Jump to content

Banner.jpg.b83b14cd4142fe10848741bb2a14c66b.jpg

freiform

Members
  • Posts

    193
  • Joined

  • Last visited

Posts posted by freiform

  1. Hi,

    there was a nice thread over at CN comparing different software with one set of raw images. And what dou you know? It was the user, not the software. People were producing amazing results using PI, PS, APP, DSS, GIMP, StarTools and whatnot. Also, different people using these exact tools were producing not so pretty images.

    You have to learn the tools. There's not really a shortcut. If you're put off by the UI, well, stick with it until you get it or use something else. If your're put off by pricing models (as am I for some of the software discussed here), use something else.

    But I cannot stress it enough, better is very subjective here... assuming there is no magic (there is none, although some tools of the AI variety make artefacts and interpretation appear very convincing), and further assuming that the tools in question implement the scientific and mathematical principles properly, it's down to the user interface, presets and/or the time you're willing to invest. I use PI and struggle constantly (but not with preprocessing ;) ). I have access to PS due to my job but mostly refuse to use it. I tried the trial of APP and didn't see how I would benefit, already owning PI, and rather spend the money on something else. 

    Highly subjective, all this. But then again, if someone comes and hands you a silver bullet, it's most likely a quack.

    CS

    Sven

     

     

    • Like 1
  2. You don't need 64-bit. Raspbian (and probably others) support LPAE, which allows to access the full 8GB with a 32-bit system. It is limited to 3 GB / process though, But I doubt that's a practically relevant limitation for 99% of use cases.

     

    Sven

  3. And it is working!

    Using James Chambers benchmarks for the SD (Toshiba M203 32 GB):

      Category                  Test                      Result
    HDParm                    Disk Read                 40.83 MB/s
    HDParm                    Cached Disk Read          40.55 MB/s
    DD                        Disk Write                7.1 MB/s
    FIO                       4k random read            2978 IOPS (11912 KB/s)
    FIO                       4k random write           115 IOPS (461 KB/s)
    IOZone                    4k read                   9293 KB/s
    IOZone                    4k write                  2083 KB/s
    IOZone                    4k random read            7767 KB/s
    IOZone                    4k random write           255 KB/s
    
                              Score: 653

    and for the SSD (Samsung 830 128 GB via StarTech USB3S2SAT3CB):

         Category                  Test                      Result
    HDParm                    Disk Read                 258.46 MB/s
    HDParm                    Cached Disk Read          281.51 MB/s
    DD                        Disk Write                141 MB/s
    FIO                       4k random read            17109 IOPS (68437 KB/s)
    FIO                       4k random write           4303 IOPS (17213 KB/s)
    IOZone                    4k read                   26985 KB/s
    IOZone                    4k write                  20848 KB/s
    IOZone                    4k random read            14347 KB/s
    IOZone                    4k random write           23584 KB/s
    
                              Score: 5969

     

    While those are nothing new, the SD card is gone. Neat!

     

    Sven

  4. This is 65 Minutes of Barnard's Loop (5x300s Ha, 4x300s OIII & SII), shot with an ASI183MM through an Nikkor 35/f1.8 on a SA. I just threw everything together quickly, no noise reduction or sharpening or anything like that. Nothing special, but I like how the Ha regions come out at even modest exposure times.

     

    1589923967653-png.139106

     

    CS

    Sven

    • Like 8
  5. 9 hours ago, kryptonite said:

    I'm interested by your Raspberry Pi comment - what OS is it running? I currently run my imaging from an old laptop, but a Pi sounds cleaner..

    I am running a Linux system and INDI (the backend, like ASCOM for Windows) and KStars/EKOS as planetarium software and integrated solution for hardware control and image acquisition. There are alternatives, though; Cartes du Ceil is also available for Linux and CCDCiel is a capturing software by the same author. For planetary/lucky imaging there is oaCapture and FireCapture, although with limited support for cameras; not every vendor offers a Linux SDK, even less offer ARM binaries (Pi's CPU architecture.).

    If you do not want to start from scratch, there's pre-made images with everything you'll need; as a first stop, I would strongly recommend AstroBerry [1].  Copy the image to an SDCard, boot up the Pi, login to its WiFi Network (or connect it to you existing WiFI or Ethernet) and off you go.

     

    9 hours ago, AntHart said:

    I’ve been looking in to this and while it is possible to ‘get’ windows on a pi it’s certainly a struggle and some say not worth the hassle. I’ve even read that it’s locked to 1gb of ram! 

    It's a tech demo at best. You cannot do anything reasonable with Windows on a RPi. Basically it's Linux or nothing.

     

    All the best

    Sven

     

    [1] https://www.astroberry.io/

  6. I am currently using a Raspberry Pi 4 and have used a RPi3 before that. Both are perfectly adequate to guide, plate solve, image, autofocus and everything else you need for imaging. There are a few drawbacks, though. The internal WiFi can be a bit weak, depending on your infrastructure that might or might not apply. Also when using it with a SD-Card, transfer rates a a bit on the slow side which effectively limits its use for planetary or lucky imaging. But that might be migitated by using a fast storage (i.e. SSD via USB3).

    Anyway, even the smallest NUC currently available, which features a dual-core Celeron CPU, outperforms the Raspberry Pi considerably. Your only issue might be how snappy eveything runs, i.e. Windows itself, start times of apps, etc. But as I can just power my device on when getting ready for imaging, I really don't care about that. Why anyone would hang an i7 with 16 GB RAM on the scope is beyond me, but to each his own, I guess.

    I just ordered a NUC with Pentium J5005 (~155€) as I want to take a look at what happens on the Windows side of things. I am not particularly fond of ASCOM, but N.I.N.A. looks interesting and develops fast and then there's Voyager's siren call of carefree automatic sessions ;)

    TL;DR

    They will all work fine for the basic tasks of equipment control and image aquisition. Maybe stay away from Atom Cpus.

     

    On 23/04/2020 at 09:43, Carbon Brush said:

    When powered on, the internal power supply takes a BIG inrush current. If the battery/powerbank/whatever cannot provide the big inrush without the supply dropping, the NUC may not start.

    It thinks the supply voltage is too low.
    This means you need a battery pack with good pulse current capability, or a mains supply with a very large output capacitor.
    Don't skimp on wire gauge connecting the power supply to the NUC and keep the length as short as possible.

    Great information, thank you!

     

    All the best

    Sven

  7. Hi,

    do you have any recommendations for an OTA for imaging with a focal length of 700 to 900 mm that will play nice with my old Vixen GP? Currently I am using a 72/432 ED refractor but I would like to get closer to some targets.

    I also have a 127/1500 Mak, but with no proper reducers available it's simply too slow. My Vixen R135S Newton is great for visual, but it has a sliding focuser. that probably won't hold hold camera and filterwheel and I doubt that I can fit an coma corrector.

    So a new scope it should be. The current contestants are

    • 130PDS, which only has 650 mm focal length but is only 4 kg at 600 mm length.
    • 150PDS wich has 750 mm but weights in at 5 kg and has a length of 750 mm, adding quite a bit to the lever.
    • 6" RC, with an 0.67x reducer for ~900mm focal length. It weighs 5.5 kg, but might be easier to handle with its 480 mm length. 

    Of course , the 6" RC with reducer will be more than twice the price of the 130 or 150 with coma corrector.

    Any thoughts or recommendations on the matter? Last year I planned on getting a new mount, but found that my GP actually performs rather well and the next tier of mounts is too heavy for my (current) taste.

    Thanks

    Sven

  8. Thank you, that is reassuring.

     

    Another point just came up; I am using a pier with a extension that takes Vixen GP level mounts. Looking at the manuals, the CEM25 might fit, but the CEM40 certainly won't (it's screwed to the tripod's head?) Is there an adapter available for that? I would I have to build one or get a new pier?

     

    Sven

  9. Thank you all for your input!

    while I am still partial to the HEQ5, the CEMs are gaining ground ;). I would go with a non-EC version, as I plan on guiding anyway. While discussing this with my SO the other day, she promted me to eliminate all but two; the HEQ-5 and the CEM40 remained.. and while practically irrelevant (not really ;)) the CEM40 is so much prettier.

    But on a more serious note, while EQMOD is really nice to have, I guess iOptrons drivers will work just as well. As far as I can see they work with INDI as well, so all is good. The one drawback of the CEMs is the missing alternative for EQDirect. I really wouldn't mind getting rid of the handcontroller to reduce the clutter. Since I've been using the GP for AP, the SynScan handset is lying in some forgotten box in a dark corner somewhere, and I haven't missed it once.

    I would really like for manufacturers to update their mounts with modified boards to allow for easy connection using USB or Ethernet. I guess Skywatcher did it with the EQ6-R, it seems to have a USB-A port now.

    The CEM40 also sports a USB-A port, can the handset be ommited and the mount be controlled directly via USB?

    @Icesheet Quite a conundrum. I feel like the HEQ-5 would provide a (relatively) safe harbour and the iOptrons as possible gambles, but with potential treasure to win ;)

    @Adreneline Adrian, how do you percieve build quality and robustness of the little CEM25?

    @Jkulin what aspect of the the CEMs are in another class compared to SW? Handling, performance, build quality?

    @Bukko Gordoen, should you get a chance to test teh CEM25P anytime soon, please report!

    @fwm891 Francis, are the CEMs supposed to be quiet?

    @TerryMcK I would fit the belt mod with the HEQ-5, the EQ6 mounts are too heavy. If I'd go that route, I would probably just get the EQ6-R.

     

    Sven

     

  10. Hi,

    I seem unable to push my Vixen GP reliably beyond 120s exposures with around 5kg load and a focal length of 430mm. Especially the Dec-axis is giving me trouble I cannot seem to get rid with adjusting the worm. As I wanted a second mount anyway to have one for visual while imaging, I am looking for opinions and experiences.

    With weight not being an issue, I would just get an EQ6-R, but I would very much prefer something below around 10kg (head).

    Currently I am considering one of the following mounts:

    - iOptron CEM25P

    - iOptron iEQ30

    - SW HEQ-5

    Stretching it a bit, I could go for an iOptron CEM40, a Vixen SX2 (which would, bare specs, have to make up for it in quality). Getting fancy, I might start looking for a used Losmandy G8, but on that deal I probably would not want to wait.

    As I am currently using a SynScan Kit with the GP, I would feel right at home with EDMOD. Also I think the HEQ-5 is tried and tested, so there should be little to no surprises.

    With the CEMs, I would be able to get the best possible weight/load ratio. But I am a bit cautious due to reported quality control issues with iOptron mounts. If you get a good one, they seem to perform very well. Or so the forums know. The iEQ30 is a more traditional design , but still significantly lighter than the HEQ-5 with rather similar load capacity.

    Any thoughts on this? Any other recommendations? Currently, I am not planning to put anything beyond 5 to 7kg on the mount. I would like to get a relatively painless experience, using mainly INDO/EKOS, on occasion or if something new pops up also ASCOM and the usual suspects with Windows.

     

    Thanks

    Sven

     

     

  11. Also, if considering PS, make sure you are not in any country the US might put an embargo on. As it nearly happened with Venezuela, Adobe might prevent access to the subscribed Software. If you don't want to be dependent on the whims of unstable minds, choose something else.

    My vote goes to PixInsight. Or use the trial of APP. The free route probably would be siril and the gimp.

     

    Sven

  12. Hi,

    I have been fishing for some Information to make sure that I am looking at the right mount ;):

    Would I be correct in assuming that the Pro version is the full featured HEQ-5 with the larger motors and corresponding logic board for full speed (i.e. 800x) slewing? And that it was sold in a SynTrek variant that didn't come with a SynScan handcontroller but a simple controller just for slewing? So getting a HEQ-5 Pro SynTrek would have full GoTo capacity when used with an EQDirect Cable and EQMod?

    Also, is the Pro distinguishable by the presence of a Guide Port, which the non-Pro doesn't have (Besides the sticker on the wedge)?

    Assuming that I am looking at a HEQ-5 Pro SynTrek, what would be a reasonable price to pay for a 7 year old specimen?

    Thanks

    Sven 

  13. Adam, Alex, thanks!

    That is reassuring, I guess. Buf of course the price of the product is (well, mostly, hopefully) reflected in its performance (no pun intended..).

    Out of curiosity; how would a image look like with the filter mounted the wrong way? I probably will have the last few hours of clear skies for teh next few weeks so I'd rather be spending the time capturing actual data  ;)

     

    Sven

  14. Hi,

    my Ha exposures show a relatively pronounced halo around Alnitak. The OIII also shows this but not as heavily and the SII looks fine. Is this regular behaviour for ZWO 31mm filters (new version)?
    Or have I mounted the filter(s?) the wrong way around? I wasn't sure when equipping the wheel because I couldn't quite reproduce the hue given on the ZWO site [1]. And so far I didn't have such bright stars in my captures so the issue did not show.

    Also, I suppose the halo not beeing concentric hints at a tilted sensor or filter?

    Any insight would be greatly appreciated.

    Sven

     

    image.thumb.png.4fa52666fc91b6366050ba9a20ec9dbf.png

    (Single frame, 120s Ha, cropped )

     

    [1]https://astronomy-imaging-camera.com/product/zwo-new-narrowband-31mm-filter

  15. Hi Dan,

    unlike CCDs, binning for CMOS chips is software only. You will loose 2 bits in depth, which might not hurt itself, but the readout noise will increase due to the quantization error. There is no reason to use a CMOS camera's binning modes unless you want to quicken focusing runs or plate solving, etc. If you are heavily oversampled, downsample after acquisition.

    Regarding exposure, there's some useful recommendations for the ASI183 on CN:

    https://www.cloudynights.com/topic/616524-sub-exposure-tables-for-the-zwo-asi183mm-and-qhy183m-and-colour-versions/

    Last night I tried high gain (200) for 120s OIII subs and on first glance the subs look good. With the low read noise at high gain of current CMOS chips and the resulting short exposure times, you can really take the pressure off a small mount ;) .

    Sven

    • Thanks 1
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.