The music is “like it” from sapiens fx.

New Zealand Helicopter Ride! from Jon Rocatis on Vimeo.

I got quite a lot of footage from various places and since I turned down the jpeg quality to the absolute minimum to get more shots on my 2GB SD card I even had the battery die before the card was full some times. I need to look into making some kind of external battery hack.

Anyway, just finished editing the best sequences in Final Cut Express and uploaded it to Vimeo.

British Columbia 2009 from Jon Rocatis on Vimeo.

The sound is garbled in a couple of places – quite weird. It is not on the source material but it seems that it happens in FCE when it converts from mp3. Must investigate.. Update: I tried converting the mp3 to wav using VLC (QuickTime will do as well) and then using that in FCE. That solved it! Maybe FCE doesn’t like VBR mp3’s or something.

The music is by HiFi Hustlers btw.

Also I put instruction timings and whether they execute in the odd or even pipe right there with the instructions – all this info is taken from the “SPU Cheat Sheet” document by Insomniac Games. Thanks for those!

I hope you will find it useful. It really was more work than I anticipated and “Pages” which I used for this is not really suited for this kind of work and it slowed to a crawl. Rather painful actually. It handles PDF import very nice though. I used OmniGraffle for the diagrams and I found it pretty much perfect!

Anyway, enjoy!

Download it here

So I bought two 500GB Seagate Barracuda 7200.11 rpm drives and formatted them for RAID usage. I selected 64K block size – I think default was 32K.

Next I copied my boot drive to the RAID setup using Carbon Copy Cloner and made the new RAID my new boot disk. There seems to be some confusion wether or not you can boot from a RAID drive – I do it now so I guess you can.

The old boot drive was a standard drive supplied by Apple – a Seagate Barracuda 7200.9 160GB . I tested it using Xbench 1.3:

Disk Test	48.12
Sequential	86.64
Uncached Write	82.30	50.53 MB/sec [4K blocks]
Uncached Write	82.89	46.90 MB/sec [256K blocks]
Uncached Read	86.32	25.26 MB/sec [4K blocks]
Uncached Read	96.43	48.46 MB/sec [256K blocks]
Random	33.31
Uncached Write	11.15	1.18 MB/sec [4K blocks]
Uncached Write	91.24	29.21 MB/sec [256K blocks]
Uncached Read	90.87	0.64 MB/sec [4K blocks]
Uncached Read	119.29	22.13 MB/sec [256K blocks]

Disk Test	120.21
Sequential	182.63
Uncached Write	305.03	187.28 MB/sec [4K blocks]
Uncached Write	323.33	182.94 MB/sec [256K blocks]
Uncached Read	78.29	22.91 MB/sec [4K blocks]
Uncached Read	362.56	182.22 MB/sec [256K blocks]
Random	89.59
Uncached Write	31.76	3.36 MB/sec [4K blocks]
Uncached Write	353.19	113.07 MB/sec [256K blocks]
Uncached Read	170.70	1.21 MB/sec [4K blocks]
Uncached Read	223.48	41.47 MB/sec [256K blocks]

So quite a bit faster at least according to Xbench. I also measured cold-boot times until the Quicksilver logo appears. Before: 2 minutes. After: 41 seconds. Nice!

Only 5 more days until she flies again!

For some reason no movie player that I’ve come across will play movies shot in portrait mode rotated properly. Maybe the information is not stored within the movie file – anyway the end result is that you have to tilt your head when watching!

I really wanted a tool that could do loss-less rotating but I’m not sure mencoder does that. Anyway it works fine.

Movie rotate:
mencoder -ovc lavc -lavcopts vcodec=mjpeg -vf rotate=2 -oac copy orgmovie.avi -o rotatedmovie.avi

It also does another thing I’ve been wanting to do: Making a movie file out of jpgs. Here is an example of some options that created a file that works in QuickTime:

mencoder mf://*.JPG -mf fps=8:type=jpg -ovc lavc -o rig.avi -vf scale=640:480 -ffourcc DX50

The Mac executable I downloaded from the MPlayer site didn’t include mencoder so I got it using MacPorts instead.

So I wrote this little Ruby script to check it out myself.

I msut say taht it is rllaey ture waht tehy say. the biran is pertty aminazg 

hree is the spcrit if you wnat to try it out yurseolf

# word scrambler

str = ARGV[0]
srand(Time.now.usec)
strs = str.split(' ')
res = ''
strs.each {|curword|
  scramble = String.new(curword)
  len = curword.length
  if (len > 3)
    begin
      sel = curword[1..len-2].split('')
      (1..len-2).each {|c|
        idx = rand(sel.size)
        scramble[c] = sel[idx]
        sel.delete_at(idx)
      }
    end while (curword == scramble)
  end
  res += scramble + ' '
}

puts res

Getting the launch tickets was pretty intense as they sell out quickly. So the first countdown we experienced was the countdown for opening of the ticket sales! I was refreshing the ticket site URL pretty often until they finally started the sale. Then it was a race to enter the required information.. typing away like crazy.. done.. clicking “next”.. Arg! Forgot to click “I have read and understand bla. bla.”. Back.. clicking the checkbox.. “next”.. Arg! My credit card had changed from Visa to a Discover card! Back.. changing credit card type… “next”. Finally! The order went through and the confirmation mail appeared in my inbox! Phew!

We visited the Kenendy Space Center 2 days before launch to check things out. They had a “Up Close”-tour which we wanted to go to but they apparently sell out quickly so we had to get tickets for the next day. And the next day again was launch day so we ended up going to KSC 3 days in a row.

But it was all worth it. Seeing the Shuttle up close on the launch pad was really cool and it was a nice thing to do before seeing it launch.

At launch day we arrived at KSC pretty early as we didn’t want to take any chances with heavy traffic or things like that. The amount of people there surprised me! There were more people that evening than we had seen the previous days – and that was during the weekend! We killed some time taking pictures in the Rocket Garden where they illuminate the rockets at night.

I think we started to get in line for the busses that would take us to the launch viewing site around 23:00 or something. There were really long lines there, but we finally entered a bus and arrived at the NASA causeway around 01:15 – about an hour before liftoff. Then it was camera gear setup time! I brought my trusty old 300D for one purpose only: doing a long exposure of the launch. Setting it up using a gorillapod was pretty difficult – the camera was sitting all the way down on the ground and with everything around the being black it was really hard getting reference points.

With all camera setup done I could finally just relax and enjoy the countdown..

Unfortunately clouds had moved in over the area during the night – and they were thick clouds as well. So the Shuttle vanished out of sight fairly quickly after launch. That was pretty disappointing of course but at around the same time I realised that we weren’t going to see it again the sound reached us. Then the disappointment immediately went away! First you hear the rumble of the main engines followed by the awesome crackling sound of the twin solid rocket boosters. That is impressive to say the least. Even though we were around 9.5 kilometers away you can still feel the sound!

“Roger roll, Endeavour…”

Above is my long exposure attempt. The horizon is a little crooked and I should have had a little more ground in the picture. But I was afraid that I couldn’t fit the entire arc in the frame – that turned out to be a non-issue..

Even with the disappointing clouds it was still an awesome experience and something that I will never forget.

The cool thing about rocket launches is that they aren’t designed to be bright like the sun or to make as much noise as possible – that is simply a by-product of what it takes to go into space! That is all kinds of awesome! To quote one of Andreas’ T-shirts: “Science works, bitches!”

See more launch pictures here!

Hopefully it will stay safe and sound for a long time…

Recently I got a new found interest in electronics after realizing what you can do nowadays with FPGAs and microcontrollers. I think it all started when I read about the Minimig – awesome project!

So I started toying around with the idea of replacing the GDROM drive with something else – a hard drive, connection to a PC or maybe a flash card. I did some googling and realized I wasn’t alone with that idea. A good thing is that the gdrom electronics are on a separate board which connects to the Dreamcast motherboard. Therefore it should be easy to replace at least physically.

Initially I thought the gdrom used some kind of scsi communication but others have already done some digging in this area and the gdrom board actually contains an IDE controller chip! That pretty much settled it. Problem is that normal IDE cables have 40 pins but the connector in Dreamcast is a 50 pin one. Since there are no other wires going to the drive module some of them must be power lines.

So I need some extra Dreamcasts that I can hack – I don’t want to risk breaking the ones I already got! I got some extras off eBay the other day. They all worked fine – except one that died when I tried booting a cd! It blew a fuse on the motherboard – I think there is something wrong with the gdrom drive – the laser sled was parked all the way at the top which is not normal..A much bigger problem surfaced today! I opened one of the Dreamcasts and realized it was different to all the ones I’ve seen before – including the ones I have seen on the net: it has no gdrom controller board! That has been integrated on the motherboard – the connector is gone and only wires controlling the drive motor is left!! This is not hacker friendly anymore! So I guess Sega did some cost reducing. That Dreamcast was produced in november 2000.. I don’t know when they started doing that – somewhere between december 1999 and november 2000 apparently. I have yet to open the last machine I bought for hacking purposes but I’m hoping it’s an older model!