Seems like I moved from

to

When I got the Canon PowerShot G10, this all changed. Grabbing a shot is now as easy as taking the camera out of my jacket pocket and taking the shot. So here is one (less dramatic) example of what I used to miss out on.

Good morning Winnipeg, G10 @ ISO80

Change, G10 @ ISO100

Abstract, G10 @ ISO80

Twin peaks, G10 @ ISO80

Last to go, G10 @ ISO 80

I am cold, G10 @ ISO80

Caught by surprise, Canon PowerShot G10 @ ISO80

The comparison is between my iPhone 3G 2MP camera phone, Olympus 850SW P&S, Canon PowerShot G10 14.7MP, Nikon D80 10.2MP DSLR and the Canon 1Ds Mark III 21.1MP DSLR. ISO was at base ISO for most shots - except for the Canon and Nikon which were at ISO200. Aperture was set to the lenses' optimum values. No tripods were used - all shots were handheld. This was done to show how each camera performed under typical scenarios where you might not have time to set up on a tripod.

Conclusions

1. There is a lifetime of difference between the 2MP iPhone and all the other cameras - indicating to me that
   1. 2MP is not enough
   2. The iPhone's camera sucks
   3. The lens on the iPhone gets dirty very easily since it is not protected - and nobody in their right mind will clean the lens before every shot so this reflects real world results.
2. There is a huge difference between the 8MP Olympus and the other more expensive cameras. The Olympus looks like a point and shoot - the others perform in a different league. But that being said, the results are pretty good considering 8MP has been crammed in to a sensor 1/2.25" in size.
3. The G10 is very close to the Nikon - it actually resolves more detail in some shots, but it cannot quite match the Nikon's dynamic range, which clearly shows how much of an advantage it is to have a 1.5x sensor vs. the 1/1.7" sensor of the G10. Still, the G10 is very close to the (much) more expensive Nikon, and I am confident that in prints up to A3 you would need to look long and hard to find significant differences.
4. The 1Ds Mark III is the best in all these tests, except maybe the last one of the fall coloured fir tree. There is a significant difference between the 1Ds3 and the Nikon and G10, however in my opinion not $7000 worth. The poor results in the last shot of the fir tree, is due to depth of field limitations and camera shake. The G10 with its built in IS allowed me to grab a perfectly sharp image without problem - the Canon could not do this and needed a tripod.

It seems like at Can$550 for the G10 one gets like in 90% of the image quality, for Can$1800 for the Nikon D80 one gets 94% there, and for a whopping Can$8500 one can get 100% of the baseline - the Canon 1Ds Mark III. All I can say - I am glad I have the 1Ds3 since I can do things the G10 can never dream of (like shooting macro at 5x life size or wide angle at 14mm or telephoto at 1400mm or action scenes at 5fps or low light shots at ISO 1600 almost noise free or... But still - the G10 is going to be a very valuable carry everywhere camera for those shots I used to miss out on cause the 1Ds3 is too big and bulky.

Add to this 3.75TiB of redundant (RAID1 and RAID5) local storage space...

Just.....

this time....

it is a little monster.

What would you do with an

8 Core 2.8GHz Xeon CPU with 1.6GHz 64-bit dual independent FSB

8GiB RAM

3.75TB Redundant Storage (RAID1 + RAID5)

NVIDIA GeForce 8800 GT with 512MB of GDDR3

30" 2560 x 1600 Cinema LCD

<math xmlns='http://www.w3.org/1998/Math/MathML' mathematica:form='TraditionalForm' xmlns:mathematica='http://www.wolfram.com/XML/'> <semantics> <mrow> <mrow> <msqrt> <mi> z </mi> </msqrt> <mo> &#10869; </mo> <mrow> <munderover> <mo> &#8721; </mo> <mrow> <mi> k </mi> <mo> = </mo> <mn> 0 </mn> </mrow> <mi> &#8734; </mi> </munderover> <mrow> <semantics> <msub> <mrow> <mo> ( </mo> <mrow> <mo> - </mo> <mfrac> <mn> 1 </mn> <mn> 2 </mn> </mfrac> </mrow> <mo> ) </mo> </mrow> <mi> k </mi> </msub> <annotation encoding='Mathematica'> TagBox[SubscriptBox[RowBox[List[&quot;(&quot;, RowBox[List[&quot;-&quot;, FractionBox[&quot;1&quot;, &quot;2&quot;]]], &quot;)&quot;]], &quot;k&quot;], Pochhammer] </annotation> </semantics> <mo> &#8290; </mo> <mfrac> <msup> <mrow> <mo> ( </mo> <mrow> <mn> 1 </mn> <mo> - </mo> <mi> z </mi> </mrow> <mo> ) </mo> </mrow> <mi> k </mi> </msup> <mrow> <mi> k </mi> <mo> ! </mo> </mrow> </mfrac> </mrow> </mrow> </mrow> <mo> /; </mo> <mrow> <mrow> <semantics> <mo> &#10072; </mo> <annotation encoding='Mathematica'> &quot;\[LeftBracketingBar]&quot; </annotation> </semantics> <mrow> <mi> z </mi> <mo> - </mo> <mn> 1 </mn> </mrow> <semantics> <mo> &#10072; </mo> <annotation encoding='Mathematica'> &quot;\[RightBracketingBar]&quot; </annotation> </semantics> </mrow> <mo> &lt; </mo> <mn> 1 </mn> </mrow> </mrow> <annotation-xml encoding='MathML-Content'> <apply> <ci> Condition </ci> <apply> <eq /> <apply> <power /> <ci> z </ci> <cn type='rational'> 1 <sep /> 2 </cn> </apply> <apply> <sum /> <bvar> <ci> k </ci> </bvar> <lowlimit> <cn type='integer'> 0 </cn> </lowlimit> <uplimit> <infinity /> </uplimit> <apply> <times /> <apply> <ci> Pochhammer </ci> <apply> <times /> <cn type='integer'> -1 </cn> <cn type='rational'> 1 <sep /> 2 </cn> </apply> <ci> k </ci> </apply> <apply> <times /> <apply> <power /> <apply> <plus /> <cn type='integer'> 1 </cn> <apply> <times /> <cn type='integer'> -1 </cn> <ci> z </ci> </apply> </apply> <ci> k </ci> </apply> <apply> <power /> <apply> <factorial /> <ci> k </ci> </apply> <cn type='integer'> -1 </cn> </apply> </apply> </apply> </apply> </apply> <apply> <lt /> <apply> <abs /> <apply> <plus /> <ci> z </ci> <cn type='integer'> -1 </cn> </apply> </apply> <cn type='integer'> 1 </cn> </apply> </apply> </annotation-xml> </semantics> </math>

It is awfully complex and verbose! This just reiterates what I held to be true for a long time now - XML is being abused. It is great for moving hierarchical data between disparate systems - even non-hierarchical data if you focus on the interoperability and compatibility it facilitates, however using it as RDBMS storage solutions or this kind of hyper verbose markup is down right a waste of space and time. When will people learn that efficiency is important?

Look at this well known equation (all from the Wikipedia entry):

In LaTeX it looks like this:

x = \frac{-b \pm \sqrt{b^2 - 4ac}}{2a}

Terse, huh?

Look at MathML:

<math xmlns="http://www.w3.org/1998/Math/MathML">
 <mi>x</mi>
 <mo>=</mo>
 <mfrac>
  <mrow>
   <mrow>
    <mo>-</mo>
    <mi>b</mi>
   </mrow>
   <mo>&PlusMinus;</mo>
   <msqrt>
    <msup>
     <mi>b</mi>
     <mn>2</mn>
    </msup>
    <mo>-</mo>
    <mrow>
     <mn>4</mn>
     <mo>&InvisibleTimes;</mo>
     <mi>a</mi>
     <mo>&InvisibleTimes;</mo>
     <mi>c</mi>
    </mrow>
   </msqrt>
  </mrow>
  <mrow>
   <mn>2</mn>
   <mo>&InvisibleTimes;</mo>
   <mi>a</mi>
  </mrow>
 </mfrac>
</math>

It is disgusting...

I had this situation on a system I am playing DBA on. Slony-I was badly broken so I had to restart from scratch. My ground zero was to drop the database and start over... But... what if you can't???

-2000 Lines Of Code