353 replies to this topic

[R Botero]

Where is the 'Wait Until Later Than' command? I can't seem to find it in the sequence editor.

Make sure you have the latest version of SharpCap. It was not there at the beginning of last year when I started playing with it to plan for this eclipse. In fact, it only became available this year to 0.1s accuracy. Last year it was only to nearest second…hence my Sequence based on pressing Run manually. 
In any case, it’s under Misc:  https://docs.sharpca...equence Listing
 

Roberto

Edited by R Botero, 17 February 2024 - 04:41 AM.

[PI_CO100]

Indeed. I noticed that the sequencer was almost there; I asked Robin and he fixed this very quickly. It is great to see a piece of software updated with requests from the users so promptly... Sharpcap is a really top quality software and there is excellent support.

Piero

[JethroXP]

Where is the 'Wait Until Later Than' command? I can't seem to find it in the sequence editor.

In SharpCap Pro (Version 4.1.11680.0), under Sequencer, open the Solar System Sequencer Planner (Experimental).  When it opens, click the Advanced Edit... button at the bottom.  That will open this UI:

 

 

The Wait until later than command is found on the Misc tab as shown above.

 

However if you are using subsecond timings don't edit in the SharpCap editor, it can only display and save whole seconds.  For subsecond timings you just edit in a text editor like NotePad.

This is an example of the syntax for the Wait Unitl Later than Command:

WAIT UNTIL LATER THAN LOCALTIME "1:40:13.1 PM"

Edited by JethroXP, 18 February 2024 - 03:40 PM.

[JethroXP]

After more testing and more scrutiny on the output logs, I've decided to abandon LIVE mode entirely in my script.  I was using it for The Diamond Rings, Bailey's Beads, Chromosphere, and Prominences, while using STILL mode for bracketing the corona.  I'm using STILL mode exclusively now.  This means that I can take far fewer photos, but I now have precise control over when they happen.

 

The problem I began to notice is that in a narrow time window such as 10 seconds for the Diamond Ring, or 6 seconds for Bailey's Beads, I could setup a LIVE capture that would get 20+ images and it would complete within the window, but when I looked at the logs what I found was that the images were captured in a very short time, like 3+ seconds, but the rest of the 10-second window was spent writing them to disk.  So I was getting a lot of images, but not across the full range of the phenomena I was trying to capture, rather they were all front loaded in the capture window while the bulk of the time was writing to disk.

 

So now I'm using STILL mode, and I've got enough time for 5x Diamond Ring, 4x Bailey's Beads, and 5x each for Chromosphere and Prominences.  I'd rather take fewer images but ensure I'm getting the exact phenomena with the proper exposure than just get a bunch of images with very little control over when they fire.

[markmanner]

After more testing and more scrutiny on the output logs, I've decided to abandon LIVE mode entirely in my script.  I was using it for The Diamond Rings, Bailey's Beads, Chromosphere, and Prominences, while using STILL mode for bracketing the corona.  I'm using STILL mode exclusively now.  This means that I can take far fewer photos, but I now have precise control over when they happen.

 

The problem I began to notice is that in a narrow time window such as 10 seconds for the Diamond Ring, or 6 seconds for Bailey's Beads, I could setup a LIVE capture that would get 20+ images and it would complete within the window, but when I looked at the logs what I found was that the images were captured in a very short time, like 3+ seconds, but the rest of the 10-second window was spent writing them to disk.  So I was getting a lot of images, but not across the full range of the phenomena I was trying to capture, rather they were all front loaded in the capture window while the bulk of the time was writing to disk.

 

So now I'm using STILL mode, and I've got enough time for 5x Diamond Ring, 4x Bailey's Beads, and 5x each for Chromosphere and Prominences.  I'd rather take fewer images but ensure I'm getting the exact phenomena with the proper exposure than just get a bunch of images with very little control over when they fire.

Hi, the fact that you saw delays writing to disk I think means that your system is too slow. I'm not getting any write to disk delays and am not seeing what you are. Is your pc fairly fast with SSD, good ram size, etc? 

[JethroXP]

Hi, the fact that you saw delays writing to disk I think means that your system is too slow. I'm not getting any write to disk delays and am not seeing what you are. Is your pc fairly fast with SSD, good ram size, etc? 

This is a MeLE Quieter 4 miniPC.  It has a quad-core Intel N100 CPU, 16GB RAM, and 256GB NVMe.  I get about 450MB/s write speeds to disk.

[R Botero]

Jason

 

That is strange.  I use a MeLE 2 for planetary imaging and in fact, that's the computer I used when I started practising for the eclipse last year using SharpCap.  I will use a faster laptop in Mazatlan but like Mark, I find no delay whatsoever in writing to disk, even with my older MeLE 2.

 

Have you run SharpCap's write to disk test in Preferences?  That may improve your speed.  Also, when running LIVE were you saving individual FITS files or SER sequences?  Note you can save SER in 16bit format which is effectively equivalent.

 

Roberto

Edited by R Botero, 19 February 2024 - 04:20 AM.

[JethroXP]

Jason

 

That is strange.  I use a MeLE 2 for planetary imaging and in fact, that's the computer I used when I started practising for the eclipse last year using SharpCap.  I will use a faster laptop in Mazatlan but like Mark, I find no delay whatsoever in writing to disk, even with my older MeLE 2.

 

Have you run SharpCap's write to disk test in Preferences?  That may improve your speed.  Also, when running LIVE were you saving individual FITS files or SER sequences?  Note you can save SER in 16bit format which is effectively equivalent.

 

Roberto

No, I didn’t even know SharpCap had a disk test, I’ll look for that and give it a try.  I am saving as FITS files, do you think SER is more efficient?  It’s still the same amount of data isn’t it?

[markmanner]

No, I didn’t even know SharpCap had a disk test, I’ll look for that and give it a try.  I am saving as FITS files, do you think SER is more efficient?  It’s still the same amount of data isn’t it?

Same data I believe, but although not tested, I bet faster to do SER than fits.  We can test it and see. After all this, it will be cloudy eclipse day of course, but we will have at least learned something about Sharpcap. 

[JethroXP]

Same data I believe, but although not tested, I bet faster to do SER than fits.  We can test it and see. After all this, it will be cloudy eclipse day of course, but we will have at least learned something about Sharpcap. 

Lol, don't jinx us!  

[R Botero]

No, I didn’t even know SharpCap had a disk test, I’ll look for that and give it a try.  I am saving as FITS files, do you think SER is more efficient?  It’s still the same amount of data isn’t it?

Each of the frames of the SER file will be a 16-bit image.   You can then use planetary software to stack them or export each of the frames as FIT files (or TIFF or PNG or JPEG) for individual processing.

 

One thing to note that you may want to try out:  I installed a separate SSD (an NVMe drive) in my MeLE 2 for image capture (and larger capacity).  I don't know if that runs faster than the native SSD that came with the PC but as I said, the computer is not the bottleneck for me even for planetary imaging.  I thought I'd mention this in case it helps.

 

Roberto

Edited by R Botero, 19 February 2024 - 11:23 AM.

[JethroXP]

Each of the frames of the SER file will be a 16-bit image.   You can then use planetary software to stack them or export each of the frames as FIT files (or TIFF or PNG or JPEG) for individual processing.

 

One thing to note that you may want to try out:  I installed a separate SSD (an NVMe drive) in my MeLE 2 for image capture (and larger capacity).  I don't know if that runs faster than the native SSD that came with the PC but as I said, the computer is not the bottleneck for me even for planetary imaging.  I thought I'd mention this in case it helps.

 

Roberto

The NVMe in my MeLE Quieter 4 is about 2x faster than the native eMMC storage so I'm using that.  Also I found this in the SharpCap help file:

4)      How fast can SharpCap save images from the camera to file

This rate can vary significantly depending on:

·         What format SharpCap is saving to (video formats like SER and AVI are much faster than saving each image to a separate file for PNG or FITS)

·         How fast your SSD or hard disk can write data

·         How full your SSD or hard disk is

 

Now I just need to find the disk speed test, still looking....

[JethroXP]

Ok, found it, it's a button called "Speed Test" on the "Filenames" tab of the SharpCap settings.  I just ran it and it shows 542 Mb/s.  But it does the test in .SER format.  I'll try .SER in my script and see how it goes.

[JethroXP]

First test ran very fast, but realized I had forgotten a "SET STILL MODE" command in the script, so despite explicitly setting the file type to .SER, the output defaulted to .PNG but it was still much faster than FITS (likely because the PNGs are 90% smaller in file size due to compression.  Now I'm wondering, would PNG be a good option?  They are 16-bit, and the compression is lossless.  With the .SER files you are capturing, how do you plan to process them?  Will you stack each one in AutoStakkert, or will you extract individual frames from them with something like PIPP?

Edited by JethroXP, 19 February 2024 - 12:17 PM.

[R Botero]

Jason

 

Glad the suggestions worked.  I don't know yet what I'll do with the files.  I may try to stack them using planetary processing methods, or at least sort them in terms of quality to then extract the best for manual stacking/composing.  It will be a nice problem to have if the weather cooperates and all this planning actually works!

 

Roberto

[JethroXP]

Jason

 

Glad the suggestions worked.  I don't know yet what I'll do with the files.  I may try to stack them using planetary processing methods, or at least sort them in terms of quality to then extract the best for manual stacking/composing.  It will be a nice problem to have if the weather cooperates and all this planning actually works!

 

Roberto

Second test complete, though this time I took the lens cap off and pointed at a light in my ceiling (still cloudy outside) and the PNG files grew in size, but they ranged from about 11MB to about 48MB vs. 50MB for the .FITS files, so overall still faster.  I did notice the issue Dr. Glover mentioned, that the first few images in a series still had the prior exposure settings.  It takes about 3 images for the settings to "catch up".  But if I can capture more images faster, does it matter?  Just when I thought I was close to getting this dialed in I'm now realizing I've got a lot more testing to do.  Thanks!

[markmanner]

First test ran very fast, but realized I had forgotten a "SET STILL MODE" command in the script, so despite explicitly setting the file type to .SER, the output defaulted to .PNG but it was still much faster than FITS (likely because the PNGs are 90% smaller in file size due to compression.  Now I'm wondering, would PNG be a good option?  They are 16-bit, and the compression is lossless.  With the .SER files you are capturing, how do you plan to process them?  Will you stack each one in AutoStakkert, or will you extract individual frames from them with something like PIPP?

Jason, for the annular eclipse last October, I took 225 SER files over the 3 hour period, each one 10 or 15 sec duration, 12fps, 8bit.   I intended to stack each SER to create a video of the entire event, but because of the appearance of annularity, it was not possible to fully automate the stacking for all 225 SERs in Autostakkert, so I picked ones at intervals pre and post annularity that I could do automatically, then manually did the ones around annularity.  For the number we'll have here during totality, 19 or so in my case, it won't be an issue to stack each manually if necessary to get a nice image. Or pick out individual ones if the stream is longer than desired for movement blur. 

[JethroXP]

Third test, this time with .SER files, and it was much faster, but I'm wondering now about how to process these given that for things like the Diamond Ring and Bailey's Beads, these are dynamic events where you want to see the frame-to-frame differences.  If they were stacked it would just blur the changes together.  Are you stacking these, or somehow extracting individual frames from the SER file?

[markmanner]

Third test, this time with .SER files, and it was much faster, but I'm wondering now about how to process these given that for things like the Diamond Ring and Bailey's Beads, these are dynamic events where you want to see the frame-to-frame differences.  If they were stacked it would just blur the changes together.  Are you stacking these, or somehow extracting individual frames from the SER file?

For fast changing features like BB or DR, I suspect that you'll just extract the sharpest individual frame from the SER file, and not stack it. For some of the longer Corona images, maybe stacking will help. The annular eclipse in Ha was fairly well suited to 10 or 15 sec SERs that were stacked. It will be interesting to see. As Roberto said, maybe you can find 10 individual frames sequentially that are good and time-synced ok, that you can stack for a litte more SNR. Maybe all we'll have is the maximum number of individual frames to choose from.

[JethroXP]

Ok, more testing, and I'm back to my previous plan, Still Mode and .FITS files.  So .SER will capture files very fast, but it still takes time to write data to disk.  And because I can only control when the .SER capture starts what ends up happening in a short 10-second window for The Diamond Ring is a bunch of images very close together, like 25 images in about 3 seconds, that then take the remaining 7 seconds to write to disk.  So I'm not getting much change over time.

 

And when I first tried this with PNG files, it was faster, but that was with the lens cap on.  Now that I'm pointing at a light, I believe PNG is actually slower than FITS, even though the files are a bit smaller, probably because some time is needed for the compression.

 

And when I tried PNG or FITS with Live Mode, I found that trying to capture a set of images, change exposure and capture another, always resulted in the first few images of the new set having the exposure value of the old set.

 

So now, with STILL mode and FITS files, I'm only getting 5 images during the 10-second window of the Diamond Ring, but the exposure is locked in, and I'm getting images across the full 10-second window which will show the change over time.  Maybe this is a limitation of my gear, but it's what I've got so this is what I'm doing.

[dizzydlz]

For those planning on capturing a constant live frame stream did you find it working better to capture X amount of frames, change setting and repeat.  Then test to make sure it stops and starts where intended. wouldn't it be ok to script to take exactly X seconds of live frames , change setting and repeat.   If a specific number of seconds of a particular live frame exposure were requested it seems the runtime of the script would be less dependent on what computer hardware it was running on.  

 

A bit of background.  For this eclipse I plan to use my ASI2600MC Duo and Williams Optics FLT 91 on AM5 and a mele 4 with Sharpcap.  Similar equipment to what some others are using in this thread. I have been working on a script in Sharpcap so this thread has been very informative, thank you all for sharing your tests it is so valuable. For the first partial stages I am dividing the C1 to C2 time into 100 periods and taking 5 seconds of live cap for every loop. That loop finishes about 46 seconds before C2. The hard part is deciding what works best for me from 25 sec before C2 to 25 seconds past C3. I see numerous methods in this thread.  1-taking still mode shots at the exact moments or 2- capturing frames constantly from just before C2 to C3. Capturing lots of frames with SER is my first preference if i can manage to start and stop the streams accurately enough. It would align better for the experience I had for  2017 eclipse in Wyoming.  I used a 8" Celestron with reducer and sony a6300 and piggybacked a sony a6000 with 2x and 300mm lens. On both I took 4k video from C1 to C4. I used a properly exposed sun for the manual setting the whole way thru. Took filters off 25 seconds before 25 seconds after and did not make any exposure changes until mid eclipse where i manually moved the iso higher to get some extra corona then moved it back.  I managed to end up getting individual frames of video with a nice diamond ring and chromosphere and flares at the 4k resolution as the movement of the moon changed the exposure right at C2 an C3.

 

I plan to piggyback the sony camera taking 4k video since I know that worked.  For the ASI2600MC after the beginning phase loop the script will start the live frame exposure capture at -20 seconds  C2 and set to take exactly X(to be determined) seconds of live frames for diamond ring, change exposure just before C2, take another X seconds live frames for chromosphere , change exposure , and another X seconds etc for prominences before then switching to bracketing the middle part of eclipse. Still testing to determine if the SEr captures are being captured exacly when the script calls for them. I suppose if there is a time delay the script could be adjusted to account of lag in the actual captures.

[Juzwuz]

I ran into that issue too, I originally wanted to capture a lot more bracketed images but the delay in changing settings is very slow.  This is why testing with your own setup is so critical.  The only way I was able to improve the speed slightly was by ensuring SharpCap was writing to the fast drive on my miniPC.  It’s a MeLE quieter 4, with the C: on eMMC storage and a data drive on NVMe.  SharpCap defaults to the C: for storage, but switching it to write to D: increased the write speed by 2x.  That helped some, but I believe the bottleneck for settings changes is the USB interface itself.

Just wondering if you have tried adjusting the High Speed Frame Cache setting to see if you can get around the bottleneck? 

 

https://docs.sharpca...Memory Settings

[markmanner]

As I noted above, I'm getting about 7.5 FPS full frame with the ASI2600MC.  I guess the SSD I'm using is fast, and I adjusted the ram usage to be 1/2 of my 32GB available ram. 

[BrianOttum]

It is really cool seeing you folks learning how to do something no one has done before.  How would you recommend I get started on my own testing?

 

Objective: manually start the bracketing once totality focus is confirmed  (low bar of just a nice totality shot, easy script, low stress)

ASI2600MC Pro, SharpStar 150mm f/2.8 reflector, ZWO AM-5, new Dell Win11 w/SSD, Kerrville

[glancey]

Those of you planning on imaging the eclipse with your ASI2600MC Pro camera, any reason to use Sharpcap for sequencing rather than ASIStudio app? ASICap has an auto-run function that allows you to create a sequence of captures with different camera settings and intervals of your choosing. No complicated scripting required.

 

And if you prefer FITS vs. AVI or SER file formatting, why not ASIAIR Plus, which also has an auto-run sequencing function. The ASIAIR also has power ports that your ASI2600MC Pro camera needs to run.

Edited by glancey, 21 February 2024 - 12:09 AM.