100% on the LCD

When are you viewing the picture you have taken at 100% magnification when looking at it on the K10D LCD screen?

Which translates to 100% view in Photoshop or any other editor.

Is it at

20X, 16X, 12X, 10X, 8X, 6X, 5X, 4X, 3.5X, 3X, 2.6X, 2.3X, 2X, 1.8X, 1.6X, 1.4X, 1.2X

?

If by this, you mean 'one pixel on the LCD corresponds to one pixel on the sensor', then if I've done the arithmetic right, it's about 7X.

Calculated as follows:

The LCD is, according to the specs, approx 210,000 pixels.

Its vertical height in pixels is therefore x, where x is given by:

x times 1.5x = 210,000. So x is about 374 pixels.

The K10D sensor is 2592 pixels high, which is about 7x the vertical height of the LCD.

The magnifications in your list refer to the linear dimensions, not the area (easily checked by noticing that at 2X, you get to see a quarter of the frame).

So you can get 1/7 of the vertical height of the frame onto the LCD at one pixel on the LCD, to one pixel on the sensor. It's the same, of course, in the horizontal direction.

It does beg the question, what's the point having >7X magnification, I suppose. Have I made a silly mistake in my sums? I've assumed that the LCD shows the whole frame, and all of the sensor pixels are visible, which is probably not entirely accurate, but I imagine it would make only a small difference to the result.

I keep meaning to do a test and actually know the correct figure - this topic might just make me get off my backside and do it. 7x sounds about right though, I use 5 or 6x as a quick check of whether the photo is sharp or not.

If 7X is the magic number then it does beg the question and partially why I posted this in the forum. I have a friend who is really particular about sharpness but I kept mentioning (like I was a pro) you had to just sharpness at 100% and not at greater than 100%. But I found it quite tricky to gauge where 100% really was.

By 100% I mean the largest (i.e. acceptable for print) I could print the full 10 megapixel file. How can I see that magnification on the LCD.

And why do they have 1.8, 1.6X, 1.4X, 1.2X? I would be quite happy to jump to 2X straight from 1.2X.

You can change the "Initial zoom display" in the Custom Menu, and go straight from 1X to 2X. I have mine set this way.

What, exactly, do you mean by this?

It's not a trivial question - if you want to print every pixel, you have to be sure that the printer is printing at a resolution that uses every pixel, on the paper size that you choose, as well as using an unresized image.

What is 'acceptable' is also dependent on viewing distance. If you're twice as far away, the same pixels can be twice as big and still look as sharp.

Then again, it depends on the lens you're using. If the resolution of the lens is good enough, then you can use every pixel and the picture will look sharp. If it's not as good, you might not be able to, and the image will look sharper smaller.

So '100%' doesn't really mean much unless you explain what you're really trying to achieve.

You refer to:

If my post earlier doesn't answer this question, then I don't understand the question. If my sums were right, you see every pixel at about 7X magnification.

Below this, the LCD isn't showing every pixel, it's doing clever software stuff to remove some of the pixels and make the rest look sensible.

Above this, the LCD has to invent pixels between the ones actually recorded, and again, it's doing clever software stuff to achieve this.

And to make things more complicated, isn't it showing the embedded small JPG when you shoot in RAW?
This JPG image is of a smaller resolution if I am right.

That would mean that the best view of the picture you've just taken is only available when you shoot in JPEG.

Please tell me this isn't so...

I read a recent review of the K20D which said that 'actual size' was 11x on the camera's LCD, which I must admit does come in useful.

I just did a check in PhotoMe and it shows the following information for a K20D PEF file.

Filesize: 15.388,6 KB

Thumbnail dimensions: 160x120 pixels
Thumbnail size: 7074 bytes

Additional Image size: 4672x3104 pixels
Additional Image size: 1138923 bytes

The question is, does the camera use the full resolution but highly compresses embedded JPG (1mb for a 14,6MP image) or does it use either the full JPG when you shoot JPEG or the RAW file when shooting RAW.

[edit]
wait there is more information....
In the manufacturers note there are entries about a preview image size and length (dimensions and size are 640x480 and 27082 bytes)

Why is there so much extra information embedded in the RAW file?
[/quit]

That's interesting. If I do the same calc as for the K10D, it comes out as 8.

230,000 pixels on the LCD, which is 391 pix high.

Vertical sensor resolution = 3104.

3104/391 = 8

Because different people use their screens in different ways. I rarely use mine to check for things like "sharpness" as I know from bitter experience that I'll be disappointed with what it shows me - I wait until the image is loaded into the PC where I can have a proper look at the technical side of things.

I do, however, use the LCD screen for checking composition in the field and having the fine increments at the start of the zoom range allows me to better judge how various crops will work ("if I get rid of the lampost will his ear still be there?" etc).

It's nice of Pentax to give us the option of starting at a higher magnification (for those that use their camera that way) but I for one would be gutted to lose the finer control that you get at the start of the range. Each to their own...

This review on DPreview of the Nikon D300 may shed some light on this matter.
http://www.dpreview.com/reviews/nikond300/page3.asp

Below you can see a real-life, same-size example of the difference in resolution between this new 921,600 dot screen and a more typical 230,400 dot screen (in this case on the Canon EOS 40D). Both cameras were set in play mode with the same image (note that the D300 doesn't use the whole screen in play unless you magnify), a shot was taken of each camera from the same distance (hence the screens were captured at the same magnification).

They talk about dots as in subpixels. So the D300 has a 640x480 pixel screen which accounts for 1920x480 subpixels which equals 921.600 dots.

The EOS40D has a 320x240 screen which gives 960x240 subpixels and that gives a grand total of 230.400 dots.

This would make a 14,6x magnification needed for a 100% crop on the K20D.

The review I mentioned was also DPReview, as quoted:-

Turn the rear dial to the right to enter magnification view, keep turning to magnify further. There are twenty one steps to an impressive sounding 32x zoom but beyond 11x the magnification is greater than 1:1, so you don't get any additional detail.

Just to add to the confusion!

I certainly agree with you. Your post was breath of fresh air in a thread that is otherwise more like I expect to see in DePressingReview. Use the screen for a quick and dirty composition/exposure check and leave the fine tuning until later. Look at the picture, not the pixels.

Duncan