My first post here, so greetings everyone and best wishes for the
holiday season.
I'm hoping to get some suggestions for how to estimate or calculate
the HFOV parameter used throughout Hugin, but more particularly for
the align_image_stack tool.
I shoot some of my macros with an Olympus E-system DSLR fitted with
two adapters, one or more extension tubes, and a reversed 50mm manual
lens, so it's not immediately obvious to me what HFOV the tool would
be best run with, nor how precise I must be to get good results. Any
help would be appreciated.
On 24 Dez., 21:32, tbransco <TFTAJLLYM...@spammotel.com> wrote:
> My first post here, so greetings everyone and best wishes for the
> holiday season.
> I'm hoping to get some suggestions for how to estimate or calculate
> the HFOV parameter used throughout Hugin, but more particularly for
> the align_image_stack tool.
> I shoot some of my macros with an Olympus E-system DSLR fitted with
> two adapters, one or more extension tubes, and a reversed 50mm manual
> lens, so it's not immediately obvious to me what HFOV the tool would
> be best run with, nor how precise I must be to get good results. Any
> help would be appreciated.
What macros do you do?
If you want to make focus stacks, the HFOV doesn't matter much: you go
in and out rectilinear and no transformation takes place. Use a rule-
of-thumb value.
On the other hand, if you were making macro panoramas, which is a bit
more unusual, you can do a 360 degree panorama, start with some
estimate for the HFOV and optimize for fov, then take the value you
get after optimizing. It doesn't sound like your setup is for
panoramas, though...
You can also meaure the value, if you know where your NPP is. Take a
picture of, say, a horizontal ruler and take a reading of how many cm
of the ruler fits in your image. Let that be w, and d the distance of
the ruler from your NPP. If v is your horizontal field of view, you
have w / ( 2 * d ) = tan ( v / 2 ), so v = 2 atan ( w / ( 2 * d ) )
I posted a much longer response that has disappeared, but it came down
to a big thank you for the excellent information you provided. Your
first point about there being no transformation carried out during the
stacking operation was a relief, and the technique for measuring the
HFOV is a keeper for later reference to be sure.
Thanks again for taking the time and sharing your insights.
> On 24 Dez., 21:32, tbransco <TFTAJLLYM...@spammotel.com> wrote:
> > My first post here, so greetings everyone and best wishes for the
> > holiday season.
> > I'm hoping to get some suggestions for how to estimate or calculate
> > the HFOV parameter used throughout Hugin, but more particularly for
> > the align_image_stack tool.
> > I shoot some of my macros with an Olympus E-system DSLR fitted with
> > two adapters, one or more extension tubes, and a reversed 50mm manual
> > lens, so it's not immediately obvious to me what HFOV the tool would
> > be best run with, nor how precise I must be to get good results. Any
> > help would be appreciated.
> What macros do you do?
> If you want to make focus stacks, the HFOV doesn't matter much: you go
> in and out rectilinear and no transformation takes place. Use a rule-
> of-thumb value.
> On the other hand, if you were making macro panoramas, which is a bit
> more unusual, you can do a 360 degree panorama, start with some
> estimate for the HFOV and optimize for fov, then take the value you
> get after optimizing. It doesn't sound like your setup is for
> panoramas, though...
> You can also meaure the value, if you know where your NPP is. Take a
> picture of, say, a horizontal ruler and take a reading of how many cm
> of the ruler fits in your image. Let that be w, and d the distance of
> the ruler from your NPP. If v is your horizontal field of view, you
> have w / ( 2 * d ) = tan ( v / 2 ), so v = 2 atan ( w / ( 2 * d ) )
Not sure I know just yet; I'm just starting out. I'm thinking still
life of any sort would be a good starting point with the gear I've
described earlier. The DOF is very shallow and the working distances
awfully close.
> If you want to make focus stacks, the HFOV doesn't matter much: you go
> in and out rectilinear and no transformation takes place. Use a rule-
> of-thumb value.
I see. Good to know, and thanks again.
> On the other hand, if you were making macro panoramas, which is a bit
> more unusual, you can do a 360 degree panorama, start with some
> estimate for the HFOV and optimize for fov, then take the value you
> get after optimizing. It doesn't sound like your setup is for
> panoramas, though...
Too true, though I suppose one could slap together a circular subject
of an appropriate diameter to allow the gear to focus, and suspend it
around the camera in some fashion. If the HFOV was a critical
parameter, this might provide the most precise estimate.
> You can also measure the value, if you know where your NPP is. Take a
> picture of, say, a horizontal ruler and take a reading of how many cm
> of the ruler fits in your image. Let that be w, and d the distance of
> the ruler from your NPP. If v is your horizontal field of view, you
> have w / ( 2 * d ) = tan ( v / 2 ), so v = 2 atan ( w / ( 2 * d ) )
Great information, Kay. I'll search around for how to find the NPP in
a reversed lens and then give this a try.
