-GPS camera. A system controlled by TIMEBOX or RPi/GPS or NTP/Meinberg would be good solutions as the time errors are small compared to typical exposure times of 50ms and longer. As to the time insertion point for the frame (start-mid-end), this is something you will need to find from documentation or experiment - or answers in this thread hopefully.
I do hope you can get some observations. Although they may not be timed with super accuracy
, as long as you know the likely timing uncertainties involved, just report those.
Tim
I would like to add to your request for an on-the-fly Signal/Noise ration (SNR) indicator in Sharpcap, as this would be helpful for timing events like this. From experience a SNR of 4 to 5 would be sufficient in stable seeing conditions. The exposure needed will depend on the telescope aperture and camera off course.
I am one of the 100+ observers hoping to record this TNO occultation, and i expect you could get a lot of responses on your question of timing. I use the QHY174m-GPS camera. A system controlled by TIMEBOX or RPi/GPS or NTP/Meinberg would be good solutions as the time errors are small compared to typical exposure times of 50ms and longer. As to the time insertion point for the frame (start-mid-end), this is something you will need to find from documentation or experiment - or answers in this thread hopefully.
I do hope you can get some observations. Although they may not be timed with super accuracy
, as long as you know the likely timing uncertainties involved, just report those.Tim
Re: occultation recording with SharpCap
Hello Markus and Tim,
Welcome to this TNO event. I hope I can record it too, my biggest worry is the weather ...
I agree with Tim's points. Markus, don't worry about SNR. The best way is to convey on site between different parameters: sufficient short exposure time to record a possible secondary event (satellite) - but test for dropped frames, long enough exposure to have a clear picture of the target star (that never disappears due e.g. seeing) - but avoid saturation. You should also have some comparison stars in your FOV. Use sensor cooling if your cam allows this. If you have found your settings you can record a short FITS sequence and do photometry, e.g. with Tangra, there you can see the SNR. However, often just before the event clouds (hopefully only high and thin ...) come into your FOV and you have to change the settings ...
You should record in FITS.
"SC-Timestamps_01.pdf" comes from me and it is still true:
Digital video cameras without GPS and QHY174GPS without/lost GPS lock:
The SC timestamp is the frame *end time* (the time, the frame is received by SC).
However, for FITS sequences, SC provides with the header DATE-OBS an estimated frame *start time*.
The time reference is the system time (with all it’s inaccuracies).
Tangra as well as PyMovie await start of exposure timestamps (in Tangra you can also choose mid and end of exposure).
Hope this helps a bit.
Christian
Welcome to this TNO event. I hope I can record it too, my biggest worry is the weather ...
I agree with Tim's points. Markus, don't worry about SNR. The best way is to convey on site between different parameters: sufficient short exposure time to record a possible secondary event (satellite) - but test for dropped frames, long enough exposure to have a clear picture of the target star (that never disappears due e.g. seeing) - but avoid saturation. You should also have some comparison stars in your FOV. Use sensor cooling if your cam allows this. If you have found your settings you can record a short FITS sequence and do photometry, e.g. with Tangra, there you can see the SNR. However, often just before the event clouds (hopefully only high and thin ...) come into your FOV and you have to change the settings ...
You should record in FITS.
"SC-Timestamps_01.pdf" comes from me and it is still true:
Digital video cameras without GPS and QHY174GPS without/lost GPS lock:
The SC timestamp is the frame *end time* (the time, the frame is received by SC).
However, for FITS sequences, SC provides with the header DATE-OBS an estimated frame *start time*.
The time reference is the system time (with all it’s inaccuracies).
Tangra as well as PyMovie await start of exposure timestamps (in Tangra you can also choose mid and end of exposure).
Hope this helps a bit.
Christian
-
admin
- Site Admin
- Posts: 13339
- Joined: Sat Feb 11, 2017 3:52 pm
- Location: Vale of the White Horse, UK
- Contact:
Re: occultation recording with SharpCap
Hi,
you can get the SNR of a single pixel from SharpCap's histogram function if you have done the sensor analysis - like this
1) Display histogram
2) Turn on the 'selection rectangle' option from the toolbar
3) make the selection rectangle a tight box around your star of interest (as small as possible) - this will mean the histogram only shows data from that part of the image
4) hover the mouse over the right hand end of histogram peak - this represents the brightest pixels in the selection area.
5) read off from the hover text the number of electrons, uncertainty in electrons and SNR for a pixel of that brightness
In the text, everything except the pixel count depends on the horizontal position of the mouse cursor on the histogram. The pixel count depends on the vertical position and is the number of pixels represented by that vertical position (not really interesting here).
I know this is not quite the SNR of the whole star flux that you are looking for, but the whole star SNR will be considerably better than the single pixel SNR, so it may still be useful.
As to the timestamps, for all cameras (except QHY GPS cameras) the SER frame timestamps are the time that SharpCap receives the frame data from the camera software. This will be at some point after the end of the frame, but how much the delay is can only be determined experimentally. Certain camera parameters (including exposure, size of capture are and any USB related parameters may affect the delay). For GPS cameras, the SER frame timestamp is the start of the frame as reported by the GPS data.
SharpCap does use the most precise way to get the computer time when capturing the timestamps for non-GPS cameras (https://docs.microsoft.com/en-us/window ... asfiletime). This avoids the 15.6ms granularity of the default time functions on Windows, so at least removes one source of time jitter.
cheers,
Robin
you can get the SNR of a single pixel from SharpCap's histogram function if you have done the sensor analysis - like this
1) Display histogram
2) Turn on the 'selection rectangle' option from the toolbar
3) make the selection rectangle a tight box around your star of interest (as small as possible) - this will mean the histogram only shows data from that part of the image
4) hover the mouse over the right hand end of histogram peak - this represents the brightest pixels in the selection area.
5) read off from the hover text the number of electrons, uncertainty in electrons and SNR for a pixel of that brightness
- Capture.JPG (17.65 KiB) Viewed 1075 times
I know this is not quite the SNR of the whole star flux that you are looking for, but the whole star SNR will be considerably better than the single pixel SNR, so it may still be useful.
As to the timestamps, for all cameras (except QHY GPS cameras) the SER frame timestamps are the time that SharpCap receives the frame data from the camera software. This will be at some point after the end of the frame, but how much the delay is can only be determined experimentally. Certain camera parameters (including exposure, size of capture are and any USB related parameters may affect the delay). For GPS cameras, the SER frame timestamp is the start of the frame as reported by the GPS data.
SharpCap does use the most precise way to get the computer time when capturing the timestamps for non-GPS cameras (https://docs.microsoft.com/en-us/window ... asfiletime). This avoids the 15.6ms granularity of the default time functions on Windows, so at least removes one source of time jitter.
cheers,
Robin
Re: occultation recording with SharpCap
thank you all very much for the hints. Now my biggest worry is the weather, as always 
Will do a short sequence of FITS images before and after the event but record a SER file for the minutes of the predicted occultation to get the best possible time resolution.
btw. a real-time SNR could be added to the seeing monitor. My intent was more in planning for events to quickly evaluate feasible observation parameters. I did the last occultation observations more than 20y ago with my friend Otto Farago+ and analog equipment (Mintron) and one of the first GPS based VTIs...
Markus
Will do a short sequence of FITS images before and after the event but record a SER file for the minutes of the predicted occultation to get the best possible time resolution. btw. a real-time SNR could be added to the seeing monitor. My intent was more in planning for events to quickly evaluate feasible observation parameters. I did the last occultation observations more than 20y ago with my friend Otto Farago+ and analog equipment (Mintron) and one of the first GPS based VTIs...
Markus
Re: occultation recording with SharpCap
Hi Markus,
I strongly recommend recording in FITS instead in SER. SER isn't made for stellar occultation observations but for planetary imaging. With SER you have much less information per frame and - if there is an issue you can loose all data. With FITS sequences you have a lot of useful header information for every single frame. In case of a hardware/software issue you have the chance that not all frames are affected. Regarding the time resolution there should not be a (remarkable) difference. After finding the target reduce the FOV, this will help to enhance the FPS rate.
Cheers
Christian
I strongly recommend recording in FITS instead in SER. SER isn't made for stellar occultation observations but for planetary imaging. With SER you have much less information per frame and - if there is an issue you can loose all data. With FITS sequences you have a lot of useful header information for every single frame. In case of a hardware/software issue you have the chance that not all frames are affected. Regarding the time resolution there should not be a (remarkable) difference. After finding the target reduce the FOV, this will help to enhance the FPS rate.
Cheers
Christian
Re: occultation recording with SharpCap
Hello Gentlemen,
I do dabble in Exoplanets but I only have enough aperture to measure the brightest stars. I have a 71mm f/5.9 refractor. Not ideal.
Last year I invested a good amount of time developing the code for a web app that I call U235-Pro. I designed it to help plan imaging sessions with a monochrome camera. I invite you to give it a test run. There is no cost to sign up.
Here is a screenshot of a project that I just put together to illustrate how you might use it. It tells me that I need an exposure of 30 seconds in order to capture a 16th magnitude star at SNR 9.21 at an altitude of 60 degrees above the horizon with the luminance filter. To achieve a total SNR of 100.92, I need to stack 120 frames. Note that the Altitude, Exposure, and Frame count fields require user input. The remaining fields are calculated:
After you sign up, click on the following link to see the project I created in that screenshot. Click on each of the tabs to see what kind of information I had to supply. You will have to do the same but only once. The properties of the telescope, camera, and target are saved to the database for future use in new projects. I created this project very quickly:
https://u235-pro.com/project/314428930102657617
This is how it works:
1. You tell it the Surface Brightness of the star. It is a function of the star's magnitude and your seeing conditions. I provide a calculator for that. You can read more about it at this link under the heading "Open Star Clusters": https://u235-pro.com/guidelines/best-practices
2. You tell it your Sky Brightness in mags per square arcsec.
3. You tell it the properties of your telescope.
4. You tell it the properties of your camera and the sensor's QE. See here for the method I use: https://u235-pro.com/guidelines/color-balance
The web app costs me nothing to run, but I am on a free tier with the cloud database provider. They used to have a very generous free tier, but then they got greedy. Feel free to use the app, but if we exceed the quota then access is blocked until the beginning of the next month. This has never happened with my current user base, but it could if more people sign up and use it regularly. We can then decide what choices we have.
EDIT: Markus, thanks for the links to IOTA and Mirametrics. I see that the star's normal mag is 11.7V but will drop to 20.4V at the midpoint of the occultation. With my equipment, a 5-second exposure is needed to capture 11.7V at SNR 10. I compared my results against the Mirametrics calculator. We are in total agreement, using an aperture of 10 pixels and a Johnson V filter. This is expected since I am using the Vega System and the flux density in the V-band. The only significant computational difference between my web app and Mirametric's calculator is that I am not asking for a photometric aperture. U235-Pro was designed for imaging, not necessarily for photometry, but in my opinion it can get you in the ballpark.
Brian
I do dabble in Exoplanets but I only have enough aperture to measure the brightest stars. I have a 71mm f/5.9 refractor. Not ideal.
Last year I invested a good amount of time developing the code for a web app that I call U235-Pro. I designed it to help plan imaging sessions with a monochrome camera. I invite you to give it a test run. There is no cost to sign up.
Here is a screenshot of a project that I just put together to illustrate how you might use it. It tells me that I need an exposure of 30 seconds in order to capture a 16th magnitude star at SNR 9.21 at an altitude of 60 degrees above the horizon with the luminance filter. To achieve a total SNR of 100.92, I need to stack 120 frames. Note that the Altitude, Exposure, and Frame count fields require user input. The remaining fields are calculated:
- u235-Pro Occultation.png (45.17 KiB) Viewed 1043 times
https://u235-pro.com/project/314428930102657617
This is how it works:
1. You tell it the Surface Brightness of the star. It is a function of the star's magnitude and your seeing conditions. I provide a calculator for that. You can read more about it at this link under the heading "Open Star Clusters": https://u235-pro.com/guidelines/best-practices
2. You tell it your Sky Brightness in mags per square arcsec.
3. You tell it the properties of your telescope.
4. You tell it the properties of your camera and the sensor's QE. See here for the method I use: https://u235-pro.com/guidelines/color-balance
The web app costs me nothing to run, but I am on a free tier with the cloud database provider. They used to have a very generous free tier, but then they got greedy. Feel free to use the app, but if we exceed the quota then access is blocked until the beginning of the next month. This has never happened with my current user base, but it could if more people sign up and use it regularly. We can then decide what choices we have.
EDIT: Markus, thanks for the links to IOTA and Mirametrics. I see that the star's normal mag is 11.7V but will drop to 20.4V at the midpoint of the occultation. With my equipment, a 5-second exposure is needed to capture 11.7V at SNR 10. I compared my results against the Mirametrics calculator. We are in total agreement, using an aperture of 10 pixels and a Johnson V filter. This is expected since I am using the Vega System and the flux density in the V-band. The only significant computational difference between my web app and Mirametric's calculator is that I am not asking for a photometric aperture. U235-Pro was designed for imaging, not necessarily for photometry, but in my opinion it can get you in the ballpark.
Brian
Re: occultation recording with SharpCap
Hi Brian, thanks for the tool. You need to provide a more detailed description of how this is supposed to work or change the UI design. It took me some time to comprehend that creating a folder does not create a new instance because this is prominently displayed. Finally I detected that I can click on the create project writing to make it happen. No indication that this is an action...
My result is that I need a 0.1s exposure to record the 11.8m star with a SNR of 10 with my equipment. That's what I wanted to know. The total duration is predicted to be 21s and the rule of thumb is that the exposure time should be smaller than 1/10 of the duration.
Markus
My result is that I need a 0.1s exposure to record the 11.8m star with a SNR of 10 with my equipment. That's what I wanted to know. The total duration is predicted to be 21s and the rule of thumb is that the exposure time should be smaller than 1/10 of the duration.
Markus
Re: occultation recording with SharpCap
Hi,
Markus, 100ms @SNR 10 sounds good - may I ask which telescope and camera?
I made a test with 150mm f3.8 (and QHY174GPS) to see the starfiled https://forum.vdsastro.de/viewtopic.php ... 669#p37956.
I think with this equipment I would have to go with about 400ms at 75% gain.
Generally I think one should work with 8" and above.
Christian
Markus, 100ms @SNR 10 sounds good - may I ask which telescope and camera?
I made a test with 150mm f3.8 (and QHY174GPS) to see the starfiled https://forum.vdsastro.de/viewtopic.php ... 669#p37956.
I think with this equipment I would have to go with about 400ms at 75% gain.
Generally I think one should work with 8" and above.
Christian
Re: occultation recording with SharpCap
Markus,
The Mirametrics calculator does exactly what you want to do with regards to planning. U235-Pro isn't designed for photometry, but Mirametrics is. It even has Load/Save buttons so you only need to key in the properties of your telescope, camera, and observatory just once. Furthermore it is a Single Page Application (SPA), meaning that the calculator doesn't need to access its web server to perform a calculation. That's important if you are at a remote site having no Internet. Simply load the page in your laptop or handheld web browser before you leave for the remote site.
I browsed the Mirametrics site. Their products are reasonably priced for what you want to do. It looks like $599 to purchase + $250 annual maintenance. If you can convince Robin to implement similar functionality in SharpCap, then that is a huge savings for you.
Brian
The Mirametrics calculator does exactly what you want to do with regards to planning. U235-Pro isn't designed for photometry, but Mirametrics is. It even has Load/Save buttons so you only need to key in the properties of your telescope, camera, and observatory just once. Furthermore it is a Single Page Application (SPA), meaning that the calculator doesn't need to access its web server to perform a calculation. That's important if you are at a remote site having no Internet. Simply load the page in your laptop or handheld web browser before you leave for the remote site.
I browsed the Mirametrics site. Their products are reasonably priced for what you want to do. It looks like $599 to purchase + $250 annual maintenance. If you can convince Robin to implement similar functionality in SharpCap, then that is a huge savings for you.
Brian