0.00
#...
·
 |
|---|
|
Hi All, I'm wondering whether it might be possible to compare notes with you on a specific aspect of astrophotography? I use a relatively low-powered laptop to operate my kit. For some time I've been feeling that it isn't really powerful enough to do the job well. But I've been finding it difficult to get to a firm conclusion. I'm thinking about maybe upgrading to a more powerful machine, but am unsure what benefits I would see if I did so, and how powerful a machine would be the best bet. I'm rather lacking in knowledge of how other people's kit performs. I've come up with a way of comparing notes using specific data. Would you be willing to collaborate in a survey, from which useful information could hopefully be collated and then shared with the group? I'd like to compare measurements of the time intervals between frames when shooting a sequence. How many seconds elapse between one exposure completing and the next one starting? I think my setup might be rather slow, and that yours might be faster. But it would be good to have some figures to compare. For information my system consists of a Windows 10 laptop with an i3 processor. My camera is a Canon DSLR and I am using the software package "Backyard EOS" to operate it. The following listing shows details for a few frames out of a set - giving file names and date/time info as recorded in the file tags. dither D1827.0011 B01 Elephant (600mm f8.1 iso800 480sec) 05/10/2022 21:55:13 D1827.0012 B01 Elephant (600mm f8.1 iso800 480sec) 05/10/2022 22:03:47 dither D1827.0013 B01 Elephant (600mm f8.1 iso800 480sec) 05/10/2022 22:12:59 D1827.0014 B01 Elephant (600mm f8.1 iso800 480sec) 05/10/2022 22:21:33 dither D1827.0015 B01 Elephant (600mm f8.1 iso800 480sec) 05/10/2022 22:30:48 D1827.0016 B01 Elephant (600mm f8.1 iso800 480sec) 05/10/2022 22:39:23 dither D1827.0017 B01 Elephant (600mm f8.1 iso800 480sec) 05/10/2022 22:48:45 With this dataset I was dithering every second frame. The exposure time was 480 seconds (8 minutes). If you look at the time difference between each of the pairs of "between dither" frames, they all work out to roughly 8 minutes and 35 seconds. I know that by switching off the downloading of images to the laptop and just saving them on the camera's memory card, the time difference will come down to about 8 minutes and 25 seconds. So of the 35 seconds between frames, roughly 10 seconds appears to be downloading time. The other 25 seconds is taken up with assorted stuff that the software seems to need to do before it is ready to ask the camera to start the next exposure. I wonder how the figures compare with those from your system? Would you be willing to share any information? Solid data if possible, together with an outline spec of your hardware and software. I'll be extremely grateful for any feedback that you might be prepared to offer! Regards, Tim |
|
0.00
#...
·
|
|---|
|
Hi Tim, I'm afraid I dont have a Canon - if it helps I can set up my Nikon and run it through the ASIAIR to compare times if you think it would help? What I would suggest if it helps is (& sorry if its stuff you've already looked at)
Nige |
|
0.00
#...
·
|
|---|
|
Hello Nige, Many thanks for your response. I think I perhaps put rather too much information into my message, which may have muddied things rather. I only mentioned my Canon DSLR as part of giving an outline description of the kit I am using. I didn't intend to focus any attention on the camera. What I am interested in finding out is whether or not a laptop with a more powerful CPU would be likely to work better. When I bought my laptop about 3 years ago, I found it difficult to properly assess what would be required. But reading around the subject I got the impression that quite a lowly machine ought to be fine. I went with that and bought a low powered one, thinking it might have advantages when running off a battery in the middle of a muddy field - something I haven't ended up actually doing very much! So I'm now thinking about whether to splash out on an upgrade. I don't think communication between the computer and the camera is a worry. It is what it is, and I don't think I can do anything to change it. In any case, if I want to I can opt to save the images just on the camera and avoid the 10 seconds of download time. I believe the other 25 seconds is mostly stuff happening on the laptop rather than communication with the camera (hard to prove - but I think it is). And it is those other 25 seconds of inter-shot miscellaneous processing time that I'm thinking might be excessive compared with the experience of others, and might be improvable on. In answer to your questions, the laptop has plenty of space on the hard disc. The disc is mechanical rather than SSD. I don't see evidence of the disc being worked very hard, but I do see signs that the CPU is a bottleneck. There's no antivirus slowing anything down - the laptop is never allowed to see the internet, so I don't have antivirus installed. The laptop is a Dell Inspiron 17-5000. When I call up the Windows information display it shows the CPU as "i3-6060U 2GHz". I'm not keen to get too tied up in the details of the spec though. I'm more interested in comparing people's experiences with broad categories - i3, i5 and i7 CPUs (or the AMD equivalents). So in summary if you are able to provide any information on the "turnaround time" that you get between shots, together with outline details of the computer power and software you're using I'd be very interested to see. At a basic level I'd really like to find out whether my experience is "typical" or whether my kit is doing things much more slowly than yours. I've probably put too many words in my message again, but I hope you can glean some sense of what I'm driving at! Thanks again for responding! Tim |
|
0.00
#...
·
|
|---|
|
Hi Tim, I connected by Nikon up to my 10 year old PC, which has a traditional Mechanical hard drive (not SSD), I controlled it with DigiCam software (as I couldn't find an ASCOM driver to use SharpCap or FireCapture) . When I do this, saving the image to the PC hard disk I get sub-second transfer & save times. I'm assuming that you don't have a delay between shots set in BackyardEOS (Sorry for asking the obvious)? If you're comfortable with techie stuff, ou could look at the logfile and see where the delay is occurring. the logfile should be at "My Pictures\BackyardTEMP\logs" you should see something like 2020-05-22 11:04:18,825 [7] DEBUG - TAKE PICTURE ::: Shutter(MANUAL) Cable(CameraUSB) Av(1.8) Duration(20.0) Iso(1600) Quality(RAW) Mirror(0) SaveTo(PC) 2020-05-22 11:04:42,606 [5] DEBUG - NIKON EVENT --> OnDeviceOnImageReady(1,type[1]) fired 2020-05-22 11:04:59,204 [CameraTakePictureOnMessageRecieved] INFO - Timespan 40.603 The key events are
In the above log file it took 17s to download a The above is taken from a thread with a Nikon user with the same sort of issue Delay - BackyardNIKON - O'Telescope (otelescope.com) Or you could post a query on the O'Telescope website for Backyard EOS Personally, I doubt your transfer time would be improved by a new laptop (unless you're looking for an excuse to buy one ). Happy to chat about this further if you want at the next meeting |
|
0.00
#...
·
|
|---|
|
Hi Nige, Many thanks for doing that experiment. The results sound interesting. Just to clarify though, when you say "sub-second transfer & save times" does that mean the complete time between the camera shutter clicking closed at the end of one exposure and it clicking open to start the next (when a set of exposures are running)? That sounds implausibly fast, so I'm not sure I've understood. What I'm looking for feedback on is that complete time gap. Are you able to elaborate and clarify a little (please)? I'm also slightly unsure about comparing laptop and desktop machines. In my experience, desktop machines, when compared with apparently similarly specced laptops, generally do things a lot faster. Might just be the particular machines I've used, but over several decades I have found that generally to be the case. Ideally I'd like to get feedback from somebody who is running a higher powered laptop. I've had a look at the BackyardEOS logfiles. Never had cause to look for those before. There's a lot of stuff in there. I found it a bit hard to decode, and haven't gleaned much from it other than seeing that the time gap between exposures is what I already knew it to be. That might be showing my lack of detective skills! I'm aware that the camera control and image handling aspects of the program run in separate threads, so the camera can in theory be fired again while the image handling thread is still working. That adds a degree of complication and obfuscation I think when reading the log file. And no I don't have any deliberate delays set - other than a 2 second "mirror delay". Your final thought seems to be at odds with the result reported from the experiment - the suggestion that buying a faster laptop wouldn't help. But the injection of that phrase "transfer time" reinforces my feeling that we're not talking about the same things. No I'm not looking for an excuse to buy a new laptop - that's why I'm asking questions - to hopefulyl only buy one if it will provide definite advantages. The other thing I've got in mind to do is to install NINA on the laptop and see how that compares. It' on my "to do" list. But I have a feeling it's the laptop rather than the software that is slow. As you suggest, I might try asking around various forums/websites. I do have in mind to do that, but thought I would try our group first. I think a chat at the next meeting might be good. Hopefully we can do that. How are you getting on with the guiding/flexure issues we talked about when we last spoke? Any progress? Thanks again for your help and feedback. I owe you a beer! All the best, Tim |
|
0.00
#...
·
|
|---|
|
Hi Tim, I've attached the log from DIGICAM and yes this is transfer of the image from the DSLR to the PC. Granted PC tend to be more powerful but mine is 10 year old Dell (i7 3Ghz with 8Gb RAM. I've attached an extract from the DigiCam Log to demonstrate - its taking 1s exposures with a 1s wait time between exposures. As you will see the time  Then in the interest of completeness, I hooked the Nikon DSLR to my ASIAIR pro, and run a series of image captures, the DSLR images (RAW) are being saved to the MicroSD card. This time I see a delay of 7 seconds (after the 1 sec exposure) to transfer the image to the ASIAIR and write it to its microSD card  So what does this all mean? The answer is I'm not sure or where to go next? Perhaps you're suggestion of installing NINA and seeing what that gives you in terms of performance is a good idea, as is posting the question on the Backyard EOS web page. Let me know if you'd like to get together to try your camera on My ASIAIR or if you want to try with an SSD memory stick (I have a couple) or any combination to try and understand whats happening Nige |
|
0.00
#...
·
|
|---|
|
Hi Tim, disclaimer - I am neither a Canon nor Backyard EOS user, so apologies if my suggestions are wide of the mark. My suggestion would be to figure out what the 25s delay is down to. I don’t think your laptop is the rate determining step, my low powered Asiair barely pauses between frames. I would try a test exposure with the camera disconnected from the laptop to determine if it is the camera taking 25s to report ready, or if it is the laptop doing something. The camera may be doing a post-capture noise or hot pixel reduction step which is triggered by the long exposure time. You may have to burrow through settings to find it, or it may be listed in a detailed exif file. I had to turn a whole bunch of things off on my camera to stop it chewing on my frames after each exposure. Otherwise possibly a guiding / settling timeout? best wishes, Richard |
|
0.00
#...
·
|
|---|
|
Hi Nige, Hi Richard, Thanks very much to you both for your messages. Nige - I'm feeling somewhat guilty, as I didn't start out with the intention of inveigling anybody into doing so much. But I'm very grateful. Your results are quite useful. But I definitely owe you a return favour, or a large drink. Please remind me of that when an opportunity arises! I wasn't familiar with Digicam, but I've looked it up. It looks like a pretty good general purpose camera control program. I'm going to download it and give it a try, out of interest. I'm still rather flabbergasted by how quickly it does things, so I think I need to see it with my own eyes. Your second test, with the Asiair, provides very useful results from my point of view. The fact that you get a seven second turnaround time between shots confirms that my setup, with its 35 second turnaround time, is definitely below par. That's good, as it means I should be able to find a way to do much better. Thank you for all your help and experimentation. It's greatly appreciated. I think you should stop now. It's time for me to get my finger out and do some experiments of my own. Richard. Many thanks for your thoughts. The fact that your Asiair "barely pauses between frames" adds to Nigel's evidence and further confirms that my setup is abnormally slow. With regard to the camera being disconnected from the laptop, it behaves quite "normally" in that state. With short exposures I can (of course) shoot several frames per second if I wish to. For longer exposures "long exposure noise reduction" might kick in - I think that's what you are talking about - but I have very definitely always had that switched off. I can for example shoot a series of long-exposure dark frames using camera and intervalometer, with the delay between frames being virtually non-existent. Guiding/dithering settling issues etc are of course a possible concern, but it's definitely not that (sorry!!). I'm very aware if/when anything like that is going on, and it's not the cause of the ever-present slow operation. Thanks for your thoughts and information. It's good to consider things and rule them out, which you have prompted me to do. Separately I was talking on the phone with Nick T about something else. I took the opportunity to quiz him on the subject. We dug into the timing information contained in a series of dark frames on his computer, and ascertained that he gets a turnaround time of about 8 seconds between shots. He's using a quite powerful laptop, with an i7 CPU, running NINA, and an Atik camera with about 5.5Mpix. So although the kit is different, his timings are similar to yours Nige, and probably not out of line with yours Richard. Time for me to do some experiments. I'm going to install and try NINA, and also try one or two other things, such as running without a USB hub in the system (although I don't believe the hub is the cause of any issues). I'll get back to you with a full report (!!) once I'm done. In the meantime, to complete the picture, if there is anybody out there who is using an i5-based laptop for their astrophotography, I would dearly love to know how well that works. Thanks again. All the best, Tim |
2.61
#...
·
|
|---|
|
Tim, By way of comparison I checked the transfer time from my ASI294MC _ Pro and compared it to when I used a Canon 450D both with a Raspberry Pi Results: (all dithered) Canon 450D to Pi variable delay on 2 min subs up to 120 secs! ASI294 MC to RPI 5 sec delay (USB2 writing to SD card) ASI294 MC to Old DEll PC SSD & USB3 no measurable delay! ie less than 1 sec The ASI294MC Pro camera includes a 256MB DDR3 memory buffer I think most of your delay is due to the camera transfer. I would assume this happens sequentially, ie the camera waits to complete the transfer before taking the next image. This will therefore vary with image size. Max |
|
0.00
#...
·
|
|---|
|
Hi Max, Many thanks for doing those experiments. They add to the picture. Although I find that 120 second delay thing rather worrying!! I've been doing my own experiments. The data transfer from camera to computer seems actually to take only about 5 seconds. The problem appears to be that the software then spends a lot of time "mucking about" before it gets around to asking the camera to start the next exposure. I tried a package called digiCamControl, which Nige pointed me to. That produces an intershot delay of about 5 seconds. As soon as the download has completed it tells the camera to start the next one, then after that it does the "mucking about" thing and about 25 seconds later the downloaded image appears on screen. I've put NINA on my laptop today and got it all working. Mostly seems good, apart from the galling fact that it seems to be even slower than BackyardEOS. Like BYEOS it does a lot of messing around before it asks the camera to fire the shutter again. Not sure what the answer is going to be!! All the best, Tim |
|
2.61
#...
·
|
|---|
|
I think the “mucking about” thing you describe has to do with buffers stacks and all that good stuff. DSLRs were not designed for time critical downloads. However astro cameras are. I found moving on from a DSLR to be a breakthrough step in imaging. (As well as for my bank balance) |
|
0.00
#...
·
|
|---|
|
Bit late to the discussion, but I have a fairly high spec (possibly over the top) set up and have no issues at all with acquisition and transfer times. In my dome I have my kit connected to an NUC mini PC; Win 11; 11th Gen Intel i5; 500Gb SSD main storage. Camera is a ZWO ASI071MC OSC. Image capture is via SGP. To make things more complicated, the images I capture are written to a MS OneDrive folder on the SSD. This means that they are more or less instantaneously uploaded to the cloud via a WiFi connection (and subsequently downloaded to my "imaging" PC indoors). This does not appear to slow SGP image capture down at all. I am also monitoring from indoors via a Remote Desktop connection to the dome NUC PC from my indoor PC. In summary, I suspect the specs of the PC used for image capture are critical for performance. An older device with a standard HDD is likely to struggle. |
|
0.00
#...
·
|
|---|
Bit late to the discussion, but I have a fairly high spec (possibly over the top) set up and have no issues at all with acquisition and transfer times. Thanks for the info. Your setup is all rather different to mine, so can't carry across directly I don't think, but your thoughts on computer specs fit with what others are saying and with what I'm experiencing. Your i5 NUC sounds good - and adds weight to the story that a decent i5 works well. Many thanks for joining in! |
|
0.00
#...
·
|
|---|
|
Hi all, The project seems to be heading towards a successful conclusion. New (actually a refurbished unit) laptop has been obtained, software installed and set up. Timing trials have been performed using three different image capture packages - BackyardEOS, NINA and digiCamControl. All give delays of around 4-5 seconds between exposures. Testing so far has been indoors only, but I've done my best to simulate a real capture session, and am hopeful the results will be replicated out under the stars sometime soon. I'm looking forward to making properly efficient use of those gaps between the clouds! For info the new machine is a Dell Latitude 7490, with an i5-8350U CPU (4 cores, 1.7GHz base speed), 16GB RAM, 512GB SSD, and Windows 11. Whilst all aspects are upgrades from the old machine, I believe the more powerful CPU is the key factor. Many thanks to you all for your feedback and help in getting to this point! Regards, Tim |
