Dark Night Focus

In astronomy we use software that can realign multiple images (don’t know how but it does). Deep Sky Stacker and Registax being just two of them. Personally I use Astro Pixel Processor.

If I was taking an Astro image I would use a special mount that can track the sky with great precision but that is going into another world.

If I was to use my stills camera, LUMIX G9, I would use high ISO and take up to a ten second exposure. For moon shots it’s ISO 200 at 500th of a second.


Sent from my iPad using Tapatalk

I use
-Live View.
-Flashlight or laser pointer.
-Manually focus.
-Find the hyperfocal distance.
-Find contrast detection.

For astrophotography preferable contrast detection, if it is possible.
In more detail, I wrote in my article about night focusing.
If there are links to this filter or information about filter share it.
I will updated my article.

Quote:

---

Originally Posted by steveg4iwr

In astronomy we use software that can realign multiple images (don’t know how but it does). Deep Sky Stacker and Registax being just two of them. Personally I use Astro Pixel Processor.

---

Yes. I am aware of DSS.

For “photography” use, as opposed to “scientific investigation” use, I am of the understanding that it is typically used to stack 5~8 frames – again maybe I am out of touch with a new loop of procedures, that’s why I asked the OP.

Additionally, the phrase underlining and bolding was purposeful in my previous -“A Night Sky scene in any location in S.A. will show a minor shift in alignments at about each 120 seconds”

It was referencing changes in alignment between some of the objects, not movement of the whole object field relative to the camera.

I didn’t do any of the maths, and probably, as I mentioned, using a W/A camera lens for 100 shots each of 10 seconds, the FoV of those captures will render any alignment shifts absolutely unnoticeable: just thinking about it now 18mm lens on an APS-C that is a lot of sky in shot.

But the predicate for my question, is, I think that the OP is not only over-thinking the exercise, as Manfred has already suggested, but I additionally think that the OP is making the task way more difficult and way too much work.

Moreover, all this extra effort may render poorer results than simply making one, good quality image and using excellent Post Production Skills to enhance it.

***

Quote:

---

Originally Posted by steveg4iwr

If I was to use my stills camera, LUMIX G9, I would use high ISO and take up to a ten second exposure. For moon shots it’s ISO 200 at 500th of a second.

---

I did however do some maths on this subject of exposure time.

No idea where Wellington East is located, but, for Adelaide, and for CLEAR Summer skies, the OP could crib to pulling around 22 seconds exposure time and get a reasonably static scene.

Some additional back of the envelope arithmetic gave me: if the sky is clear in S.A. (no idea of what Bushfire Smog is there at the moment and the OP uses the F/3.5 he has available, then likely he’d be pulling ≥ISO3200.

Obviously, if the OP wants Star Trails, then that’s a different kettle of kippers . . . that’s another reason why I asked for clarification of the meaning of [eliminates any] “movement issues”.

WW

This again is getting way of theme.... but if you have an interest have a look at https://project-nightflight.net/projects009.html

This is just one direction I will be taking so I'm not overthinking or whatevr, I'm just trying to see if anyone has experience with the Star cross lens as another optional focussing tool and is it worth persuing in that endeavour.
Brian

Quote:

---

Originally Posted by bjk896

This again is getting way of theme.... but if you have an interest have a look at https://project-nightflight.net/projects009.html

This is just one direction I will be taking so I'm not overthinking or whatevr, I'm just trying to see if anyone has experience with the Star cross lens as another optional focussing tool and is it worth persuing in that endeavour.
Brian

---

Had you posted this article at the beginning, it would have been very useful in trying to describe what you are trying to do.

1. You are trying to do whole sky images using a reasonably wide focal length (18-55mm) on a crop frame camera.

2. Rather than using tracking hardware to compensate for the earth's rotation, you are looking to use the alignment functionality in software to compensate motion by using a large number of individual shots that have a short enough duration be be motion free.

3. You are looking at using a standard dark-frame technique to remove the fixed sensor noise component in these captures.

It has been a very long time since I did astrophotography using a long focal length reflector telescope; but the focus techniques you have written about were primarily used when using those very long (400mm and longer) focal lengths with relatively small apertures. Pulling a sharp focus with that gear was challenging and specialized focusing aides helped.

At the fairly short focal lengths with fairly fast lenses like way you are using, it still strikes me as overkill. Depth of field tables / hyperfocal tables at the recommended apertures in the article (at 18mm you should stop down to at least f/5 and at 55mm at least f/8; i.e. at least 1 f-stop below wide open), so long as you are close, the image will be sharp. Look at the following for guidance:

https://www.cambridgeincolour.com/tu...htm#calculator


https://www.cambridgeincolour.com/tu...calculator.htm


Some of the focus techniques we have suggested should be more than adequate at the range of focal lengths you are looking at. As I wrote before, you are definitely overthinking this one.

I would be far more concerned about a clear view of the sky (not sure where in the world you are located - Wellington East South is not that helpful), light pollution, a very sturdy tripod and head (you definitely don't want and sag during the exposure), etc. Pulling a sharp focus with that camera and lens combination is going to be relatively easy.


I have not used a star-cross filter in years; but all they effectively do is create a star-burst appearance which gives you a larger, but dimmer, object to focus on. With a long focal length and the ability to isolate on your target, I can see how they might be useful, but with the type of shot you are trying to take, you are going to see so many star-bursts that focusing is going to be challenging. Sort of like trying to focus on a single tree in a large forest from a distance. These filters are largely obsolete in the digital days as we use either software or small apertures to create this effect.

Quote:

---

Originally Posted by Manfred M

... At the fairly short focal lengths with fairly fast lenses like way you are using, it still strikes me as overkill. Depth of field tables / hyperfocal tables at the recommended apertures in the article (at 18mm you should stop down to at least f/5 and at 55mm at least f/8; i.e. at least 1 f-stop below wide open), so long as you are close, the image will be sharp. ...

---

I agree about the overkill. In another universe some time back, there was confusion about infinity focusing with wide-angle lenses. I came to a conclusion that there is a certain hyperfocal distance beyond which there is no visible improvement in the sharpness of distant objects.

For example, the conventional DOF calculator for Manfred's 18mm lens at f/5 and 1.5 crop gives the hyperfocal distance as approximately 3 meters. But Merklinger has found that distant objects can be less than perfectly in focus with such a setting:

http://www.trenholm.org/hmmerk/TIAOOFe.pdf

If, however, we were to set the CoC to 1 pixel or less, that would give a focusing distance of virtual infinity because one can not resolve finer than approx one pixel (Rayleigh, Abbe, et al). For example with that same lens and using a pixel pitch of 6um as the CoC, you get a hyperfocal distance of 11 meters! That may seem ridiculously small when shooting Astro stuff but my theory is quite sound, IMHO.

The formula that I use is Distance = F^2/(p*N)+f' where F = marked focal length, p = pixel pitch (or some factor thereof), N = f-number and f' = effective focal length.

Ted - what you have written makes sense, but the issue I have is that minor camera movement; head droop, ground movement or wind can quickly wipe out all of the benefits of the technique. Even camera shake from shutter / mirror actuation is going to play a role in the proposed exposure times, so the camera needs to be set to shoot in mirror up mode.

Quote:

---

Originally Posted by bjk896

This again is getting way of theme. . .

---

That’s your opinion, given now twice and trust it will be respected now and in the future.

Another view would be that one came here to harvest both general and specific assistance, from a wide range of experience, garnered by the exchange of information: I have found this approach renders a more profitable outcome, both in the short and the long term.

***

Quote:

---

Originally Posted by bjk896

. . . I'm just trying to see if anyone has experience with the Star cross lens as another optional focussing tool and is it worth persuing (sic) in that endeavour.

---

Specifically:
> Yes I have with telescopes.
> Yes, it had worthwhile uses for that purpose.
> I have no experience with star crossed filter use as a focusing tool with wide angle~short telephoto focal length lenses; therefore cannot comment on that specific point. Additionally, I read the article you linked and also the pdf download in it, twice, I could not reference the use of star crossed filters in the technique described the pdf article.

However, as you written, “This is just one direction I will be taking”, then you’ve already decided to use the filter, so you’ll find out the worth, in due course, anyway, and perhaps you’ll be kind enough to report on that outcome.

Good luck with you venture.

WW

Quote:

---

Originally Posted by Manfred M

Ted - what you have written makes sense, but the issue I have is that minor camera movement; head droop, ground movement or wind can quickly wipe out all of the benefits of the technique. Even camera shake from shutter / mirror actuation is going to play a role in the proposed exposure times, so the camera needs to be set to shoot in mirror up mode.

---

Indeed, Manfred, camera movement is certainly an issue when shooting distant objects. However, my point was that such an effect on an image would be virtually the same - whether the lens is focused at my "virtually infinite" distance or it is focused at the true infinity setting (the one where incoming parallel rays actually focus at the sensor - not necessarily at the mark on the lens).

I suppose I'm agreeing with you, by the way. There would be no benefit other than not having to be too picky about the focus setting when shooting stars, as you have already said in a previous post.

Quote:

---

Originally Posted by xpatUSA

Indeed, Manfred, camera movement is certainly an issue when shooting distant objects. However, my point was that such an effect on an image would be virtually the same - whether the lens is focused at my "virtually infinite" distance or it is focused at the true infinity setting (the one where incoming parallel rays actually focus at the sensor - not necessarily at the mark on the lens).

I suppose I'm agreeing with you, by the way. There would be no benefit other than not having to be too picky about the focus setting when shooting stars, as you have already said in a previous post.

---

Ted - I think we are agreeing with each other and are just having trouble saying just that! :D

Quote:

---

Originally Posted by bjk896

…..The idea is that the 4 star points can give a good indication of focus with me refining the manual focus ring it until the crosses are crisp. Some say to use screen magnification to help it all along.
Has anyone had success with this technique and maybe a process to streamline the time taken and also mark or identify the focus ring position for later shots.

---

Hi Brian,

I have never used a 4 point cross lens/filter.
For manual focusing I use a Bahtinov mask, cheap, easy and reliable, but challeging on a little DSLR screen, screen magnification is helpfull. I use a laptop where I can clearly see what I am doing, but my eyes are probably less accurate than yours…:)
Depending how critical you are about pinpoint stars, Keep in mind that you might do some refocusing (influenced by temperature) during the night depending on the duration off the session.

Typically I refocus every 30 minutes or when temperature change is 1°C.

Personally I don't lock the focus ring. Putting a mark can be usefull but IMO only as a reference.

HTH.

Quote:

---

Originally Posted by rudi

Typically I refocus every 30 minutes or when temperature change is 1°C.

---

Given that the coefficient of thermal expansion of aluminum (the material in the lens that likely has the highest factor or thermal expansion) is approximately 23.6 x 10-6 (°C)-1 chances are that one will throw the focus off rather than anything else. This means a 1m long piece of aluminum will shrink or grow 0.0000236m for every degree C of temperature change. None of my lenses, especially the wide angle ones have nearly that much metal in them.

I'm not quite sure regarding your logic. I've done long exposure shooting in the evening / night with an intervelometer that ran over close to two hours and see no loss of clarity or focus when I look at the first frame to the last frame, in spite of a fairly significant drop in temperature over the period.

Quote:

---

Originally Posted by rudi

. . . For manual focusing I use a Bahtinov mask, cheap, easy and reliable, but challeging on a little DSLR screen, screen magnification is helpfull.

---

I concur, a Bahtinov Mask is useful in some situations, by the way, mentioned earlier here: [LINK].

Certainly, as you mentioned a small screen can be problematic, additionally, from my experience, arguable not much use at FL = 18mm and Aperture F/3.5 on APS-C Format.

When using a W/A Lens I have found a Bahtinov Mask can work quite usefully at Apertures between F/1.4~F/2.

***

Quote:

---

Originally Posted by rudi

Typically I refocus every 30 minutes or when temperature change is 1°C.

---

I disagree with that procedure. Manfred has explained why, with the Maths. Additionally I've field tested it to show Students that it certainly can create a mess for stacking.

***

Quote:

---

Originally Posted by rudi

Personally I don't lock the focus ring. Putting a mark can be usefull but IMO only as a reference.

---

I concur: good practice and yes a reference point marking can be useful, especially if the lens focuses beyond Infinity.

WW

Quote:

---

Originally Posted by rudi

Typically I refocus every 30 minutes or when temperature change is 1°C.

---

We often read of thermal expansion within standard camera lenses affecting critical focus in astro photography but to date I've never seen any tests or evidence where this has been demonstrated.

Whilst the coefficient of expansion for each material is known and measurable I wonder if any designer has actually calculated how each individual component reacts in relation to each other component:eek:

Quote:

---

Originally Posted by Stagecoach

Whilst the coefficient of expansion for each material is known and measurable I wonder if any designer has actually calculated how each individual component reacts in relation to each other component:eek:

---

While I can't comment on what lens designers do, it is certainly common practice in mechanical and civil engineering work where heat and cold are important operating parameters. It can be as simple as an automatic belt tension control on a conveyor belt to keep the belt taut under normal operating conditions to engines or refrigeration equipment that undergoes significant thermal expansion or contraction during normal operations.

Camera gear specs include maximum and minimum operating temperatures, so that suggests that some of the camera and / or lens components are temperature sensitive and the equipment has been designed to operate within that temperature band.

Quote:

---

Originally Posted by Manfred M

While I can't comment on what lens designers do, it is certainly common practice in mechanical and civil engineering work where heat and cold are important operating parameters. It can be as simple as an automatic belt tension control on a conveyor belt to keep the belt taut under normal operating conditions to engines or refrigeration equipment that undergoes significant thermal expansion or contraction during normal operations.

Camera gear specs include maximum and minimum operating temperatures, so that suggests that some of the camera and / or lens components are temperature sensitive and the equipment has been designed to operate within that temperature band.

---

Yes it's certainly standard practice in machinery design otherwise we would be getting a lot of seizures:)

Whilst a camera will have a specified design min/max temp working range this will be with respect to operation. It may very well be that temperature difference over the specified working range has absolutely no affect on 'focus' accuracy (once set) on a basic camera lens.

Surely someone somewhere has undertaken a controlled test to prove/disprove the 'theory'.

I can't say I specifically tested for this because I'm quite certain that the theory is wrong.

A couple of years ago, I went downtown Ottawa to shoot a pano in conditions that were below -25C (well below the camera / lens operating specs). It was a busy spot, so it took me about 40 minutes between the first shot and the last one. The camera was carried from the car to the site in a reasonably well insulated camera bag. I set the focus once and did two full single row panos without focusing. The camera was likely just above 0C when I started and have no idea as to how cold it got by the end, but I did not refocus for the two series I shot other than the original setup. Sharpness was even throughout the entire series (shot 100% on manual with a tripod, so on automation was used; manual ISO, manual shutter speed, manual aperture and manual focus).

https://live.staticflickr.com/704/30...e310304f_k.jpg

Quote:

---

Originally Posted by Stagecoach

. . . Whilst the coefficient of expansion for each material is known and measurable I wonder if any designer has actually calculated how each individual component reacts in relation to each other component . . .

---

Canon certainly has, at the least, acknowledged that there is a difference.

Different expansion and contraction rates of individual elements, within the lens, has been cited as one reason for some of their lenses to be able to focus beyond Infinity.

(N.B. There are other reasons for lenses being able to focus beyond Infinity).

As a general comment, this feature is usually found on:
> Long FL Lenses
> Large diameter barrel lenses, especially those with a Fluorite Element.

(Noted also, nowadays, some W/A lenses are an inverted telephoto design and many of these lenses are designed to focus to beyond Infinity.)

I haven't used Nikon for a while, but a document easily to my hand now, from Nikon, also mentions this design factor, though does not specify the different rates of expansion of various lens elements:

“The reason for this [a lens being able to focus beyond infinity] is that in extremely hot or cold conditions the effective focal length of the lens may change due to the effect of the temperature on elements in the lens, which causes expansion or contraction compared to optical performance at regular temperatures. Depending on the construction of the lens elements, some lenses will be affected more than others. To compensate for this, the option to focus with the lens beyond the infinity focus position is available on some lenses.”

REF: Nikon Support Services; Article ID 000027772; Published 12/21/2016; “Why is it possible to focus manually beyond infinity on some lenses?”

***

My Post #33, mentioned Manfred's maths which referred to Aluminium (i.e. Lens Barrels).

In my Post #33, an implicit conclusion could be made that change in the size of the Aluminium lens barrel being so insignificant, was the reason for my disagreeing with "refocus every 30 minutes or when temperature change is 1°C"

Such a conclusion, is slightly flawed.

Specifically, I disagree with "refocus every 30 minutes or when temperature change is 1°C"
because:

1. a change of 1°C in the Ambient Temperature will have no real world effect on expansion and contraction of anything in a lens.

2. in reasonable shooting conditions, there is absolutely no reason to refocus every 30 minutes.

WW

Quote:

---

Originally Posted by William W

Specifically, I disagree with "refocus every 30 minutes or when temperature change is 1°C"
because:

1. a change of 1°C in the Ambient Temperature will have no real world effect on expansion and contraction of anything in a lens.

---

Doing some math and pretending that Manfred's 18mm example-lens is just a block of aluminum, or aluminium in Oz:

The co-efficient of expansion of aluminum is 23.1 um/m/degK from which the focal length changes by 0.4um for one degree of temperature rise.

At an aperture of f/5, I can't imagine an average sensor noticing that either. :D

Quote:

---

Originally Posted by xpatUSA

Doing some math and pretending that Manfred's 18mm example-lens is just a block of aluminum, or aluminium in Oz:

The co-efficient of expansion of aluminum is 23.1 um/m/degK from which the focal length changes by 0.4um for one degree of temperature rise.

At an aperture of f/5, I can't imagine an average sensor noticing that either. :D

---

I can't imagine a focus mechanism that can focus at that level of precision. In fact, I suspect the overall manufacturing tolerances for the various machined or moulded parts would at best be in the order of ±0.03mm / 0.001" as a best case. In fact I would not be surprised if the manufacturing tolerances are not a lot looser than that.

That being said; the operational specs on my camera body say that it can be used from 0 °C - 40°C / 32 °F - 104°F; we are looking at expansion of 40 x 0.4 = 16 µm; still a very small number.