Plate solving stored image problem: Page 2

Plate solving stored image problem

w7ay

AstroDude42 My favorites are the Rosette and the Sadr Region.

In that batch, my favorite would be the monochrome ASI2600, Olli. It was done with the luminous filter from a cheap set of 36mm Antlia LRGB filters, while my Chroma filters were on order. (You do see some effects of aperture vignetting on a couple of bright stars.)

My Chroma filters came is just two weeks ago, too late to beat the rainy season, unfortunately. I got a complete LRGBSHO set, all with the same glass thickness, so there should be no problem switching without refocusing -- thus reducing one of the problem of using monochrome cameras. The cleanliness of the simple monochrome image floored me, and convinced me that the monochrome ASI2600 does actually have enough dynamic range at gain of 100. Lots of shades of gray in that image.

Anyway, even the Bayer images look much better on a 27" 5K LG display (LG made it just for Apple, so it is the same as the panels in the 27" iMacs), than on a 24" 4K LG display (non-Mac version).

The 4K LG display is not too shabby to start with (the top of their non-Apple line, actually), but moving the image over to the 5K display shows just how much dynamic range is missing from the cheaper display. It also means that the images themselves already have more dynamic range than the 24" display.

Chen

AstroDude42

w7ay In that batch, my favorite would be the monochrome ASI2600

I feel very tempted to buy the 2600MM as well. But currently it is out of my budget, as i would have to buy new filters too. And my wife would not be too happy if i buy more astro gear at that price point 😆

w7ay I got a complete LRGBSHO set, all with the same glass thickness, so there should be no problem switching without refocusing.

I hope it turns out as expected. I got some Baader SHO filters in front of my 1600MM Pro – they are all of the same thickness, but their individual focus point is different in the range of a few 100 steps on the EAF. The worst is the SII filter by about 1000 steps off.

w7ay [snip] but moving the image over to the 5K display shows just how much dynamic range is missing from the cheaper display.

Agree, i got a 5k iMac as well and its display is such a joy to use!

w7ay By the way, allow me to dispel one of the myths about auto-guiding and round stars.

Thank you, great explanation! And nice image of comet Swan 🙂

w7ay I don't even use Photoshop nowadays, but switched to Affinity Photos many years ago (Affinity Photos can now read FITS).

I process my images in PixInsight and finish them in Photoshop/Camera RAW (with the help of some filters from Nik Collection and TopazLabs). I tried out Affinity Photo but never really got used to the Persona based workflow. I know it's a powerful tool – and as of late for Astro too. I guess i should try it out again sometime.

Olli

w7ay

AstroDude42 The stars in all images look perfectly round to me.

By the way, allow me to dispel one of the myths about auto-guiding and round stars.

First, recall something from probability theory -- if you have two orthogonal axes x and y, and each one has a Gaussian probability distribution, then the combined distribution has the so-called Jointly Gaussian distribution. If x and y have the same error statistics, the result looks like a Gaussian whose argument is x²+y². Notice that this is perfectly radially symmetric!

So, as long as the errors have zero mean, the result will simply bloat a star. It will not elongate a star if P(x) = P(y) and both have zero means.

So, just because you see round stars does not mean that the guide error is small. It simply means that the guide error is symmetrical.

And sometimes, elongated stars is even desirable :-P

http://www.w7ay.net/site/Images/SWAN.tif

Comet Swan 2006, Vixen 80mm, Canon 20d, manually stacked using Photoshop (I had more patience when I was younger :-).

I don't even use Photoshop nowadays, but switched to Affinity Photos many years ago (Affinity Photos can now read FITS).

Chen

w7ay

AstroDude42 But currently it is out of my budget, as i would have to buy new filters too.

Yeah, rule of thumb is to pay more for filters than you pay for the camera :-).

But the filters can be reused.

their individual focus point is different in the range of a few 100 steps on the EAF

I wonder if that has more to do with the OTA itself instead of the filter thickness; i.e., how "APO" is the telescope. One day, perhaps manufacturers will design optics that are not full APO, but have perfect convergence at OIII and H-alpha wavelengths, ha ha.

Agree, i got a 5k iMac as well and its display is such a joy to use!

I agree that short of professional quality monitors, the iMac 5K are nothing short of stunning.

The LG panels for iMacs may even be further sorted. The one in my 27" LG display is probably from the reject bin :-). I worked closely with the display group at Apple back in the days (many of my patents are on how to make standard LCD displays look better :-).

My spouse uses the 5K iMac. But I don't like integrated solutions, so my primary computer right now is a M1 Mac Mini with a 27" and a 24" display (I need two displays to do Xcode development). My previous system is a 2009 Mac Pro (dual Xeon) with two Apple Cinema Displays. For most things, the little M1 Mini beats out the Xeon based MacPro. Astro Pixel Processor appears to be almost twice as fast on the M1.

Chen

w7ay

AstroDude42 The worst is the SII filter by about 1000 steps off.

I was just thinking about this again, Olli ....

I don't know what your EAF scale is. With my R&P focusers, I see about a 4 µm to 4.2 µm drawtube travel per EAF step. (You can measure it easily by taking 1 thousand or two EAF steps and measure the drawtube travel. It is worth measuring and keeping in your notebook for future references).

If yours is about 4 µm per step (EAF attached to SCTs tend to have many more steps per mm), 1000 EAF steps would represent a whopping 4mm of travel! As they say in German, it is unmöglich!

I assume you are using a Bahtinov mask right? Who knows what the ASIAIR does in autofocus with the halos that are more prevalent with narrow band filters. It could be measuring changes to the halos, instead of the actual Airy point spread function. The Bahtinov works in the diffraction (Fourier) domain, and is always right.

Perhaps you have the typical EAF-to-SCT focuser set up which have EAF scales in the hundreds of EAF step per µm?

Glass thickness is definitely not possible, since it would represent 12mm worth of glass thickness difference for a constant refractive index of 1.5. I also cannot imagine 12mm worth of axial change due to chromatic aberration.

What is your EAF scale?

Perhaps you are measuring with an SCT, or a belt driven helical focuser?

If all the Baader glass have the same thickness and glass refractive index, you should get no focus shift. At the same time, I can't imagine such a large focus shift due to chromatic aberration, either, especially since SII is so close in wavelength to H-alpha. So, it is mysterious why you are seeing a focus shift.

Speaking of the EAF, I have special EAF mounting plates fabricated (Web machine shop in Seattle, Washington that is a subsidiary of a German company; they have a CAD tool for macos - draw it, submit drawing [with money :-)] and back comes a plate after a week).

I couldn't stand the fact that the ZWO mounting has the EAF stick down below the dovetail plate if the R&P focuser is mounted horizontally. This is my plate on the FSQ-85:

Notice the plate with a large hole to accommodate the EAF shaft.

Here is different view of another EAF my FOA-60Q:

Notice that with my Dual Saddle plate, I cannot let the EAF stick below the dovetail.

My EAFs are all "upside down." At this point, between 12V and 5V EAF, I have about 7 of them (perhaps more -- I lost count).

Chen

AstroDude42

w7ay I wonder if that has more to do with the OTA itself instead of the filter thickness; i.e., how "APO" is the telescope. One day, perhaps manufacturers will design optics that are not full APO, but have perfect convergence at OIII and H-alpha wavelengths, ha ha.

I'm sorry, i confused the offsets. Ha and OIII are in a relative short range being about < 100 steps apart. SII is about 400 steps away from Ha. EAF is able to focus them all correctly when imaging in SHO. Side note: SII produces huge halos on bright stars where the others do not, so i guess its quality is worse than that of Ha and OIII and that may be the cause to having a bigger offset.

When i shoot LHaRGB and it comes to Ha, EAF is often not able to focus because the focus point of Ha is too many steps aways from the focus point of the other filters. I guess It is related to the V-curve ASIAir is producing. This is my perception: let's say the focus point for LRGB is around 5,000 absolute steps (just fictious values, i need to measure the actual values), at the beginning of creating the V-curve, ASIAir adds a fixed amount of steps to extend the range in which it might find the focus point for the current filter – let's say it adds 1,000 steps, so it begins at 6,000 absolute steps counting downwards in subsequent tries. If Ha's focus point is at 6,200 absolute steps, it would lie outside the range of the V-curve and the focus process eventually fails. Now that i think of it, i guess it would be better to start with Ha so ASIAir has a bigger chance to find focus for the LRGB filters, but this of course would just solve it for a single autorun session, not for plan mode, when imaging several targets successively.

Btw: all filters are from the same manufacturer and have the same thickness of 2 mm. And i activated the reverse option in the EAF settings in the ASIAir app.

w7ay I don't know what your EAF scale is. [snip] What is your EAF scale?

Unfortunately, me too 🙂 At least at the moment. It's on my todo list to measure it.

w7ay I assume you are using a Bahtinov mask right?

Yes, i do for the initial setup. But during a whole night of imaging, i let ASIAir do its autofocus routine multiple times (usually on filter and temperature change).

w7ay Perhaps you have the typical EAF-to-SCT focuser set up which have EAF scales in the hundreds of EAF step per µm?

I can tell, that on the SpaceCat, EAF has a maximum range of 10,000 steps and it won't let me change this using the ASIAir app. On my SCT i have a different EAF which had a maximum range of 60,000 steps by default.

w7ay Speaking of the EAF, I have special EAF mounting plates fabricated

I was thinking of that too when i was evaluating if i could add a Crayford Focuser to my SCT. For those i looked at, i had to find a special solution as the ZWO plates would not work out of the box. In the end i gave up because the backfocus of the SCT is so tight that it's hardly possible to achieve proper backfoucs with a Crayford in the image train. Today, i'm using the ZWO SCT adapter plate to attach the EAF:

And on the SpaceCat i use the DeepSkyDad solution:

Both are the 12V EAF models. I have one spare 5V EAF but until now i did not use it.

Olli

w7ay

AstroDude42 And on the SpaceCat i use the DeepSkyDad solution:

I use Pavle's ring gear and pulley gear, but have my own mounting scheme with again, a custom EAF plate. This is from my archive:

EDIT: FWIW, the EAF scale is 2.0 µm per EAF step.

Chen

w7ay

AstroDude42

Same kind of idea here (and ring gear also from Pavle) for the Askar ACL200:

The EAF scale (typical with belt drive helical focusers) is a tiny 0.44 µm per EAF step.

Chen

AstroDude42

w7ay Same kind of idea here (and ring gear also from Pavle) for the Askar ACL200:

I like that the small cylinder (is it called a well?) which is attached to the EAF and moves the belt is made of metal. The plastic one in my solution feels so cheap and from time to time i have to tighten the screws. It ruined at least two imaging sessions for me due to falling off because the screws went loose.

Olli

w7ay

AstroDude42 I like that the small cylinder (is it called a well?) which is attached to the EAF and moves the belt is made of metal.

Oh, you can buy the metal pulleys from Amazon; there are plenty of them.

However, Pavle told me that the 3D printed ones have a little less backlash. I used the 3D printed one for the WhiteCat, but a Metal one with the ACL200. I don't really see any difference.

I think Pavle uses 15 teeth pulley (don't have the WhiteCat close by to check, but you can count yours). You can choose a different EAF scale by using a pulley with different number of teeth per revolution, but you may need to buy a different belt if you change the pulley size.

If you want to replace your 3D printed one, make sure you use an MXL pulley (do not use GT2 pulley since the pitch is slightly different -- MXL is 2.032 mm per tooth). Pavle uses 6mm wide belts, so the pulley's length needs to match that to keep the belt from wandering around, and the bore of the pulley should be 5mm to match the EAF shaft diameter.

Chen

AstroDude42

w7ay However, Pavle told me that the 3D printed ones have a little less backlash. I used the 3D printed one for the WhiteCat, but a Metal one with the ACL200. I don't really see any difference.

Ok, then it probably isn't worth the effort. At least with a regular pulley. I would buy a metal pulley immediately that fits perfectly onto the rotating gear coming out of the EAF, taking the flattened part into consideration. I guess that would stick perfectly and reduce backlash as far as possible. @tech@zwo i would buy one if you offer it. 🙂

w7ay If you want to replace your 3D printed one, make sure you use an MXL pulley

Thank you! If i ever replace it, i will take this into regard.

Olli

w7ay

AstroDude42 [zwo] i would buy one if you offer it.

Why???!!! They would just buy the same ones that you can get at Amazon, and then mark up the price to resell to you.

Chen

AstroDude42

w7ay As i wrote, if it fits perfectly on the EAF and ideally has no backlash i would be sold. The important aspect for me would be if it had one screw making it less of a pain to attach it to the EAF and its hole being flat on one side, so it matches the shape of the EAF’s rotating part. This way it can not slip off the EAF.

I tried to scribble it on my phone. Maybe it gets more clear what i mean. Please excuse my drawing „skills“ – every child can draw better than i can. 😆

the green line is the EAF‘s barrel
the red part is the pulley
the yellow part is the screw

The pulley would not have any room to rotate around the EAF‘s barrel.

Olli

w7ay

AstroDude42 Maybe it gets more clear what i mean.

Nope, perfectly clear, Olli :-).

But if you make that inner red D shaped hole a full circle, the yellow screw will still hold the green shaft just as tightly as a D-hole, as long as you do it properly.

To securely fix a pulley (or the flexible coupler), first loosen the grub screw that is opposite to the flat side of the shaft completely. You then tighten the grub screw on the flat side as much as you can, making sure that the coupler or pulley can rotate freely as you tighten that grub screw. This seats the round side of the shaft along the full length of the hole. Only after the grub screw on the flat side of the shaft is completely tighten that you tighten the grub screw on the non-flat side; it is really a security screw -- should not be load bearing.

By doing that, you get the same result as your smiley face diagram :-). If you don't tighten the flat side grub screw first, the two grub screws will be holding your shaft in place, instead of letting most of the shaft rest on the surface of the hole, and the shaft will be floating and wiggling between the two points.

Chen

AstroDude42

w7ay But if you make that inner red D shaped hole a full circle, the yellow screw will still hold the green shaft just as tightly as a D-hole, as long as you do it properly.

I think that is how my current pulley should work. Unfortunately it doesn’t or the manual is misleading. The three screws are placed at 0, 120 and 240 degrees around the pulley. The manual states not to put any screw on the flat side of the EAF‘s barrel. I had at least two imaging sessions where the pulley went loose and the stars went out of focus during an autofocus run because the EAF‘s torque and the belt had more strength than the screws.

w7ay By doing that, you get the same result as your smiley face diagram :-)

Ok, i will give it a try. Hopefully you are right. It was no joy for me to fiddle around with those tiny screws. My blood pressure went into some unhealthy range. 😆

Olli

w7ay

AstroDude42 My blood pressure went into some unhealthy range.

Hey, my doctor prescribed Losartan daily for me :-) :-).

Yeah, loosen the 120º and 240º grub screw so they don't stick into the hole. Tighten the 0º grub screw on the flat surface of the EAF shaft.

Then, slightly tighten the two other grub screws.

One problem with Pavle's pulley is you really can't tighten too much before risk of cracking the pulley. Perhaps get an aluminum one from Amazon. It might help your blood pressure :-).

Chen

AstroDude42

w7ay Yeah, loosen the 120º and 240º grub screw so they don't stick into the hole. Tighten the 0º grub screw on the flat surface of the EAF shaft.

Will do! I don't know why i just followed the manual like a zombie. 😅

w7ay you really can't tighten too much before risk of cracking the pulley.

Yes, that is exactly the same here with mine. The manual states it clearly several times: "after screwed in too strong comes broken" 😆 That's why i thought one out of metal would be more durable.

w7ay It might help your blood pressure :-)

Oh, i try to keep it low. But it's proportional to the amount of technical flaws i experience in my life. 😆

Olli

w7ay

AstroDude42 But it's proportional to the amount of technical flaws i experience in my life.

You'd better stop using the ASIAIR, then :-) :-).

Unfortunately, I cannot give you a useful link to an Amazon part, since the one I bought has been discontinued. However, if Pavle's EAF mount has a way for you to adjust the tension of the belt, the diameter (and thus number of teeth in the pulley) is not that critical. Just get any metal one that has number of teeth within 20% or so. Then recalculate the EAF scale to change your Focuser Step sizes in ASIAIR.

I still have two spare pulleys from Pavle that I did not use, and I labeled the bag as 20T, so that must be his standard pulley. The metal one I used in the ACL200 (and I also have two unused spares) has 15 teeth (15T)

It is probably best to leave the 3D printed one alone and wait until a metal one arrives. If you crack the 3D printed pulley, you may pop a blood vessel :-).

Chen

w7ay

AstroDude42

By the way Olli, here is another ring gear that Pavle 3D printed for me. Sigma 135/1.8 ART lens.

Chen

AstroDude42

w7ay You'd better stop using the ASIAIR, then :-) :-).

No, it isn't that bad. 🙂 I actually like the ASIAir. It's a better experience for me to have it all in one system. I don't like the clutter when having n applications open on Windows to achieve "the same".

w7ay Unfortunately, I cannot give you a useful link to an Amazon part, since the one I bought has been discontinued.

No worries, i guess there are plenty of them available. I just have to search.

w7ay Just get any metal one that has number of teeth within 20% or so.

Isn't it a bit more restricted, since the distance of the pulley's teeth need to be balanced with the belt and the gear on the scope which has a dependency to the belt? I might be wrong though.

w7ay If you crack the 3D printed pulley, you may pop a blood vessel :-)

Haha! Definitely 😃

Olli

w7ay

AstroDude42 Isn't it a bit more restricted, since the distance of the pulley's teeth need to be balanced with the belt and the gear on the scope which has a dependency to the belt?

You might be conflating two parameters. One is the pitch of the belt (mm per tooth) -- the pulley must have the same pitch, and Pavle uses the belts (and gears) with the MXL pitch. So you must get MXL pulley, and not the GT2 (or other) pulleys.

The other independent parameter is the number of tooth in the pulley; and that will determine the gear ratio. I.e., if you have a pulley with 20 teeth, and the ring gear has 120 teeth, then the ratio is 6 turns of the EAF shaft per one turn of the helical focuser. As long as you are not exceeding the torque of the motor (by using a pulley that is too large), all is well, since you simply compensate for the gear ratio by using different EAF step sizes. A 20 tooth pulley is usually denoted by 20T, for example. So, 20T MXL would mean a 20 tooth pulley, with 2.032mm tooth pitch.

However, when you pick a pulley with different number of teeth, it also means that the diameter of the pulley changes. This is where the belt tensioner comes in. The tensioner must have enough range to accommodate the difference in radius of the pulley, if you have different number of teeth than the original.

SteveZ

We only hijacked Olli's original topic because he is the originator of the thread, and wouldn't complain about changing the topic :-).

If you want to veer off the topic, it is probably best to start another thread of your own, because many people will miss your questions on this thread, since 99% of the people are already avoiding the chit-chat :-P.

Olli and I share many common equipment (primarily the RainbowAstro mount). Most of the people on this forum has not even heard of the RainbowAstro :-)

Chen

« Previous Page Next Page »