Calculation of a required focal length using a crop of known focal length.

[digitalirony]

I'm not sure if this is the correct forum for this question but as a Canon user it is the forum I'm most familiar

with.<br>

<br>

I have a 300mm lens and used it recently to take some photos of a <a

href="http://en.wikipedia.org/wiki/Cricket">cricket match</a>.<br>

<br>

The lens was not long enough to frame the photos of the action as I would have liked but as I was (severely)

cropping the image to the framing I desired I was wondering...<br>

<br>

If the medium jpeg file I recorded is 2816 pixels x 1880 pixels and I know the dimensions of the cropped image

that would give me the focal length equivalent that I would have liked, say 1408 pixels by 940 pixels, can I use

this to calculate the approximate equivalent focal length of the lens that I would need to capture the cropped

area as a full (APS-C) framed image?<br>

<br>

I've tried to work out the maths but got stuck with the angle of views to pixels business. Trigonometry is

clearly not my strong point.<br>

<br>

Tamron has a Focal Length Comparison Tool at <a

href="http://www.tamron.com/lenses/learning_center/tools/focal-length-comparison.php">http://www.tamron.com/lenses/learning_center/tools/focal-length-comparison.php</a>,

which I have looked at and approximated that the length required would be about 500mm but it is hard to determine

using their tool.<br>

<br>

Has anyone got a quick and dirty way of doing the calculation?<br>

<br>

<img title="cropped framing" src="http://www.digitalirony.com/blog/uploaded_images/2008-06-28/1451s.jpg"

width="400" height="267">

[rainer_t]

-- "Has anyone got a quick and dirty way of doing the calculation?"

 

focallength * originalsize / croppedsize

 

300mm * 2816 / 1408 = 600mm

[digitalirony]

Wow, that was easier than I though.

 

I didn't realise that it was a straightforward linear relationship - doh!

 

Many thanks.

[rainer_t]

-- "I didn't realise that it was a straightforward linear relationship - doh!"

 

Well in reality it isn't, but usually the results are good enough.

[sarah_fox]

Jon wrote: "I have a 300mm lens and used it recently to take some photos of a cricket match."

 

Wrong lens! You need a macro to photograph crickets! :-P

 

Sorry... couldn't resist!

[bobatkins]

I have a field of view calculator on this page - http://www.bobatkins.com/photography/technical/field_of_view.html

 

A simple linear realtionship is fine for telephoto and even normal lenses, but for wideangles the math is a little more complex if you want accurate results.

[digitalirony]

Thanks Rainer, Sarah & Bob, for helping me to work it out.

 

There's no hope of me getting access to a 500mm or 600mm lens any time soon so I'll have to keep cropping. Might be able to borrow a 100mm - 400mm zoom which would be interesting.

 

The real problem is the DOF. although I'm sitting quite some distance from the pitch (about 100m), and the people in the background are about an equal distance as well at 300mm @ f/5.6 it is still quite difficult to separate the players from the audience (I would say fans, but this is cricket!).

 

It looks worse on the downsized images but I'm guessing that's why the pros are all using 600mm f/4.0 lenses.

[chris_adams11]

Something else you could do is shoot full res jpegs which would give you more pixels to crop from.

[sarah_fox]

You thanked me for my help, and all I did was crack a lame joke! ;-) OK, here's some help, maybe...

 

I don't know what you're using, but others have reported good results with the 70-200L and a (Canon) 2x teleconverter. If you've got the 70-200, you're halfway there.

 

I was going to suggest a cheap Russian 500/8 mirror, but you'd have to space that with a 12mm tube to mount it on most Canons, which would put the max focus at about 21.3 m -- not long enough for you. Perhaps use a teleconverter instead of a tube to achieve the spacing? Then image quality would suffer quite a bit.

 

If you're a hobbyist and tinkerer, you can adapt to a big telescope with adapters available on auction. I had considered doing that with a 10" telescope of mine. It's maybe not the best image quality, but you can't beat the light gathering power!

 

If you want a real solution with quality optics, I'm afraid it's really going to cost you! The 100-400 is probably the closest thing within the grasp of mere mortals.

[digitalirony]

Hi Chris,<br>

<br>

Although we are drifting off topic here, I catually record RAW+Medium JPEG. I use the JPEGs as a guide and the smaller files save space on teh CF card and my HDD. If the image is worth printing, I work back from the RAW file. If I'm just doing some small prints for someone else (friend or relative) the Medium JPEGs are large enough.<br>

<hr>

Hi Sarah,<br>

<br>

Again drifting off topic, I looked into your suggestion of a mirror lens and there are a few that come in canon fit that are cheap. But a bit of research suggested, the crops from the 300mm end of my zoom lens blown up twice the size would be of comparable if not better quality.<br>

<br>I agree with you that the 100-400 is probably my most realistic change of getting closer to the action (it is cricket, but I suppose you can still call it action). I'm going to see if I can convince someone to let be borrow one for a day.<br>

<hr>

Once again, many thanks to everyone who contributed answers to my original question.

[sarah_fox]

Hi Jon,

 

The Russian mirrors are actually supposed to be reasonably sharp. Their principal failing is that they're not coated. However, they are also rather simple and have very few surfaces to flare.

 

By the way, if you want to get some idea how fast a lens you will need to blur out the "audience" to your satisfaction, check into the object field method of computing depth of field. Harold Merklinger did a series of four articles on the subject that are quite interesting. You can easily google them. I think I'm understanding that the crowd is twice as far from you as the action you're trying to photograph. You would need a HUGE aperture (about the width of someone's head) to completely isolate the action -- not likely to happen with any lens you're going to find. Half that diameter would yield figures that are recognizeable as people but still pleasantly blurred. That would be 100mm aperture, give or take. So for a 400mm lens, that would be f/4. Thus, I think the 100-400 f/4L would probably do what you want it to do.

 

Good luck!