[ Houdini Cookoff ] 02

Yes, finally I got my act together to acquire the plate, as well as capture the image sequences required for creating a hdri map.

First off is the location.

A nice window sill, with a nice cloudy sky providing beautiful soft shadow. It's a little tough to see, but there are four dots of blu-tack setup in a 12x12cm grid - this will be to ensure ease of camera matching the plate, and visible by the shooting camera at all times.The cup specifically chosen to be easy to model in Houdini (i.e. straight sides), and has its bottom blu-tacked to the sill so as not to move during the stirring action.

For the actual shoot though, I'd put a white box as a makeshift reflector in order to lessen the contrast of the shadow side of the cup.

In terms of image acquisition for the hdri map, I'd found a great recommendation by a colleague - Ball Finials in a chrome finish! A whole lot of these can be found at John Lewis (in the UK). The surface is far from perfect, but it is very solid, and more than adequate for little projects like these.

From the reflection above, I am actually quite close to the sphere; barely few feet away. Usually for HDRIs, you'd want to be far away so as to present less of the photographer in the ball, and being further away will better simulate an orthographic camera. Sadly, due to physical space constraints, I brought out my 200 and 300mm full sized SLR lenses but I simply could not find the space. I just hope these env maps will work! I only need them for reflection and diffuse lighting. Obligatory link to Paul Debevec's website on light probes.

Stills were acquired in two positions, from the shooting camera's POV, and one with the lens axis parallel to the ground. For each shooting position, I'd shot at iso 100, F2.8, 1/2000 to 1/4, giving me wide range of exposures to pick from. Each of these positions were shots twice, with different white balances (Cloudy and Custom) as I know my raw converter software does not do well with the white balance of the Lumix LX5.



Above is the plate :3 Well one of them anyways. Haven't yet decided which to use. I shot it overlength, and will only use at max 15 seconds, or less - I only have 4gb of ram on my home machine, and tests so far are already hitting the ceiling. Not really wanting to buy more ram as I rarely hit the ceiling.

The first few seconds was to allow the camera to settle down after I pressed the shutter, and the next few is to allow the milk carton/bottle to fill up the tea. Next, I stick the end of a spoon into the cup, stir, and pull out the spoon.

Originally, I'd thought to leave the spoon in, but I do not think I have the time to properly integrate a live action spoon, hence the stir and remove. I will probably need to create a stand in cg spoon element, and hopefully the motion blur seen in the plate above won't cause too much issues. I'd also blu-tacked the spoon at the 5cm mark to give me a frame of reference when I'm matching the animation. *Not* looking forward to that.

Well now that I've got plates and ldr images for the hdr map, no excuses....

Daaw!

I've been quite inspired recently by OCRemix, and I hope to do some arrangements of my own for the good old games back in the day. As such, I've built a beginner's DAW.
The signal chain is quite straightforward; microphone -> audio interface -> laptop, and my main application (for now) would be Garageband.

For microphone, I decided (after much, much going around in circles) to go with the good old Shure SM57. The logic follows that it is a very hard wearing microphone, plus it's a dynamic and will help reject ambient noise as my room is not properly treated.

Audio interface is an Onyx Blackjack. There are several interfaces at the same price point, and this was chosen mainly as the preamps are supposedly of very high quality. There is no midi in/out, and that is alright as I do not think I will be connecting any devices that use traditional midi connections at the moment.

The mini keyboard is the Akai LPK25, a really small usb midi keyboard that I think would be fine for use with garageband. Having used larger M-Audio keyboards with weighted keys, the keys on the LPK feel like a cheap casio. In addition, the velocity feels non linear. Can't be helped as it is very affordable. Still, I had alot of fun playing tunes on it in garageband. Some songs with extended ranges annoy with the octave break though :)

Sound wise, I haven't tested my saxophone on it. Based on vocal tests, this is a very different sound from my Zoom H1. The H1 is very clear, but doesn't seem to have as much bass. The XLR signal chain at the moment sounds very smooth, with the right amount of bass. I actually thought my voice wasn't too bad through the sm57.

Can't wait for the weekend when I can go all out and check out this setup on the saxophone :) Still working on my Houdini cookoff project - was doing fluid sim tests - and was playing with the above whilst the sims were cooking away :)

Pst. You know that chick?

Yeah this is supposed to be a photography blog? So where's the photographs? Not putting up the majority of my stuff lately as I'm prepping a new website. Or should I say, was prepping a new site. With the projects I'm working on at home and learning sax, doing photography. I'm pooped.
This was snapped with my dear Lumix LX5 on the way to a funk gig. As an attendee :)

Wore through my R-Strap!

Ok so I've been using the Blackrapid R-Strap for several years now, and I love it. I used to use it quite heavily during weekends - but not much after picking up the saxophone - and finally today after an afternoon of shooting, I wore through part of the connecting lugs so that the clip is bent out.

This is not a crit on Blackrapid's part - but just that I do take a heavy toll on my equipment (you need to see how badly worn out my 16-85VR is - rubber has fallen off, paint is mostly gone, hood cracked in at least 3 places....

Will definitely get the updated version as I think the r-strap system is simply the best camera strap for travel, wildlife or event photography. Yes, I use to average 20+, sometimes almost 40 hours worth of wildlife/travel photography in the weekends. I blame the lack of photography nowadays on the saxophone. :P

Macbook dropped!

Omg! I've turned into a mac fanboi.

Maybe.

Nah.

Ok so I'm hoping to be producing backing tracks for my saxophone down the line, and linux is just not cutting it for me - mainly just setting it up takes *forever* to find the dependencies et al for music production. I got it working in the end, but seriously, spending 3 nights just to set it up, vs me walking into the apple store and laying down a piano track under 5 minutes (this is no joke) in garage band is a no brainer: I want a mac to work on my music. No windows machines please, thanks.

Linux will be for my main machine, houdini, raw processing etc, and I'll get a low end macbook of sorts to run my "music production".

Given that the white macbook has been around for awhile, I was hoping to see apple release a new model, sadly it seems the new macbook air has supplemented it.

The air only has a SSD for storage, and I'm just not sure if that's a good thing for audio. I'm probably going to go for a used mac off ebay, something enough to run garageband and logic, or perhaps maybe just bite the bullet and save up for a low spec macbook pro.

Will need a good think about. Not relishing the thought of running osx again though - I ran ubuntu on my last mbp ;-) Best linux experience ever. Seriously.

Zoom H1 - battery drain >.>

Exhibit 1 above is my Zoom H1 portable recorder. Issue: Major battery drain even when switched off.

This is a known issue with some H1 recorders, and actually, mine (got it in February) has been perfectly fine, till the last month or so when it started eating up batteries at an alarming rate (drains a new energizer battery in about 5 days).

Hopefully, I can get it repaired. At the same time, I'm also looking at a new recorder released by Tascam, the DR-05. Not only do reviews indicate it's much better built than the H1, it seems to have a better spl rating (125 vs 120), built in tuner and just looks cooler to me :) If Zoom's UK distributor does not get their act together, I'll probably pop this on the devil's junkyard and use what I can get from the H1 to contribute to the DR-05.

EDIT: Yay! They finally got back to me for an exhange. Sweet.

Life Drawing, Sketching, thoughts.

 
I've been doing life drawing since April 2010, and only started learning how to sketch a few months before that (see my drawing blog). Last week, I'd used up my A3 sketchpad that I use for my life drawing sessions, and this afternoon I popped into Cowling and Wilcox to get more paper. Inside, I was agonizing which to buy - a £10 pad, or the £6 pad.

I then realized, that I actually took more than a year to finish up a pad, so just bought the £10 pad (It was a heavier grade of paper, plus very nice covers). In essence, this is a very enjoyable, yet affordable hobby.

Basically, I spend £30 quid for 6 drawing sessions - this lasts me a month and a half - or £5 per session. Add on a bit for transport. Pencil lead is not exactly cheap, as I use lead holders - however out of my entire box, I've only used.... 2 pieces so far.

That's for cost of attending life session, on a day to day basis I spend from 15 minutes, to sometimes hours if I visit a museum or zoo over the weekend.

Sketching for me serves many purposes, but my key reason to actually learn how to sketch was to conquer my fear of drawing.

Or rather, how much I hated sketching in general.

Rewind to a much younger age, I was never strong in the arts - usually the lowest grade in the class. Does not help self esteem there, I assure you.

Even during my uni days, I've tried learning on my own, and attending life drawing sessions, but seriously, telling a complete beginner to draw a crunched up ball of paper without instructions... doesn't really help. needless to say that did not last long.

Fast forward to London, I decided that I would learn again, and see where it takes me. This was also a time of personal pain for me, and I really needed things to distract myself. It worked out far better than I thought.

First, I think the main thing I learnt, is how impossible it is to achieve perfection. I am only human. Attempting to reach perfection is one thing, but stressing out because something is not perfect is not the way to go. I think I have been stressing out on that at work, at photography, at lots in my life.

Just learning on my own, day to day, taught me about how to get better - simply by practice. I used to believe that I would never ever draw. I now believe I can achieve whatever I care to set my heart upon.

Other beneficial side effects (pun intended) include meeting new people. Drawing at museums often see the stranger wander up to exchange a few words, and almost always, they are friendly and positive. I remember I had an Italian sculptor wander up to me once. He was on hols, and told me to visit Genoa, as there were a multitude of amazing sculptures there. I still need to take him up on it.

Weekly drawing sessions also helped me conquer my fear of speaking with people. Maybe it does not show, since I have little fear of teaching classes or giving talks - but one on one with someone? Engage the escape hatch please. Over this time here, I've come to appreciate how the people in my life drawing sessions are still friendly, even though I am a living icon on taciturn-iness (hey I invented a word!)

Of course, I have a long way to go, being introverted and focusing on myself and a narrow range of subjects do not help.

Going slightly off topic, once again playing the saxophone was yet another distraction - again, personal issues - and once again, it is a blessing in disguise. Not only have I found a music instrument I really like (flute, piano, NAF, keyboard etc did not last too long over the years), a fair few amount of people I've met here in the UK play musical instruments, and it makes for an easier time to connect. Down the road, I too hope to be able to join in local bands or just get good enough to jam at.

I think this is due to drawing, it showed myself that I can improve, abeit slowly. And these two seem to form some sort of positive reaction loop, confidence or improvements in one area gets me going in another.

Should I even be putting this down?

Hell yeah, one thing I learnt from all of the above. Live. Live now. Carpe Diem.

Framestore's work on Deathly Hallows Part 2

Had this article pointed at me this morning :) Yay! Watery goodness. I was assigned 4 shots where I was involved in the main simulations, one of which is in the article - the one with Ron and Hermoine looking up, walls of water on either side of them.

[ Houdini Cookoff ] 01

Project time! More or less recovered from my last project (Deathly Hallows Part 2), so I thought I'd tackle this little project that SideFX has got running.

The competition's supposed to be about creating a food related project in Houdini - I know tea is not strictly food. I'm not in it for the prizes, more like a challenge for myself. Plus, I've always wanted to do it :P

So, the video above shows what I hope to achieve with Houdini, brown tea getting lightened up with a fair bit of milk.

Currently, I only got the reference above, and still plotting how deep I want to take this. For one thing, I do not plan to spend my waking non-work hours on this, so I need to keep it down to quick sims and tricks. What I definitely will need includes a backplate, with an empty cup. The plate will be basically a hand stirring the spoon, which means I'll need to track a cg spoon into place.

Thankfully, the spoon only shows the top tip of it, so I do not need to like be uber correct. One thing I noticed with the spoon is how the meniscus reacts to it when laid down to rest; I will probably have myself remove the spoon after stirring in the plates; one less thing to worry about.

The milk bottle I have two minds about it - half of me is thinking go full CG, get a HDRI env map, let PBR handle the rendering. The other half says, use practical, and get someone to track in the milk carton/bottle. That's still up in the air, and I need to get it locked down before I shoot the plates.

I'll still want a HDRI map of the scene for the top surface of the water. I also need to keep this short, as the machine I work on at home has very little hdd space left, and only 4gb of ram :P

Simulation wise, I think I will be looking hard at combining both the ripple solver and fluid solver *and* smoke solver. My gut feel (and experience) from running the fluid solver tells me that it will be all splashy when the poured liquid hits - not like the ref in which the surface stays pretty calm.

So I'm thinking using voxel fluids to get the shape, then merge the SDF volumes to create a single geo surface.

Smoke I'm thinking would be used below the surface, and we can trace into it. Or, pre-render, then project onto the water mesh. 6 weeks to go, got to get the camera track and even animation done in the next week and a half, then I can focus on simulations, the last 2 weeks would probably be final renders and comp.

Note to self: the cup moves when being stirred; recommend blu-tack to keep it locked down. I do not want have sims inside moving geo :D

Everything is a Remix

http://vimeo.com/25380454

Side Effects Cookoff!

http://www.sidefx.com/cookoff/

I'm so looking forward to this! Hopefully, I will be able to get something out in time. It has to be food related, however my idea is not exactly food - mixing milk into tea. *yeah everyone tells me it's not food* but you know what, even if it doesn't qualify for the competition, I'm still going to try to get something going. It's always possible to learn new techniques from just trying out new ideas.

Already have some reference footage of tea being made ;-) Let's go!

GW

Houdini: Foam Pass for level set fluids

I've been using Exotic Matter's Naiad for water simulations, and apart from the main simulations, other layers are included, such as foam particles floating on the surface, as well as particles ejected away from the surface to simulate spray.

This experiment is an attempt at creating foam off a Houdini Level Set Fluid simulation (Note: Naiad uses FLIP fluids.)

Emission is basically scattered points on a  mesh that was converted from the fluid's main level set ("surface" volume: with a convert sop you can get a mesh). Two parameters are used to control emission regions; curvature (measure SOP) and velocity (lookup into velocity volumes).

These points are then used as for emission in POPs; this particle layer is relatively fast to simulate because it's all pops! Inside pops, there is a fake "advection" stage where the particles sample the velocity volumes from the base simulation, and uses that to move each particle to its new position.

This however will result in particles moving into positions below the surface, or a bit too high above the surface. Tests indicate this can be improved by a high resolution level set, but we all know how expensive those are ;-)

Instead, a VOP POP was created to move the points based on the gradient of the base simulation's distance field, and points are moved accordingly to whether they are negative or positive (i.e. negative distances moved towards the surface following the gradient vector; positive distance are again moved, but against a negative distance vector).

I have no idea about those spots that hang on the ceiling - there is a cull pass that removes particles if they are a set distance away from the distance field. Since those particles are hanging there, I'm assuming there are remnants of the level set that do not have velocities affecting them and not dissipated.

That's it for this experiment, maybe I'll look into it another day.

GW

Houdini: Arrow system using Quaternions

Got inspired the last few days to design an "Arrow" system and finally took Friday evening and Saturday morning to figure out a good way to create a system that allows stuck particles (arrows) to change their orientation based on the underlying deforming geometry.

The main concept is getting an origin frame of reference (in this case, it's the primitive normal of the collided polygon that the particle hit, as well as the vector of the first two vertices. These will give a reference frame.), as well as a constantly updating frame of reference (the prim normal and vector again, from the deforming primitive).

Originally, I had used the Align VOP to perform this calculation, however, the align vop simply returns a transform from one vector to another vector via the shortest path. In this case, it is basically a rotation with an axis that is the cross product of the two vectors. What we want is a proper rotation, so I used the look at vop to generate rotation matrices from the two frames of references.

Next, these were converted to Quaternions, and then I used the SLERP function (Spherical Linear Interp VOP) to obtain yet another quaternion that blends the original frame orientation to the final deformed frame.

The issue I encountered was that the SLERP function returns the initial orientation quaternion when both the initial and target orientations are the same - this is obviously the case during the collision frame; both reference and initial orientation are exactly the same.

This, we do not want; as during the collision frame, I want my arrow's normal vector to be operated on by a Quaternion of xyzw = {0,0,0,1}, not the quaternion for the initial frame. Hence, by subtracting the initial frame's quaternion from both the initial frame as well as the target frame (this is done by doing a quaternion multiplication of the conjugate of the initial frame's quaternion against both initial and target quaternions before putting them into the SLERP).

This seemed to work, from the examples above. :)

Woo! Slowly attacking quaternions. And Houdini just makes it so simple.

Other issues I encountered is during rendering: proper motion blur for the arrows. With a stuck particle, the "v" velocity vector is actually locked in place whilst the particle is with the pstate of 16, hence during renders, this is uber incorrect.

Initial tests using to Attribute Transfer velocity vector (calculated using a Trail SOP) did not work out as it was very inconsistent. What I did then, was to do a look up into the geometry (the collision POP stores the posprim attribute, allowing us to find out which primitive to query), and allow the hit arrow to take the velocity of the underlying geometry.

GW

Saxophone Log: May 2011

Here's another saxophone log from this month. I'm playing from a play along book called "Dig It!" for Saxophone.

What's nice about this is that just one book caters for Alto, Tenor and Soprano as well! The Tenor and Soprano scores are similar, just that the soprano will need to be played one octave below the score. In addition, there are multiple parts for the different saxes, to be played together e.g. alto + 2 tenors. Good stuff!

Side Effects Software: Why use Houdini?

And more question such as:
How many companies have a new daily build accessible from their website, containing bug fixes and feature improvements?
Where can you meet the developers in person and talk to them?
How often do representatives visit you in person?

Well SESI does all that, and more. A few weeks ago, I put up the robotic servo arm controlled by Houdini. It's just a proof of concept that went overkill, but seemed to generate a fair bit of interest, so much that the code and concept may be used for other projects.

Fast forward to last night, it seems like I had failed to test for keyframed animation support, and I fixed the otl right up. The developers at SESI took it to actually rebuild my entire digital asset from the ground up, pythonized, so while it looks the same on the outside, the guts are totally revamped.

http://www.youtube.com/watch?v=TImyQ7VnS_M

http://forums.odforce.net/index.php?/topic/12916-houdini-arduino-robotic-claw/

Next project I'm working on will involve an ADXL335 accelerometer, what will I use with it? As always, these are proof of concepts I hope to learn from to build on for bigger projects :)

Saxophone Log!

To prove that I have no life, here's me on my monthly saxophone log.

Houdini + Arduino + Robotic Claw

Ok so I've been amiss from updating this blog. I am still snapping photos left right and center, but I've not been putting them up here (mainly on facebook). This is because I'd planned to have dedicated website for my photography, and a photography blog as well. The website is up, but... the content is not ready :D Will be coming probably in June, after I finish my current project and get some free time.

In the mean time, I am very busy at work (crunch time now, for the next 6 weeks at least), and when I'm not at work I'm working on other stuff, some of it which is houdini related. Like today's post!

So I've been working on Arduino for several months now, just on and off when I get a free moment, to keep busy. (Else my mind wanders). I think microprocessors are awesome, but back in the day when we had to use massive programmers and UV light to wipe the memory, it was not as accessible as the Arduino, which only requires a USB port to provide power, transfer programs from the pc to the microprocessor and do other cool stuff, like what I did above.

Source code and the HDA can be found on odforce.

I'm slowly getting back into doing Houdini as well (where do I find the time for this, when I'm learning how to draw, learning the sax, doing my photography, houdini, work and other stuff?!) - have been inspired by lots of work lately, and want to create some of them by my own hand :3

Looks like someone's laptop needs a ram upgrade. DDR3 anyone?

GW

More camera related nonsense - Nikon P7000 vs Canon G12

Basically, Nikon makes crappy handling P&Ses!

Ok to be fair, I'm mainly comparing Nikon and Canon, Sony's handling is even worse off for me.

So on my way back today, I decided to ponder if a P&S would work for me instead, and decided again to drop by Jessops to check out the Panasonic LX5. Seems like it's a very popular camera, as even the display model was sold :-/

Nevermind, I'll check out Nikon's top of the range P&S instead. Costing about an entry level DSLR, (a used D200 is cheaper), it feels really nice, and I was going, "Oh wow, look at all these controls! Gotta be a dream to shoot with!" Sadly, it took me a fair bit of fiddling to actually figure out basic stuff like trying to change the ISO or white balance. Still, it feels very good and well proportioned in my paws.

Tried a few other Nikon P&Ses but I'll not bore you to death further.

Popped over to the Canon section, and lo and behold, the luscious G12 was there. Immediately, I saw how easy it was to manage the camera - ISO was on it's own dial, even with Auto ISO - I may actually prefer having a dedicated ISO dial on my dslr - and EV comp was it's own button, and another dial for easy PASM stuff. I thought it was very well designed, and the handling was superb. Menus were also easy to get into. Folding screen + optical viewfinder? Big big pluses.

Gotta try the LX5, however if it's menu layout is like the TZ8 - I'd say I'd pick either the G12 or LX5, as they are cameras for different purposes - especially the LX5 with the low zoom range and fast lens. P7000? Another top of the line P&S I will probably not care about.

GW

Checked out a Nikon D5000 today

So I went past the Jessops along New Oxford Street (in London, UK for newer readers) seemingly by accident today, and decided to pop in to give the D5000 and D3100 a try after dismissing them the last time.

(Side Note: At the moment, I've more or less disqualified the GF1 as a choice, mainly due to the insane distortion of the normal zoom and battery life. Maybe down the road when I have loads of disposable £££ I may get it, however right now, I want something I know I will use.)

As luck would have it, the D3100 was not on the shelves, sold away perhaps? Only had the D5000 to try. As mentioned in a previous post, I put the D5k and D3.1k at a firm no just by virtue of their lack of a top LCD screen. A screen I use very very often on my D200/300 for settings.

First off, the AF on the D5000 was bloody bloody slow. And no AF points showed up. A quick tour of the menu indicated the auto area AF. Switching over to single point AF resolved any speed issues. I've owned the 18-55VR in the past, and am quite familiar with its focusing speeds. I wonder if this auto-area AF is similar to the close priority dynamic AF on the D200 - the AF was so much faster on the D200 in similar light. But who knows. I am fine with the single point and dynamic AF on the D5000 + 18-55VR, it's definitely not 17-55 AF-S speeds, but snappy enough inside the store.

One big issue concerned me - the seeming inability of the camera to separate the shutter button and auto focus. I was able to set the custom function to allow the AE/AL lock button to be used as Af-on, however, did not find a custom function to disable AF on the shutter release. As with the D90, the AF selector is just waay off from where my thumb rests.

There is a possibility I'm just asking too much from a budget, entry level dslr designed for beginners :) Some plus points include a very nice +/-2 stop bracketing ability. I'm pretty sure even my D300 has a maximum of 1 stop between brackets, though it's possible to do a 9 frame bracket on the D300 vs 3 on the D5000. That's a low priority thing anyways. Tested live view AF - a joke, seriously when compared to the GF1, or even my Lumix FS7. Also, the AF point layout is similar to the D200, plus point there. And there is even the overlaid grid which I really like. Viewfinder is of course, super dim but we're talking about a small dslr with a pentamirror, not pentaprism and 100% view which costs like 3 times more.

Also, the obvious lack of AF motor. In general though, looks promising. I'm just reeling at the idea of missing the usage of the 24/2.8 and 50/1.8 on this baby. However, as memory serves, I've never actually used either of these lenses when out on tour even though I've brought them along so..... perhaps I should just bite the bullet and get the D90.... or D5000 + a 35/1.8DX to satisfy my urge for a fast prime (hahaha 2.8 is fast for prime? lols)

On a crop body, the 35mm is about 52.5mm equivalent field of view, which I am just not a big fan off :-/ Even when I was shooting film, on my Contax G, even though the 45mm/2 Planar produced the most amazing feel, I still preferred a wide normal; hence the 35/2 was the one mostly mounted.

End of the day? I still have no bloody idea. I've still got a week or so to decide. Ah well.

Getting a smaller camera - D90? GF1?

So my D200 has been dead for several months now, and I've just started the hunt for a smaller camera.

Why?

Basically, whilst my D300 is super awesome kickass, for general day to day use, and just basic travel purposes - it's just to damned big and heavy. In addition, it's not inexpensive, and I just don't like swinging around something that expensive everywhere I go. The D200 IS expensive... back in the day.Thankfully I got mine used for a bargain.

So I've made a trip down to Jessops today, twice actually, to check out the cameras on display.

For the Nikon lineup, I tried the D5000, D3100 and D90.
Other cameras I've tried included the Sony NEX series, Olympus PEN, one of the Samsungs and the Panasonic Micro 4/3s.

Nikon side, the D3100 and D5000 were immediately off my shortlist. When I first grabbed the camera, I went... erm... where is my top lcd? Can't change my settings quickly without that! The back LCD is where its all at, but it's just too damned cluttered. Nope.

D90 wise, apart from me nitpicking certain aspects, seemed like a good choice:
- Push button + control dial for ISO, WB, Qual, Drive etc. Very much like the D200.
- Rear screen is awesome
- Good size and weight
- Shares batteries with my D300
- AF system is similar to the D200, so no issues there
- Viewfinder, though I don't have my D300 with me, seems bright enough. Don't need no 100% viewfinder.
- Motor for AF lenses! 
- Relatively quiet shutter. My D200/D300 sounds damned loud in comparison. Could this be due to the slower shutter release time?

Nitpicks:
- The main one: The 4 way rocker which is used to select the AF points during shooting is just not lining up with my right thumb. This has to do with the physically smaller camera. The D200/D300 has the AF selector perfectly where my thumb rests on the back of the camera, not on the D90. It is much lower.
- No dedicated AF-On button. I think the AE lock button can be configured for that, but then how am I supposed to lock exposure? :P

Really minor nitpicks
- Flash sync of 1/200. Bleh.
- Bracketing features weak compared to D200. But I don't bracket much so...
- Uses SD cards (yuck), but can't be helped.
- RGB histogram doesn't show flashing highlights
- The name D90 is just not as inspirational as D200 or even D5000

Olympus:
E-PL1 has so-so autofocus speeds. Really, really did not like the menu layout.
E-P2 - again so-so AF speeds. Controls much better, but menus still weak.

Sony's NEX
Af speeds seem good, and the handling of the camera is surprisingly good. However, the menus seem to expect the user to be a total n00b. Nah. Also, the 16mm lens is an odd choice for their fixed prime. 24mm equiv fov is quite wide and not easy to use as a single lens.

Samsung:
Only tried the Samsung NX100, not a fan of the interface. Controls of camera body were quite sparse too. Eh...nah.

Panasonic:
Well last to be mentioned. Because I really enjoyed the camera. I tried the G2 and GF1, and thoroughly enjoyed the GF1. The on screen display has symbols indicating e.g when in manual mode, to push in the rear dial to switch between controlling aperture and shutter speeds. ISO has a dedicated button, as with AF and some other settings that I forget.

AF speeds are fast and responsive, and I enjoyed the AF tracking mode. I mean, look. I didn't even read the manual, it just worked thanks to the blinking icons!

Body is also very well built, feels like a nice expensive piece of kit, whereas the D90 really feels plasticky.

20mm/1.7 is an amazing lens. Just for that lens alone - giving a 40mm QL17 equivalent FOV - is a tempting idea.

At the moment I'm not sure if I should get either, get a used D200, or just don't do anything. The D90 will allow use of all my lenses and flashes, and apart from the SD cards, a straight fit into my system. And it can be the lightweight counterpart I could use on a day to day basis.

The Panasonic OTOH is just so damned tiny. Even with the zoom lens on it's minuscule, even compared to say, D300 + 24/2.8D.

Ah well. I'll see what happens.