xs2eps gives inacceptable results in Scilab 5.1.1

classic Classic list List threaded Threaded
4 messages Options
Matthias Blau Matthias Blau
Reply | Threaded
Open this post in threaded view
|

xs2eps gives inacceptable results in Scilab 5.1.1

Dear list,

in Scilab 5.1.1, xs2eps creates eps files where curves are rendered in
an (IMO) inacceptable "jaggy" manner. This is extremely annoying as it
gives the impression of more noise than actually present in the data. In
this regard, Scilab 4.1.2 gave much better results.

As an example, I attach 2 screenshots of diagrams I made with Scilab
4.1.2 and 5.1.1. (identical code, except for the legend where the handle
structure has changed). Besides the jaggy curcves, you will also notice
that axis and grid lines are thicker in the 5.1.1. version which I also
dislike.

I hope this problem can be fixed soon, as publishing diagrams is a key
feature of any scientific software. I won't convince any of my students
to use Scilab if the results of their work look the way they do right now.

Thanks for any help,
Matthias


Screenshot_scilab412.png (90K) Download Attachment
Screenshot_scilab511.png (102K) Download Attachment
Francis Drossaert Francis Drossaert
Reply | Threaded
Open this post in threaded view
|

RE: xs2eps gives inacceptable results in Scilab 5.1.1

This is indeed a serious problem in Scilab 5.x.x. I hope this can be
rectified ASAP. The xs2eps creates an ugly monster compared to the
function in Scilab 4.x.x. The other export options (unless I am
overlooking something) are not suitable for reproduction in Journals,
where typically 600dpi are required.

Francis


-----Original Message-----
From: Matthias Blau [mailto:[hidden email]]
Sent: 16 May 2009 16:22
To: [hidden email]
Subject: [scilab-Users] xs2eps gives inacceptable results in Scilab
5.1.1

Dear list,

in Scilab 5.1.1, xs2eps creates eps files where curves are rendered in
an (IMO) inacceptable "jaggy" manner. This is extremely annoying as it
gives the impression of more noise than actually present in the data. In
this regard, Scilab 4.1.2 gave much better results.

As an example, I attach 2 screenshots of diagrams I made with Scilab
4.1.2 and 5.1.1. (identical code, except for the legend where the handle
structure has changed). Besides the jaggy curcves, you will also notice
that axis and grid lines are thicker in the 5.1.1. version which I also
dislike.

I hope this problem can be fixed soon, as publishing diagrams is a key
feature of any scientific software. I won't convince any of my students
to use Scilab if the results of their work look the way they do right
now.

Thanks for any help,
Matthias


This email and any files contained therein is confidential and may contain privileged information.  If you are not the named addressee(s) or you have otherwise received this in error, you should not distribute or copy this e-mail or use any of its content for any purpose. Please notify the sender immediately by e-mail if you have received this e-mail in error and delete it from your system

Ronald Michaels Ronald Michaels
Reply | Threaded
Open this post in threaded view
|

RE: xs2eps gives inacceptable results in Scilab 5.1.1

Hi -

This thread came at a good time for me.

Here is how I am now getting high res graphics.

Draw the graph in Scilab with plot() or some other graphics function
call.

In the Scilab graphics window pull down File->Print.

Select the print to file option and print to myfile.ps .

Then start GIMP.  Open myfile.ps .  In the Import from PostScript
window, set resolution = 1000, or some other number that suits your
need.  Click import.  Use GIMP to rotate, crop, etc.  Then Save as
myfile.png, or whatever.

The best thing I can say about this is that it works - WFM.

Ron

On Mon, 2009-05-18 at 09:27 +0100, Francis Drossaert wrote:
> xs2eps


Matthias Blau Matthias Blau
Reply | Threaded
Open this post in threaded view
|

Re: xs2eps gives inacceptable results in Scilab 5.1.1

Dear Ronald,

thanks for your help. Unfortunately, this didn't solve my problem.

Anyway, I found a (similarly cumbersome) way to get "non-jagged" curves.
This is how it works:
- use xs2fig to export to fig
- then in xfig, change the "join style" manually for every curve
- unfortunately, xs2fig is not good in exporting fonts, so I had to
change all fonts manually as well
- then export from xfig to whatever you want

So, this works.

But hey Scilab team, is this how you want people to work seriously?
Again, this may not be a bug in the strict sense, but it makes Scilab
unusable for people doing science (isn't that Scilab's target group?),
so please!! fix this! It can't be that hard.

Best regards, Matthias


Ronald Michaels schrieb:

> Hi -
>
> This thread came at a good time for me.
>
> Here is how I am now getting high res graphics.
>
> Draw the graph in Scilab with plot() or some other graphics function
> call.
>
> In the Scilab graphics window pull down File->Print.
>
> Select the print to file option and print to myfile.ps .
>
> Then start GIMP.  Open myfile.ps .  In the Import from PostScript
> window, set resolution = 1000, or some other number that suits your
> need.  Click import.  Use GIMP to rotate, crop, etc.  Then Save as
> myfile.png, or whatever.
>
> The best thing I can say about this is that it works - WFM.
>
> Ron
>
> On Mon, 2009-05-18 at 09:27 +0100, Francis Drossaert wrote:
>> xs2eps
>