Proper "rule" for pinpoint stars with M43

Hi all, I've read about various rules for achieving pinpoint stars on a nontracking mount but there's been some disagreement about which is correct.

I've read about the following:

the 600 rule.....on a 28mm equivalent focal length, this means a 20 sec exposure

the 500 rule......on a 28mm equivalent focal length, this means a 15 sec exposure

the 400 rule......on a 28mm equivalent focal length, this means a 15 sec exposure

the 200 rule......on a 28mm equivalent focal length, this means an 8 sec exposure

the 100 rule......on a 28mm equivalent focal length, this means a 4 sec exposure.

The particular part of the sky I'm interested in is near the celestial equator, the constellation Orion.

I did 8 sec exposures, they give me fairly round stars, but 15 sec exposure shows a lot more stars, however the dimmer stars below the belt seem to show some trailing (curiously, the brighter stars, like the ones in the belt, don't show trailing- maybe because brighter stars appear larger anyway?) I tried to go longer than 15 sec exposures, but 20 sec exposures don't show me many more stars than 15 sec (although trailing seems to be about the same), and 30 sec shows too much trailing and at that point light pollution is become too much of a factor.

With my trust old 10 mp E-520 I found the sweet spot for Orion to be 15 sec exposures at 28mm equivalent focal length for 1:1 viewing (100% pixel level) on a 19" 1280x1024 monitor, will it be the same for my 16mp E-PL6?

If so, then maybe this still applies?

https://www.dpreview.com/forums/post/56527938

I think the formula 209/FocalLength is correct for M43, but 250/FocalLength is close enough, especially if you are using JPEGs.

I made a spreadsheet calculation where I set the constraint that the star image should move the diameter of the CoC (circle of Confusion). For M43 I used a .015mm CoC, which works out to be just under 4 pixels for a 16MP sensor. I used the the maximum angular velocity, which occurs at the celestial equator, or approximately the suns path.

I get the result that the shutter speed should be faster than 209/FocalLength for an M43 camera.

But what about ISO and aperture?

Andrew Butterfield wrote:

But what about ISO and aperture?

I wouldn't use ISO of 3200 or higher. It isn't so much because of noise, but the dynamic range issue. Lower dynamic range at higher ISO makes all stars appear to be white. I find star colors come out really well at ISO lower than 3200 but once you hit that mark, the star colors get extremely dull. There's a big difference even between ISO 2500 and 3200 in terms of star colors. And it's shutter speed independent. In other words star colors show really well at exposures of both 2 sec and 15 sec at ISO 1600. But once you get to ISO 3200, regardless of the shutter speed, all stars appear white or nearly so.

To answer your question more succintly, ISO of 800 and 1600 are where you want to be. As for aperture (f number), as fast as you can get.

Also, as far as short star trails are concerned, I think that issue can be fixed with software? DSS or Sequator maybe?

..is where the Astro Photography section shines!

https://www.dpreview.com/forums/1059

yeah, astro...

However, I'll add my experience - I got perfect stars with 25 second exposure, 12 mm (24 mm EFL), f2.8.

phil from seattle wrote:

yeah, astro...

However, I'll add my experience - I got perfect stars with 25 second exposure, 12 mm (24 mm EFL), f2.8.

Thanks- was this with Orion and pinpoint stars even when viewing 1:1?

I have limited experience, but have used the 8/1.8, 12/2, 15/1.7, and 17/1.8 for stars. From what I've seen it is more complicated than a simple linear formula. I maintain:

https://docs.google.com/spreadsheets/d/1-5JkMAHVcVV4zLsEH8xlxoSeuhYURLoJtuJn57N9ues/edit?usp=drive_web

to compare m4/3 lenses and found that a non-linear formula based on angular area of view (AA) seemed most acurate. Doing a regression from the few data points gave me a quadratic equation:

(-1.6*(AA*AA))+(12.7*AA)+(4.5)

I might be over thinking this but....

Giiba wrote:

I have limited experience, but have used the 8/1.8, 12/2, 15/1.7, and 17/1.8 for stars. From what I've seen it is more complicated than a simple linear formula. I maintain:

https://docs.google.com/spreadsheets/d/1-5JkMAHVcVV4zLsEH8xlxoSeuhYURLoJtuJn57N9ues/edit?usp=drive_web

to compare m4/3 lenses and found that a non-linear formula based on angular area of view (AA) seemed most acurate. Doing a regression from the few data points gave me a quadratic equation:

(-1.6*(AA*AA))+(12.7*AA)+(4.5)

I might be over thinking this but....

Hm now I have some numbers to plug in. I was trying to use the sensor density and angle of view for starters but it looks like it might be more complicated for that. Let's use the kit 14-42 lens for starters at the wide 14mm end.

OutsideTheMatrix wrote:

phil from seattle wrote:

yeah, astro...

However, I'll add my experience - I got perfect stars with 25 second exposure, 12 mm (24 mm EFL), f2.8.

Thanks- was this with Orion and pinpoint stars even when viewing 1:1?

I've not shot when Orion was up but no trails at all.

phil from seattle wrote:

OutsideTheMatrix wrote:

phil from seattle wrote:

yeah, astro...

However, I'll add my experience - I got perfect stars with 25 second exposure, 12 mm (24 mm EFL), f2.8.

Thanks- was this with Orion and pinpoint stars even when viewing 1:1?

I've not shot when Orion was up but no trails at all.

Thanks, I viewed your spreadsheet and came up with the following: 8 seconds at 14mm focal length (28mm equivalent) should be enough to achieve pinpoint stars.  Have you found the same?

OutsideTheMatrix wrote:

phil from seattle wrote:

OutsideTheMatrix wrote:

phil from seattle wrote:

yeah, astro...

However, I'll add my experience - I got perfect stars with 25 second exposure, 12 mm (24 mm EFL), f2.8.

Thanks- was this with Orion and pinpoint stars even when viewing 1:1?

I've not shot when Orion was up but no trails at all.

Thanks, I viewed your spreadsheet and came up with the following: 8 seconds at 14mm focal length (28mm equivalent) should be enough to achieve pinpoint stars. Have you found the same?

I see 15s when I punch in the FL and aperture, with a score of 183.

I'd be curious to know how it works out as I've very little idea of what the score really means photographically. I use a standardized formula so these score should be comparable across formats. So far my best results have been with the 15/1.7, I don't find the 8/1.8 picks up more stars but I've only gotten the fisheye out once for clear skies.

phil from seattle wrote:

yeah, astro...

However, I'll add my experience - I got perfect stars with 25 second exposure, 12 mm (24 mm EFL), f2.8.

I found 25s was a touch too long when pointed at the celestial equator, but gave decent results for Orion and good results for Polaris.

My spreadsheet numbers seem to be a reasonable max for the equatorial zone and can be uped as you point your field of view towards the poles.

I'm wanting to buy a tracking mount to use with my longer lenses for better spot details.

Fair enough, All my night sky shots have been at 50 deg north.

Giiba wrote:

OutsideTheMatrix wrote:

phil from seattle wrote:

OutsideTheMatrix wrote:

phil from seattle wrote:

yeah, astro...

However, I'll add my experience - I got perfect stars with 25 second exposure, 12 mm (24 mm EFL), f2.8.

Thanks- was this with Orion and pinpoint stars even when viewing 1:1?

I've not shot when Orion was up but no trails at all.

Thanks, I viewed your spreadsheet and came up with the following: 8 seconds at 14mm focal length (28mm equivalent) should be enough to achieve pinpoint stars. Have you found the same?

I see 15s when I punch in the FL and aperture, with a score of 183.

I'd be curious to know how it works out as I've very little idea of what the score really means photographically. I use a standardized formula so these score should be comparable across formats. So far my best results have been with the 15/1.7, I don't find the 8/1.8 picks up more stars but I've only gotten the fisheye out once for clear skies.

What does the score of 183 mean?  My very unscientific test involved just looking in the LCD of the camera post capture, but I magnified the view to 7x as I was told that simulates a 1:1 100% view on a 19" 1280x1024 monitor (not sure about that.)

My conundrum is this- I tried 8s, 15s, 20s and 30s.  30s has too much trailing so that's out.  8s has the least but 15s captures a lot more stars than 8s.  20s doesn't seem to capture more stars than 15s but the amount of trailing is the same.  15s/20s don't show trailing in the brighter stars of Orion (like in the Belt) but the dimmer stars seem to show trailing.  I think that's because the trailing isn't as perceptible in the brighter stars because they also appear to be larger?  If a post processing program like DSS or Sequator can fix minor trailing during stacking then I should probably stick with either 15 or 20s?

Also is there anyway to mathematically predict the amount of trailing in terms of pixels? Like (for example) is 15s at 14mm (28mm equivalent) going to result in less than 5 pixels of trailing?  3?  I want to keep trailing down to 3 pixels or less.

Giiba wrote:

phil from seattle wrote:

yeah, astro...

However, I'll add my experience - I got perfect stars with 25 second exposure, 12 mm (24 mm EFL), f2.8.

I found 25s was a touch too long when pointed at the celestial equator, but gave decent results for Orion and good results for Polaris.

My spreadsheet numbers seem to be a reasonable max for the equatorial zone and can be uped as you point your field of view towards the poles.

I'm wanting to buy a tracking mount to use with my longer lenses for better spot details.

Me too- my only issue with tracking mounts is achieving polar alignment (ugh.)

If 25s is too long for 12mm then 20s will be too long for 14mm?

phil from seattle wrote:

Fair enough, All my night sky shots have been at 50 deg north.

Thanks Phil.  Are you referring to your latitude or the declination of the celestial objects you are shooting?

Giiba wrote:

I have limited experience, but have used the 8/1.8, 12/2, 15/1.7, and 17/1.8 for stars. From what I've seen it is more complicated than a simple linear formula. I maintain:

https://docs.google.com/spreadsheets/d/1-5JkMAHVcVV4zLsEH8xlxoSeuhYURLoJtuJn57N9ues/edit?usp=drive_web

to compare m4/3 lenses and found that a non-linear formula based on angular area of view (AA) seemed most acurate. Doing a regression from the few data points gave me a quadratic equation:

(-1.6*(AA*AA))+(12.7*AA)+(4.5)

I might be over thinking this but....

Giiba, I see that you put 14/3.5 in there! So do you find there is about the same amount of trailing with the 7-14 at 14mm and the 14-42 at 14mm?

If you could do so, would you do one for the 14-42 at 42mm and the 40-150 at 40mm please?  Thanks!

OutsideTheMatrix wrote:

Giiba, I see that you put 14/3.5 in there! So do you find there is about the same amount of trailing with the 7-14 at 14mm and the 14-42 at 14mm?

If you could do so, would you do one for the 14-42 at 42mm and the 40-150 at 40mm please? Thanks!

I was just toying for the score. As far as the spreadsheet is concerned 14 is 14 is 14 but in reality there can be slight fov differences between lenses of the same focal length. But that shouldn't change max time significantly.

40mm, 42.5,mm and 45mm are all 6s after rounding so 42 will be as well. I plan to add the 40-150/2.8 only the score would differ based on aperture.

Here is what 17mm at 20 seconds looks like facing due south. Although I see some movement of the stars for me this is an acceptable result. This is a minimally processed OOC jpeg. Using 12mm at 20 seconds as the OP said the star motion is minimal.

The limitation on the one end is star motion but on the other it is minimum exposure and noise. I most often shoot either the 12-40 at 12mm or the Oly F1.8 8mmFE. I generally shoot 20 seconds with the former because I know it is safe and 20 seconds also with the 8mm FE because I really don't need more exposure. With the 12mm I shoot at ISO 2000 to 3200 and with the FE at ISO 1000 to 1600 (with the latter being ideal).

Giiba wrote:

OutsideTheMatrix wrote:

Giiba, I see that you put 14/3.5 in there! So do you find there is about the same amount of trailing with the 7-14 at 14mm and the 14-42 at 14mm?

If you could do so, would you do one for the 14-42 at 42mm and the 40-150 at 40mm please? Thanks!

I was just toying for the score. As far as the spreadsheet is concerned 14 is 14 is 14 but in reality there can be slight fov differences between lenses of the same focal length. But that shouldn't change max time significantly.

40mm, 42.5,mm and 45mm are all 6s after rounding so 42 will be as well. I plan to add the 40-150/2.8 only the score would differ based on aperture.

I was wondering how much the difference would be between the 42/5.6 of one lens versus the 40/4 of the other lens.  I thought your score might relate to how many stars can be captured based on the aperture of the lens