No But it's a LOT more complicated than that

An RGB image (let's take a TIFF) has three bytes per colour, per pixel. So that's a total of 24-bits per pixel (8 red, 8 green, 8 blue).

However, a JPEG isn't that simple. It's actually split into a YCbCr - Luminance and two colour channels. The equation that mulches the RGB information into the YCbCr throws away most of the blue (takes only 11% IIRC). So JPEG can't contain that much information.

Added to that is the fact that the in-camera JPG came from the RAW data in the first place

And RAW is 12-bits per photo-site. Each photo-site is either R, G, or B. There are twice as many G as R or B. They're arranged in a square. The "missing" information is guessed by a demosaicing algorithm. This is done in camera for JPEG, or on PC for RAW.

Bottom line is that you have a LOT more info from RAW (at least 4 (EDIT:16) times as much, but more when you account for the loss of the blue channel). And the K20D has 14-bit RAW

Confused now?
Matt

LOL!
Or how time works

Sorry to go off topic!
Matt

Actually, not THAT difficult

All the code is contained in Dave Coffin's dcraw package. It's a single C file that is easily understood and contains pretty well all the RAW formats of every camera that ever was

You'll find it an education
http://cybercom.net/~dcoffin/dcraw/
http://en.wikipedia.org/wiki/Dcraw

Although the reassembly of image data back into a RAW file would be interesting, to say the least, the actual structure of the file and everything is contained within dcraw is pretty easy to follow.

You cannot, with absolute certainty, say that a RAW file is unedited. That's why some of the larger cameras include digital signature mechanisms within the camera (think Nikon do it in their top end) - the code creates a mathematical fingerprint within the camera that actually proves the image hasn't been altered at all (And it ain't cheap either!)

Matt

When working with concerts in these conditions I do the following:

1. Set Manual mode
2. Set the metering to Spot (the dial on the left hand side)
3. When someone comes on stage, zoom in, point the very centre to their forehead and press the Green button (this will set the exposure)
4. If this is too slow, then increase the ISO rating (on firmware 1.3 you should be able to hold in the OK button and dial in the ISO on the front wheel). Then repeat step 3.

If you use Matrix metering the camera will try to make the generally-black stage into a mid grey, and consequently everyone will be over-exposed

...and shoot RAW too - because you can easily adjust the exposure after the event

Hope that helps!
Matt

RJ,
Glad to help

Maybe you could point Pentax Germany to this thread and my findings
I reckon an all metal pin would be better than plastic (though it may not be good for CE and safety approvals). Certainly Pentax could make an improvement in the mechanism!!

Just be careful how you pack your camera & flash!
Hope you get it fixed quickly
Matt

Here you go:
http://www.pbase.com/mattmatic/af540_pin

To repeat: Absolutely no liability is assumed or implied! I am providing these images for information purposes only.

Matt

Ok... have some things to report

WARNING: Do not try this yourself. It will invalidate your warranty

Took apart the lower section (undo the four screws). The whole bottom plate can be removed - there is a small connector that can be disconnected making it much easier to work on.

First of all I tried Super-Lube (from Loctite) - a teflon based oil. Not much difference. Removed the metal bracket that houses the metal pin and connected plastic "L-shaped" piece. Tried various things but couldn't get it 100%.

Then, took the hot shoe off. Had to be very careful as this is a bit tricky!

Now, when looking at it I can see that the tolerances are very, very slightly off. The plastic pin that is actuated by the locking ring is fractionally too short and is rubbing on the part closest to the centre pin. (With the hotshoe completely disassembled you can see that the plastic pin is definitely not in the centre of the hole).

Shaved the smallest amount off the inside of the plastic "L-shaped" piece (used a scalpel with the blade at 90 degrees and gently took the tiniest amount off).

Reassembled and now it's all fine


To answer the original problem - it's possible that the plastic "L-shaped" piece is broken and is not lifting the pin. It isn't that difficult to replace, to be honest. However, if it is all under guarantee it is best to get Pentax to repair it - they should be able to do it while you watch.

However, when they have fixed it (before you leave!) check that the pin will always move when you rotate the locking lever on the flash. It may get stuck as mine and amoringello's have

I'll post some photos in a moment on pbase...
Matt

Ok, now I have my AF540 in front of me...

The pin that locks into the camera is connected to a little pin that you should be able to see from the front of the flash. Between the locking ring and the body there should be a silver pin that moves up and down when you rotate the locking ring.

If this mechanism is broken, then the only way "in" is to undo the four screws at the bottom of the flash and pull out the pin (and then fix it). Looks like a visit to Pentax is needed

And looking at my flash... the pin is stuck in the "released" position. It will not descend to lock onto the camera. Doh!

Matt

Hi,
On the AF540 (and AF360) there is a locking pin on the hotshoe.
It is located on the top left of the hotshoe. The pin is linked to the rotating lock knob - I think you can see the mechanism on the right hand side of the flashgun. There is a "bump" in the rotating lock knob that should push up the pin.

You may want to gently try moving the flash gun while unlocking the knob. It may help if you turn the camera and flash upside down (to let gravity help).

Do NOT force it - the pin enters a small hole in the hotshoe of the camera!

Hope that helps!
Matt

Hyram,
Get to grips with everything first
Then... when you feel limited by JPG (if you do), then switch to RAW (or RAW+JPG).
Matt