On Thursday, 29 November 2012 at 13:43:46 +0800, RizThon wrote:
>> Indeed.  This is what has been puzzling me.  There are two different 8
>> mm fisheyes available for Olympus: the relatively expensive 8 mm f/3.5
>> from Olympus, and the 8 mm f/3.5 from various rebadgers (Bower,
>> Samyang, Rokinon).  The former costs about $800 and has a full 180°
>> diagonal angle of view.  The latter costs about $300, and from the
>> specs state an angle of 139.3° on Four Thirds.  From what I've read it
>> will only give a full diagonal 180° on APS-C cameras.
>
> I feel like the full 180° *diagonal* angle of view from Olympus and the
> 139.5° (as it's not specified it should be horizontal,

It's specified in the descriptions of other rebadged versions of this
lens.  Go to (for example)
http://www.bhphotovideo.com/c/product/769514-REG/Rokinon_FE8M_O_8mm_Ultra_Wide_Angle.html
and read:

  Ultra wide-angle 8mm fisheye lens with exaggerated perspective and
  approximately 180° angle of view, for dramatic effects

  Ultra-wide 139.3° diagonal field-of-view for 4/3 size image formats

The Bower badged lens is also available for Canon (APS-C).
http://www.bhphotovideo.com/c/product/635178-REG/Bower_SLY358C_SLY_835C_8mm_f_3_5.html
states:

  180° diagonal angle of view on APS-C image format.

> which is what is normally used) from Bower ends up being almost the
> same (you computed for a fullframe fisheye 144°)

Yes, but diagonal values look better for selling the product :-)

>> Since my last message I've been reading a bit about fisheyes, and it
>> seems that, like Hugin, there are various projections.
>>
>> http://en.wikipedia.org/w/index.php?title=Fisheye_lens&action=edit&section=12
>> gives some information, though I'm still trying to digest it.
>
> Each lens displays what it "sees", and there are usually distortions
> compared to reality. So even with "normal" non fisheye / non very wide
> angles, you have distortions. Those distortions are directly linked to the
> way the lens is built, and thus different lenses with the same FOV will
> have different distortions (even if they are not really noticeable).

Did you read the article?  It specifies numerous different
projections.  And the difference between 140° and 180° can't be
attributed just to distortion.

>> At least one of the images appears to show a horizontal angle of
>> more than 180°.  If anybody knows more details, I'd be interested.
>
> For me a real fisheye lens should have at least a diagonal of 180°
> for FF, or a full 180° circle for circular (or at least quite close
> to 180°). It's possible that some vendors use the term fisheye just
> because of the look of the pictures you can take.

Clearly it's related to the projection.

Greg
--
Sent from my desktop computer.
Finger [email protected] for PGP public key.
See complete headers for address and phone numbers.
This message is digitally signed.  If your Microsoft MUA reports
problems, please read http://tinyurl.com/broken-mua

Attachment: pgp6dWjPXRutb.pgp
Description: PGP signature

Reply via email to