[43S] Display myth crasher



#2

Just to prevent anybody going banana, the picture below shows a 400 x 240 pixel display mockup. This is not natural size but the pixel numbers fit. Feel free to downscale it to 59mm display width preserving the pixel number to get a realistic impression of size and resolution.

d:-)


#3

Your picture is RGB/8 bit per gun, i.e. 256 shades of grey in B/W. Very nice.


#4

Watch it! It's a mockup only!

Eric, does it feature a greyscale or not? (And please look into your mailbox.)

d:-/

P.S.: Seems I've to do pixel design again - but hope dies last.

Edited: 5 Feb 2013, 6:17 a.m.


#5

No greyscale. Strictly B&W.


#6

I had feared that :-/

Apparently that's a work pursuing me. Anyway, it must be done. I need two sizes of characters so far, resulting in the following display for FIX 4:

Enough screenwork for today.

d:-)

Edited the display to implement some minor refinements.


Edited: 9 Feb 2013, 5:01 a.m. after one or more responses were posted


#7

Thank you very much for this mock-up/proof-of-concept, Walter!

BTW a thought provoking (?) Adobe OpenSource (yes, with sources!) sans-serif font family with proportional and monospaced characters of high typographic quality:

http://store2.adobe.com/cfusion/store/html/index.cfm?store=OLS-DE&event=displayFontPackage&code=1959

http://sourceforge.net/adobe/sourcesans/wiki/Home/

http://store2.adobe.com/cfusion/store/html/index.cfm?store=OLS-DE&event=displayFontPackage&code=1960

http://sourceforge.net/projects/sourcecodepro.adobe/files/

Hopefully, greek and cyrillic character sets will follow...

Martin

Edited: 6 Feb 2013, 6:58 a.m.

#8

Hi Walter,

Ich Dummkopf! Now I understand your technique: There isn't necessarily need for a monospaced stack font, because the 'lining tabular figures' have always the same width, even in a proportional font. One have only to construct a 'digit grid', which depends on the display format, e.g. FIX 7 with a grouping comma---a very rough sketch:

Martin

(Ich denke übrigens, daß man sich hinsichtlich der Formatierungen vielleicht etwas vom umfangreichen typographischen Regelwerk zum Setzen von Zahlen befreien darf, zumal kaum Platz für die volle Exponentialschreibweise wie $6.62606957\times 10^{-34}$ sein wird und - soweit ich verstehe - keine sog. "Lehrbuch-Darstellung" wie im HP 300s vorgesehen ist.

Flügges Handbuch der Physik, durchaus auch typographisch ein Maßstab, setzt interessanterweise in keiner der drei Sprachen die Nachkommastellen in Dreiergruppen; Forssman/de Jong schweigen sich dazu im Mathematik-Kapitel ihrer 'Detailtypographie' aus ...

Andererhand erhöhen die Gruppe ja die Lesbarkeit. Könnte man das Exponential-E nicht E-leminieren oder E-rsetzen? Kleiner Abstand und echte 'superscript figures'? ;-)


#9

400 pixels width: 12 pt font size at 172 dpi (Adobe Source Sans Pro)

#10

Good morning!

Looks really a lot better than expected. Any chance for backlighting? Red? Orange? Inverted? With A nixie-tube font maybe :-)

Regards
max


#11

Quote:
With A nixie-tube font maybe :-)
Ha! Yes! That's what feels wrong here: The font. I'd very much prefer a fixed width font.

#12

Quote:
Ha! Yes! That's what feels wrong here: The font. I'd very much prefer a fixed width font.

And proper super/subscript, maybe?

Massimo


#13

Buon giorno Massimo,

Quote:
And proper super/subscript, maybe?

As you've seen with the softkeys, proper super and subscripts need extra height. Besides that, no problem.

d;-)

#14

The display raises hopes! (around 172 dpi---b/w, no greyscale, if I understand right?)

Not only the (main) font is really an important issue, but also the overall typography: For the stack a monospaced font, of course, with good proportions, no serifes, not to small (sub- and superscript may become a problem), execellent readabilty, plain design, technical touch... Why not segmented characters? Just to give an idea (esp. AEG Geascript 111):

http://www.aegmis.de/PRODUKTE/LCD/SEGMENT/tabid/587/language/en-US/Default.aspx

Greek and cyrillic character sets are a separate challange.

Font(s) for menues and the keyboard etc. should more or less match. It will be a complex task (display format of numbers...), even for a good typographer with experiences in technical design... (BTW: Ask TI or Casio etc. how to make it wrong.)

Just my opinion.

Martin


Edited: 5 Feb 2013, 5:58 a.m. after one or more responses were posted


#15

Hallo Martin,

The LCD will feature 400 x 240 plain square pixels. Segmented fonts work for macroscopic sheets of plastic - they look quite nice but won't work with pixels AFAIK.

d:-/


#16

Hi Walter,

That's true, but I think one can rasterise them. The display resolution should be high enough for a smooth appearance.

Best regards,
Martin

Edited: 5 Feb 2013, 5:43 a.m.


#17

Without font hinting, the rasterized image looks like this:

Quite unfair, I admit ;-).


#18

Thanks - it's quite obvious where work is necessary.

d:-)


#19

I don't know if a decision has been made about it already, but I'd like to suggest switchable font sizes for the main stack display. At least three different sizes would be nice, where at least the smallest (monospaced) font should allow for more chars per line.

And except for the menus, where a proportional font may gain some char display space, I'd use monospaced fonts only.

#20

And have a look at HrastProgrammer's HP-1xC emulators for the HP-48/49:

http://www.hrastprogrammer.com/emulators.htm

The emulation of the segmented digits (here only 7 segments for the Voyagers, but even 16 won't be enough) etc. is IMHO very readable and has the touch of the "scientific instrument" the 43S should become.

Firstly, I will continue my investigation with fonts like AEG Geascript...


Greetings from the West-end Tower in Mainhattan!

Martin

#21

Addendum: A proportional font for the stack may be an option (strings etc.). For the softkey labels I recommend it. And for the keyboard it's a must!

#22

But does it do Chinese?!?!?!?!? :-)

TW


#23

Tim, what do you know of Chinese?

d;-)

#24

The display will do Chinese just fine.

#25

It seems fixed width to me. The space as thousands separator may look strange to some, personally I like it more than dots, commas or apostrophes.

Just my AR$ 0.10

(edited to correct previous typo in "...more than...")

Edited: 6 Feb 2013, 10:47 a.m. after one or more responses were posted


#26

Quote:
The space as thousands separator may look strange to some, personally I like it more tha dots, commas or apostrophes.

That narrow space is even demanded by ISO 31-0 - see the footnote on p. 39 of the WP 34S manual.

d:-)


#27

Good point! ISO requirements are (or should be) stronger than our personal preferences!

As I said, I think it's more clear to use narrow spaces than punctuation symbols as separators (at least I don't feel comfortable with the USA-style usage of commas as thousands separator). Here is usual to use comma as radix and dots as thousands separators, but I don't like such an arrangement either.


#28

Quote:
... I think it's more clear to use narrow spaces than punctuation symbols as separators (...). Here is usual to use comma as radix and dots as thousands separators, but I don't like such an arrangement either.

I fullheartedly concur. Let's go ISO!

d:-)

#29

Quote:
Any chance for backlighting? Red? Orange? Inverted? With A nixie-tube font maybe :-)

You can get whatever you want on an emulator - compare the WP 34S.

d;-)


#30

Quote:
You can get whatever you want on an emulator - compare the WP 34S.

I have 0 (in words: zero) interest in emulators. Never had. I am a collector of hardware, not software :-)


#31

Quote:
I have 0 (in words: zero) interest in emulators. Never had. I am a collector of hardware, not software :-)

And we have 0 (in words: zero) interest in color screens. Never had. At least I'm interested in science, not gimmicks.

d:-)


#32

Quote:
At least I'm interested in science, not gimmicks.

Luckily, science and gimmicks do not exclude each other. Most gimmicks owe their existence to science!

And just imagine the pictures of the Hubble Space telescope in black and white :-(

#33

If there was backlighting, it would be white. However, there won't be backlighting as it would slash the battery life.


#34

Hi Eric,

Indeed, low power consumption should have priority in designing a reliable device for field/laboratory/... use. Beside many many other hardware and software weaknesses e.g. of TIs nSpire CX series---high-school kid's stuff in my opinion---the short battery life is a deal-breaker. BTW there's IMHO no need for color graphics in a scientific calculator---unless wie are thinking about a 51G-RPL-like machine... ;-)

But may it possible to include a simple hardware switch for backlighting in the 43S design? In field/cockpit/... use it may be very beneficial, I imagine, and the operator decides when to "stress" the batteries with activating the backlight.

Martin

Edited: 6 Feb 2013, 3:45 a.m.


#35

Firstly, EXCELLENT work guys!

Secondly, a selectable backlight as stated by Martin would be ideal, but not necessary as my 41, 42', and WP34s all work on the flight deck.

Cheers


#36

Good morning!

Quote:
?..but not necessary as my 41, 42', and WP34s all work on the flight deck...

On _your_ flight deck maybe, but unfortunately not on mine :-(
The few times I ever need to calculate anything in the aeroplane at night, I either use my company-supplied Blackberry (the lighted keyboard is very useful!) or my iPad. Sometimes I wonder why they don't build a simple calculator function into the 100.000$ FMS?

Max

#37

It's not technically impossible, but the transmissivity of the display is only 0.25%. A backlight would be very expensive (in addition to the already extremely high cost of the LCD), and you'd have to run the backlight very bright, and drain the batteries in no time flat.

On the other hand, Harbor Freight sells a perfectly fine LED frontlight that will run for hours on three AAA batteries, for under $5 each (quantity two).


#38

OK, so making a long story short: There will be no backlight.

d:-)

#39

Quote:
On the other hand, Harbor Freight sells a perfectly fine LED frontlight that will run for hours on three AAA batteries, for under $5 each (quantity two).

And, if you keep your eye out for their coupons, it can be FREE. I have several, and they are quite bright.

#40

Excellent!


#41

I'm guessing that the little top line which shows the time & date will also have the annunciators. If there is room for more than the 41's flags 00-04 that would be nice. I've wished for one more but i probably couldn't keep track of any more than seven total anyway.


#42

Quote:
I'm guessing that the little top line which shows the time & date will also have the annunciators. If there is room for more than the 41's flags 00-04 that would be nice.

You were guessing correctly :-) Please see the last page of UI.pdf for the annunciators defined (there will be a few more, I guess). Whatever space will be remaining may well be used for flags. Anyway it's a bit early for talking about remaining screen space IMHO.

d:-)

#43

Quote:
Feel free to downscale it to 59mm display width preserving the pixel number to get a realistic impression of size and resolution.

This will not make much sense, I fear. If you keep the number of pixels constant, the size on your computer display will stay the same as well since its resolution (dpi) does not change.

400 pixels on 59 mm are 172 dpi. Unless your PC-display does not feature at least this resolution, there is no way to tell how the 43s display will look. A regular 24" PC-display with Full-HD resolution is able to resolve roughly half of this. So all we can tell by now is the fact that this will be a very high resolution screen.

Dieter


#44

How about using a printer?

d:-)


#45

This would require access to a printer with true (full color) 172 dpi printing resolution. No dithering, no other interpolation due to a limited number of printing colors. Maybe a decent dye-sublimation system would do.

You also could take the file to a photo lab and have it printed on a digital minilab - with three RGB lasers on "real" (light-sensitive, chemically processed) photographic paper. That's probably the most simple real-world solution.

Dieter

#46

Quote:
this will be a very high resolution screen.

Yes it is. Eric showed his DIY5 calculator at the HHC2012 conference. When he passed it around, some, including myself, thought we were looking at a mock-up which had a photo of the stack. Not until you pressed the keys, did you realize the "photo" was the working display - the numbers were SO crisp and clear with very high contrast. It truly is amazing how good it looks. Good job Eric and Richard.

Matt

#47

I'm sorry to diverge this thread, but which calculator is the 43s based on?


#48

No problem. Search for DIY5 (by Eric & Richard).

d:-)


Possibly Related Threads...
Thread Author Replies Views Last Post
  Trivial news of the 43S (no Prime) Walter B 117 7,971 11-22-2013, 03:26 AM
Last Post: Raymond Del Tondo
  What exactly is the WP 43S Harold A Climer 5 808 11-16-2013, 03:58 PM
Last Post: Jonathan Cameron
  Another wishlist for 43s Andrew Nikitin 3 575 07-14-2013, 09:25 PM
Last Post: Andrew Nikitin
  Some random wish list items for the 43s Marcel Samek 18 1,758 07-11-2013, 05:35 PM
Last Post: Paul Dale
  News from the 43S Walter B 78 4,578 07-08-2013, 10:33 AM
Last Post: Victor Koechli
  [43S] Display and top row, refined Walter B 20 1,746 04-24-2013, 01:36 AM
Last Post: Walter B
  [43S] Display Walter B 67 3,876 03-17-2013, 09:22 AM
Last Post: Walter B
  [43s] : How the solver will be implemented Miguel Toro 3 604 03-14-2013, 06:09 PM
Last Post: Walter B
  [43S]Top row poll (until Monday, 4.2.13, 13:13 UTC) Walter B 175 10,454 02-04-2013, 09:38 AM
Last Post: Andrés C. Rodríguez (Argentina)
  [43s] General questions fhub 7 787 01-31-2013, 11:04 AM
Last Post: Walter B

Forum Jump: