Atlantis-Created Kindle File 3 times size of MS Word to KDP

[Robert]

If you go to http://grin.hq.nasa.gov/ABSTRACTS/GPN-2000-000642.html, you can download the picture in 4 different sizes. The “small” picture is already big enough for an eBook. It weighs 367,937 Kb.
I downloaded and opened that “small” JPG picture in Paint.NET. I saved it as an 8 bit PNG picture (see attached “GPN-2000-000642.png”). “GPN-2000-000642.png” only weighs 147,893 Kb. This is about 2.5 lighter than the original “small” picture.

[rstroud]

Well, I guess we are all playing around with this image.

Probably the best thing is for C.J. to use PhotoShop Elements (PSE), since she feels comfortable with it, and apparently the program makes what it does somewhat invisible to the user.

In fact, I wonder if PSE is converting the resolution of that picture from 300dpi to 96 or even 72dpi as part of its "Save for Web" process.

I say this because I downloaded the 7.1M picture and worked with it. I used the FREE IrfanView (and its optional free plugins). I cropped the picture to make it 3:4. I was never able to get a file size even close to what C.J. reported until I reduced the resolution of the original from 300dpi to 96dpi (the Windows standard).

I reduced the resolution to 96dpi. I also reduced the physical size to about 34% of what it had been, which in inches would be a reduced size of about 8x11. (I really do not know how large the picture is supposed to be.)

Then, finally, I used the "Save for Web" option in IrfanView. This gave a file about 45K or so in size, but it also reduced the resolution to 72dpi, and I wanted 96dpi. I changed the resolution back to 96dpi, and got a file that was 89.8K. It looks as good as the others here.

My point is that the "Save as Web" feature in PSE may not be quite so magical as one might think. It probably just reduces the resolution by a lot.

PSE is a good program. I had version 2 years ago. However, I read somewhere several months ago that the latest version (11) basically acts like spyware in some way or another. That sounds weird, and I honestly do not know if this is true. However, I basically dislike Adobe because of its ungodly high prices. Most people like earlier versions of PSE. I do not have the program on my main computer, though my laptop has version 10 on it.

Of course, it is really important to see how an image with a reduced resolution looks on an e-reader, whether Kindle or something else.

[cjseasyaspie]

Roland,

I've been doing pretty much the same thing you described... trying many different things, trying to find the best way to create the best picture files for the Kindle.

I would like to find free software for my readers if I could find one that does just as good a job as Adobe Photoshop Elements does, but if one exists, I have not found it.

The most important thing... as you suggest... is the actual outcome... what the picture looks like in the finished book... rather than the specs, so I don't worry much about the specs.

The picture I uploaded here -- which, as you say, is very lightweight -- fills the screen when zoomed, and is stunning when viewed on the very high resolution screen of my new iPad 4.

I don't think it can get much better than that.

[rstroud]

C.J.,

Okay, let me try to post this picture. If it works, and you are interested, I will show you how to do this with the free IrfanView and the free plugins. It only takes about five easy steps.

But I will have to wait till tomorrow to explain how to do this (it is about 3 A.M. now). Let me know if you are interested. The picture is the same size in pixels as yours, and the file size is about 90K. It could be a hair sharper if you let the file size go just a little larger. This was an enormous picture. You can adjust the contrast of the picture with IrfanView.

Best ...

[Robert]

A few general remarks.

First, image “quality” is an extremely highly subjective notion. Here is from http://www.tech-faq.com/picture-quality.html:

Picture Quality

Picture quality is a general term used by the viewer to rate the inherent quality of the image. It is mainly subjective, meaning that each viewer rates the picture quality differently. Some people may think that an image is high enough quality, but you may have an individual with higher standards that may say the image lacks quality or sharpness. Picture quality is used in a general sense.

With digital images, many people are able to rate and categorize images based on bits of information or pixels that create an image. For instance, if you had two digital images, one being 1 mega pixels (1 million pixels) and another image having 5 mega pixels of information in the image, most people would state that the image with more mega pixels has the higher picture quality. And this would be true in lots of circumstances. However, if the 5 mega pixel image was taken in complete darkness and the 1 mega pixel image was taken using regular light, most people would agree that although an image may have more mega pixels, that doesn't necessarily mean that it has a better picture quality.

Picture quality for the most part is determined by resolution and a balance of colors which is usually termed contrast. Resolution refers to the amount of detail in a picture. Usually the more detail in a picture, the higher the resolution. For instance, a 5MP (mega pixel) image has more resolution than a 1MP image.

Besides an image having more bits of information, the sharpness or colors in a picture should also be vivid. Meaning that the viewer should be able to see an image clearly and differentiate the numerous colors from one another in the image. You don't want to look at an image that is muddy, cloudy or out of focus. You want to see clearly defined objects with sharp colors. Contrast is important to this process.

Contrast or sometimes called contrast ratio, it is the specific amount of brightness or luminance between two colors, one of these colors is always the color black, the other colors can vary, usually green, blue or red. Contrast helps a viewer perceive picture quality. If an image has a high contrast ratio, the colors will be sharp and vivid. If the image has a low contrast ration, the image will probably be muddy or out of focus.

For most people the quality of picture quality can easily be improved or reduced when taking digital images or when it is being shown on a HDTV set. Digital images can be easily augmented to add picture quality. For instance, with digital images, you can usually set the resolution level and even add the amount of colors that an image can have. Some digital cameras or HDTV sets can produce millions upon millions of different colors, with the end result of increasing the overall picture quality.

Now even though the notion is often highly subjective, image “quality” can be assessed more objectively from different factors, such as noise, tone reproduction, contrast, color accuracy, distortion, etc. You might want to take a look at this Wikipedia page: Image quality factors.

Finally, and most importantly to me, the pictures you uploaded are 4:3 pictures. This is OK. But they are also 1200x1600 pixel-wide. This is 16.667x22.222 inches. How can we expect such pictures to display in an iPad with a 9.50 inches screen, or a Kindle with a 6 or 9.7 inches screen? Obviously, the built-in software will reduce the picture size one way or another. It seems to me that it would be much preferable to choose original picture sizes that will naturally fit within the target screens without special adjustments. Here is from eBook Formatting Tip: Resize Your Images:

If you want your eBook to display content in an attractive way across all the eReaders, there are a few simple tips you’ll want to follow before your file is converted into the ePUB format.

Today’s tip: Resize your images outside of the Word doc before you insert them. Many people drag an image from their desktop into the Word document FIRST and then adjust the size (using the little square scrollers that appear in the corners when you click on the image). But here’s the catch: adjusting the size of the image in Word only changes how it APPEARS. As far as the document is concerned, the image is still the original size. So if you dragged a giant .jpg into Word and then adjusted the size, when the Word doc is converted to ePUB and displayed on an iPhone or Nook, it will display in its original size and cause all kinds of annoying formatting issues for your reading device.

To avoid this formatting issue, adjust the size of the image first! Then move it into the Word doc. Once you’ve moved the resized image into Word, it cannot be enlarged beyond the actual size. You can make it appear smaller, but no bigger. This allows you to get a feel for how the image at its actual size will display in relation to the text.

Here’s a helpful hint from our ePUB conversion guidelines:

Resize large images to 300 pixels high if you would like them to display in-line with text. Do all image resizing outside of the document, then reinsert them before saving. All images should be 72 – 300 DPI (300 DPI preferred).

Cover and full-page images: 590 pixels wide by 750 pixels high.
Logos or simple images: 75 – 100 pixels high

You might also want to take a look at Images in iPad eBooks, and Anatomy of an iBooks page.
Here is from the first page:

It probably makes more sense to decide on the size of an image by taking into account the full area possible. As I mentioned in Anatomy of an iBooks page, the usable area on a (vertical) iBooks page is about 560 pixels x 760 pixels. So, an image that measures 288 pixels wide will take up about half a page.

And from the second:

Finally, I’ve decided that the maximum width of an illustration with wrapped text (until we can apply dynamic sizes), should be between 200 and 250 pixels, so that there’s enough space for text around the edges.

If you’re adding a full page illustration, make sure it's no bigger than 600 x 860 pixels, or else it will be divided between pages (ew!)

HTH.
Cheers,
Robert

[cjseasyaspie]

Robert and Roland,

Thank you both for your input.

One thing I'm beginning to understand... or at least, recognize... is that there is plenty of reason for all the conflicting information "out there".

Perhaps it is the merging of print and digital... which are different... and also the same.

An Internet search for "what is the difference between PPI and DPI" gets many hits... from:

" there is no difference... to...

"there absolutely IS a difference!"... to...

"neither matter".

The only thing I believe for sure after that search is "many people use the terms PPI and DPI interchangeably".

Roland, you mentioned that Apple wants covers to be at 300 (DPI or PPI, I don't remember which you said) resolution.

That puzzled me because Amazon says "save your file at 72 DPI".

Then, I saw my (saved for web) Saturn image described this way:

72 screen resolution

300 print resolution

Just like that... one under the other, on the same panel.

I recorded the numbers on paper... so I know I'm not mis-remembering them... but I didn't record the source.

I'm going to see if I can find it and report back.

[Robert]

Here is from What is the difference between DPI (dots per inch) and PPI (pixels per inch)?:

Dots per inch is fairly easy. A screen has so many dots (each comprising R, G and B elements) per inch of screen. It's the same on paper. A 1200 dpi printer can lay down 1200 dots in one linear inch. In describing screen detail or printer output, dots per inch is the correct term.

PPI is where the confusion comes in. An image has so many pixels. Its metadata contains an output size in inches, cm, mm, M&Ms, whatever. It's the width in pixels divided by the output width in the metadata that "per inch" comes from. So the same image with different metadata may be 72 ppi, 150 ppi or 8000 ppi. The image information is the same; all that's changed is the metadata.

A quick and easy demo that somewhat illustrates the point is to make some marks on a piece of elastic, say five to an inch. Stretch the elastic to twice its length. The number of marks hasn't changed, even though the "marks per inch" is now 2.5.

You can see this in Photoshop if you turn off Resample Image and change the size. The ppi value changes to reflect how small the pixels must be reproduced in order to hit the measurement value in inches/cm/mm etc. Note that in this case the Pixels fields are disabled. You can't change those values unless you resample.

Mass confusion entered in when image pixels were mapped to screen dots in web browsers. A 200 pixel image shows up as 200 pixels in a browser. How large it is, measured with a ruler, depends on the dots per inch of the screen. The image metadata might say it's 200 ppi or 72 ppi or 1 ppi, it will still occupy exactly 200 screen dots. The world remains fixated on "72 ppi for the web," so the question of "what's the right resolution for web images" keeps coming up, and the correct answer, "it doesn't matter," keeps being supplied ad nauseam.

If you're still with me, there's one last step that brings the two together.

A 720-pixels-wide image at 10 physical inches wide has a resolution of 72 pixels per inch. If you print it on a 1200 dpi printer, there will be 1200 dots per inch on the paper, but the image is still 72 pixels per inch. That's why it looks like crap. On the other hand, a 7200 pixels wide image printed at 1 inch wide will exceed the resolution of our 1200 dpi printer. Photoshop (let's say) and the printer driver decide which pixels to throw away and which to actually print. Some of the printed dots will be averaged among adjacent image pixels, but, regardless, some of the image information has to be thrown away. The output will be 1200 dpi, but the resolution of the printed image will have been reduced to at most 1200 dpi by the software.

CJ, please, read carefully the last paragraph above. It explains why the DPI and PPI figures can differ widely for the same picture. You might also want to take a look at the following pages:

Difference Between DPI and PPI
DPI and PPI
PPI vs. DPI: what’s the difference?
What is the difference between DPI and PPI?

HTH.
Cheers,
Robert

[cjseasyaspie]

Thanks, Robert... some really good stuff there... and also your previous post about image quality.

I will dig in now and study a bit.

Meanwhile... to answer a question from one of your earlier posts:

Finally, and most importantly to me, the pictures you uploaded are 4:3 pictures. This is OK. But they are also 1200x1600 pixel-wide. This is 16.667x22.222 inches. How can we expect such pictures to display in an iPad with a 9.50 inches screen, or a Kindle with a 6 or 9.7 inches screen? Obviously, the built-in software will reduce the picture size one way or another. It seems to me that it would be much preferable to choose original picture sizes that will naturally fit within the target screens without special adjustments. Here is from eBook Formatting Tip: Resize Your Images:

I'm sure that would be best where possible... especially considering all the many different screen sizes on the market.

But, I'm speaking only of Kindle and Kindle applications for other devices, such as the iPad.

The Kindle software formats the screen according to the aspect ratio.

That's the reason the aspect ratio of my test file... 3:4... works for the "portrait" screens of the older Kindles... and, turned around, the same aspect ratio... 4:3... works for the "landscape" screen of the much larger iPad 4.

[Robert]

The Kindle software formats the screen according to the aspect ratio.

CJ, as I understand things, the Kindle software formats the display. The screen format itself and the associated PPI are hard-coded in the hardware. If your pictures are too big for the target screen, there is bound to be some form of adjustment, even if they are 4:3 pictures. Here is from E-ink Introduces Mobius, a New Flexible Display Technology:

Sony’s writing slate has a display resolution of 1200×1600, or about 150 ppi (pixels per inch). That is a higher pixel density than the screen on the laptop I am using at the moment, but it is also far less dense than the crispness that we have come to expect on mobile devices – even E-ink devices.
The Kindle Paperwhite, for example, comes in at 212 ppi, while the Kobo Aura sets the new standard at 300 ppi. And even the older ereaders like the Nook Glow have a screen with 166 ppi.
I think that most of the leading consumer gadget makers will look at the Mobius tech and pass on using it on smaller screens because they will choose higher resolution over ruggedness. But the tech might show up on rugged devices like sports, construction, and military equipment – markets where durability is valued more than screen resolution.

So PPI is hardware-defined, and is about “pixel density”. Here is from Computer displays:

The PPI of a computer display is related to the size of the display in inches and the total number of pixels in the horizontal and vertical directions. This measurement is often referred to as dots per inch, though that measurement more accurately refers to the resolution of a computer printer.
For example, a 15 inch (38 cm) display whose dimensions work out to 12 inches (30.48 cm) wide by 9 inches (22.86 cm) high, capable of a maximum 1024×768 (or XGA) pixel resolution, can display around 85 PPI in both the horizontal and vertical directions. This figure is determined by dividing the width (or height) of the display area in pixels by the width (or height) of the display area in inches. It is possible for a display’s horizontal and vertical PPI measurements to be different (e.g., a typical 4:3 ratio CRT monitor showing a 1280×1024 mode computer display at maximum size, which is a 5:4 ratio, not quite the same as 4:3). The apparent PPI of a monitor depends upon the screen resolution (that is, the number of pixels) and the size of the screen in use; a monitor in 800×600 mode has a lower PPI than does the same monitor in a 1024×768 or 1280×960 mode.
The dot pitch of a computer display determines the absolute limit of possible pixel density. Typical circa-2000 cathode ray tube or LCD computer displays range from 67 to 130 PPI, though desktop monitors have exceeded 200 PPI and contemporary small-screen mobile devices often exceed 300 PPI.

Also you might want to take a look at eReader and Print Book Formatting Aspect Ratios. Note that 4:3 is the traditional aspect ratio for print books. This is why it is the preferred ratio for a lot of eReaders.

[DaleDe]

While all of the redoing of images is interesting in this thread it is not the solution to the problem the OP needed. The fact is the KindleGen program is actually designed to create a file three times the size of the ePub.

The latest versions of KindleGen place three files inside the one file. The three files are:
1. The epub source file.
2. A mobi files in original mobi format
3. A KF8 (Kindle format version 8) file for the newer devices and latest firmware on some older devices.

Amazon is doing this be design. The uploaded file to Amazon containing all of these files will be divided up by Amazon for the end user but knowing which device the user owns they will deliver the right version. To understand this check http://wiki.mobileread.com/wiki/KindleUnpack which is the description of a program that can be used to take one of these files apart. There is also a forum on mobileread that can be used to discuss this.

When submitting to Amazon KindleGen by itself is the best choice but for your own use you may want to take the file apart to achieve the smallest file size for your device. The source file can always be removed leaving a file only twice the size of the original.

Dale

[DaleDe]

While all of the redoing of images is interesting in this thread it is not the solution to the problem the OP needed. The fact is the KindleGen program is actually designed to create a file three times the size of the ePub.

The latest versions of KindleGen place three files inside the one file. The three files are:
1. The epub source file.
2. A mobi files in original mobi format
3. A KF8 (Kindle format version file for the newer devices and latest firmware on some older devices.

Amazon is doing this by design. The design includes this in a transparent way so that reading devices will not even see the hidden files but will pick the version they know how to read. The uploaded file sent to Amazon containing all of these files will be divided up by Amazon for the end user by knowing which device the user owns they will deliver the right version. (The source is always removed by Amazon.) To understand this check http://wiki.mobileread.com/wiki/KindleUnpack which is the description of a program that can be used to take one of these files apart. There is also a forum on mobileread that can be used to discuss this.

When submitting to Amazon KindleGen by itself is the best choice but for your own use you may want to take the file apart to achieve the smallest file size for your device. The source file can always be removed leaving a file only twice the size of the original.

Dale