Kindle Optimizer‎ > ‎

Kindle-1 vs. Kindle-2

There has been much discussion of the Kindle-2 being different than the Kindle 1 because of variations in how the fonts are rendered, with anti-aliasing, or font smoothing taking the blame for the problems.

I have recently been provided with some screen captures comparing the digital data from the K1 and the K2 when rendered in the smallest text, size 1. These samples are shown below

Kindle 1
Kindle 2
(Digital data courtesy "DD" from KindleBoards)

These images represent the screen data, not the screen as it appears to the eye. As we shall see, this is an important distinction.

This simple comparison proves that the fonts are identical between the K1 and the K2. Every aspect of the characters is the same. They have the same size and the same anti-aliasing (font smoothing). They are, pixel-for-pixel, the same. The only differences lies in the word spacing, but this is inconsequential to the clarity of the fonts.

So why have there been so many complaints from users of the Kindle-1 who upgraded to the Kindle-2?

The only remaining difference is how that digital data shows on the screen. Just as you can look at the same DVD player on a number of different televisions and get different results, you can put the same digital data on a variety of Kindle 2 screens and get different results.

Some have misinterpreted my findings as a claim that there is no difference between the K1 and K2 - this is not the case! Obviously, people have noted a difference in what they see on the screen, there is no denying this fact. What I have done is definitively ruled out the speculation that the difference between the K1 and K2 is the result of "greater number of gray shades" or "font smoothing".

What are the real differences between the K1 and K2?

While I do not have a K1 to experiment with and measure, I do have two K2's. As shown on this page, there are differences between Kindle 2's. Some render pixels sharply, and with high contrast, while others do so less well. This explains (partly) why there is such a wide variation in user experience between people using the same device. Some people have no problem with their K2 while others experience very low contrast text and significant ghosting (afterimages when switching between pages).

Based upon user feedback from owners who have both Kindle-1's and Kindle-2's, the displayed contrast varies considerably. Andrys, from the Kindle World Blog, has taken pictures showing that her K2 is clearer than her K1. Others have bemoaned the lack of clarity of the K2's screen compared to the K1.
Andrys was also kind enough to take some comparison pictures of her two units and the results were intriguing. While not definitive, from the appearance of the text on the K1, it looks like the K1 may render some of the gray pixels in the font as black! This would obviously make the fonts look much darker. For example, in the 'a' below, the dark gray pixels looked black. This broadens the font and makes them easier to read. It does make them slightly "chunkier" looking, and some users prefer the finer strokes of the K2 screen.

The fact that some K2's are better than some K1's while others are worse points strongly to the differences being unit specific. However, it does appear that a greater number of K2's are less clear than the K1's.

As shown above, the underlying data being sent to the screens are identical. As such, the only differences that remain lie in the Kindle 2's E-Ink display and the hardware/software driving it. These differences result in a lower contrast display on some Kindle-2's.

Kindle 2- Unit-A
Kindle 2 - Unit-B

Referring again to the above images comparing microscopic images taken of two Kindle 2 units, we see considerable variation in the contrast of the displayed characters. They should be identical but they are not. The  Unit-A has pixels that are less clearly defined than those of Unit-B. This translates into a "fuzzier", lower contrast when viewed at normal size. Keep in mind that Unit-A, my original Kindle 2, subjectively looks "pretty good" under visual inspection. But it is noticably lower contrast than my second kindle 2. Now imagine that you are a user who received a third unit with even poorer pixel definition. The screen would look very washed out.

Conclusion: Large variations exist in the Kindle-2's screen contrast

Whatever Amazon / E-Ink changed in order to obtain 16-gray levels for the new Kindle-2 screen has resulted in a larger variation in pixel sharpness than was experienced with the Kindle-1's. This should have been caught by quality control. Had all Kindle-2's been shipped with as sharply defined pixels as Unit-B in the photo above, there would have been considerably fewer complaint. The "sun fading" issue is another issue that should have been caught by quality control. My conclusion is that the quality control efforts at E-Ink and Amazon have been deficient.


I believe that the image ghosting problem that exists on some units is directly related to these issues. My Unit-B shows barely visible ghosting, while Unit-A shows considerable ghosting. This is logical. Ghosting, by definition, represents incomplete screen refresh. That is, the Kindle tells the screen that it wants a "white" pixel, and what shows is a "gray" pixel. This also shows why those people can improve their display by pressing Alt-G. Each press of Alt-G makes more of the E-Ink pixels closer to their intended gray level. On good units, all of the pixels immediately get to their intended value while on poor units, the pixels are more "sluggish". Apparently, the Kindle-1's display was more consistent in its ability to render pixels at their intended gray value. This may have to do with the changes in the technology required to give the units a 16-gray level display. Or it may be totally unrelated.

Final Conclusions

Based on the analyzed pixel data, there is no difference between the underlying pixel rendering algorithms on the Kindle-1 and Kindle-2. However, there are differences between how the digital pixel values are placed on the screen OR the quality control for the Kindle-1 was much better than for the Kindle-2.

Important note:

Do not misinterpret these conclusions as indicating that there are no differences between the K1 and K2! Users have clearly seen real differences between their units. However, these differences are not due to changes to the font smoothing algorithms. They are due to how the pixel data get onto the screen.

The remaining need for alternative fonts

in spite of these findings, it is clear that many users find the native Kindle fonts to be "too light" for their tastes. Replacement of the native fonts with bolder fonts has proven to positively improve the reading experience for a number of readers (myself included) whether the unit has an excellent screen or not. Based on user feedback, even users with marginal screens find the Kindle display to be "a pleasure to use" with the bold fonts. For Amazon, it makes good business sense to give users a selection of reading fonts and the ability to add their own fonts so that they can tailor the units to their own reading preferences. Taking this basic step would remove the primary customer complaint that remains with this groundbreaking reading device.

 Discussion of the background of the K2 being darker than that of the K1 - Discussion thread on Kindleboards, with pictures