I'm just about to buy the DS1000s, but... - Page 5

badbob001 · 01-27-2010, 11:12 AM

I'm guessing screen refresh involves these three steps:

Time to process command (ie: register click, open document, access requested content, etc... all limited by cpu, ram, and storage bandwidths)
Time to pass complete screen image to e-ink controller (in irex's case, the delta controller) and for it to process.
Time for e-ink paper to update

When you open a document on a traditional display, steps 1, 2, and 3 can start in parallel: as the document is being loaded, the screen can start showing the application interface and when more of the document is loaded, more of the document can be displayed. It's like viewing a web page where things load in chunks, giving you a sense of progress.

On an e-ink device, the above steps usually need to occur in serial (one before the other). For a full page refresh, such as with a page turn or when opening a document, the e-ink controller needs to have all the screen elements in place before it can finally tell the e-paper what to display. There is capability for partial refresh, but that involves changing small parts of the same page and the original image can be degraded in the process. I guess one can think of e-ink like printer paper... you print to it when you are ready.

I suppose if device makers think components that are 'fast enough' for traditional display devices is good with an e-paper display swapped in may need to rethink things. If they can get steps 1) and 2) to be really really fast, then the physical slowness of step 3) will be less noticeable.

plisken · 01-27-2010, 11:15 AM

@Shaggy, not I am not comparing those speeds. I am responding to your statement, that every eradear has the same page turn time because of the e-ink delay. E-ink delay can be easily measured (I mean, top of it) by measuring response time of "please wait". Or IOW -- the long time of page turn on DR1000 is not caused solely by e-ink refresh time.

> Still not entirely sure what you're saying. You mean that it took the DR half
> an hour to turn a page?

Yes, or saying more exactly -- it took more than half an hour. How much I don't know (to be fair, for 2.8GHz Core2Duo computer the page turn of such pdfs is not an easy task as well, however the problem is computer is responsive while page turning, while DR1000 is not, and what worse there is no cancel button).

DR1000 didn't freeze, if it did, it wouldn't shutdown.

@badbob001

You forgot step after (1) -- rendering the page.

Shaggy · 01-27-2010, 11:33 AM

Quote:

Originally Posted by plisken

E-ink delay can be easily measured (I mean, top of it) by measuring response time of "please wait". Or IOW -- the long time of page turn on DR1000 is not caused solely by e-ink refresh time.

The point I am making is that there are multiple values for "E-ink delay". A full screen refresh takes something like 3-4 times longer than a partial screen refresh. The eInk refresh time is generally the largest component of the overall page turning time.

Quote:

> Still not entirely sure what you're saying. You mean that it took the DR half
> an hour to turn a page?

Yes, or saying more exactly -- it took more than half an hour. How much I don't know (to be fair, for 2.8GHz Core2Duo computer the page turn of such pdfs is not an easy task as well, however the problem is computer is responsive while page turning, while DR1000 is not, and what worse there is no cancel button).

It sounds like there is something about your document that caused the viewer to either crash or hang. If I had to take a guess it would be something to do with running out of memory trying to process that particular page. If that document is stressing a 2.8GHz PC, then it's not much of a surprise that it's causing a problem on a much smaller device.

Quote:

DR1000 didn't freeze, if it did, it wouldn't shutdown.

Actually, yes it would.

badbob001 · 01-27-2010, 11:46 AM

Quote:

Originally Posted by plisken

@badbob001
You forgot step after (1) -- rendering the page.

For simplicity, I was trying to lump everything that a normal non-e-ink device would do into step one and have step two and three be what is added/replaced from using e-ink. So I suppose rendering would be part of step 1.

plisken · 01-27-2010, 12:03 PM

@Shaggy

> The eInk refresh time is generally the largest component of the overall page
> turning time.

Sorry, but I cannot agree -- taking the fastest page turn as the bottom time (i.e. nothing is better) I would be just happy to have DR1000 and I would tell everyone how cool the device is.
But on average DR1000 thinks way longer, and it is obviously not e-ink refresh time.

> If that document is stressing a 2.8GHz PC, then it's not much of a surprise
> that it's causing a problem on a much smaller device.

I am not surprised by that. I am surprises that someone didn't anticipate that fact -- and she/he didn't provide:
a) wait dialog with "cancel" button
b) configurable auto-skip (let's say "if page turn takes more than 5 minutes, give up and show sorry page")

(a) and (b) are pretty common sense countermeasures when you know that run time can hit user badly.

01-27-2010, 11:12 AM	#61
badbob001 Fanatic Posts: 556 Karma: 1102020 Join Date: Sep 2009 Device: Kindle Keyboard (rip), Kindle Voyage, Fire Tablet 10 '17, iPad '19	I'm guessing screen refresh involves these three steps: Time to process command (ie: register click, open document, access requested content, etc... all limited by cpu, ram, and storage bandwidths) Time to pass complete screen image to e-ink controller (in irex's case, the delta controller) and for it to process. Time for e-ink paper to update When you open a document on a traditional display, steps 1, 2, and 3 can start in parallel: as the document is being loaded, the screen can start showing the application interface and when more of the document is loaded, more of the document can be displayed. It's like viewing a web page where things load in chunks, giving you a sense of progress. On an e-ink device, the above steps usually need to occur in serial (one before the other). For a full page refresh, such as with a page turn or when opening a document, the e-ink controller needs to have all the screen elements in place before it can finally tell the e-paper what to display. There is capability for partial refresh, but that involves changing small parts of the same page and the original image can be degraded in the process. I guess one can think of e-ink like printer paper... you print to it when you are ready. I suppose if device makers think components that are 'fast enough' for traditional display devices is good with an e-paper display swapped in may need to rethink things. If they can get steps 1) and 2) to be really really fast, then the physical slowness of step 3) will be less noticeable. Last edited by badbob001; 01-27-2010 at 11:36 AM.

01-27-2010, 11:15 AM	#62
plisken Enthusiast Posts: 25 Karma: 10 Join Date: Jun 2009 Location: Poland Device: iRex DR1000S	@Shaggy, not I am not comparing those speeds. I am responding to your statement, that every eradear has the same page turn time because of the e-ink delay. E-ink delay can be easily measured (I mean, top of it) by measuring response time of "please wait". Or IOW -- the long time of page turn on DR1000 is not caused solely by e-ink refresh time. > Still not entirely sure what you're saying. You mean that it took the DR half > an hour to turn a page? Yes, or saying more exactly -- it took more than half an hour. How much I don't know (to be fair, for 2.8GHz Core2Duo computer the page turn of such pdfs is not an easy task as well, however the problem is computer is responsive while page turning, while DR1000 is not, and what worse there is no cancel button). DR1000 didn't freeze, if it did, it wouldn't shutdown. @badbob001 You forgot step after (1) -- rendering the page. Last edited by plisken; 01-27-2010 at 11:17 AM.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
When I get a little money I buy books; and if any is left I buy food and clothes.	brecklundin	News	2	10-10-2009 02:21 PM
Sell Buy My Diamond Ring So I Can Buy A Kindle!	DaisyMama	Flea Market	0	03-03-2009 04:14 PM
Iliad or BeBook - to buy or not to buy, that is the question...	orwell2k	Which one should I buy?	11	10-19-2008 08:09 PM
Iliad or BeBook - to buy or not to buy, that is the question...	orwell2k	iRex	2	10-08-2008 11:03 AM

01-27-2010, 12:03 PM	#65
plisken Enthusiast Posts: 25 Karma: 10 Join Date: Jun 2009 Location: Poland Device: iRex DR1000S	@Shaggy > The eInk refresh time is generally the largest component of the overall page > turning time. Sorry, but I cannot agree -- taking the fastest page turn as the bottom time (i.e. nothing is better) I would be just happy to have DR1000 and I would tell everyone how cool the device is. But on average DR1000 thinks way longer, and it is obviously not e-ink refresh time. > If that document is stressing a 2.8GHz PC, then it's not much of a surprise > that it's causing a problem on a much smaller device. I am not surprised by that. I am surprises that someone didn't anticipate that fact -- and she/he didn't provide: a) wait dialog with "cancel" button b) configurable auto-skip (let's say "if page turn takes more than 5 minutes, give up and show sorry page") (a) and (b) are pretty common sense countermeasures when you know that run time can hit user badly.

Advert

Advert