Confocal Microscopy List

Image Processing Software

Classic

List

Threaded

26 messages Options

Michael Herron

Re: Image Processing Software

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

In my experience the more an application costs the less robust it is. I have always assumed that this is because a high price means a smaller user base with fewer people banging on it.

Sent from my iPad Nano

On Sep 30, 2011, at 8:39 AM, Daniel James White <[hidden email]> wrote:

DrSmithMartin@gmail.com

Re: Image Processing Software

In reply to this post by Daniel James White

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dare I dip a toe into this hot debate… I will start with the proviso that we
use both commercial and freeware in our labs and they all do have their
strengths.

The point I wanted to address was the generalisation that open source
software is slightly less robust and has less nice GUIs than the commercial
counterparts. I think that I would not use it as a generalisation but say it
is a trend and I have a theory as to why that may be true. If you think
about the time you take coding software you can basically put that time
into: 1. making it do more functions, 2. making it look prettier or be more
intuitive or 3. making it more stable.
A lot of open source projects are small and written by the end user. In
those cases points 2 and 3 are of lower concern because the coders know how
it works and can sidestep bugs. So in those cases it makes more sense to
invest more time in making it do more things.

Also if we go back to evolution 101 we need to consider the selective
pressures in this competitive landscape. Commercial software has to be
rounded and polished to a certain standard or it is not acceptable to the
market, does not sell and no money comes in. So a programmer in a commercial
company may want to write new features but his boss tells him he has to fix
the bugs first. The open source programmer does not have that pressure so
can do more of what they find fun and enjoyable.
Now I am sure that the larger open source projects which have funding for
full time paid programmers can exert pressure on the coders which by then
are effectively employees so that is why they start to get more robust. They
in effect re-introduce the pressure that is present in commercial companies
to fix bugs. Now I am sure pride also motivates open source coders but I
think in the end ‘cash > pride’.

On the GUI side I expect that there are similar pressures. Commercial
solutions have to look pretty so they can spend resource on GUI and workflow
experts. I imagine getting funding for a usability person in a university
funded open source endeavour is a lot harder than getting funding for a
coder to write a new function which furthers our research. And besides the
end result is free so the people who download it cannot complain too much.
So in conclusion I think that there are some pressures which explain why it
may not be unfair to categorise open source software as typically buggier
and less easy to use, but that pressure lowers once the software begins to
get financial funding, so much so that you see things like FIJI and Firefox
being very useable.

But coming back to advice suitable for the list rather than a full thesis I
think you do just have to work out which software will help you to progress
your research in the most cost effective manor. And since its Friday and the
list has a more relaxed feeling I will end sharing an albeit cheeky quote I
quite liked from a visiting sales rep from a software company last year:
“Remember, free software is only free to people who’s time is worth
nothing.”

All the best
Martin

On Fri, Sep 30, 2011 at 3:39 PM, Daniel James White <[hidden email]>wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Hi Mark,
>
> On Sep 30, 2011, at 12:38 PM, Scott, Mark wrote:
>
> > Yes it was intended as a generalisation, most open source project
> software is slightly less robust, more buggy and lacks as nice a UI as it's
> more commercial counterparts.
>
> Do you have objective data to back that statement up, or is it just your
> feeling or experience?
> I think it does a diservice to the many well maintained, well designed,
> stable and user friendly open source softwares that are used in our
> discipline and many others.
>
> Sure, there is crap open source software too, but there are also very poor
> expensive softwares and also very good ones.
>
> My objection is the labeling of open source software with an automatic
> lower quality than commercial software.
> This isn't really fair, and each case should be made on its own merits.
>
> > That wasn't to say that there are exceptions to the rule
>
> What rule? Where is the data to prove any such rule?
>
> > and some do look nice. Firefox isn't a realistic example though since it
> was designed to compete with other web-browsers of the time which clearly
> wasn't going to work as a legitimate competitor if it was a clunky box with
> pop-up windows and an unorganised UI.
> >
> > I also was not condoning other applications over ImageJ, I was simply
> suggesting that it is worth not discounting other commercial applications
> just because ImageJ is free if those commercial software package encourage
> users to use them more, do more work and produce results faster and easier
> than then having to learn ImageJ from scratch.
>
> |I totally agree.
> However, the big picture view, in my experience, is that the "Ease of use
> and fast results, / no need to develop your own tools" draw of commercial
> software is often an illusion. There are very strong cases where commercial
> software is very very good at a certain thiung.. eg Huygens for
> deconvolution. But if you dont happen to hit one of those specific
> applications and you need something just a little or very different, an open
> software platform where you can chain together existing tools and add your
> own special little bits pays off big time.
>
> ImarisXT is a good thing as it enhances a closed platform by leveraging an
> open one! Smart!
>
> > It seems certain members of the imaging community are rather touchy in
> defence of their beloved ImageJ and leap to its defence if anyone dares say
> anything other than that it is the only option.
>
> Did anyone say its the only option? Certainly many folks think it's avery
> good option. But commercial software can also be a good fit if it really
> does exactly what you want already. Beware the specter of "vendor lock-in"
> though......
> Give a man a fish and he can eat for one day. Teach a man to fish and he
> can feed his family for a lifetime.
>
> > I for one find that while Volocity has a nice user friendly familiar
> UI, it is still rather unintuitive to use, but at the same time some users
> much prefer this.
>
> Every human is different. Some people are narrow minded and just want a
> "working " solution now.
> Others are smarter and understand the value of open platforms being very
> compatible with doing good science,
> and are not interested in how pretty a software looks, but rather in what
> it can do with not too much effort.
> >
> > Using ImageJ if your users are comfortable with it is fine, but using it
> just out of refusing to use other applications and "forcing" users to learn
> ImageJ because it is what you personally like is dangerous and I was simply
> suggesting that all points are addressed rather than dismiss other options
> outright just because people who have input into ImageJ refuse to
> acknowledge other applications even exist.
>
> I think I just suggested this is not the case. There is room for open
> source and proprietary software in this ecosystem.
>
> > I have seen non-computer savvy people be forced into using non-windows
> based platforms simply out of programmers refusing to use windows and this
> normally resulted in less productivity due to people not being comfortable
> with the other platforms.
>
> Perhaps in the short term, with short attention spans, this can be the
> case.
> In the long run, investing in open platforms does pay off.
> There are plenty of good reasons to use windows, but there are also some
> very good reasons that
> the folk setting up facilities and providing infrastructure avoid it at
> nearly all costs... these guys are looking at the big picture.
>
> Vendors might provide a certain closed platform or operating system on the
> computers attached to their microscopes, believing that this is what most
> users are familiar and comfortable with.
> You can ask any user of an Applied Precision DeltaVision microscope system
> if this actually matters.
> I think its a false assumptions. Users like what works and don't care about
> the computer Operating System very much.
>
> As a systems admin I very much like the Linux based systems over the window
> systems, as the latter provide a much more troublesome environment to keep
> in good shape, easily administered, and virus free etc.
>
> > All I was suggesting was that there are options and opting for ImageJ
> because it's free or opting for other software because (as you say) it looks
> pretty, doesn't make it the right choice. The end decision should come
> after considering the user needs and software requirements over all aspects,
> not just pricing and certainly not just looking nice, if it is for a
> multi-user facility.
>
> Absolutely right!
> I think we agree on much more than you might have thought!
>
> cheers
>
> Dan
>
>
>
> >
> > Cheers
> > Mark
> >
> >
> > -----Original Message-----
> > From: Daniel James White [mailto:[hidden email]]
> > Sent: 27 September 2011 08:19
> > To: Confocal Microscopy List; Scott, Mark; [hidden email]
> > Subject: Re: Image Processing Software
> >
> > Hi Guys,
> >
> >> Date: Mon, 26 Sep 2011 12:20:32 +0000
> >> From: "Scott, Mark" <[hidden email]>
> >> Subject: Re: Image Processing Software
> >>
> >>
> >> Hi Venkat,
> >>
> >> As many have already suggested ImageJ (or FIJI) is a very nice option
> since=
> >> it will do almost anything you could require of it and it is free.
> Howev=
> >> er, it is a rather cut down UI typical of open source software
> >
> > Thats a rather sweeping generalization, and possibly a little inaccurate?
> > There are very many open source free softwares that have very nice
> > extensive and feature complete user friendly graphical user interfaces.
> > Ever heard of the Firefox web browser? Seen Icy? Seen PyMol and VMD?
> >
> > Certainly imageJ has a low complexity tool bar only primary interface,
> > and that is to make it less confusing and easier to use for novices.
> > There are not too many clickable items, so its not too scary for a new
> user,
> > but tons of functionality is there in the menus and others tools,
> > so in total the user interface is very very rich.
> >
> > ImageJ2 project is working hard to improve the GUI,
> > which does have its limitations,
> > and the Fiji project has added many GUI/functionality enhancements like
> the command finder/launcher.
> >
> > There is nothing to stop anyone of us making whatever GUI for imageJ2 we
> desire,
> > that will fit the needs of a certain class of users perfectly.
> >
> >> and as such =
> >> isn't the most user friendly of available software - and as it is for a
> fac=
> >> ility it is important that your users feel comfortable with using the
> softw=
> >> are (yes it might be feasible for you and I to learn how to do something
> bu=
> >> t not all microscope users are as comfortable with this thought).
> >
> > In our experience, anyone who works in a research lab is smart enough
> > to figure out how to use imageJ, given a bit of basic training,
> > and some tips and tricks, and good documentation, eg the great new
> ImageJ manual,
> > and the websites sites of Fiji and ImageJ Documentation Wiki.
> >
> > I tend to warn out users that just because an expensive software "looks"
> nice, doesn't mean its going to
> > be the best for for every job, or in fact even easy to use, once you get
> into the details.
> > Often you don't know what is really happening as the documentation is
> limited or missing.
> > Not the case for open source softwares.
> >
> > Just because imageJ spits windows all over the place (which can indeed
> get confusing),
> > doesn't necessarily put it out of reach of beginners in image processing,
> > in fact, because its free and open source, and easy to use,
> > we prefer to use Fiji for all our teaching, and as the general platform
> for image processing and analysis in our institute.
> >
> > cheers
> >
> > Dan
> >
> >
> >
> > Dr. Daniel James White BSc. (Hons.) PhD
> >
> > Leader - Image Processing Facility,
> > Senior Microscopist,
> > Light Microscopy Facility.
> >
> > Max Planck Institute of Molecular Cell Biology and Genetics
> > Pfotenhauerstrasse 108
> > 01307 DRESDEN
> > Germany
> >
> > +49 (0)15114966933 (German Mobile)
> > +49 (0)351 210 2627 (Work phone at MPI-CBG)
> > +49 (0)351 210 1078 (Fax MPI-CBG LMF)
> > chalkie666 Skype
> > http://www.bioimagexd.net BioImageXD
> > http://fiji.sc Fiji - is just
> ImageJ (Batteries Included)
> > http://www.chalkie.org.uk Dan's Homepages
> > https://ifn.mpi-cbg.de Biopolis Dresden Imaging
> Platform (BioDIP)
> > dan (at) chalkie.org.uk
> > ( white (at) mpi-cbg.de )
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
>
> Dr. Daniel James White BSc. (Hons.) PhD
>
> Leader - Image Processing Facility,
> Senior Microscopist,
> Light Microscopy Facility.
>
> Max Planck Institute of Molecular Cell Biology and Genetics
> Pfotenhauerstrasse 108
> 01307 DRESDEN
> Germany
>
> +49 (0)15114966933 (German Mobile)
> +49 (0)351 210 2627 (Work phone at MPI-CBG)
> +49 (0)351 210 1078 (Fax MPI-CBG LMF)
> chalkie666 Skype
> http://www.bioimagexd.net BioImageXD
> http://fiji.sc Fiji - is just ImageJ
> (Batteries Included)
> http://www.chalkie.org.uk Dan's Homepages
> https://ifn.mpi-cbg.de Biopolis Dresden Imaging Platform
> (BioDIP)
> dan (at) chalkie.org.uk
> ( white (at) mpi-cbg.de )
>

Johannes-P. Koch

Re: Image Processing Software

In reply to this post by Daniel James White

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi,

On Sep 30, 2011, at 12:38 PM, Scott, Mark wrote:
> I also was not condoning other applications over ImageJ, I was simply suggesting that it is worth not discounting other commercial applications just because ImageJ is free if those commercial software package encourage users to use them more, do more work and produce results faster and easier than then having to learn ImageJ from scratch.

Actually, it can be seen as advantage to learn ImageJ (or other
packages) from scratch. Not knowing what you are actually doing to your
data while hitting any black-box button might be convenient but in the
end you might have to face corrupted/wrong data. However, some
commercial softwares might be more restrictive and do not like
explaining all about their algorithms (which is somewhat understandable...).

I learned pretty much back then, when I had/wanted to study the maths
and the principles of softwares such as Huygens or ImageJ.

Surely, some users prefer glossy UIs and rather not "play around with
parameters".

The point is - and I totally agree with Dan - it is all about the person
using the software (intuitive vs non-intuitive is a rather personal
thing) AND the goal you want to achieve with the software. I wouldn't
buy Imaris just for cropping and merging images, nor would I stick to
ImageJ for high-end decon...

Johannes

--
Mag. Johannes-P. KOCH
Department of Biochemistry and Cell Biology
MFPL, University of Vienna
Dr. Bohrgasse 9/5
A-1030 Vienna
Austria

phone: 0043 1 4277 52809
fax: 0043 1 4277 9528

mail to: [hidden email]

Vitaly Boyko

Re: Image Processing Software

In reply to this post by DrSmithMartin@gmail.com

The story is much simpler. Most of the com
*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi Martin, Mark and others.

The story is much simpler. Most of the commercial software is 5 or more years behind the current developments in the filed. The open source covers much broader spectrum of users.

It is a sort of obvious that private companies have to make money thus that follow the Majority. Advanced science is a Minority.

The rest is just emotions...

I am in the field for over 14 years, believe me.

All the best,

Vitaly

________________________________
From: "[hidden email]" <[hidden email]>
To: [hidden email]
Sent: Friday, September 30, 2011 11:11 AM
Subject: Re: Image Processing Software

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dare I dip a toe into this hot debate… I will start with the proviso that we
use both commercial and freeware in our labs and they all do have their
strengths.

The point I wanted to address was the generalisation that open source
software is slightly less robust and has less nice GUIs than the commercial
counterparts. I think that I would not use it as a generalisation but say it
is a trend and I have a theory as to why that may be true. If you think
about the time you take coding software you can basically put that time
into: 1. making it do more functions, 2. making it look prettier or be more
intuitive or 3. making it more stable.
A lot of open source projects are small and written by the end user. In
those cases points 2 and 3 are of lower concern because the coders know how
it works and can sidestep bugs. So in those cases it makes more sense to
invest more time in making it do more things.

Also if we go back to evolution 101 we need to consider the selective
pressures in this competitive landscape. Commercial software has to be
rounded and polished to a certain standard or it is not acceptable to the
market, does not sell and no money comes in. So a programmer in a commercial
company may want to write new features but his boss tells him he has to fix
the bugs first. The open source programmer does not have that pressure so
can do more of what they find fun and enjoyable.
Now I am sure that the larger open source projects which have funding for
full time paid programmers can exert pressure on the coders which by then
are effectively employees so that is why they start to get more robust. They
in effect re-introduce the pressure that is present in commercial companies
to fix bugs. Now I am sure pride also motivates open source coders but I
think in the end ‘cash > pride’.

On the GUI side I expect that there are similar pressures. Commercial
solutions have to look pretty so they can spend resource on GUI and workflow
experts. I imagine getting funding for a usability person in a university
funded open source endeavour is a lot harder than getting funding for a
coder to write a new function which furthers our research. And besides the
end result is free so the people who download it cannot complain too much.
So in conclusion I think that there are some pressures which explain why it
may not be unfair to categorise open source software as typically buggier
and less easy to use, but that pressure lowers once the software begins to
get financial funding, so much so that you see things like FIJI and Firefox
being very useable.

But coming back to advice suitable for the list rather than a full thesis I
think you do just have to work out which software will help you to progress
your research in the most cost effective manor. And since its Friday and the
list has a more relaxed feeling I will end sharing an albeit cheeky quote I
quite liked from a visiting sales rep from a software company last year:
“Remember, free software is only free to people who’s time is worth
nothing.”

All the best
Martin

On Fri, Sep 30, 2011 at 3:39 PM, Daniel James White <[hidden email]>wrote:

DRESDEN

> > Germany
> >
> > +49 (0)15114966933 (German Mobile)
> > +49 (0)351 210 2627 (Work phone at MPI-CBG)
> > +49 (0)351 210 1078 (Fax MPI-CBG LMF)
> > chalkie666 Skype
> > http://www.bioimagexd.net BioImageXD
> > http://fiji.sc Fiji - is just
> ImageJ (Batteries Included)
> > http://www.chalkie.org.uk Dan's Homepages
> > https://ifn.mpi-cbg.de Biopolis Dresden Imaging
> Platform (BioDIP)
> > dan (at) chalkie.org.uk
> > ( white (at) mpi-cbg.de )
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
>
> Dr. Daniel James White BSc. (Hons.) PhD
>
> Leader - Image Processing Facility,
> Senior Microscopist,
> Light Microscopy Facility.
>
> Max Planck Institute of Molecular Cell Biology and Genetics
> Pfotenhauerstrasse 108
> 01307 DRESDEN
> Germany
>
> +49 (0)15114966933 (German Mobile)
> +49 (0)351 210 2627 (Work phone at MPI-CBG)
> +49 (0)351 210 1078 (Fax MPI-CBG LMF)
> chalkie666 Skype
> http://www.bioimagexd.net BioImageXD
> http://fiji.sc Fiji - is just ImageJ
> (Batteries Included)
> http://www.chalkie.org.uk Dan's Homepages
> https://ifn.mpi-cbg.de Biopolis Dresden Imaging Platform
> (BioDIP)
> dan (at) chalkie.org.uk
> ( white (at) mpi-cbg.de )
>

Johannes Schindelin

Re: Image Processing Software

In reply to this post by Scott, Mark

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dear Mark,

On Fri, 30 Sep 2011, Scott, Mark wrote:

> Yes it was intended as a generalisation, most open source project
> software is slightly less robust, more buggy and lacks as nice a UI as
> it's more commercial counterparts.

You are a scientist. Where is the evidence?

> It seems certain members of the imaging community are rather touchy in
> defence of their beloved ImageJ and leap to its defence if anyone dares
> say anything other than that it is the only option.

While I enjoy the occasional flame-war revolving mostly about feelings,
I'd like to point out that it is not about "ImageJ", or "cost".

The major point is that we do science. The major point is that if you
cannot explain why a certain thing does what it does, that is not science.

Yes, if it is an obvious algorithm such as "segmentation by threshold" or
"Gaussian Blur", it is not an issue if I cannot see the code.

However, if you click on a button and the result looks nice, but there
is no explanation what exactly was done and why it should make sense
as to the scientific question asked, then that is art, not science.

After all, if you publish a scientific result about, say, a fly mutant,
you make that fly mutant available to other scientists. And it is
important that you do.

Ciao,
Johannes

Scott, Mark

Re: Image Processing Software

In reply to this post by Vitaly Boyko

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi again,

I think the original point has been somewhat lost, I saw a post about which software to get for a multi-user facility and simply suggested that the most important thing to consider is your users needs and expertise levels and get software that will fulfill these criteria.

I have used software from custom written command line unix applications to expensive commercial software and everything in between, and I personally feel comfortable learning new software and as has been disussed, it is good to know what you are doing and why it happens. But, for everyday common useage for biological scientists in my case I know that some people feel more comfortable using certain platforms than others and at the end of the day this should be your priority unless you have very independant users or a lot of time to dedicate to helping them learn more complex packages.

Been a fun discussion though none-the-less.

Cheers
Mark

________________________________________
From: Confocal Microscopy List [[hidden email]] on behalf of Vitaly Boyko [[hidden email]]
Sent: 30 September 2011 16:19
To: [hidden email]
Subject: Re: Image Processing Software

The story is much simpler. Most of the com
*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi Martin, Mark and others.

The story is much simpler. Most of the commercial software is 5 or more years behind the current developments in the filed. The open source covers much broader spectrum of users.

It is a sort of obvious that private companies have to make money thus that follow the Majority. Advanced science is a Minority.

The rest is just emotions...

I am in the field for over 14 years, believe me.

All the best,

Vitaly

________________________________
From: "[hidden email]" <[hidden email]>
To: [hidden email]
Sent: Friday, September 30, 2011 11:11 AM
Subject: Re: Image Processing Software

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Dare I dip a toe into this hot debate… I will start with the proviso that we
use both commercial and freeware in our labs and they all do have their
strengths.

The point I wanted to address was the generalisation that open source
software is slightly less robust and has less nice GUIs than the commercial
counterparts. I think that I would not use it as a generalisation but say it
is a trend and I have a theory as to why that may be true. If you think
about the time you take coding software you can basically put that time
into: 1. making it do more functions, 2. making it look prettier or be more
intuitive or 3. making it more stable.
A lot of open source projects are small and written by the end user. In
those cases points 2 and 3 are of lower concern because the coders know how
it works and can sidestep bugs. So in those cases it makes more sense to
invest more time in making it do more things.

Also if we go back to evolution 101 we need to consider the selective
pressures in this competitive landscape. Commercial software has to be
rounded and polished to a certain standard or it is not acceptable to the
market, does not sell and no money comes in. So a programmer in a commercial
company may want to write new features but his boss tells him he has to fix
the bugs first. The open source programmer does not have that pressure so
can do more of what they find fun and enjoyable.
Now I am sure that the larger open source projects which have funding for
full time paid programmers can exert pressure on the coders which by then
are effectively employees so that is why they start to get more robust. They
in effect re-introduce the pressure that is present in commercial companies
to fix bugs. Now I am sure pride also motivates open source coders but I
think in the end ‘cash > pride’.

On the GUI side I expect that there are similar pressures. Commercial
solutions have to look pretty so they can spend resource on GUI and workflow
experts. I imagine getting funding for a usability person in a university
funded open source endeavour is a lot harder than getting funding for a
coder to write a new function which furthers our research. And besides the
end result is free so the people who download it cannot complain too much.
So in conclusion I think that there are some pressures which explain why it
may not be unfair to categorise open source software as typically buggier
and less easy to use, but that pressure lowers once the software begins to
get financial funding, so much so that you see things like FIJI and Firefox
being very useable.

But coming back to advice suitable for the list rather than a full thesis I
think you do just have to work out which software will help you to progress
your research in the most cost effective manor. And since its Friday and the
list has a more relaxed feeling I will end sharing an albeit cheeky quote I
quite liked from a visiting sales rep from a software company last year:
“Remember, free software is only free to people who’s time is worth
nothing.”

All the best
Martin

On Fri, Sep 30, 2011 at 3:39 PM, Daniel James White <[hidden email]>wrote:

DRESDEN