BigTIFF

[Battle]

I working on some large panoramas, and have run into the file size limit of TIF > 4GB.  

I see from 2018.0 that BigTif is possible.

The release announcement says

* Nona, verdandi and hugin_stacker can now write BigTIFF images (for files >4 GB). The output of BigTIFF has to manually activated on the command line. It is not set automatically. (For a complete panorama in BigTIFF you will probably also need enblend/enfuse from repository. The last released version of enblend does not yet support writing BigTIFF files.)

I need a bit of help figuring out how to make this happen.  I'm using OS X 2018.0.0, but have a Windows box available if necessary.  

Thanks,

Battle

[T. Modes]

For nona, verdandi and hugin_stacker add the switch --bigtiff to the command line. The development versions from enblend/enfuse require --parameter=export-bigtiff

The easiest way to use this would be to create a user defined output sequence.

Thomas

[Battle]

HI Thomas,

I hate to be a kludge, but I'm not finding any development versions of Enblend.  Maybe I don't know where to look.  Looks like the last version at source forge is 4.2 in 2016.  

I see that in regular Hugin Stitcher tab I can add command line options for enabled, but I don't see a way to do that for nona.  So that means using terminal I suppose.  Is there a tutorial on command line?  

Thanks,

Battle

[Greg 'groggy' Lehey]

On Monday, 17 September 2018 at 16:30:17 -0700, Battle wrote:
> HI Thomas,

> I hate to be a kludge, ...

My guess is that you were looking for a different word. To quote OED,
a kludge is "a hastily improvised and poorly thought-out solution to a
fault or ‘bug’."

> I see that in regular Hugin Stitcher tab I can add command line options for
> enabled, but I don't see a way to do that for nona. So that means using
> terminal I suppose. Is there a tutorial on command line?

There is one at
https://wiki.panotools.org/Panorama_scripting_in_a_nutshell, but it's
horribly out of date. I've been meaning to update it, but I've always
found something more rewarding to do. In particular, it makes
reference to obsolete programs.

--
Sent from my desktop computer.
Finger groo...@gmail.com for PGP public key.
See complete headers for address and phone numbers.
This message is digitally signed. If your Microsoft mail program
reports problems, please read http://lemis.com/broken-MUA

[T. Modes]

Am Dienstag, 18. September 2018 01:30:18 UTC+2 schrieb Battle:

HI Thomas,

I hate to be a kludge, but I'm not finding any development versions of Enblend.  Maybe I don't know where to look.  Looks like the last version at source forge is 4.2 in 2016.  

I know of no binary development version of enblend. You need to compile enblend for yourself or using the internal blender.

I see that in regular Hugin Stitcher tab I can add command line options for enabled, but I don't see a way to do that for nona.  So that means using terminal I suppose.  Is there a tutorial on command line?  

No need to use the terminal. As I already wrote an user defined output sequence would be sufficient and can be simply called from the GUI.

Thomas

[T. Modes]

@Greg

Am Dienstag, 18. September 2018 05:46:34 UTC+2 schrieb Groogle:

There is one at
https://wiki.panotools.org/Panorama_scripting_in_a_nutshell, but it's
horribly out of date.  I've been meaning to update it, but I've always
found something more rewarding to do.  In particular, it makes
reference to obsolete programs.

Sorry, but is wrong. The main part with the description is up to date. The only one obsolete program in the main part is autopano-sift-c - but this is only shown as alternative; the mentioned main program cpfind is still part of Hugin. All other program are still there. Maybe in the list at the end there are some obsolete programs, but not in the main part.

The next time be more precises what you are referring instead of make such general and diffuse statements.

Thomas

[Battle]

So if I use the builtin blender instead of enblend I can use the —big tiff switch and that will output BigTIFF?

[Battle]

Thomas,

Thanks for your replies so far.  I'm looking at the user defined output sequences, and am not quite sure where to put the command line and parameters for big tiff.  

from user defined menu the executor file for normal smart blend reads as follows:  

[Step0]

Description=Remapping images

Type=remap

Arguments=-r ldr -m TIFF_m

[Step1]

Description=Blending with smartblend

Type=merge

Input=all

Program=smartblend

Result=%prefix%.tif

Arguments=-o %result% %input%

WrapArgument=

Presumably they go in the Arguments line...

In [Step0] it doesn't look like nona is specifically called as the remapper.  But presumably I would add --bigtiff as follows

Arguments=--bigtiff -r ldr -m TIFF_m

And for [Step1] similarly 

Arguments=--bigtiff -o %result% %input%

or if using normal enblend u.d.o. script --parameter=export-bigtiff

Am I on the right track here?

Thanks,

Battle
On Monday, September 17, 2018 at 2:29:40 PM UTC-4, T. Modes wrote:

[Greg 'groggy' Lehey]

This is mainly addressed to Thomas Modes, but since it's not the first
time I've seen him behave like this, I think it's time to come out in
the open.

On Tuesday, 18 September 2018 at 10:00:34 -0700, T. Modes wrote:
> Am Dienstag, 18. September 2018 05:46:34 UTC+2 schrieb Groogle:
>>
>> There is one at
>> https://wiki.panotools.org/Panorama_scripting_in_a_nutshell, but it's
>> horribly out of date. I've been meaning to update it, but I've always
>> found something more rewarding to do. In particular, it makes
>> reference to obsolete programs.
>
> Sorry, but is wrong. The main part with the description is up to
> date. The only one obsolete program in the main part is
> autopano-sift-c - but this is only shown as alternative; the
> mentioned main program cpfind is still part of Hugin.

Sorry, but is wrong. I spent quite a bit of time going through this
page a while back, and it contains numerous issues. Apart from
autopano-sift-c, in the main text there are also references to
autopano-sift, match-n-shift, ptoinfo, ptscluster, ptoclean,
ptovariable and panostart, as well as to the pano.mk file that no
longer exists.

In the "Full list of command-line [sic] tools", only 5 of the first 10
tools still exist. I also seem to recall that some of the options are
also no longer relevant, though I could be wrong here.

This list also contains the line:

autopano - Keypoint matcher, part of autopano-sift-c and not to be confused with autopano

Well, I'm confused. Beginners certainly will be.

Maybe you have a different definition, but for me this is very out of
date.

I had intended to bring this page up to date. Based on this attack,
you will understand that I am now disinclined to do so. It will
remain out of date until somebody else who recognizes the problem also
finds time and inclination to do something about it.

> The next time be more precises what you are referring instead of
> make such general and diffuse statements.

Next time, before answering, check your facts. You come across as
offensive and inaccurate, not for the first time. This attitude also
greatly reduces people's willingness to cooperate in the project. A
few years back a whole lot of people were involved in the project.
Now you seem to be the only one. It would be easy to get the
impression that your attitude has something to do with this.

Greg

[T. Modes]

Hi Battle,

Am Mittwoch, 19. September 2018 03:37:14 UTC+2 schrieb Battle:

In [Step0] it doesn't look like nona is specifically called as the remapper.  But presumably I would add --bigtiff as follows

Arguments=--bigtiff -r ldr -m TIFF_m

And for [Step1] similarly 

Arguments=--bigtiff -o %result% %input%

or if using normal enblend u.d.o. script --parameter=export-bigtiff

Am I on the right track here?

You are on the right track. But I don't think that smartblend supports BigTIFF.

So you need to run instead the internal blender. This can be combined with the nona call.I'm attaching an executor file for BigTIFF output of the normal panorama using the internal blender (aka verdandi).

Thomas

[T. Modes]

Am Mittwoch, 19. September 2018 06:00:08 UTC+2 schrieb Groogle:

Sorry, but is wrong.  I spent quite a bit of time going through this
page a while back, and it contains numerous issues.  Apart from
autopano-sift-c, in the main text there are also references to
autopano-sift, match-n-shift, ptoinfo, ptscluster, ptoclean,
ptovariable and panostart, as well as to the pano.mk file that no
longer exists.

Next time, before answering, check your facts.

Have you checked your facts?

Match-n-shift, ptoinfo, ptoclean, ptovariable and panostart are part of the Panotools::Scripts. There are still available, are up to date and work with the current Hugin version. They are not obsolete like you claim.From the linked wiki page:

<quote>

Panotools::Script for manipulating project files

<snip> Many of the command-line tools described in this document are written with this library.
</quote>

Remaining autopano-sift-c, autopano-sift and ptscluster,

From the wiki page:

<quote>

Hint: Other control points generator, such as autopano-sift, autopano-sift-c, panomatic and match-n-shift <snip>

</quote>

<quote>

Other control point cleaners are ptscluster and ptoclean. cpclean has largely replaced ptoclean

</quote>

These are mentioned in a hint/footnote. So when there are some smaller references to older programs in footnotes, you declare the whole page "very out of date."

So it seems I have another definition of "up to date" than you…

[Greg 'groggy' Lehey]

On Wednesday, 19 September 2018 at 8:22:08 -0700, T. Modes wrote:
> Am Mittwoch, 19. September 2018 06:00:08 UTC+2 schrieb Groogle:

Nein, er schrieb das am Mittwoch, 19. September 2018 14:00:03 +1000.
Er hat mit UTC+2 nichts am Hut. Wo die 5 Sekunden Versatz entstanden
sind, ist mir ein Rätsel.

>> Sorry, but is wrong. I spent quite a bit of time going through this
>> page a while back, and it contains numerous issues. Apart from
>> autopano-sift-c, in the main text there are also references to
>> autopano-sift, match-n-shift, ptoinfo, ptscluster, ptoclean,
>> ptovariable and panostart, as well as to the pano.mk file that no
>> longer exists.
>
> Next time, before answering, check your facts.

I checked my facts.

> Have you checked your facts?

Yes.

> Match-n-shift, ptoinfo, ptoclean, ptovariable and panostart are part
> of the Panotools::Scripts. There are still available, are up to date
> and work with the current Hugin version. They are not obsolete like
> you claim.From the linked wiki page:

I stand corrected. The document is not only out of date, it's also
incomplete. The location of those programs is not described in this
document, and that's what we're talking about.

> <quote>
> Panotools::Script for manipulating project files
> <snip> Many of the command-line tools described in this document are
> written with this library.
> </quote>

Which? This is incomplete. It's also in the wrong place. Let's take
an example:

Is:
match-n-shift - All in one wrapper around align_image_stack and autopano-sift-c
Should be:
match-n-shift - All in one wrapper around align_image_stack and
autopano-sift-c, part of <link>Panotools::Script</link>

This section also needs sorting. align_image_stack is a forward
reference.

> Remaining autopano-sift-c, autopano-sift and ptscluster,

Wouldn't that have been enough?

>> From the wiki page:
>
> <quote>
> Hint: Other control points generator, such as autopano-sift,
> autopano-sift-c, panomatic and match-n-shift <snip>
> </quote>

Yes, I saw that. Your point? It's a reference to a programs that are
no longer in Hugin and commands that are part of an unmentioned script
library. For me, that's an indication that the document is
incomplete.

> <quote>
> Other control point cleaners are ptscluster and ptoclean. cpclean has
> largely replaced ptoclean
> </quote>

Again, I don't see what point you're trying to make here.

> These are mentioned in a hint/footnote.

The only other reference to ptoclean is in the list of commands. The
link takes you to a page https://wiki.panotools.org/Ptoclean, which
doesn't exist. Your point?

> So when there are some smaller references to older programs in
> footnotes, you declare the whole page "very out of date."

No, "horribly out of date". To that I add "incomplete and
misleading".

> So it seems I have another definition of "up to date" than you???

Yes, so it seems!!! Your definition seems to be "defendable" rather
than usable. I don't accept it. You have found what you consider to
be a sufficient excuse. It may be sufficient to you, but it isn't to
me, and I assert that it isn't to a beginner. The document is broken.

Your continuous obnoxious behaviour shows that you don't understand,
or that you don't want to. And I note that you avoided answering any
of my comments on the effect on the project of this attitude. As a
result, I plan to look for alternatives to Hugin. This wouldn't have
happened if the project team were interesting in improving the
current situation.

[Bruno Postle]

I feel like I ought to respond to this.

I agree that Thomas does sometimes react to criticisms of Hugin as
personal attacks, and responds overly defensively in a way that
isn't required, sometimes no response is required - the particular
wiki page in discussion is something I created in a burst of
enthusiasm and then failed to keep up-to-date - I'm very conscious
that this page wastes people's time by listing all the tools without
recommending which tools to use.

Some history needs correcting: Hugin was pretty much moribund
before Thomas joined. Pablo had stepped back, in part because being
a sole maintainer wasn't much fun for him, we had a few successful
Summer of Code projects, but these ran out of steam because there
was nobody to manage them - do you see a pattern here? Though we
would all love for free software to be a self-organising
collaborative sharing experience, the reality of most free software
projects is that one person is overburdened and does most of the
work.

Having said that, we would love to see more contributors, there is
lots to do (I'm embarassed that there is a libpano13 release ready
to go that should have been out years ago, maybe today).

--
Bruno

[John Muccigrosso]

Bruno, 

Thanks for the context and history.

I have to say that while I find Hugin a great tool, it's often more difficult to use than it seems like it should be. (A lot of the interface is very difficult to understand without a good understanding of what's going on in the background, IMO.) Some of the discussions here about usability have also run into problem that there aren't many developers capable of making the changes and, more to the point, those developers haven't agreed about the need for such changes.

Without piling on here, I want to say that I'm grateful for Thomas' work on Hugin, and I'd urge him to take seriously the comments that repeatedly crop up here about the tone that his comments take...or at least the tone that others perceive in them. I suspect that some of that is due to language issues, but that certainly doesn't explain all of it.

A grateful user,

John