Frame size vs magnification

[thegableguy]

As an amateur / semi-professional photographer of the better part of a decade, I understand the difference between focal length and crop factor. Though the FOV is the same, there is a significant difference between a photo taken with a full frame DSLR using a 50mm lens and one taken with a crop sensor DSLR using a 35mm lens. I absolutely understand the difference.

However.

I have two full frame DSLRs and one crop sensor. They're all 24 megapixels.

When using a telescope focused to infinity, surely the difference between the two is literally exactly the same as increasing the magnification..? I keep having people insist upon me that they're not the same thing. I contend that with these cameras having the same resolution but different size sensors they are EXACTLY the same as increased magnification.

Please explain why I'm wrong!

[Merlin66]

Chris,
I think the problem (if there is a problem) is related to your photographic experience. The application of a camera to a telescope and the resulting image has nothing to do with crop factors etc.
What different sized chips affect is the field of view recorded, not the "magnification" of the image within the field.
The focal length (usually fixed) for the telescope determines the plate scale of the image (arc sec/ pixel).
Try CCDCalc as a means of getting your head around the issue. It allows you to input various sensors and view the probable outcome with sample images of well known astronomical objects.
http://new-astronomy-ccdcalc.software.informer.com/1.5/

Hope this helps.

[sharptrack2]

Hi Chris,

I'm sure the experts will chime in anytime now, but this is my understanding of the difference...

It has to do with the area each pixel covers (pixel size), you may have the same number of pixels but a cropped sensor will have a smaller effective pixel size, effectively increasing pixel density, making the image appear larger for a given frame.

I'm not pursuing astrophotography but have been around long enough to gleaned a few concepts. I think this same principle applies to the example you gave.

All I can add now is that I hope I'm close to teh mark and haven't create a pool of confusion... ;)

[sharptrack2]

Merlin to the rescue!! :rofl::thumbsup:

[thegableguy]

I'll have a look at that website, but perhaps you can tell me what difference there would be between a 24MP crop sensor DSLR on a 1000mm telescope and a 24MP full-frame DSLR on a 1500mm telescope. As far as I can tell, there won't be any difference (besides vignetting etc which is beside the point I'm making).

If they are indeed identical in the images they take, surely it's not unreasonable to call the effect magnification.

[bojan]

There is no crop factor! This (totally useless IMHO) "invention" was introduced to "help" people transition from "standard" 24x36mm film size to digital sensors which were smaller in the beginning... and it only introduced additional confusion.

The only relevant thing to AP that exist is "scale" of the camera.. pixels/arcsec.

Or FOV.

[thegableguy]

Right, because there should only be one size of sensor for all DSLRs...? Not sure I understand your frustration with what I'd say is a pretty apt term. I think both have their uses.

Anyway. Doesn't really answer the question. What is the difference between those two examples in my below post, and how is it different to increased magnification?

[thegableguy]

I guess what I'm asking is how an increase in pixels / arc second DOESN'T translate directly to an increase in magnification. Everyone suggests they're these totally different things that have nothing to do with one another, but I honestly don't see the difference.

[bojan]

Hi Chris,
OK, in my last reply I was too hasty and not precise enough, my apologies..
The answer to your question about difference between say 24Mp sensors (cropped and FF) and 1000mm and 1600mm FL (respectively) is; there is no difference in terms of arcsec/pixel size (expressed in /um or nm for example). Please note, this number has dimension (angle/length).
However, this doesn't tell us the resolution ("sharpness" of the image) of those two systems, in general it will not be the same (It would have been, if we were dealing with ideal optics and infinitely small wavelength of incoming light).

In real world, the resolution (arcsec/pixel) will be determined by the linear aperture (diameter) of the lens, it's FL and (of course) it's optical quality and physical pixel size of the sensor.

In AP, term "magnification" should not be used, because it is non-dimensional number, reserved for visual magnification, which is defined as ratio between FL's of objective (primary) and eyepiece. It also tells us how much the angular size of the object observed through telescope appears to us compared to it's angular size when observed with unaided eye.

How do you define "magnification"?

If we keep F/ number the same, aperture of the telescope will be 60% larger.. resulting in (theoretically) 60% better resolution.
Now, if physical pixel size is larger than angular resolution of the optical system, there will be no improvement (you will only be able to record ~1/2 dimmer stars in the same exposure time (because amount of photons collected is dependent on square of the aperture, not it's linear size)

[Camelopardalis]

Chris, I can understand your frustration...

Increasing the pixels per arc second increases the resolution and not the magnification.

The difference being that the angular dimensions of an object at the focal plane remain the same as captured through your telescope regardless of sensor/pixel size, you're just sampling it differently with different sized pixels. If you could measure the imprint on the sensor of the object your imaging, it would be the same regardless of sensor.

Magnification alters the apparent dimensions of the object at the focal plane. If I change the 10mm eyepiece for a 5mm, I double the magnification and apparent size of the object, and also make it less bright. You can measure this by projecting the image onto a piece of paper (as many do for solar viewing).

Subtle, but not the same :D

[Merlin66]

Maybe easier for you....
Say you want to image Jupiter which is 30 arc sec diameter.
The physical size of the "image" of Jupiter at the focal plane will depend on the effective focal length.
The only way to "magnify" or increase the size of the image is to increase the effective focal length.
Example
1000mm fl = 145.4 micron (0.2 arcsec/micron)
1500mm fl = 218.2 micron (0.137 arcsec/micron)
2000mm fl = 290.9 micron (0.1 arcsec/micron)

The size of the pixel will determine the plate scale (quoting a 24 Mpixel array doesn't in itself mean anything.)

So depending on the pixel size say 5 micron then the sampling/ plate scale would be:
1000mm fl = 1 arcsec/pixel
1500mm fl = 0.685 arcsec/pixel
2000mm fl = 0.5 arcsec/pixel

That's about it.

[thegableguy]

Okay, thanks everyone. I think I'm getting it. Basically it's just semantics from what I can tell.

Firstly, here's a post I wrote about the difference between magnification and focal length:

https://m.facebook.com/story.php?story_fbid=27177076319876 1&id=204996913209480

I used the word magnification recently in a different IIS thread when saying I'll miss the magnification of the crop sensor and, as always, had someone insist that it's not magnification, it's this that and the other...

Look, I get what you're all saying. I understand the difference in terms and how the word "magnification" doesn't quite fit in the context of AP. And yet, I still think (and haven't yet been presented with reason to think otherwise) that, all things being equal, a crop sensor DSLR offers a 1.5x magnification over a full frame. I don't see it as an incorrect way of expressing it.

Does that make me just plain wrong? Wouldn't be the first time...

[Merlin66]

Chris,
You say:
""And yet, I still think (and haven't yet been presented with reason to think otherwise) that, all things being equal, a crop sensor DSLR offers a 1.5x magnification over a full frame.""
In your Facebook page you say:
"" I can take a huge wide image and crop out 95% of it; what's left will be low resolution""
That's not true. The remaining image as recorded remains unchanged.

On what basis?
The size of the image (the usual measure of "magnification") or the number of pixels in the image (just a sampling rate)
or the physical size of the image as it appears is each frame (just a Region of Interest cropping exercise - the image remains the same)

Help us to help you...
What's the measure you're applying?

[thegableguy]

Okay, let's try this way:

Photo A is 24MP.
Photo B is also 24MP.

Photo A is taken through a 1500mm telescope.
Photo B is through a 1000mm telescope.

Photo A camera is full frame.
Photo B camera is crop sensor.

What is the difference between the two photos? Same exact FOV, same exact resolution... how are they different?

If they're the same, why can't it be said that Camera B offers 1.5x magnification over Camera A?

[Camelopardalis]

Guilty as charged. Apologies if you took it the wrong way. In future, I'll resist the temptation to help newbies in the acerbic nature for which I'm known and loved...

Semantics or otherwise, there are technical differences as described above.

Get it :P

[thegableguy]

Acerbic is fine! I'm similarly adored for my own acerbic wit. I bring witheringly scornful joy wherever I go.

Just wish I understood how I'm wrong about this.

[bojan]

Chris,

As you put it below, They are the same.




The "magnification" is meaningless term as far as this optical system is described.

[thegableguy]

Okay, cool - I think that's where we finally fetch up against the wall. I'd say the term magnification is the perfect way to describe the difference, as the image from one is identical to the other except, well, magnified 1.5x; you disagree. I can't say I fully understand why, but maybe I don't need to.

Well that was a fun way to procrastinate from today's admin..! Cheers all.

[Merlin66]

Chris,
If the full frame (assuming 24 x36 mm) has 24Mp and the cropped frame (APS-C, 15 x 23 mm) the same pixel count, the pixel size in the APS must be smaller.
Using the 1500mm fl, with the full frame the FOV would be larger than the same using the APS. at 1000 fl.
If you used the full frame at 1000mm the FOV would be larger than the FOV in the APS.
You are comparing an image 15mm wide with an image 24mm wide.
You need to compare them at the same focal length.
As Covington says about comparing full and APS frames in his "Digital SLR Astrophotography", p17:
"This has nothing to do with zooming (varying the focal length) in the normal sense of the word."

[tempestwizz]

From what I can understand, everybody is right. It's just terminologies and semantics making things seem more complicated than they are.
For the same final sized picture dimensions, an APS sized sensor of 22 by 15 mm will produce what appears to be a magnified image over what a full frame 35 by 24 mm produces.

If both sensors are 24 megapixel, then the pixel dimensions for the APS must be smaller than for the full frame sensor. Accordingly, the apparent resolution for a particular object on the output pictures would seem better for the APS than for the full frame.

For a full frame sensor to produce the same resolution of a particular object, a lens with a focal length 1.6 times longer would need to be used.

Magnification is a relative term. An APS image appears to be magnified by 1.6 times more than a full frame image through the same lens. But, the APS image does not show as wide a field of view as the full frame.

The practical ramifications of the above:

The apparent increase in resolution for the APS sensor for a particular target object comes at a cost. The number of photons per unit of time hitting each of the smaller pixels is less. Hence more exposure is required to achieve the same apparent brightness. The longer exposure can increase noise produced in the sensor, and or allow tracking errors to blur the potentially higher resolution image.
One more compromise amongst the many others in this field.

HTH

Brian

[thegableguy]

Brian FTW.

Great answer mate. Thank you. Helped clarify a few small points I'd not considered, like pixels being smaller means they'll be inherently noisier.

Cheers again. Happy to let it rest there.

[barx1963]

I think the word magnification should only be used in relation to visual. If you think about an object in the sky it has a "size" which is expressed in terms of angles. So the moon for example is 30' wide. If I look at it in a scope and it appears to be 10 degrees wide, then it is reasonable to talk about it being magnified 20x. Personally I dislike using the term as it conveys the idea to non astronomy people that the purpose of a scope is to magnify, which of course is only a partial answer.
If I stick a camera on the scope, take an image and then put it my computer screen in what sense is it "magnified". The image is not bigger than the moon (unless my screen is a few thousand kms wide) and you cannot really quantify the image size in terms of angle, unless I can state the conditions under which it is being viewed ie. I could state that is this image is viewed on a at such and such a size and I am sitting at such and such distance from the screen then the image is so many degrees wide. That is obviously silly o that is why for imaging we are usually more interested in talking about true field of view and resolution. At least that is I think about it!!
When I joined the local camera club here I was totally confused when they all talked about using a crop factor camera to get more "magnification" or increase their focal length, which simply made no sense to me. Using a smaller sensor, to my way of thinking made no more difference than loading an image into PS and cropping it. And the resolution is set by the combination of focal length and pixel size, sensor size is irrelevant. Sensor size is only relevant in discussing the size of the field of view.

Note: as I always say when offering my silly opinions in imaging matters, I am happy to be contradicted by more experienced people!!

Malcolm

[thegableguy]

Look, I really do get the whole magnification-only-applies-to-visual stance. I know where you're coming from. I know the actual magnification of any given image depends on the screen or print size, and that resolution & cropping further complicates it.

All I'm talking about is the example of two cameras with the same megapixel resolution, but different size sensors. By any test, in any context, you can reasonably say that the crop sensor DSLR will produce a more magnified image than the full frame.

That's all I'm saying. Say the term magnification doesn't apply if you like. I respectfully disagree. It's close enough to the definition of what magnification is for me.

[Merlin66]

Chris,
OK.
Next time you're at the telescope take a image of say a galaxy (or planet) with both cameras ,the full frame and the APS.
Measure the image size of the galaxy/ planet (in micron) - you'll find it to be the same.

[thegableguy]

Of course it will be. Never suggested otherwise.

What do you think is easier to say and understand:

"This photo is a magnification of that photo"

"This photo is a higher-resolution but cropped version of that photo"

I'm going with the first one. Because whatever tiny difference there is absolutely doesn't matter.

[Camelopardalis]

Referring to your earlier reply and this...it's not so much about the "wrong" answer, it's just more a case of "accepted" or commonly used terminology. When referring to deep space objects, folk tend to reference them by resolution i.e. 1.5"/pixel rather than the M-word. Even when it comes to planetary, few refer to "Jupiter captured at 1000x", even though it is calculable, most just list the optical configuration used to capture it.

In your A or B example, all things being equal, they'd give the same result, but it's just not as simple as that. As has been mentioned, pixels are not equal. The resolution of the bigger scope is better than the smaller scope. Then there's the atmospheric conditions to consider, which could make all the difference or make them practically indistinguishable. My money would be on the bigger scope and bigger pixels every time though :D

[thegableguy]

Again, I'm not using the M word as a description of what's being imaged. I really do fully appreciate the myriad reasons why that doesn't work. Honest! I really, really do.

I'm using it to describe as simply as possible the difference between the two cameras, saying one camera will produce an image that is essentially a 1.6x magnification of the other.

Meh. I'll stick with this terminology until I get sick of people correcting me. Gimme another six months and I'm sure I'll come around...

[bojan]

Yes it was..

The physical image formed on FF sensor from 1,5m FL lens is bigger compared to smaller sensor and shorter lens.
However, after sampling, it has exactly the same number of pixels.. So it's not magnified, it is exactly the same.
Or magnification is exactly 1.

However, if we had a film photography, we could have been talking about magnification of 1.5x

[thegableguy]

Yes, the image would be the same, 1:1, with the two different focal length telescopes. With the same telescope, the image from the smaller (but same resolution) sensor would, for all intents and......

Ahhhh,surely you get my point by now. Broken record time. All good. I think we understand each other, we just express ourselves differently.

[bojan]

Yes :)