Very confused on cropping!! Related to HS10

[RoysLaw]

You're confused because you are calculating it wrong. Here's how it should be calculated:

• 126mm on an APS-C sensor (1.5 crop factor) = approx 190mm
• 720(mm) / 190(mm) = approx 3.8
• (4368 X 2912) / 3.8 = approx 1150 x 766

Any 1150x766 portion of the original image will equal to a 720mm crop and will have the same FOV as the original HS10 image, (best resized to 1150x766 also for comparison).

Hope this helps and doesn't add more to the confusion. ;-)

Just a quick glance at your calc but if the scale factor is 3.8 then doesn't that apply to both width and height and you should divide 3.8 squared ?
Excuse me if I'm just muddying the waters!

Yes you are right but unfortunately it is very hard to express calculations in here without making it confusing. Just consider it as an equational summary. The calculation or answer is right though and I don't think it is difficult to follow what I was trying to express.

P.s. to get the forward slash (for division) use a backslash in front of it.

It actually works without as long as there is a space before and after. I may have deleted the spaces on that one when I did some edits. Glad you caught it though as that one expression made it confusing without it LoL! Thanks! =)

...

 

[zebarnabe]

Simple... we multiply for accounting for APC cropping factor, and 18MP to 10MP cropping... that goes like: 250mm x 1.6 x 1.8 = 720mm

to better understand what cropping does to focal length equivalents, is better to first explain what a full frame is.

A full frame is the term used to indicate the usual used area taken with a ratio 3:2 on a 35mm film, this area is a physical rectangle with 36mm per 24mm... using pitagoras theorem on can easily get the diagonal size of 43.26mm (i'm rounding here)

Canon APC size has a crop factor of 1.6, that means diagonal of the image is 27.04mm, that is obtainable when dividing 43.26 per 1.6 ... or full frame size is obtained by multiplying 27.04 per 1.6...

I'm not sure that Canon APC has these measurements as the crop factor can be applied to horizontal or vertical size instead of diagonal, but i would say it is a good guidance since diagonal represents lens image circle...

When cropping a 18MP digital image to 10MP, you're dividing both width and height by 1.8, to prove that the diagonal is indeed divided by the same ratio, let's do some more math....

let's consider that pixel is square and has 1 unit length of size
let's h be the height and w the width in pixels

in a 18MP image we have the following equation to count pixels:
h * w = 18
diagonal in pixel units is given by:
sqrt(h^2+w^2) = d
that is equivalent to:
h^2+w^2 = d^2

when applying the crop to 10MP we obtain the following to count the pixels
(h/1.8) * (w/1.8) = 10
to prove the previous equation:
h*w * 1/1.8 = 10 18 * 1/1.8 = 10 18 = 10 * 1.8 18 = 18 Correct!

proceeding to diagonal size calculus, let's consider nd as the new diagonal:
(h/1.8)^2 + (w/1.8)^2 = nd^2
h^2*(1/1.8)^2 + w^2*(1/1.8)^2 = nd^2
( h^2+w^2) * (1/1.8)^2 = nd^2, applying h^2+w^2 = d^2,
d^2 * (1/1.8)^2 = nd^2
d^2/1.8^2 = nd^2
(d/1.8)^2 = nd^2, since all values we are dealing with are positive,
d/1.8 = nd, or
d = nd*1.8

so when applying a crop on a 18MP to 10MP you are effectively applying a crop factor of 1.8 to diagonal...

and so 250 * 1.6 * 1.8 = 720!

This is a lot more easier to explain using a geometric demonstration * sigh *

 

[Mister T]

You're confused because you are calculating it wrong. Here's how it should be calculated:

• 126mm on an APS-C sensor (1.5 crop factor) = approx 190mm
• 720(mm) / 190(mm) = approx 3.8
• (4368 X 2912) / 3.8 = approx 1150 x 766

Any 1150x766 portion of the original image will equal to a 720mm crop and will have the same FOV as the original HS10 image, (best resized to 1150x766 also for comparison).

Hope this helps and doesn't add more to the confusion. ;-)

Just a quick glance at your calc but if the scale factor is 3.8 then doesn't that apply to both width and height and you should divide 3.8 squared ?
Excuse me if I'm just muddying the waters!

Yes you are right but unfortunately it is very hard to express calculations in here without making it confusing. Just consider it as an equational summary. The calculation or answer is right though and I don't think it is difficult to follow what I was trying to express.

P.s. to get the forward slash (for division) use a backslash in front of it.

It actually works without as long as there is a space before and after. I may have deleted the spaces on that one when I did some edits. Glad you caught it though as that one expression made it confusing without it LoL! Thanks! =)

Sorry, I did say I'd had a quick glance, and yes your answer is right at just under 1MP, and yes it is hard to write the squared bit into the equation and make it clear - I've just tried several ways and the clearest I could manage was:

(4368 / 3.8) x (2912 / 3.8) = 1150 x 766
Sorry for being so pedantic

Not sure many HS10 users would agree that the HS10 Image should be resized from 10MP to 1MP for the comparison! Yes I know that's fine for most prints but I don't think many pixel peepers will agree ... (ducks his head and runs for the trenches) ...
I would suggest a compromise resolution of, say 3 or 4MP, for comparison.

The interpolated DSLR image might look soft but, in my experience, a bit of PP in Photoshop and you can easily convince people that the image is actually of a higher resolution (in the same manner that all digital cameras kid us by having, for example, 10M light cell receptors but really only 2.5M pixels of data - as each pixel uses a red, a blue and 2 green cells - if my understanding is correct) .

 

[Mister T]

so when applying a crop on a 18MP to 10MP you are effectively applying a crop factor of 1.8 to diagonal...

and so 250 * 1.6 * 1.8 = 720!

This is a lot more easier to explain using a geometric demonstration * sigh *

Ok, I'd assumed that a 250mm lens on the Canon was the 35mm equivalent focal length (i.e. the crop factor of 1.6 had been applied and the actual focal length of the lens was 250 / 1.6 = 156mm)

It's strange that Canon don't use 35mm equivalent focal lengths like other camera manufacturers so you can easily compare them.
My mistake...

 

[DaLiV]

Simple... we multiply for accounting for APC cropping factor, and 18MP to 10MP cropping... that goes like: 250mm x 1.6 x 1.8 = 720mm

may be easier to say: as HS 10 have 1/2.3 sensor - it crop factor is 6 and lenses is 4-120 mm ?

think all knew next:

• persist 3 types of lie : Lies, damned lies, and statistics ....

and all that mathematics workaraunds with cropping factors, image resizings, that we do now - is statistics ...

 

[zebarnabe]

Statistics are a diferent subject, this is just an scientific approach to understanding what cropping actually does in terms of focal length...

But... i keep saying that i would love to see a photo of the moon (because is a good subject to check for CA and sharpness) with some good 250mm lens (that cost more than a single HS10 ) from 550D ...

But you are right.... calculations aside what matters is the final result

*checks flickr for 550D shots*

 

[tdkd13]

What you say Canon does is what everyone does. None of the DSLR makers use effective focal length in 35mm equivelant terms on the lens or the camera. They all use true focal length. If you take the exact same lens, say a Sigma 200mm lens and put it on a Nikon APS camera the 35mm eqiv focal length is 200 x 1.5 = 300mm, put it on a Canon and the equv focal length is 200 x 1.6 = 320mm. The lens focal length never changed and what is printed on the lens will be 200mm.

Not sure where you have been seeing 35mm equivelant focal lengths but it probably wouldnt be on an actual lens. Even the bridge and point and shoot cameras print the actual focal length on the lens. In this picture of an S100fs it says "f=7.1-101.5mm" right on the lens, that is the true focal length. Below that is a Canon SX200 which says "5.0 - 60mm" again, that is the true focallength.





Ted

It's strange that Canon don't use 35mm equivalent focal lengths like other camera manufacturers so you can easily compare them.
My mistake...

--
http://photobucket.com/albums/y260/tdkd13/

 

[tdkd13]

I think the moon is a poor subject to test for sharpness unless the two cameras being tested were shot side by side. The reason is the biggest cause of soft moon pics is atmospheric conditions and those change minute to minute. The best way to test sharpness is actually test targets, but fur and feathers work pretty well.
Ted

Statistics are a diferent subject, this is just an scientific approach to understanding what cropping actually does in terms of focal length...

But... i keep saying that i would love to see a photo of the moon (because is a good subject to check for CA and sharpness) with some good 250mm lens (that cost more than a single HS10 ) from 550D ...

But you are right.... calculations aside what matters is the final result

*checks flickr for 550D shots*

--
http://photobucket.com/albums/y260/tdkd13/

 

[maisatomai]

Statistics are a diferent subject, this is just an scientific approach to understanding what cropping actually does in terms of focal length...

But... i keep saying that i would love to see a photo of the moon (because is a good subject to check for CA and sharpness) with some good 250mm lens (that cost more than a single HS10 ) from 550D ...

But you are right.... calculations aside what matters is the final result

*checks flickr for 550D shots*

You mentioned the moon which is my interest. The photo below is shot by me with TCON 17 but it actually not optimal as the TCON 17 pushes back my HS10 lens and I only shot at 87.4mm which I think is 21X.





Let me show you the link to 200mm DSLR moon shot. I was criticized many times for pointing out overexposed moon shot but the links before is the first 5 I got from google image (I am sorry tdkd13 but I accept your previously argument about moon shot partially. I still cannot accept the fact that all images shown on google image are wrongly exposed shots.).

200mm moon shot:











300mm moon shot:





(Great lens 300mm f2.8)







400mm moon shot:

 

[maisatomai]

Hi, I referring to a FF camera in this example

 

[Martyn_Bannister]

This one came out reasonably well. There are a few more at

http://www.flickr.com/photos/48607572@N04/sets/72157623923211536/

 

[tdkd13]

Of the batch of moon shots you just posted, most are not wrongly exposed, I'd say #2,6,7 and 14 are a bit off, a few more might be off by a half to a full stop. My main point all along with you had been exactly as stated in my post a few minutes ago. Moon shots are just too adversely affected by atmospheric conditions to be good judge of comparing cameras, unless shot side by side correctly. We have both seen what can be produced by the HS10 and by DSLR's shot correctly, no need going back there is there? I would easily capitulate the point that the HS10 does the best job in the $500 price range.

There are just so many better ways to judge this. That said, the HS10 is certainly capable of getting good shots of the moon, if you feel the HS10 offers the best solution for shooting the moon I wont argue any more.
Just for the record here is the first HS10 moon shot I found on Google images:



Ted
--
http://photobucket.com/albums/y260/tdkd13/

 

[Kim Letkeman]

Not sure where you have been seeing 35mm equivelant focal lengths but it probably wouldnt be on an actual lens. Even the bridge and point and shoot cameras print the actual focal length on the lens. In this picture of an S100fs it says "f=7.1-101.5mm" right on the lens, that is the true focal length. Below that is a Canon SX200 which says "5.0 - 60mm" again, that is the true focallength.

And , the information isn't useful at all for common comparison unless you know the crop factor ...

--
http://kimletkeman.blogspot.com
http://letkeman.net/Photos

 

[Kim Letkeman]

Let me show you the link to 200mm DSLR moon shot. I was criticized many times for pointing out overexposed moon shot but the links before is the first 5 I got from google image (I am sorry tdkd13 but I accept your previously argument about moon shot partially. I still cannot accept the fact that all images shown on google image are wrongly exposed shots.).

Most of those actually are wrongly exposed ... very bright or very dark. A few really nice images, but mostly a blur fest.

The fact is that most people look at a technical subject like the moon and assume that anything they get must be the best they can get. That's just plain wrong ... it takes time and many attempts to get good at moon shots. And one must cull the crappy shots and not post them for the world to see.

Because we all know from experience that, when poor work is posted, someone is going to use it out of context to illustrate some point or other. We all do it. The key is to avoid broad generalizations. The fact is that you can find many poor moon shots from almost any camera. Does that mean that none of them can do the job?

So here is my answer ... I've posted this before, but just to remind you that it can be done, but requires considerable processing to pull out this level of detail.



So yes ... the HS10 takes a very good moon shot.

--
http://kimletkeman.blogspot.com
http://letkeman.net/Photos

 

[maisatomai]

Ted, the HS10 image is surely overexposed and the google image search "HS10 moon" gave me this



I not arguing HS10 produced the best moon shot. You got my point wrong. A look at your gallery won my respect honestly.

My argument this time is about many claiming that a 200mm DSLR beat HS10 hands down. I just showing various moon shot at different focal length to show it is not necessary true.

 

[maisatomai]

In statistic, people called it sampling bias right? I am not depending on 1 photo, but more than 15.... All these photo are from different people... I can continue my results and show you 100 different moon shot and I can tell you most of them are almost the same quality. It will be faster if you browse through 100 pages of google image yourself.

I accept the fact that many people here are very good and experienced photographers. But I cannot accept the fact that the other 100 people out there holding DSLR are newbies who does not know anything about exposure. Hope you see the point I am trying to make. One of the "poorly" taken moon shot owner even has a web site teaching people how to take moon shot.

Btw, I have seen great moon shot, liquidstone standard. But seriously, how many such photo can you found. I need to thanks TED for showing me such nice photo. I really amazed.

Once again, I not arguing on the fact that HS10 takes great moon photo (I am saving up to buy a Sigma 500mm f4.5 and FF after seeing photo shown to me by TED). I am arguing on the fact claiming "DSLR with 200mm" beats HS10 in all aspect.

Let me show you the link to 200mm DSLR moon shot. I was criticized many times for pointing out overexposed moon shot but the links before is the first 5 I got from google image (I am sorry tdkd13 but I accept your previously argument about moon shot partially. I still cannot accept the fact that all images shown on google image are wrongly exposed shots.).

Most of those actually are wrongly exposed ... very bright or very dark. A few really nice images, but mostly a blur fest.

The fact is that most people look at a technical subject like the moon and assume that anything they get must be the best they can get. That's just plain wrong ... it takes time and many attempts to get good at moon shots. And one must cull the crappy shots and not post them for the world to see.

Because we all know from experience that, when poor work is posted, someone is going to use it out of context to illustrate some point or other. We all do it. The key is to avoid broad generalizations. The fact is that you can find many poor moon shots from almost any camera. Does that mean that none of them can do the job?

So here is my answer ... I've posted this before, but just to remind you that it can be done, but requires considerable processing to pull out this level of detail.



So yes ... the HS10 takes a very good moon shot.

--
http://kimletkeman.blogspot.com
http://letkeman.net/Photos

 

[maisatomai]

Mr Kim Letkeman, went to your blog and looked at the moon taken by HS10. It is really nice. I think is the nicest moon I ever seen from HS10!!!

 

[zebarnabe]

Yes .... you are correct, sharpness charts and fine textured subjects give a better benchmark and have more controlled variables

 

[Kim Letkeman]

In statistic, people called it sampling bias right? I am not depending on 1 photo, but more than 15.... All these photo are from different people... I can continue my results and show you 100 different moon shot and I can tell you most of them are almost the same quality. It will be faster if you browse through 100 pages of google image yourself.

Actually, it will be faster if you stipulate to my assertion that most people don't know how to shoot the moon

I accept the fact that many people here are very good and experienced photographers. But I cannot accept the fact that the other 100 people out there holding DSLR are newbies who does not know anything about exposure.

The evidence is very clear.

Hope you see the point I am trying to make. One of the "poorly" taken moon shot owner even has a web site teaching people how to take moon shot.

No, I don't see the point at all. Less than competent people try to teach all the time. Any member of the FTF has witnessed that numerous times.

Btw, I have seen great moon shot, liquidstone standard. But seriously, how many such photo can you found. I need to thanks TED for showing me such nice photo. I really amazed.

Here's another one for you from Michael Stecker's page here:

http://mstecker.com/pages/moonjrf_myref.htm

Once again, I not arguing on the fact that HS10 takes great moon photo (I am saving up to buy a Sigma 500mm f4.5 and FF after seeing photo shown to me by TED). I am arguing on the fact claiming "DSLR with 200mm" beats HS10 in all aspect.

Actually, the HS10 can take a great moon photo. It's not easy, but it can be done.

Whether the HS10 is a bit better than a dSLR with a 200mm lens or not really comes down to skills, the quality of the air and the specific dSLR. There are no hard and fast rules with something requiring such a long lens. But ... the dSLR with consumer zoom or short prime certainly gives the HS10 a run ....

Here's a D300 with a 300mm prime and a teleconverter ... this is a 100% crop:



The D300 with a hand held shot from a 300mm consumer zoom ... the HS10 cannot touch this hand held ...



How about a rather short 180mm and a teleconverter?



--
http://kimletkeman.blogspot.com
http://letkeman.net/Photos

 

[Kim Letkeman]

Mr Kim Letkeman, went to your blog and looked at the moon taken by HS10. It is really nice. I think is the nicest moon I ever seen from HS10!!!

Thanks ... that's why I am so sure that the HS10 is good at this specific task.

--
http://kimletkeman.blogspot.com
http://letkeman.net/Photos