Dale,

On September 27, 2010 01:32:48 am Dale Beams wrote:
> APPro beats Hugin in alignment hands down.

IMHO it is more correct to state that "APPro operated by Dale Beams beats 
Hugin operated by Dale Beams hands down".

I bet that Hugin operated by Bruno Postle would beat APPro operated by Dale 
Beams on quality of stitch based on exactly the same input images. 


> In truth [...]

Truth? your limited experience, maybe.  If this is what works for you, buy 
APPro.  The experience of other users is completely different [0].

I don't think that the trial of APPro is meant for the use that you, Emad, and 
maybe others are doing.  IMHO it is morally important to respect proprietary 
software as much as it is important to respect Open Source software.

 
> Hugin has a problem with feature creep.

With funding from Google, Agnos, Nodal Ninja, and last but not least the 
community we had a total of 20+ student projects in the past 4 years.  Plus 
many contributions from interesting smart people, too many to mention.  For 
you these contributions may be features creep.  For me they, and the people 
associated with them - mentors, students, other contributors - are what make 
Hugin interesting.

Hugin's problem is that *consumers* like you project their *wishes* on it.  


> In truth, Hugin met
> it's initial objective (0.7.x) but for whatever reason which users don't
> understand and don't have the programming experience to examine code,
> Hugin isn't currently accomplishing this goal.  I've tested it and posted
> my results to the list.

Stop it with your ludicrous "truth".  Hugin does not have any objective.  
People do.  You have objectives.  I have objectives.  We team up when our 
objectives are aligned.  We negotiate and compromise on the inevitable small 
conflicts of interest.  When the objectives are too different and it no longer 
make sense to compromise, we part ways.

So Hugin is not accomplishing *your* goal?  I feel sympathy for you, but I 
don't think that this is a problem for Hugin.  Why should your objective be 
more important than contributors' objectives?


> btw, I use ubuntu,  built my own LFS system in 2003, had some introduction
> to C++ programing in college, wrote some interesting scripts, worked on
> OS400 CISC to RISC conversions, etc. and enjoye a good GUI.  I sit between
> worlds, user/tester/dev/  I'm an oldie who doesn't need to prove anything,
> but would enjoy just using the software.

Good for you, I don't question your computing skills.  I do question the 
integrity, completeness, methodology, and ultimately results and purpose of 
your "test".

You claim it is between Hugin and APPro.  I claim it is first and foremost 
between you and yourself.

You don't provide input images nor PTO files, only screenshot.  Frankly 
speaking, both screenshots look like a crappy stitches to me, but I don't have 
enough information to judge what is the cause.

You claim it is the control point detection.  Are you aware that Hugin is not 
a CP detector (yet) [1]?  Hugin only provides an interface to CP detectors 
installed on your system.  You don't even mention which one you are using.

There are many reasons for a crappy stitch.  Could be parallax problems in the 
input images.  You don't provide them to verify.

Could be indeed CP detection like you claim.  Could also be bad optimization.  
The result of optimization depends also a lot on the operator.  But without 
providing a pto file and input images it is not possible to help you.

Last but not least, you claim it is Hugin.  Recent threads on this list tells 
me that it is probably *your build* of Hugin and related tools; and that you 
had all sorts of problems following instructions that have been leading to 
successful builds thousands of (silent) times.  Before claiming any validity 
for your "test results", you want to exclude bad build as a factor as well, by 
testing against other builds, possibly on different machines / from different 
users.

No, you don't have to prove anything indeed.  But don't call your personal 
experience "truth", and don't claim your two snapshots to be a conclusive test 
of anything else but your ability to take snapshots and post them on the web.

The only claim you can make at this stage is that APPro works better for you.  
If you are happy with the result you obtained with APPro, buy a license and 
happy stitching.

Yuv

[0] http://www.flickr.com/groups/hugin/
[1] http://panospace.wordpress.com/2010/09/26/the-quest-is-over/

Attachment: signature.asc
Description: This is a digitally signed message part.

Reply via email to