Sampling: What Nyquist Didn't Say, and What to Do About It

On 21/12/10 20:55, D Yuniskis wrote:
> Hi Grant,
>
> On 12/21/2010 12:13 PM, Grant Edwards wrote:
>> On 2010-12-21, D Yuniskis<not.going.to.be@seen.com> wrote:
>>
>>>> I prefer long lines, reads much better.
>>>>
>>>> Then you're in a small minority. Cognative research shows that long
>>>> lines are harder to read for pretty much everybody else.
>>>
>>> Agreed -- though I may have some "inherent" issue with "minimizing
>>> head/eye motion" (e.g., my workstations are designed so I don't have
>>> to move my head to see different monitors).
>>>
>>> Short lines (e.g., 2 or 3 column format) read *much* quicker (IMO)
>>> than long ones. Just the difference between a paperback and a
>>> "hard-bound" tome alone is significant (though I wonder if that isn't
>>> because the actual *physical* width of the text is smaller?).
>>
>> The research that I've read over the years indicates that both
>> physical width (specifically the angle subtended by the line), and the
>> number of characters had an effect.
>
> The former seems to coincide with my experience. I had
> assumed it "trumped" the latter. (?) I would have to
> think about how/why "character count" would be significant
> (barring the obvious correlation that, for a given "angle",
> more characters means *smaller* characters).
>

The basic physiological issue here is the time taken to read the line - 
if there is too much in it, then you have "forgotten" where the line 
started by the time you want to move to the next line, even if you can 
still "see" the start of the line.

So both the subtended angle (which is obviously dependent on the reading 
distance) and the line's contents matter.

It also depends somewhat on the type of text.  If there are a lot of 
long and technical words, it's important that the line is long enough 
that these are not often broken.  But if the text is slow to read, or 
boring, lines should be short or you will again lose track of the line 
starts.

>>> With short line lengths, you never lose sight of the left edge
>>> of the column. So, moving to the next line requires less
>>> visual (subconscious) "hunting".
>>
>> Right. The longer the line, the harder it is to jump to the beginning
>> of the next line without getting "lost" or making false starts on the
>> wrong line.
>
> Setting the type "ragged right" actually helps with readability.
> But, it doesn't "look pretty".
>

Ragged right is certainly better than rampant hyphenation or wildly 
varying spacing to justify the text.  Good typesetters for short-line 
publications like newspapers will work with the editors to change text 
to fit, so that it can be readable /and/ pretty.

> DTP is a lot more "art" than one would think. You *assume* it's
> just a matter of getting everything to fit on the page and
> making it "look pretty". In fact, making something that is
> "easy to read" (i.e., so that it doesn't interfere with the
> *content*) can be very difficult.
>

There is also a lot more "science" to typesetting than most people 
think.  Even if you don't want to use TeX or its friends, I'd thoroughly 
recommend reading the TeXbook to see how Knuth thinks about typesetting.

> Knuth (?) commented that once you start designing fonts, it is
> hard NOT to look at EVERY font that you encounter with a
> critical eye. I.e., you stop seeing things as "words on paper"
> but, instead, start inspecting individual glyphs, kerning, etc.
> The same is true of DTP -- you look at why someone opted for
> "big caps", or outdents, or...
>

I know the feeling.  When you've used TeX or LaTeX, and especially if 
you've read the books explaining the background, it is hard not to think 
about it.  I can usually spot a (La)TeX'ed document immediately - there 
is just far more attention to the small details than you get with other 
programs.  It is often easy to see when someone has used a professional 
program like Frame Maker rather than an amateurish word processor, but 
the difference is not as great unless it is done by a very skilled 
typesetter.  I am often asked to proof-read documents at work - it is 
sometimes hard to concentrate on the relevant issues rather than glaring 
double-space errors or font issues.

> Always fun to see newbies using dozens of "fonts", colors, etc.
> Starts to read like: "LeAvE $1,o0o,00o iN a PapuR bAg uNDer
> tHe BRidgE at 5tH & MaIn iF yOu EVeR wANt tO sEe yOuR..." :>

On 2010-12-21, D Yuniskis <not.going.to.be@seen.com> wrote:

> Always fun to see newbies using dozens of "fonts", colors, etc.
> Starts to read like:  "LeAvE $1,o0o,00o iN a PapuR bAg uNDer
> tHe BRidgE at 5tH & MaIn iF yOu EVeR wANt tO sEe yOuR..."  :>

And then along came HTML and they could put all that mess on top of a
nice busy background image!

-- 
Grant

On 2010-12-21, David Brown <david.brown@removethis.hesbynett.no> wrote:

> There is also a lot more "science" to typesetting than most people 
> think.  Even if you don't want to use TeX or its friends, I'd thoroughly 
> recommend reading the TeXbook to see how Knuth thinks about typesetting.
>
>> Knuth (?) commented that once you start designing fonts, it is
>> hard NOT to look at EVERY font that you encounter with a
>> critical eye. I.e., you stop seeing things as "words on paper"
>> but, instead, start inspecting individual glyphs, kerning, etc.
>> The same is true of DTP -- you look at why someone opted for
>> "big caps", or outdents, or...
>
> I know the feeling.  When you've used TeX or LaTeX, and especially if 
> you've read the books explaining the background, it is hard not to think 
> about it.  I can usually spot a (La)TeX'ed document immediately - there 
> is just far more attention to the small details than you get with other 
> programs.

With (La)TeX, the defaults pretty much always produce a good-looking
document.  And, if you're not happy with the results, you can always
tweak things to make it worse.  [At least that's usually what happens
when I decide I want to start changing things.]

-- 
Grant

On a sunny day (Tue, 21 Dec 2010 13:57:09 -0700) it happened D Yuniskis
<not.going.to.be@seen.com> wrote in <ier3pv$a0q$1@speranza.aioe.org>:

>Hi Jan,
>
>On 12/21/2010 12:18 PM, Jan Panteltje wrote:
>> On a sunny day (Tue, 21 Dec 2010 19:12:35 GMT) it happened Jan Panteltje
>> <pNaOnStPeAlMtje@yahoo.com>  wrote in<iequ7b$m90$1@news.datemas.de>:
>>
>> PS
>> I prefer it this way:
>>   http://panteltje.com/panteltje/pic/gm_pic/
>
>That forces you to look *past* a photo for the next bit of text.
>I don't "interrupt" the reader unless the interruption is
>"worthwhile".  I.e., a reader should be able to keep reading
>and ignore photos, illustrations, tables, etc. if they aren't
>"significant enough".

Yes, I should perhaps just have provided a link to the big pictures,
and put the text next to it.
Something to consider for the next project (that is already progressing in a nice way):
 ftp://panteltje.com/pub/sc_pic/xscpc.gif
Uses a photo multiplier tube and scintillation crystal:
 ftp://panteltje.com/pub/PMT/PMT_1_img_2435.jpg
Once the project is finished, I will use your suggestions to make that web page.
Then there is already the next project materialising.. but that is a VERY cold project,
super conducting cold actually.


>For example, I like to try to arrange photos/tables/etc.
>at the top or bottom of a column so they user can conveniently
>ignore them.  OTOH, doing this rigidly, results in a boring
>(visually) presentation.
>
>In the newsletter I mentioned (elsewhere), for example, I
>took liberties with which "articles" were on each page.
>And, did some significant editing to those articles to
>cause images to fall where I wanted them.
>
>When I do technical presentations, I try to silently impose
>a structure on each page so the reader knows where to look
>for things.  E.g., if I have "screenshots" in the document,
>then I might opt to "always" put them in the same relative
>position on a page -- so the user's "muscle memory" directs
>him to the text or image, as appropriate (without having to
>"hunt")

Yes, on my website I sort of start from the point of view that if somebody is REALLY
interested they will need all data they can get.
So I try to provide the essential info.
And I assume some real knowledge how to use it.
It is not a training or electronics education, something like that.
Although I just got a letter from the chamber of commerce that I was registered as business training centre, can you imagine.
Wonder hat gave them that idea :-)

Hi Jan,

On 12/21/2010 3:13 PM, Jan Panteltje wrote:
> On a sunny day (Tue, 21 Dec 2010 13:57:09 -0700) it happened D Yuniskis
> <not.going.to.be@seen.com>  wrote in<ier3pv$a0q$1@speranza.aioe.org>:
>
>> On 12/21/2010 12:18 PM, Jan Panteltje wrote:
>>> On a sunny day (Tue, 21 Dec 2010 19:12:35 GMT) it happened Jan Panteltje
>>> <pNaOnStPeAlMtje@yahoo.com>   wrote in<iequ7b$m90$1@news.datemas.de>:
>>>
>>> PS
>>> I prefer it this way:
>>>    http://panteltje.com/panteltje/pic/gm_pic/
>>
>> That forces you to look *past* a photo for the next bit of text.
>> I don't "interrupt" the reader unless the interruption is
>> "worthwhile".  I.e., a reader should be able to keep reading
>> and ignore photos, illustrations, tables, etc. if they aren't
>> "significant enough".
>
> Yes, I should perhaps just have provided a link to the big pictures,
> and put the text next to it.

For HTML, I'd put a reduced resolution/size image "in-line"
with the associated text.  And, a link *on* the image so you
could "click for bigger picture".

[I consider this too much work -- which is why I don't
create HTML documents!  :> ]

While I despise PDF's, they are an effective way
of controlling content *and* presentation.

["No matter how much you dislike pickles, they are,
after all, the only thing you can do with CUcumbers!"
(a bizarre reference that I suspect few will recognize)]

Hi Jan,

On 12/21/2010 1:49 PM, Jan Panteltje wrote:
> On a sunny day (Tue, 21 Dec 2010 13:50:30 -0700) it happened D Yuniskis
> <not.going.to.be@seen.com>  wrote in<ier3dg$96o$1@speranza.aioe.org>:
>
>> <grin>   This is what I was "competing" with:
>>
>> Before:
>> http://www.fkbcl.org/f/Fall_2009_Newsletter.pdf
>
> # wget http://www.fkbcl.org/f/Fall_2009_Newsletter.pdf
> --21:44:41--  http://www.fkbcl.org/f/Fall_2009_Newsletter.pdf
>             =>  all_2009_Newsletter.pdf'
> Resolving www.fkbcl.org... 69.90.45.41
> Connecting to www.fkbcl.org|69.90.45.41|:80... connected.
> HTTP request sent, awaiting response... 404 Not Found
> 21:44:42 ERROR 404: Not Found.

<Grrrr>  My fault cutting and pasting.

http://www.fkbcl.org/f/Fall_2008_Newsletter.pdf

(Fall 200*8* precedes Spring 2009  :<  I was no longer with the
group in Fall 2009 -- which is why the *next* newsletter didn't
show up until Spring 2010!)

>> While I am an advocate of FOSS (though don't run Linux), I am
>> not a zealot.  I *gladly* avail myself of the tools that are
>> available under (e.g.) Windows.  Especially if they make my life
>> easier or my "product" better!
>
> Yea, but I burned my XP disk, as it wasted too much time.
> So far Linux has always helped me out, did not want to put
> too much time into that DTP stuff back then,
> I write my web pages in html with a text editor.

The only "big" DTP (i.e., hundreds of pages) I've done were with
Ventura (starting in the "GEM" days).  I could coerce VP to do
a lot more "clever" things -- but, it required more hand-holding
(though I suspect some of that had to do with the state-of-the-art
at the time).

When Corel started mucking with VP (in particular, when they
replaced the TEXT files that VP used with "wacko prorpietary
format" files -- which were impossible to "patch"), I went
looking for a replacement tool (e.g., Quark, Frame, etc.).

FrameMaker doesn't let me play all the layout tricks that VP
could be coerced into doing.  But, it adds some other
capabilities that VP didn't have (e.g., a much nicer equation
editor).  For small publications (e.g., newsletter, my "notes"
series, etc.) it gets out of my way and lets me get the job done
very quickly.  E.g., most of the time spent on that newsletter
(discounting taking pictures, rounding up articles, etc.) was
spent editing other peoples' prose (it is amazing how uninspired
some folks are as writers  :< ) and gluing together the panoramic
photos.  The actual "publishing" was probably just a few hours.

>>> At very old age, if I get that far, when all I can do is press a key,
>>> maybe I should write a book.
>>> Although these days it is probably 'twitter'.
>>> :-)
>>
>> <frown>
>
> Yes, I frown on that too, but the younger generation seems to use it a lot.

<shrug>  We'll see what they say later in life (I take privacy
issues considerably more seriously than the young-uns)

On 12/21/2010 03:21 PM, D Yuniskis wrote:
> [...]
> For an example of this, see the center of page 4 at
> http://www.fkbcl.org/f/Spring_2009_Newsletter.pdf
> also note the fancy photos!

Wow, that looks fantastic! I admire a piece of good
typesetting, at least as far as my eye can see, and
this close to the top of my scale.

> The software that does this is BFM as far as I am concerned!

BFM?
-- 
Randy Yates                      % "My Shangri-la has gone away, fading like
Digital Signal Labs              %  the Beatles on 'Hey Jude'"
yates@digitalsignallabs.com      %
http://www.digitalsignallabs.com % 'Shangri-La', *A New World Record*, ELO

On a sunny day (Tue, 21 Dec 2010 15:52:40 -0700) it happened D Yuniskis
<not.going.to.be@seen.com> wrote in <ieraii$p7j$1@speranza.aioe.org>:

><Grrrr>  My fault cutting and pasting.
>
>http://www.fkbcl.org/f/Fall_2008_Newsletter.pdf
>
>(Fall 200*8* precedes Spring 2009  :<  I was no longer with the
>group in Fall 2009 -- which is why the *next* newsletter didn't
>show up until Spring 2010!)

Looks like a scan to me, page 2 text is not horizontal, grey background, no colors.
Yours is a zillion times nicer.


>The only "big" DTP (i.e., hundreds of pages) I've done were with
>Ventura (starting in the "GEM" days).  I could coerce VP to do
>a lot more "clever" things -- but, it required more hand-holding
>(though I suspect some of that had to do with the state-of-the-art
>at the time).
>
>When Corel started mucking with VP (in particular, when they
>replaced the TEXT files that VP used with "wacko prorpietary
>format" files -- which were impossible to "patch"), I went
>looking for a replacement tool (e.g., Quark, Frame, etc.).

Corel did very strange things, they once made a Linux distro,
and I bought it,
In that distro they redirected all error messages to /dev/zero,
so if something did not work you would not know about it.
Was on my system for a VERY short time (hours),
before it was replaced by Suze IIRC.


>FrameMaker doesn't let me play all the layout tricks that VP
>could be coerced into doing.  But, it adds some other
>capabilities that VP didn't have (e.g., a much nicer equation
>editor).  For small publications (e.g., newsletter, my "notes"
>series, etc.) it gets out of my way and lets me get the job done
>very quickly.  E.g., most of the time spent on that newsletter
>(discounting taking pictures, rounding up articles, etc.) was
>spent editing other peoples' prose (it is amazing how uninspired
>some folks are as writers  :< ) and gluing together the panoramic
>photos.  The actual "publishing" was probably just a few hours.
>
>>>> At very old age, if I get that far, when all I can do is press a key,
>>>> maybe I should write a book.
>>>> Although these days it is probably 'twitter'.
>>>> :-)
>>>
>>> <frown>
>>
>> Yes, I frown on that too, but the younger generation seems to use it a lot.
>
><shrug>  We'll see what they say later in life (I take privacy
>issues considerably more seriously than the young-uns)

These days we are all an open book to gov :-)

Hi David,

On 12/21/2010 2:06 PM, David Brown wrote:
> On 21/12/10 21:21, D Yuniskis wrote:
>> On 12/21/2010 12:12 PM, Jan Panteltje wrote:
>>> On a sunny day (Tue, 21 Dec 2010 12:11:41 -0700) it happened D Yuniskis
>>> I means what those DTP programs sometimes
>>> produce, and then they use variable spacing between the WORDS
>>> to get even line length... terrible.
>>
>> Things are much better with modern DTP tools. In years past,
>> to pad a line to full length, you had to insert *whole*
>> spaces. For fixed width fonts (e.g., courier), this can be
>> "like fingernails scraping on a chalkboard".
>>
>> Note that most tools will also play games balancing column
>> lengths (if you so choose). Some will even use *variable*
>> line SPACING to achieve this!
>
> I can see the point of that - by reducing the vertical line spacing, you
> are reducing the area of the large space and thus its visual effect. I
> am not sure I like it, however - I thing the line spacing change is
> distracting and the space is still too big.

I think feathering is intended for more "artistic" presentations.
E.g., "I have a two column-inch box that I want 'THIS BIT OF TEXT'
to *fill* -- stretch/compress it as necessary to achieve that goal".

So, in multicolumn layouts, it just looks silly -- you don't want to
(usually) "spread out" (vertically) one column just so it is a
certain vertical size (there are other tricks that FM will employ
to balance the columns without feathering).  Instead, you usually
want to "synchronize" the text baselines of adjacent columns
(otherwise, the visual effect is very unsettling).

Recall that your document is not one long string of characters
(with images/tables interspersed).  Rather, it may be composed
of many *discrete* "flows" (FM-speak).  E.g., when I want to past
a lengthy bit of code into an article, I create a separate "flow"
for that code -- with it's own layout rules.  Then, I tell FM
which "frames" the flow should occupy.

So, I can start the flow in a 3"x4" box on page 1 and continue it
in a 3"x6" box on page 7 (if that made sense).  With this in mind,
you can see how some publications (e.g., advertisements!) might
want to use things like feathering in particular "frames".

> Sometimes there is nothing that can be done to make the typesetting look
> good. The answer in a case like this is to slightly re-write the text
> until you get a good fit - not to massage the spacing to give a slightly
> less bad fit.

You use both techniques.  You can adjust the kerning and spacing within
lines (and character sequences).  E.g., highlight a section of text and
then interactively squeeze or stretch the characters within that
highlighted region.

The downside of short line lengths -- especially in technical
publications -- is that you tend to end up with "big words"
(and/or words that are glued together like "and/or"  :> ).
This doesn't give you many places to "insert whitespace".
So, if you *don't* set it "ragged right", you frequently
end up with things like:
       |the    counterclockwise    rotation|
Your only remedy is to let some other word(s) onto the line
and hyphenate them (I *really* dislike hyphenation!  With
narrow columns, it becomes VERY frequent)

> The tool in question (FrameMaker) seems to do a better job than word

Word processors are crap.  I have yet to find a need for one
*other* than writing single page correspondence (I do even
that in FrameMaker since I don't want to deal with yet
another program that does "something similar").

> processors, and does a reasonable job of the hyphenation, but it has a
> lot to learn from TeX. The typesetter (person rather than program) has

Ah, well... I'll gladly refund the money that I (wasn't) paid!  :>

> missed a few points too - though again, it is typeset far better than
> most publications these days, and it looks very nice.

Hi David,

On 12/21/2010 2:19 PM, David Brown wrote:

[attributions elided]

>>>> With short line lengths, you never lose sight of the left edge
>>>> of the column. So, moving to the next line requires less
>>>> visual (subconscious) "hunting".
>>>
>>> Right. The longer the line, the harder it is to jump to the beginning
>>> of the next line without getting "lost" or making false starts on the
>>> wrong line.
>>
>> Setting the type "ragged right" actually helps with readability.
>> But, it doesn't "look pretty".
>
> Ragged right is certainly better than rampant hyphenation or wildly
> varying spacing to justify the text. Good typesetters for short-line
> publications like newspapers will work with the editors to change text
> to fit, so that it can be readable /and/ pretty.

The problem is that rewriting takes a disproportional amount of
time.  When I write technical documentation, I often DELIBERATELY
use boilerplate prose.  The point being to let the user ignore
the repetitious aspects of the document and concentrate on that
which *differs* from one "topic" to the next.

E.g., like consistently-written man(1) pages (you don't even see
the cruft parts of the sentences that just serve to provide
"proper grammar" for the *MEAT* of the sentence).

>> DTP is a lot more "art" than one would think. You *assume* it's
>> just a matter of getting everything to fit on the page and
>> making it "look pretty". In fact, making something that is
>> "easy to read" (i.e., so that it doesn't interfere with the
>> *content*) can be very difficult.
>
> There is also a lot more "science" to typesetting than most people
> think. Even if you don't want to use TeX or its friends, I'd thoroughly
> recommend reading the TeXbook to see how Knuth thinks about typesetting.

When I did my first DTP project, I canvassed the state of the
art for tools, texts, etc.  In addition to _The TeXbook_ (volumes
A through E, hard copy), I have an assortment of texts on the
(legacy) printing process, composition, style guides, etc.

>> Knuth (?) commented that once you start designing fonts, it is
>> hard NOT to look at EVERY font that you encounter with a
>> critical eye. I.e., you stop seeing things as "words on paper"
>> but, instead, start inspecting individual glyphs, kerning, etc.
>> The same is true of DTP -- you look at why someone opted for
>> "big caps", or outdents, or...
>
> I know the feeling. When you've used TeX or LaTeX, and especially if
> you've read the books explaining the background, it is hard not to think
> about it. I can usually spot a (La)TeX'ed document immediately - there
> is just far more attention to the small details than you get with other
> programs. It is often easy to see when someone has used a professional
> program like Frame Maker rather than an amateurish word processor, but
> the difference is not as great unless it is done by a very skilled

Chase down the "before" and "since" links I mentioned in one of these
posts to see how "others" have tackled the same newsletter.  <frown>
It's disturbing (in any field) when people *think* something is
"easy" just because it *looks* like it SHOULD be easy!

I am always amazed at the various part of my brain that are "missing".
E.g., I can't draw a living creature.  But, I can draw landscapes,
buildings, plants, etc. in very good detail (proper multippoint
perspective, etc.).  I.e., that portion of a human brain that
can draw people is MISSING in my case!  :>  OTOH, if you were
to show me a drawing of a person, I could immediately tell you
what was wrong with the drawing and how to correct it!

(I suspect a similar phenomenon explains why so many people can't
"design from scratch" -- but can patch 'till the cows come home!)

> typesetter. I am often asked to proof-read documents at work - it is
> sometimes hard to concentrate on the relevant issues rather than glaring
> double-space errors or font issues.

For me, it is spelling (though I don't use a spell-checker -- to
force myself to be *better* at it!).  And, particularly, getting
names correct (e.g., all of the names of the "sponsors" in that
newsletter).  Thankfully (regrettably?), I only obsess about it
in formal documents...

>> Always fun to see newbies using dozens of "fonts", colors, etc.
>> Starts to read like: "LeAvE $1,o0o,00o iN a PapuR bAg uNDer
>> tHe BRidgE at 5tH & MaIn iF yOu EVeR wANt tO sEe yOuR..." :>