Camera timing offset adjustment

Anything that doesn't fit into any of the other forums
Eric Smith
Posts: 4
Joined: Thu Jul 24, 2025 3:19 pm

Camera timing offset adjustment

#1

Post by Eric Smith »

Hi Robin,

I know this has been a previously asked question but it looks like no resolution was published.

I use SharpCap for remote asteroid occultation observations via an automated script and other observatory control software. I principally use a planetary camera (ASI462MM) and using a GPS calibrated NEXTA device I have measured a very accurate camera delay value of 17ms per frame. This delay is calculated at the center of the rolling shutter frame and by using a GPS/PPS clock on the PC and NTP V4 I can adjust the system clock to be exactly 17 ms slower than UT. This allows me to have precise timing in the ADV file frame headers and achieve what I need.

However, I would prefer to have this adjustment made at the time of capture using an adjustment value for each camera profile if possible. This would allow me to use multiple camera offsets without needing to alter the NTP config file each time. There is also a new NTP / PPS timing device "Timebox 2" that allows you adjust the system clock. I own one of these devices and found them not to be remote observatory grade and frequently loose lock requiring a full system reboot to restore.

Another reason for requesting this feature is to allow users to have multiple timing sources for backup. This is possible with NTP V4 by specifying a range of NTP servers to backup the local GPS / PPS timing device. The problem I have with this backup configuration is that you can only adjust the local GPS/PPS server in hardware and not the backup NTP time servers.

Just another note to add that In the occultation community there is a growing number of users that are keen to use CMOS cameras due to the latest models having superior sensitivity to analogue cameras, but the timing delay's have proved to be a challenge for user to compensate for.

I would be very happy to help with the development and testing of this feature if you feel you have the time to integrate it in your software. I own a NEXTA calibration device that digitally time stamps the video frames captured and can calculate the capture delays to sub millisecond accuracy.

Many thanks for your consideration in this request.

Best regards
Eric Smith
User avatar
admin
Site Admin
Posts: 17580
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Camera timing offset adjustment

#2

Post by admin »

Hi Eric,

an interesting request - let me fire some questions at you so that I can be sure that I understand it...

* Where do you see the user interface to adjust the offset? I am guessing in amongst the camera controls on the right so that it is available and will be saved in camera profiles.

* What is the expected range and accuracy needed for the offset? Is millisecond accuracy sufficient, or does it need to go sub-millisecond?

* You are presumably aware that the delay for CMOS cameras is likely to vary based on things like any 'USB Speed' or similar setting that affects the camera clock speed and possibly the height of the ROI in use.

In a similar area, I have a sample of an IMX585 camera with GPS functionality here at the moment - that has a rolling shutter rather than the global shutter in the IMX174, but in this case SharpCap can retrieve the row delay from the camera SDK and store that in the headers of captured FITS files so that you can (in theory) work out the timings of any row in an image.

cheers,

Robin
Eric Smith
Posts: 4
Joined: Thu Jul 24, 2025 3:19 pm

Re: Camera timing offset adjustment

#3

Post by Eric Smith »

Hi Robin,

Thank you for the prompt reply and your interest. To answer your questions...
* Where do you see the user interface to adjust the offset? I am guessing in amongst the camera controls on the right so that it is available and will be saved in camera profiles.
Yes that would be the preferred place to store the settings. Each camera setting can effect the delay depending on the ROI, binning and system performance.
* What is the expected range and accuracy needed for the offset? Is millisecond accuracy sufficient, or does it need to go sub-millisecond?
For asteroidal occultations the submitted values are to three decimal places meaning milliseconds would be sufficient. For a PC disciplined clock with GPS/PPS timing applied, the accuracy is around +- 0.5ms, for other devices without PPS timing (NMEA) the accuracy is less due to USB serial timing delays etc. I am unsure there is any device currently available to confidently measure to microsecond values and maintain that accuracy over a few thousand frames. I am sure that the worst clock in the imaging train is the camera's internal clock accuracy and that can change with temperature. Possibly less for a cooled CMOS camera but I doubt it in reality because the cooling is for the sensor and not the onboard clock oscillator.
* You are presumably aware that the delay for CMOS cameras is likely to vary based on things like any 'USB Speed' or similar setting that affects the camera clock speed and possibly the height of the ROI in use.
Using the NEXTA device you can precisely measure this.

The recommended method to measure the delay is to use the shortest exposure you can measure a LED display on the NEXTA device. For global shutters this is the same wherever the display is read on the frame, but for rolling shutters you need to account for the position of the LED display in the frame. The way I calculate this is to position the NEXTA display in the centre of the screen and directly read the computed delay. This works for me because I always image the occulted star in the centre of the frame.

This is probably the most important factor when considering where to place the offset adjustment. Placing it in the camera profile will allow a user to calculate the delay's for each ROI. With my tests, the delay changes with vertical row sizes, but less with horizontal crops. I assume this is because the camera reads the entire row despite the crop size of the X axis.

The other most noticeable effect on the measured delay is binning. The trade off in binning can actually increase the delay's measured for cameras that do not support hardware binning. This varies depending on the camera performance and is not always intuitive. I have a range of CMOS cameras and have tested most of these with the NEXTA device and found with planetary cameras binning increases the overall delay, but with larger DSO cameras like the ASI533MM Pro, hardware binning mode makes a huge difference in performance. (I am sure you already know this anyway).
In a similar area, I have a sample of an IMX585 camera with GPS functionality here at the moment - that has a rolling shutter rather than the global shutter in the IMX174, but in this case SharpCap can retrieve the row delay from the camera SDK and store that in the headers of captured FITS files so that you can (in theory) work out the timings of any row in an image.
Interesting! I am unaware of a GPS Rolling shutter camera available on the market. The IMX585 is a terrific sensor and would be a nice camera to use for occultations. Having the row read times is a great addition to being able to calculate the exact timing depending of row position. If this camera has the GPS timestamp synchronised to the first row and the user enters the row the occulted star is being displayed, this would allow for very accurate time stamping. This of course is dependent on the user knowing the row position the star before recording the event, and this can change over time.

For owners of the the global shutter cameras like the 174MM (I have access to an early cooled version of this camera), simply knowing the time offset of the USB download and other small system delays is all you need.

All of this is never going to replace the GPS/PPS time inserted values for accuracy like the DVTI or ASI174 GPS cameras, but for anyone entering into the occultation detection area using their existing cameras, this would be a really useful way to bring the timestamps recorded by SharpCap much closer to the realtime the event occurred.

Just a final note to say that any user induced timing delays should be added to the camera settings file to ensure the user doesn't forget the timing was adjusted. If this was in the FITS header it would also be very useful so the user can have a permanent record of user delays. I use the ADV format for recording occultations due to it's superior performance and smaller file sizes, I am not aware if it possible to add this into a user retrievable format, but I guess you would know this if it is possible.

Many thanks again and I hope this has helped you to understand in a little more detail how I use SharpCap for occultation work.
User avatar
admin
Site Admin
Posts: 17580
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Camera timing offset adjustment

#4

Post by admin »

Hi,

thanks for the comprehensive reply - that all makes sense and I think I can build something based on that. I think I will probably put something in the main SharpCap settings to allow this to only be enabled by those who need it - otherwise it's just another control in the list that almost everyone will either ignore or be confused by :)

At the moment, the GPS equipped IMX585 is not in production yet - I think the manufacturers developed it as a technology test and are deciding whether to productize it. I will mention to them that the idea of such a product was interesting to the occultation community.

One last thing... would a range of +/- 1s in total be sufficient for the offset?

cheers,

Robin
Eric Smith
Posts: 4
Joined: Thu Jul 24, 2025 3:19 pm

Re: Camera timing offset adjustment

#5

Post by Eric Smith »

Hi Robin,

Many thanks for the positive response to my request. It will make a huge difference to the way we and others use CMOS cameras for high speed video recordings.

If placing it in the main settings section is the best way for you to implement the changes, then that's fine. I would imagine +/- 1s range would suffice for most uses in occultation / planetary imaging. If the system inserted more than a second in downloading a frame I wouldn't imagine it being a suitable solution for occultation or high speed imaging anyway.

A couple of quick questions if I may.

Will you be able to add the offset info in the camera settings file and/or FIT's header?

Finally, if the settings are going to be globally applied to the capture logic based on the master settings, would this setting be available to change in the sequences we use to control the capture based on the camera we choose to connect?

Thank you again for this new addition, I will pass on the good news to the occultation group I am a member of.

Eric and Deborah Smith, Gergal, Spain
procyon12
Posts: 267
Joined: Tue Jan 14, 2020 11:32 am

Re: Camera timing offset adjustment

#6

Post by procyon12 »

Hi,

This is a very interesting approach.

From SODIS's perspective, however, I am a little concerned about whether users will understand and use the tool correctly. This should be taken into account as far as possible during implementation. I also would therefore like to see the offset values applied appear in the camera settings file and also in the metadata of the commonly used recording formats.

Cheers

Christian Weber
TimHaymes
Posts: 65
Joined: Sun May 28, 2017 8:11 am

Re: Camera timing offset adjustment

#7

Post by TimHaymes »

Hi Robin,
Agreed, an interesting development.

You wrote that a "585-GPS" is/might be in development. Do you have any more info? I presume you may have the 585 colour (i havnt seen the mono). What sort of tests have you done (if any) regarding a comparison with IMX174 (e.g. sensitivity).

I dont have the 585mc but its in my basket.

Thanks very much
Tim
(Abingdon) with QHY174mGPS
User avatar
admin
Site Admin
Posts: 17580
Joined: Sat Feb 11, 2017 3:52 pm
Location: Vale of the White Horse, UK
Contact:

Re: Camera timing offset adjustment

#8

Post by admin »

Hi folks,

thanks for the feedback. To clarify...

* There is likely to be an on/off switch in the main settings - without turning this on, the new 'timestamp offset' control will never show, so it avoids confusing the majority of users

* The new control will be in the camera settings on the right, so it will get included into profile save/load operations, etc.

* I will add the value of this control to both the camera settings file and to FITS and ADV headers. TIFF will also get the info if the 'Astrotiff' option is on. SER, AVI, PNG, etc do not receive rich header information, so in those cases the camera settings file will need to be relied on

* I am wondering if a notification pop up (orange bar underneath the menu) might be a good idea when this control is set via loading a profile, just to remind you that a saved value of this has been loaded

* To try to minimize confusion I will includ a relatively length tooltip on the control (or maybe text in the settings) to clarify that 'A positive value will make the recorded times later by the specified amount, use this if the true measured time of the frame is *later* than your PC time. A negative value will do the opposite and make the recorded times earlier - use this if the true frame time is *earlier* than your PC time'.

* On the subject of the IMX585, I haven't done a sensitivity test relative to the IMX174, but I suspect that the IMX585 will win on sensitivity and read noise, but loses out because of the smaller pixel size and rolling shutter. Both of the negatives can be dealt with (binning and row timing offsets). I believe that they use a similar generation of sensor technology to the IMX533 and 571, both of which are excellent sensors.

cheers,

Robin
procyon12
Posts: 267
Joined: Tue Jan 14, 2020 11:32 am

Re: Camera timing offset adjustment

#9

Post by procyon12 »

Hi all,

Robin, you have thought everything through very well (as we know you do) and have very good ideas for implementation. In this respect, my concerns have been taken into account. I would very much welcome a notification pop-up.

Many thanks

Christian
Eric Smith
Posts: 4
Joined: Thu Jul 24, 2025 3:19 pm

Re: Camera timing offset adjustment

#10

Post by Eric Smith »

Hi Robin,
I agree that it would be useful to have a message in orange at the bottom to indicate the feature is turned on, thank you.

Thanks for the full clarification. It sounds like a perfect solution.

Eric.
Post Reply