PDA

View Full Version : Night / Astro Help ?



sanger
09-05-2017, 10:00pm
As an example....Would a 10mm-20 lens @ 10mm f3.5 let in more light than an 18-55 @ 18mm f3.5.
Does that make sense ?

ameerat42
09-05-2017, 10:27pm
Night/Astro... They are not equivalent shooting conditions.

F-stops: give an indication of the intensity of light hitting a sensor for ANY lens.
But - in fact, different lenses at the same f-stop TRANSMIT different amounts of light
(because of construction factors, etc.)

Now, all of the foregoing certainly holds for "extended objects" - those that have "area".
But, if shooting stars, you have to consider what happens to "point sources" of light.
Stars do not have "area". That is (in practical terms), you cannot "magnify" the size of a star.
For these, simply the diameter of the lens aperture determines their brightness at the sensor.

And then there are complications for "skylight", "nebulae", and "other factors I can't think of just now".

So for an answer to your question, you'll see there are in fact two: "Yes" for "point sources", and "No"
for "extended sources".

And for those "mathematical moments" you're bound to have, use the following formula to work out what
size aperture you get for a lens at any given f-stop:

A = FL/(f-stop)

A is Aperture, FL is focal length.

And now "Good Night/Astro":p

sanger
10-05-2017, 2:58pm
Night/Astro... They are not equivalent shooting conditions.

F-stops: give an indication of the intensity of light hitting a sensor for ANY lens.
But - in fact, different lenses at the same f-stop TRANSMIT different amounts of light
(because of construction factors, etc.)

Now, all of the foregoing certainly holds for "extended objects" - those that have "area".
But, if shooting stars, you have to consider what happens to "point sources" of light.
Stars do not have "area". That is (in practical terms), you cannot "magnify" the size of a star.
For these, simply the diameter of the lens aperture determines their brightness at the sensor.

And then there are complications for "skylight", "nebulae", and "other factors I can't think of just now".

So for an answer to your question, you'll see there are in fact two: "Yes" for "point sources", and "No"
for "extended sources".

And for those "mathematical moments" you're bound to have, use the following formula to work out what
size aperture you get for a lens at any given f-stop:

A = FL/(f-stop)

A is Aperture, FL is focal length.

And now "Good Night/Astro":p

I think that means maybe :)

Cage
10-05-2017, 3:15pm
As an example....Would a 10mm-20 lens @ 10mm f3.5 let in more light than an 18-55 @ 18mm f3.5.
Does that make sense ?

Depends how you want to quantify it.

A 10mm lens on a FF camera would have a FOV of 130° and will collect light from a much bigger slice of sky than the 18mm's 100° FOV.

I'm guessing that the amount of light hitting each pixel on the sensor would be similar for both lens.

piczzilla
10-05-2017, 5:59pm
Hi Sanger, not sure about the amount of light, but I've always thought wider is better for astro because of the rule of 600 (unless you're after trails)

arthurking83
12-05-2017, 6:10pm
Others have already discussed the pro and cons of each focal length factor, but I wanted to mention T-stops

T-stops are usually referred too for video camera lenses(ie. video work), and is relevant to your query.

All lenses are made unequally. Even tho they sound equal, they usually aren't .. in terms of actual light transmission through the lens and to the sensor.
Aperture value is really only a guide, and what you've asked here about the amount of light at f/3.5 for those two lenses is a transmission factor ... not really an aperture related factor.

So while the aperture value only partially helps to determine the light transmission through to the sensor, it's only a partial factor.
The glass used, how many elements, the quality of those elements, the types of coatings used on those lens elements .. are they glass at all? many lenses use plastic lens elements .. as an example the 18-55 kit lenses already referred too.

So there are more factors than simply aperture values going on.

The best way to figure this out is to use a well known, and respected review site that does do transmission testing on lenses. (some do, others don't!).

One of the things I do like about DxO is that they measure this and have a consistent method of doing this(so far as I know) .. but that's not overly important.
They have tested many lenses, so area rich source of info with respect to some lens test data.
(ps. I still don't understand their PMpix ratings and some other suchlike data, so they can come across as confusing .. and not everyone agrees that they have a good source of info .. but to get back on topic ) ..

I've had a look at a few lenses you're probably interested in.

1/. Sigma 10-20 f/3.5[Dx]
2/. Nikon 18-55 VR(old version)[Dx]
3/. Nikon 18-55 VRII[Dx]
4/. Nikon 18-105VR[Dx]
5/. Nikon 18-35 f/3.5-4.5[Fx]

Sigma 10-20 definitely transmits better than any of those lenses below it. Nothing to do with focal length or aperture, almost certainly due to Sigma's use of whatever glass and coatings.
This lens loses no light on a D3200(closest I could find to your model camera). In fact by their measured data, it's actually brighter than the specs state!(very rare). At f/3.5(and 10mm) it's T-stop is actually f/3.3
At f/3.5 and 18mm it does lose a very slight 0.3Ev tho.

** a T-stop is an adjusted transmission value to indicate how much actual light gets through compared to the aperture value. It doesn't mean that the actual aperture value of the lens is a T-stop, it's just that the lens design only transmits that much light relative to the aperture value.

Nikon 18-55VR loses a bit of light(which is usual even for the super high end lenses too mind you!) .. at T-stop 3.8. So that is, while the aperture value may well be a true f/3.5 because the glass in the lens is so lossy in it's ability to transmit, the effective light value of that aperture value is f/3.8. Your blur and DOF will still be the same, but the exposure will be off by that difference(t/3.8 vs f/3.5)

Nikon 18-55VRII loses about the same amount of light as the older model, with a slight difference in that it loses more as you zoom up the focal length range. ie. zoom in more, you lose a little more T-stop. The old model VR lens loses less T-stop at about the 35mm focal length range, and then loses more again as you approach the 55mm focal length.

Nikon 18-105VR has a T-stop of 4 at 18mm. Compare that to the 18-55's even, and now the lens is producing over 0.5Ev of lost due to the lens elements.
It's not a big difference, but it's a difference.
A note too: there are almost certainly two reasons for this situation with this lens. first is that it's a larger zoom range. more lens elements, to achieve this, more lost light through those elements. Previously I said that the aperture may be a true f/3.5 or whatever, but this is usually not true. Most lenses are a bit optimistic with their actual aperture size. Many f/2.8 lenses may actually be f/3.0 lenses. There are allowances in the way such tech specs are calculated to allow manufacturers to state that an actual f/3.0 lens is an f/2.8 lens. So aperture will have an actual impact on this T-stop factor .. but usually it's the lens elements.

An interesting lens is the Nikon AF-S 18-35 f/3.5-4.5(note! not the older AF model of this lens).
This lens transmits the same light factor as the stated aperture value up to about 28mm, the final step in the focal length range to 35mm it loses very little. So while it's aperture at 35mm may be f/4.5, it's Tstop is actually 4.8
But at the wide end it's pretty much true.

In a partial conclusion it appears that the Sigma 10-20 f/3.5 is the better lens of the lot as it's actually brighter in transmission than it's stated specs imply! :th3:
But, the reason for that being a partial conclusion is that we haven't yet introduced vignetting into the equation.
You probably know what vignetting is, and you may even add a bit to your portraits and landscapes in PP, but you want as little of it from the actual lens.
Astro can be a good reason for wanting this from a lens, but also if you do stitched panorama jobs too. Vignetting kills the stitch due to differences in exposures at the corners.

Why I added the expensive 18-35 AF-S 3.5-4.5.
Where the Sigma loses about -2Ev at the corners at 10mm but dropping back to only -1Ev at 17mm(closest to 18 they have for that lens) .. the AF-S 18-35mm only loses less than -1Ev at the 18mm end. Basically due to the fact that it's an Fx format lens.

When you're doing astro and you want the wide field of view, usually you want all of that image to be as good as you can capture it.
So vignetting should also be taken into account.

Hopefully that all makes some sense to you .. it's not a simple topic to understand either!