I am unsure about the "canonical" way to scale images to the desired size. my current understanding is, that `scale=' should do what I want in a way portable across `context' incarnations. so currently I use commands like \externalfigure[image.png][scale=750] to adjust the image size to my taste. my problem: the same document looks completely different regarding image size in the produced pdf with standalone installations on osx64 and linux-64. actually, on oxx64 I have to use something like `scale=1500' where on linux-64 `scale=750' seemingly does about the same. questions: * what am I missing? why does the same document compile differently? I can only guess that `context' (or luatex?) has two different opinions of dpi resolution on the two machines when producing the pdf? where can I control/check this? * if `scale=' is not the way to achieve invariant and unambiguous size of images embedded in the document, what is? `width=XXX cm'? i.e.: how is this supposed to be done correctly(TM)? thx/joerg -- Using Opera's revolutionary email client: http://www.opera.com/mail/
Am 2017-02-17 um 15:20 schrieb j. van den hoff
I am unsure about the "canonical" way to scale images to the desired size. my current understanding is, that `scale=' should do what I want in a way portable across `context' incarnations. so currently I use commands like
\externalfigure[image.png][scale=750]
to adjust the image size to my taste.
my problem: the same document looks completely different regarding image size in the produced pdf with standalone installations on osx64 and linux-64. actually, on oxx64 I have to use something like `scale=1500' where on linux-64 `scale=750' seemingly does about the same.
questions:
* what am I missing? why does the same document compile differently? I can only guess that `context' (or luatex?) has two different opinions of dpi resolution on the two machines when producing the pdf? where can I control/check this?
* if `scale=' is not the way to achieve invariant and unambiguous size of images embedded in the document, what is? `width=XXX cm'? i.e.: how is this supposed to be done correctly(TM)?
I can’t tell you anything about the difference in calculations, otherwise it depends on what you want to achieve: I always use width= or height=, because I need specific sizes of my images. Or I define maxwidth=\textwidth. I never had a case where the actual scaling would have been important - maps might be a use case. Greetlings, Hraban --- http://www.fiee.net http://wiki.contextgarden.net GPG Key ID 1C9B22FD
On 2/17/2017 3:20 PM, j. van den hoff wrote:
I am unsure about the "canonical" way to scale images to the desired size. my current understanding is, that `scale=' should do what I want in a way portable across `context' incarnations. so currently I use commands like
\externalfigure[image.png][scale=750]
to adjust the image size to my taste.
my problem: the same document looks completely different regarding image size in the produced pdf with standalone installations on osx64 and linux-64. actually, on oxx64 I have to use something like `scale=1500' where on linux-64 `scale=750' seemingly does about the same.
questions:
* what am I missing? why does the same document compile differently? I can only guess that `context' (or luatex?) has two different opinions of dpi resolution on the two machines when producing the pdf? where can I control/check this?
* if `scale=' is not the way to achieve invariant and unambiguous size of images embedded in the document, what is? `width=XXX cm'? i.e.: how is this supposed to be done correctly(TM)?
just use width=4cm and so ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl -----------------------------------------------------------------
On Fri, 17 Feb 2017 15:43:29 +0100, Hans Hagen
* if `scale=' is not the way to achieve invariant and unambiguous size of images embedded in the document, what is? `width=XXX cm'? i.e.: how is this supposed to be done correctly(TM)? just use width=4cm and so
understood, will do (thanks to henning, too, for answering). in any case, I have settled for `width={fraction}\textwidth' as the most convenient solution. but if you don't mind explaining: _why_ is `scale' causing me a problem in the first place? what is the actual intended (and good) use of this parameter? .. -- Using Opera's revolutionary email client: http://www.opera.com/mail/
On Fri, 17 Feb 2017 17:04:27 +0100
"j. van den hoff"
On Fri, 17 Feb 2017 15:43:29 +0100, Hans Hagen
wrote: * if `scale=' is not the way to achieve invariant and unambiguous size of images embedded in the document, what is? `width=XXX cm'? i.e.: how is this supposed to be done correctly(TM)? just use width=4cm and so
understood, will do (thanks to henning, too, for answering). in any case, I have settled for `width={fraction}\textwidth' as the most convenient solution. but if you don't mind explaining: _why_ is `scale' causing me a problem in the first place? what is the actual intended (and good) use of this parameter?
scale= is EXTREMELY useful when combining external figures and one wants to retain uniform linewidths, text pointsizes, etc. Using width= (or height=) in such a case will lead to very poor results, indeed. Alan P.S. Personally, I find that scale=1000 meaning 1 is an unfortunate left-over from the previous century - prehistoric days. In the chemistry module rewrite, we take abs(scale)>10 in units of 1000, just to confuse things. (Hans: I would suggest dropping this...).
On Fri, 17 Feb 2017 17:51:11 +0100, Alan Braslau
On Fri, 17 Feb 2017 17:04:27 +0100 "j. van den hoff"
wrote: On Fri, 17 Feb 2017 15:43:29 +0100, Hans Hagen
wrote: * if `scale=' is not the way to achieve invariant and unambiguous size of images embedded in the document, what is? `width=XXX cm'? i.e.: how is this supposed to be done correctly(TM)? just use width=4cm and so
understood, will do (thanks to henning, too, for answering). in any case, I have settled for `width={fraction}\textwidth' as the most convenient solution. but if you don't mind explaining: _why_ is `scale' causing me a problem in the first place? what is the actual intended (and good) use of this parameter?
scale= is EXTREMELY useful when combining external figures and one wants to retain uniform linewidths, text pointsizes, etc. Using width= (or height=) in such a case will lead to very poor results, indeed.
OK, thanks. will try to remember that. -- but in this case (important/useful parameter...) I am still wondering, why the same value (say: scale=750) leads to totally different figure size in the pdf output for two different machines/OSes? I understand that scale=1000 means "original size" but in want sense? it seems to dependent on the machine/engine's idea of dpi resolution or something like that ... joerg
Alan
P.S. Personally, I find that scale=1000 meaning 1 is an unfortunate left-over from the previous century - prehistoric days. In the chemistry module rewrite, we take abs(scale)>10 in units of 1000, just to confuse things. (Hans: I would suggest dropping this...).
-- Using Opera's revolutionary email client: http://www.opera.com/mail/
On 2/17/2017 6:05 PM, j. van den hoff wrote:
On Fri, 17 Feb 2017 17:51:11 +0100, Alan Braslau
wrote: On Fri, 17 Feb 2017 17:04:27 +0100 "j. van den hoff"
wrote: On Fri, 17 Feb 2017 15:43:29 +0100, Hans Hagen
wrote: * if `scale=' is not the way to achieve invariant and unambiguous size of images embedded in the document, what is? `width=XXX cm'? i.e.: how is this supposed to be done correctly(TM)? just use width=4cm and so
understood, will do (thanks to henning, too, for answering). in any case, I have settled for `width={fraction}\textwidth' as the most convenient solution. but if you don't mind explaining: _why_ is `scale' causing me a problem in the first place? what is the actual intended (and good) use of this parameter?
scale= is EXTREMELY useful when combining external figures and one wants to retain uniform linewidths, text pointsizes, etc. Using width= (or height=) in such a case will lead to very poor results, indeed.
OK, thanks. will try to remember that. -- but in this case (important/useful parameter...) I am still wondering, why the same value (say: scale=750) leads to totally different figure size in the pdf output for two different machines/OSes?
I understand that scale=1000 means "original size" but in want sense? it seems to dependent on the machine/engine's idea of dpi resolution or something like that ...
or identifying libraries used ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl -----------------------------------------------------------------
On 2/17/2017 5:51 PM, Alan Braslau wrote:
On Fri, 17 Feb 2017 17:04:27 +0100 "j. van den hoff"
wrote: On Fri, 17 Feb 2017 15:43:29 +0100, Hans Hagen
wrote: * if `scale=' is not the way to achieve invariant and unambiguous size of images embedded in the document, what is? `width=XXX cm'? i.e.: how is this supposed to be done correctly(TM)? just use width=4cm and so
understood, will do (thanks to henning, too, for answering). in any case, I have settled for `width={fraction}\textwidth' as the most convenient solution. but if you don't mind explaining: _why_ is `scale' causing me a problem in the first place? what is the actual intended (and good) use of this parameter?
scale= is EXTREMELY useful when combining external figures and one wants to retain uniform linewidths, text pointsizes, etc. Using width= (or height=) in such a case will lead to very poor results, indeed.
Alan
P.S. Personally, I find that scale=1000 meaning 1 is an unfortunate left-over from the previous century - prehistoric days. In the chemistry module rewrite, we take abs(scale)>10 in units of 1000, just to confuse things. (Hans: I would suggest dropping this...).
sx= sy= iir small scales are also seen as fractions ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl -----------------------------------------------------------------
On Fri, 17 Feb 2017 18:42:25 +0100
Hans Hagen
sx= sy=
iir small scales are also seen as fractions
Like for the chemistry module, this can be confusing. Also, sx=,sy= muck with the aspect ratio, don't they? As we know, there is also an inconsistency in the handling of figure scaling. Consider: \startMPcode draw externalfigure("cow") xsized 5cm ; draw rawtextext("\externalfigure[cow]") xsized 5cm ; draw rawtextext("\externalfigure[cow][width=5cm]") ; \stopMPcode The first will draw a distorted, square 5cmX5cm cow, The second and third will both draw an undistorted, rectangular 5cmX3.635cm cow.
or identifying libraries used
which assume different default dpi. Can this be normalized in the case of file formats (such as pdf) that do not explicitly set the dpi? Alan
On 2/17/2017 7:06 PM, Alan Braslau wrote:
On Fri, 17 Feb 2017 18:42:25 +0100 Hans Hagen
wrote: sx= sy=
iir small scales are also seen as fractions
Like for the chemistry module, this can be confusing.
Also, sx=,sy= muck with the aspect ratio, don't they?
As we know, there is also an inconsistency in the handling of figure scaling. Consider:
\startMPcode draw externalfigure("cow") xsized 5cm ; draw rawtextext("\externalfigure[cow]") xsized 5cm ; draw rawtextext("\externalfigure[cow][width=5cm]") ; \stopMPcode
The first will draw a distorted, square 5cmX5cm cow, The second and third will both draw an undistorted, rectangular 5cmX3.635cm cow.
indeed but that has to do with compatibility ...
or identifying libraries used
which assume different default dpi. Can this be normalized in the case of file formats (such as pdf) that do not explicitly set the dpi?
luatex teies to figure it out from the tags in th efile ... there has been some improvements so maybe older luatex and new ones are different Hans ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | www.pragma-ade.nl | www.pragma-pod.nl -----------------------------------------------------------------
On Fri, 17 Feb 2017 19:06:30 +0100, Alan Braslau
or identifying libraries used which assume different default dpi. Can this be normalized in the case of file formats (such as pdf) that do not explicitly set the dpi?
if possible I would say this could help "pedestrians" like myself some: during document creation and processing I nowhere do see any of that. I just get to severly different pdf documents depending on which machine I compile it (it was a factor of 2 different figure size in my case (extending well below page limits ...)). from a user perspective this is of course highly unwelcome (and feels a bit like what is happening with MSword to this day, probably ;-)) -- Using Opera's revolutionary email client: http://www.opera.com/mail/
participants (4)
-
Alan Braslau
-
Hans Hagen
-
Henning Hraban Ramm
-
j. van den hoff