Hi all,

I just joined this group. Last weekend I tried out astrophotography with a non-wide lens for the first time. I went to a relatively dark site for the occasion (I currently live in Tokyo, which is not the darkest place on earth). I'm still learning how to do both the photography and processing parts. Here is my first attempt with the pentax-M 200mm/4, 8 frames. No dark frames or flat frames used, astrotracer on, a total imaging time of about 3 minutes.

I have more shots with a tamron 500mm mirror lens, but I think it does actually not resolve more than the 200mm. I'll confirm that asap.

Please tell me what to do to improve!

PS: in the lower right and left frames something seems to have gone wrong with either the stacking in DSS or with the astrotracer. Does anyone know how to avoid this?

Not bad for a first attempt and much better than my first attempt that I showed. You have the 4 major nebulae in the frame showing up. You may want to increase the saturation as they all look a bit washed out and gray
What f-stop were you using? That lens needs to be run at f/5.6 or so and you are stopped down some (the diffraction spikes on the bright stars)

I assume that this wasn't on an equatorial and that you were using astrotracer as those types of artifacts usually are from that. They come about from the geometric distortion from the lens so that the stars aren't in the exact same relative positions in the frame from one shot to the next. It is usually most pronounced in the corners and edges of the frame which is what is seen here and it is happening in all 4 corners and along the edges just not as pronounced. With an equatorial mount the relative
positions don't change from the distortion which is why I assumed you were using astrotracer. One thing DSS that can help alleviate this is the stacking method. Instead of using just a regular average use kappa-sigma with a sigma of 3 and 5 iterations. This will eliminate outliers which will help, but not eliminate, some of the distortion artifacts.

Since you mention you are in a really bright area and traveled to a darker one I would also suggest getting some form of a light pollution filter. Since you are just starting out and working with camera lenses you don't need to get a special purpose one but a Hoya Red Intensifier filter does wonders. I use one when I am out and have one in the drop in holder on one lens and have a 77mm one that I use with stepdown rings on my other lenses. They aren't too expensive so if you decide you don't want to continue you aren't out a lot of money. Another one to consider is the NiSi Natural Night filter which is a more dedicated light pollution filter but is substantially more expensive. I can't speak to how effective it is as I don't own one but it looks like it should provide better results.

Thanks for the reply! If I remember the lens was at 5.6, but I'm not 100% sure.


Regarding astrotracer causing the effects seen around the corner: wouldnt you expect the stars becoming smears instead of getting multiple copies? Some smearing in the corners can also be seen in the images of course. I had somehow expected the distortion on the 200mm to be negligible, but that assumption was not based on much.

Regarding saturation, that should be easy enough to fix. I guess I have not yet developed a good feel of what the colours actually are.

The filter situation I have to think about. The spot I traveled to was really quite dark, so I think it might not be necessary there. I'd rather instead take many more frames (but from in-camera checking I thought the 200mm results were no good, so I quickly stopped trying). On the other hand, close to where I live the situation regarding light polution is so bad that I think there is no use even trying, filters or not.

Thanks for the advice!

Nice image considering it's only 3 minutes. Dark skies help! As to the doubling of stars towards the corners, I have seen the same thing in Sequator if I forget to enable "reduce distortion effects." I usually select the "lens" setting, which is the minimum correction but if that doesn't work there is "complex." I don't recall if DSS has a similar capability. If not, then give Sequator a try.

Yes and there is some smearing but it is pretty light. However the geometric distortion is always worse in the corners and edges and that 200 is an old design so it isn't as well corrected as modern ones.

No it doesn't. If one wants to throw money at the problem I understand that Astro Pixel Processor (APP) does a really good job correcting things and will even stack images taken with different cameras and lenses successfully. It seems that with astrophotography throwing money at the problem is often a very valid solution to problems.



I don't know where you are exactly in Japan but it looks to be a rather long haul from Tokyo out to a nice dark area. I don't know if you are familar with the bortle scale but it is a good rough way of classifying light pollution. I'm in a bright bortle 8 with a 9 about a mile away, thankfully it isn't huge but there is sky glow all the way to the zenith. As for what can be done in a heavily light polluted area I got this earlier this year from my backyard using a light pollution filter. I didn't spend a lot of time processing the image so there is a lot more that could have been brought out. Also I didn't know that lens all that well at that point so there are some issues that show up. I was shooting wide open plus since it is a long lens with a huge slab of glass it takes a long while to come to temperature and I waited no where near as long as I needed to. I also didn't have a bahtinov mask for that lens at that time so the focus was probably slightly off. Stopping down .7 stops to f/3.5 with that lens cleans up the CA and star bloat while letting it come to temp for a couple of hours solves the coma and I figured that out at a later date when I spent a night testing it. This lens is truly wonderful with very little distortion and I will image things as the drift from one end of the frame to the other minimizing the number of times I have to recompose and even with doing that I don't get smeared stars or duplicates.

It was a bit of a trip, it took about 4.5 hours to get to a somewhat dark sky, and even then the horizons were still somewhat bright. According to the light pollution map it was bortle 3. I'll investigate if it is doable to get better skies with less travel time, but it might be quite challenging.

Your results in a very bright area are very impressive. I'm not really planning to do much at home. Its class 8-9, and I don't even have a backyard (this is tokyo after all). But its a good excuse to get out of the city when time permits.


I'm wondering if using hugin/enfuse to align and stack would be doable. It can handle distortion. I'll give it a shot when I next have time. For now, if I throw money at the problem, its probably for better glass.


In the meantime I managed to have a first attempt at stacking the pictures from the 500mm tamron mirror lens. It didn't do great, but partially it seems that I didn't get astrotracer to work too well, the stars are slightly elongated. According to suggestions above I have increased saturation. 17 shots of 10 seconds, so again slightly under 3 minutes total exposure. No dark frames or flat frames, although I think the K-1 did do its automatic dark frame subtraction.

I'm still slightly squeamish at the massive amount of contrast, saturation and exposure compensation I have to use. I might give sequator a try too, it looks interesting. Its too bad it and DSS both don't run in linux.

As before, please let me know what to change for next time!

That I don't know. I'm not sure if it would be able to align star images or not. I've stacked images that were aligned with hugin using imagemagick with great success when doing some super resolution shots but never astro shots. I may give it a try as the last time I was out the O-GPS1 flaked out after a couple of shots so most of what I captured was not tracked well at all. I think it is time for a new battery. I would think that DSS would run in Linux using WINE as it seems like a program that isn't tied closely to system calls. Worse comes to worse convert windows to a VM and run it in virtualbox on linux.

With the 500mm shot it almost seems like the white balance is a little too much to the magenta side of things.

That is rough I can get to a 3 in about 2.5 hours and in 45 minutes can get to a dark 5 almost a 4. I still use a light pollution filter in dark 5 place near me as it does provide a noticeable benefit.

Turn that off and either shoot your own dark frames at the end as you pack up or just shoot more light frames. However expect very little benefit unless you shoot a lot of light frames and have tens of dark frames. I also doubt their benefit at lower ISOs and shorter shutter speeds as the systematic dark current becomes less pronounced. Add in that you aren't on an equatorial so aligning the images would tend to shift the dark current pattern around and it would become more like regular noise.

Out of curiosity how are your focusing? If you don't have one see about getting a bahtinov mask for each of the long lenses you plan on using. I have one for my 300mm and my 400mm that I 3d printed. I generated the SVG file using this page from there I converted the SVG to a STL file with an extruded of 3mm using this page and then 3D printed the masks at my local library.

Those old 200/4 lenses are actually a pretty good lens. Later this week an even older 200/3.5 Takumar will be arriving to me on loan and I plan on doing a comparison between it and my S-M-C 200/4 Takumar for astro. Don't discount your SMC M 200/4 when I say it can provide very good results there are images to back that up. Granted that was off of an equatorial mount by someone who is more capable at astro imaging and editing than I am but show what it can do.

Editing is the hard part. The individual who posted the nebulae in Orion image I linked to gave me a link to a series of introductory astro image editing videos. I will dredge them up when I get home and provide the link. I never use contrast or exposure compensation when editing images with most of my edits being levels and curves. Also you may want to look into upping your ISO some so that what little data is there gets spread across all 14 output bits as much as possible before it gets recorded digitally. I think I really need to write up a beginners guide to astro image editing as I am finally starting to get some reasonable competency.

Some nice work there, and hey, big shout-out to another BBAR user! I was never able to get much usable off the Tammy with AstroTracer; I managed to get this on a borrowed iOptron SkyTracker though. That lens is so slow that you really have to fight to get enough photons on the sensor to stretch. Still, I think your closeup Orion is quite successful there.

If you find you get really, super, head-over-heels into this hobby there are options for even the brightest areas. This Veil Nebula image and this Elephant's Trunk one were not shot with a Pentax, but both were in the city. In fact the Veil was in a brightly-lit parking lot! The secret is so-called "narrowband" imaging, using filters that admit only a few nanometers' width of the spectrum at a time. It can be done with a DSLR but it's much, much, MUCH easier to accomplish with a dedicated astro camera.

As for processing, yeah, you have no choice but to wreak some crazy things on your data for deep-sky objects! Experienced deep-sky imagers say that processing is at least half of the battle. Here is a link to a Photoshop method that quickly gets you in the ballpark for deep-sky objects.

I think the idea of a beginner's guide to editing would be great. If you're interested in understanding the basics of the art and science, and also getting some direct instruction in editing astrophotographs, I highly recommend Charles Bracken's The Deep-Sky Imaging Primer. There are certainly other good books out there, including some by Jerry Lodriguss and one that's very popular in England called Making Every Photon Count. I happen to really like Charlie's work.

That is again quite impressive for a 3-minute shot, especially with a mirror lens. Is that an f/8 model? (There are some f/6.3 ones but I'm not sure who makes or made them.)

Regarding automatic dark-frame subtraction, with 17 sub-exposures there is no need for it, as stated. You merely need to use kappa-sigma rejection when stacking, which will delete hot pixels, cold pixels, satellite trails and even airplane trails. That would have given you in this case a few additional minutes of sub-exposure in the same amount of time it took you to do the 17 sub-exposures.

Thanks for the replies. This was indeed with a 500mm/8 mirror. The tamron 55BB has a decent reputation as being sharper than most mirror lenses, although there are (more expensive) ones by zeiss and leica. Its hard to find any astro work done with them though, maybe they are not particularly good at infinity. I've not used the lens much, so I don't really know its strengths and weaknesses yet.

I got lucky yesterday and could see orion in the tiny part of the sky that was visible from my apartment. So I thought to just try shooting the nebula from a bortle class 8-9 environment. I took more shots (I think it was 28*10s frames after throwing away a handful), and still ended up with more noise and having to do more aggressive processing (contrast in rawtherapee to its max value of 100 for example). One of the problems I found is that I can't seem to get astrotracer to work for less than 10 seconds. As with my earlier shots, this seems to be too short. I'd prefer to try 5 seconds. However, the fact that I got any results at all does make me more interested in investigating filters for shooting from Tokyo. Possibly one of the brightest places in the world.


Unfortunately the K-1 also doesnt seem to want to do interval shooting with astrotracer, so I might have to invest in an external intervalometer.


I didnt have my 200mm at home, but I tried with the 100mm as well. With actual dark frames (23) and even more frames than the 500mm (around 65 * 20s). But it didnt work out well at all. I'm not sure why. It did well in the dark spot last weekend.

500mm shot (unfortunately its a screenshot, I don't have the file readily accessible here, I'll replace it with the real thing later. But there is not much more detail in the real thing than in this screenshot)



So some questions: is there a way to do astrotracer with shorter times than 10s? And is there a way to get it to take repeated shots without an external intervalometer?

EDIT: I've now tried using both DSS and Sequator. Sometimes DSS does not seem to produce an image in its viewer after stacking, but it usually still produces the file. This confused me at first. I've not yet been able to develop a preference for one or the other.

Re less than ten seconds, you can disable the timed exposure. When you do the K-1 will act like it normally does in Bulb mode wrt exposure time. You could count out, say, five seconds ("one thousand and one, one thousand and two,...") or use an external intervalometer. You'll need the latter anyway to do interval shooting with Astrotracer regardless of how the exposure time is controlled.

But...if you simply let an intervalometer do the work of taking successive exposures you may run into two issues. First, the subject will move across the frame, eventually leaving it. Not a problem with wide angle or even short telephoto but definitely a problem with long telephoto starting at 200mm. With 500mm on full-frame the subject should really be reframed at least every second shot. (With 300mm on my K-3II APS-C I reframe every second or third shot.) Second issue is walking noise. If you let the subject drift in RA between shots that can accentuate what is called walking noise, resulting in "wormy" patterns in the background sky. If you reframe often this goes away.

I am again impressed by what you have done with the 500mm in the city this time. Given the light pollution you faced I think you did an excellent job with the processing.

Re Sequator, I far prefer it to DSS. It is much faster and does a great job with alignment.

Thanks for the advice. I'll check out getting an intervalometer. I think in shots like this, where the subject is quite a bit smaller than the frame with the tamron and 500mm, it should be possible to do quite a few more 5s shots than 2 or 3 when only considering the subject movement. At least last evening I only reframed much less, those were 10 second each shot, I'm planning to do 5 second per shot or less, hoping to get a sharper image.

Regarding walking noise, I'll have to read up on that, I have no idea what that is, but I think I have already encountered it. I did see some small structure in one of my stacked exposure, similar to what I'd usually associate with the sharpening applied in rawTherapee. Maybe that was walking noise?


I'm also considering getting another lens. Whereas the 55BB mirror lens is generally thought of as pretty decent for a mirror lens, its not considered comparable in sharpness and contrast to normal lenses. I'm currently considering the 645 and 67 300mm lenses, as they seem to be cheaper and of similar quality as the full frame variant. Someone is also selling an A* 67 400mm/4, but its not clear what condition that is in. One of the perks of living in Japan: the used market for lenses is good here! I'll also try more with the 200, but I'll have to wait for clear skies first.

True, if you are doing 5 second exposures you will need to reframe half as often as with 10 second exposures.

Re walking noise, it is due to a correlation over successive frames of the color "blotchiness" in the sky background. That blotchiness is the result of the way demosaicing algorithms treat low level noise. In single exposures it looks like faint green and magenta "worms" in the sky background. If your frame drifts in the same direction, over time those "wormy" demoisaicing artifacts correlate and produce streaks (so I misspoke above). Here is one (particularly bad) example: Colored Streaks in Integrated Result - DSLR & Digital Camera Astro Imaging & Processing - Cloudy Nights

If you reframe often you are effectively randomly dithering because you will never end up in exactly the same place twice. The worms average out instead of correlating.

Re telephoto lenses, one thing you aren't seeing with your mirror lens is chromatic aberration (CA). If you buy one of those vintage telephotos you mentioned, you can expect to see plenty of CA and you'll need to deal with it. Stopping down helps and so too, in my experience, does use of a UV(0) filter. However, in the end you'll still need to deal with it in post processing and it may not be possible to get rid of all of it. You can avoid it (or most of it) if you buy an ED lens but of course it will be more expensive.

I have posted examples here of what I have done with an SMC Takumar 300mm f/4 lens (purchased from Japan, of course!) on a K-3II and using Astrotracer. My processing was anything but perfect. In particular, I think it is possible to reduce the CA even more:
An Early M42 - Astrophotography - PentaxForums.com
Test: M31 with SMC Takumar 300mm f/4 - Astrophotography - PentaxForums.com
Astrotracer with a 300mm lens - Astrophotography - PentaxForums.com

The walking noise looked a bit different from what I expected, and I don't think I've seen it yet. But I'll keep my eye out. Especially with the 500mm, not having a guided / equatorial mount, reframing is anyway necessary now and then, I guess in this case its a good thing. I guess this could also be avoided if we could somehow get astrotracer and pixelshift to work together (ie, while rotating the sensor with the stars, also do the (0,0), (0,1), (1,0) and (1,1) pixel offset frames, but I guess that is such a ridiculously niche application of the two pentax technologies that it will never happen. But I'd be curious to know if on an equatorial mount pixel shift might get rid of some of the walking noise. Might be an interesting experiment for someone who has access to both a mount and a pixel-shift capable camera.

On your picture with the 300mm/4, I see the same as what I got with the D-FA 100mm macro: purple stars. In my case this only showed up after increasing contrast and saturation to high levels, which I guess can be unavoidable. With the pentax-M 200mm, I did not have that problem, but I also didnt push the contrast and saturation quite that far as I had good dark skies and a larger number of frames. I know basically nothing about postprocessing (I usually adjust exposure and white balance in rawtherapee, switch off sharpening and might push shadows or reduce highlights. I never use lightroom/photoshop or similar programs), so I guess I have some learning curve ahead.


The vintage lenses I have all have ED elements. Especially the A* 300mm 645 is not very expensive. It is of course larger than the full frame variants, does not come with a tripod mount, and needs a somewhat expensive 645-KP adapter, so its not all perfect. The M* 67 300mm does have a tripod mount, a wider barrel, otherwise more or less identical dimensions, but is a bit more expensive and harder to find. Then there is the A* 67 400mm/4, of which there is currently a cheapish version for sale, but the condition is not well described, so I should probably stay very far away. The Tamron adaptalls 300mm/2.8 of course also exist.


Anyway, after having played around with longer lenses and the astrotracer only twice, I probably really should focus on improving technique before starting to buy lenses. Especially if they are lenses I am unlikely to take on other trips (ie non-mirror long glass). Instead getting filters and learning more about postprocessing should probably be a priority. LBA is so easy to trigger though. While going through the forums I've also developed a desire for the Takumar 200mm/3.5 (for non-astro use probably). Will it ever end?

Why I never use the Tamron 500mm anymore:

1) It's Bog-awful slow.
2) It is a pitiless bee-yotch to focus
3) Did I mention that it was slow?
4) My mother-in-law gave me some Christmas money and I splurged on a Stellarvue 70mm.
5) Which is more like f/4.8.

The tiny scraps of light coming off the back of the Tammy make it difficult to find and focus on targets, though you have done an AMAZING job at that. Do consider a Bahtinov mask, you can focus on the brightest star in view and then repoint to your target. One of the Tamron's excellent points is its exquisitely-damped focusing mechanism, it's butter-smooth but doesn't move if you don't nudge it. But its depth of field is essentially nil, so over an evening it will probably drift out of focus due to temperature changes.

And, hey, for astro at least you don't have to worry about donut bokeh :-)