Kindle DX 2 screen contrast improvement analysis

The improved Kindle DX 2 screen contrast is its big selling point. Amazon hasn’t really revealed too much about what led to this improved screen contrast.

Well, let’s take a closer look and see what factors might have helped improve the Kindle DX 2’s screen contrast.

Kindle DX 2 Screen Contrast Improvement – possible Factors

Here are the 3 factors that we can identify right away –

  1. Actual hardware improvements. eInk say the Pearl screen has 10:1 contrast as compared to the 7:1 contrast in previous versions. Quite frankly it doesn’t really seem like the hardware improvement by itself is 43%. In terms of measurements we have this from Bruce Wilson’s comment at Teleread

    From density numbers alone – white is a little whiter, black is a lot blacker.

    Old White Kindle DX 1 (6 months old):
    white area density = 0.46, Lab = (65.8, -2.3, 0.6)
    black area density = 1.30, Lab = (26.6, -1.0, -2.2)

    New Graphite Kindle DX 2:
    white area density = 0.42, Lab = (68.2, -2.4, 0.9)
    black area density = 1.58, Lab = (18.5, -0.1, -3.6)

    I used a Datacolor Spectrocolorimeter model 1005. “Lab” is a color space measurement like RGB, only for print.

  2. Graphite Casing. Amazon have implied this is not factored into the 50% better contrast – However, it’s clear after playing around with the Kindle DX 2 that the graphite casing has quite an important role in making the screen look better.  
  3. Speckling on the Screen. There are very tiny speckles on the screen of the Kindle DX 2 when you zoom in. Click on the last photo on the Kindle DX 2 Photo page to see this speckling. When photos have noise like this added to them it improves their contrast – It’s hard to believe there could be any other reason speckling would be added to the screen of the Kindle DX 2.

We also have two additional possibilities –  

  1. Software improvements. Kindle software upgrades have improved Kindle screen contrast in the past by making the text bolder and it’s possible that Kindle DX 2 comes with some software improvements. Kindle DX 2 comes with firmware version 2.5.5 and it makes you wonder if that firmware version includes screen contrast tweaks.
  2. Additional changes in the screen hardware. There’s a very interesting mention in the official Kindle forum that the Kindle DX 2 screen is noticeably whiter if you tilt it a little rather than look at it straight on. For my Kindle DX 2 this is true – It’s noticeably whiter when tilted a little. Is this by design? Is this a byproduct of the new screen technology?  

The former is very, very likely while we understand too little about the latter to factor it in.

Breaking down the supposed 50% screen contrast improvement

After shooting a lot of Kindle DX 2 videos and taking a lot of photos and comparing screens in all sorts of lighting conditions it seems to me –

  1. Compared to Kindle 2 Global – Kindle DX 2 screen is 25% to 30% better normally, 30% better in sunlight, and 30% better when Kindle DX 2 and Kindle 2 are both tilted a bit. 
  2. Compared to Kindle DX 1 – Kindle DX 2 screen is 40% better normally, 45% better in sunlight or when both are tilted a bit.

There isn’t really a 50% improvement in screen contrast. It’s 40% to 45% when compared with Kindle DX 2 and 25% to 30% when compared with Kindle 2 Global. 

Furthermore it seems that this 40% improvement is broken down into –

  1. Half due to hardware improvements. If the spectrocolorimeter readings are correct hardware improvements might be responsible for as much as three-quarters of the improvement.
  2. A quarter due to the graphite casing. 
  3. A quarter due to the speckling.

The Kindle 2 Global screen is much closer to the Kindle DX 2 ‘better hardware screen’ than the Kindle DX 1 screen. This might be due to software tweaks or International Kindle 2s getting better screens or perhaps my Kindle 2 global was an exceptionally good version.

How did 50% screen contrast improvement and a graphite case and speckling and possible software improvements add up to 40%?

Well, it seems that eInk messed up and Amazon did as much as they could to make up for it.

Seriously – Look at the videos and photos. If you happen to have any of the earlier Kindles and the Kindle DX 2 compare them in various lighting conditions. If eInk’s claim is valid and there’s a 50% screen contrast improvement then it means that the graphite casing and the speckling and the software improvements (if any) contributed minus 10%.

The far more likely case is that eInk did a terrible job with their screens and improved just 20%. Then Amazon did a lot of brainstorming and came up with the graphite case and the speckling design for the screen and software improvements to get to 40%.

Amazon better hope Pixel Qi or Qualcomm Mirasol deliver color eInk screens soon because Amazon can’t keep compensating for eInk’s inadequacies with software upgrades and smart design decisions. The new Kindle DX 2 has managed to use almost every design and software trick possible to improve screen contrast (we’re including font sharpness improvements in the Kindle 2.5 upgrade). It’s had to because the actual screen technology from eInk isn’t improving fast enough.

Quick Summary 

Yes, Kindle DX 2 has a noticeably better screen. No, eInk isn’t responsible for all of the improvement. If eInk really would have improved their eInk screens 50% we would be looking at 70% to 75% better screen contrast on the Kindle DX 2.

Is the quality of an eReader reflected in every part of it?

The Contrast Blog has a very interesting post claiming that the thickness of a napkin reflects the quality of the restaurant.

 A friend in the restaurant business told me about a survey that showed a massive correlation between category of napkin and customer satisfaction. 

The napkin represents a degree of care, preparation and devotion that goes above and beyond asking if they want fries with that.

The post uses the phrase – Quality is Fractal.

My understanding of it is that they’re saying that when a product is very good there is a level of committment to excellence and quality that shows up in every aspect of the product.

The overall post and idea look quite impressive.

The problem is they are just an attempt to write a fancy story

The Contrast Blog post is rather imprecise, as opposed to being wrong, for quite a few reasons. 

It uses the wrong example and the wrong context to make its point –  

  1. A survey that a ‘friend’ told the author about is as undependable a source of data as you could find.
  2. It doesn’t address causation – We don’t know that napkin had anything to do with customer satisfaction. Perhaps most restaurants with doors of metal have really high customer satisfaction – Does that mean quality is metallic?  

It assumes there is time and incentive to polish every aspect of a product –

  1. Any good product always prioritizes things that are most important to customers over things that are less important to customers.  
  2. If you were to study a part that is not very important to customers you might find a lack of polish – However, we don’t know whether that means there is a lack of polish in general or whether it means that the effort and polish was put into things that are more important.

Any product has limited development time. You have to focus on the most important features and aspects. That necessitates that the not so important features can’t be perfect.

The post assumes that beautiful appearance means beautiful function –  

  1. There’s a difference between a product that does its work well and a product that gives the appearance it does its work well (whether it does or not). 
  2. The post makes this implicit assumption that the two (looking good, being good) are interlinked – see the photos and the examples and it’s clear.

The post is the perfect example of the mistake it’s making

The Contrast Blog post has pretty pictures and a brilliant sounding idea and a sexy catchphrase.

All that’s missing is the hard research to back up their hypothesis.

They are focused so much on making their hypothesis look accurate and pretty they forget to actually prove it.

Quality being Fractal – It’s likely but not a given

Let’s say we look at one aspect of a product and it’s not very good.

  • The first test is whether it’s an essential aspect – If it’s not then it means that it doesn’t reflect on the eReader.

You could look at the Kindle’s tiny keyboard keys and the Nook’s difficult to open case and if quality really were fractal then that would mean both are terrible eReaders. But they’re not – they’re both excellent.

  • The second test is whether every core aspect of an excellent product is excellent – It’s not.

Let’s take the iPhone since the author cites the example of Apple for excellence – the quality of the calls is far from excellent. Does that make it a terrible phone – No. It’s still excellent because it has excellent strengths that override its weaknesses.

The most accurate statements we can make are –

If one small part of a device is of excellent quality then it makes it likelier that the device will be of excellent quality than of poor quality.

An excellent product’s core parts are usually of excellent quality.

Quality is fractal makes the mistake of posing as a tautology. It’s not.

Why coin ‘Quality is Fractal’?

Basically – the desire to find a great short-cut.

Quality is Fractal sounds beautiful – there are mathematical undertones and a sense of universality and it encourages us to make assumptions and not dig in and do hard research and make hard decisions.

However, it’s a shortcut that’s dangerously capable of steering us in the wrong direction – especially since the easiest things to notice usually have nothing to do with the quality of the device.

Take an eReader. Your first experiences will be –

  1. The Ads you see.
  2. The visual design of the eReader itself.
  3. The store and website sales-copy and images.
  4. The packaging.
  5. Holding it and seeing it in your hands.
  6. Turning it on for the first time. 
  7. Opening a book for the first time.

If a company were to channel all their energy into these things what would suffer?

Well, the things that you do most often on the eReader –

  1. Read – which depends on screen quality and such.
  2. Turn Pages.
  3. Find Books and Buy Books.
  4. Search for information and do reference look-ups.
  5. The usability in general.

If we focus on making a device look like it’s a great reader and making non-essential features excellent because Quality should be Fractal we run the risk of running out of time to make it a great eReader for reading books.   

eReaders and Design Trade-offs

eReaders actually suffer from the sort of ‘Quality is Fractal’ mind-set that looks oriented products propagate.

A product can go in one of three directions –

  1. Utter focus on doing what it’s supposed to very well.
  2. Utter focus on looking like it does what it’s supposed to very well.
  3. Utter focus on both doing what it’s supposed to very well and looking like it does what it’s supposed to very well.

It’s almost impossible for eReaders to pull off 3.


  1. It’s a new area.
  2. Economies of Scale and Profits have not yet been established.
  3. Technology is still evolving.
  4. User feedback is still being collected and acted upon.
  5. Lots of other reasons.

Where does that leave eReaders?

In a very interesting position –

  1. eReaders are remarkably good at doing their core function i.e. reading ebooks. That’s why 93% of owners are happy (very satisfied or somewhat satisfied).
  2. eReaders are at the same time remarkably bad at looking like they’re the hot new happening gadget.

Which leads to a remarkable paradox –

  1. eReaders are becoming very popular – because they are great for reading.  
  2. eReaders are confusing lots of people – because they don’t LOOK like what people imagine a successful product ought to look like.

eReaders capture the experience of reading a book very well while failing almost completely to look impressive.