thomas_rutter

I can't really speak for the Olympus but I have the Nikon D40. It is a good camera, but a few things to note:

- The kit lens isn't a particularly good one; you'll want to factor in the cost of a better one. I got the Nikon 18-70 AF-S (which was, ironically, the kit lens of the D70). Some prefer the Nikon 18-200 AF-S with VR which will cost you more - probably more than the body. Or there are Tamron lenses I guess.

- The D40 can only autofocus with AF-S lenses, not AF. Not sure if this is the same on the Olympus.

For more troubleshooting you might want to determine if it happens when you shoot JPEG, or if it happens when you use the Nikon-supplied RAW converter. Looking at the image and knowing a little big about Bayer interpolation, I would suspect the RAW converter. However, I could be wrong.

I use the technique mentioned by Martin C above, but I use 'Lens Blur' tool. It actually adjusts the radius of blur according to the mask. I'm not sure if this tool is available in your version of photoshop though. I have CS2.

I do the following. Not sure how it compares to other methods - your comments are welcome.

 

I duplicate the current layer. I run a high pass filter at 0.3 pixels, then boost the contrast on that resulting image by about 60 (in brightness/contrast). Then I run a second high pass filter at about 1.5 pixels, and then I set the blend mode to Overlay. Then I tweak the opacity of that layer until it looks as sharp as possible without going too far.

The camera (or in the case of RAW, the RAW converter) needs to apply interpolation to the image before saving it because of the Bayer filter that the sensor uses. Without any sharpening applied, the image would look very soft as it comes straight out of the Bayer interpolation process.

 

All of the in camera sharpening settings for JPEG will add additional sharpening to this soft interpolated image, with one exception: there may be one setting which claims to add 'no sharpening at all' which will be the softest of all the settings.

 

The standard sharpening setting on the camera does add a fair bit of sharpening to this soft interpolated image. It is a trade-off between appearance of sharpness and faithfulness to the original interpolated image from the sensor. I usually like to set my in-camera setting a couple of notches below the default (when using JPEG) and sharpen further in photoshop.

That looks like a fairly large spot so it should be fairly easy to blow it off. Don't brush it, and don't blow using your breath because it can contain small particles like saliva. Always blow using a blower with no brush and never touch the sensor with anything. If it continues to be a problem take it to a camera servicer you trust.

 

You'll probably find that using a wide aperture will make the spot less visible.

 

When changing lenses try not to minimise the chance of dust getting into the camera or the inside surface of any of the lenses.

It isn't possible to use a physical filter to desaturate colours while leaving the contrast the same. However, decreasing contrast will also dull colours by a proportial amount. So, for example, a contrast filter or fog filter. But I wouldn't say that it would get colours 'as color-less as possible'.

 

I would encourage you to experiment in Photoshop using layers to replicate what you want, because you might find that it is not what you are looking for.

Yes, a JPEG file will lose quality each time it is 'saved'. So it is best to open the JPEG and save it as another format like PSD while editing, only outputting to JPEG when you need a compressed version for the web or a print lab, but keeping the PSD. That way you only ever re-save it as JPEG once if at all.

Wow that last one of yours (DSC_0536sm3.jpg) looks great.

It looks like the type of scene where it is very hard to get the white balance looking natural, because of the blue sunlight streaming in the window and the yellow indoor lights. The white balance chosen seems to compensate for the blue sunlight, knocking out a lot of blue.

 

It's hard to adjust for natural skin tones in this because one side of her is lit by the blue sunlight and the other side is lit by the yellow indoor lights. Also, the red channel is blown out on her nose and the right side of her face. You could probably fix it better if you have it in RAW.

 

I tried to adjust the white balance a bit, but I think that in times like this where daylight is fighting with indoor light it is largely personal preference.<div></div>

It looks like the stuff is actually on the sensor. That's exactly what dust looks like when on the sensor - due to the infra-red filter that sits over the sensor, it is a millimetre or so above the sensor hence the blurriness. It will look much more blurred at wider apertures and may look a little sharper at narrow apertures.

 

Those dust specks are huge though, and the bigger a dust particle, the easier it usually is to blow off. So my guess is that something tougher than just dust is stuck to the sensor.

 

I'm sure you could probably see these particles on the sensor with your bare eyes, in the right light.

 

I wouldn't touch the sensor with anything myself; get it cleaned professionally. But you can first try blowing it. Use a blower brush always, don't blow your spit or remnants of your lunch onto it which are permanent!

<img src="http://img524.imageshack.us/img524/8505/ryangel3.jpg" alt="">

<p>

http://img524.imageshack.us/img524/8505/ryangel3.jpg

<p>

I created a duplicate of the background layer and applied a gaussian blur of radius about 88. I inverted that layer and set its mode to 'Linear Light'. I set its opacity around 25%.

<p>

I created a curves layer for darkening the dark areas and I applied a mask to this layer with a fairly large brush so it mainly just operated on the top right of the image and the spot between their heads.

<p>

I created a color balance layer to subtly adjust the colour balance.

<p>

I created a master curves layer to give contrast to the whole thing, particular because the second layer made the contrast pretty flat.

Come to think of it, the Pro1 digital and Pro1 non-digital are both slimline. And both have the same number of coatings as far as I know. So now I don't know.

One person I asked about digital filters said that they are slimmer and thus less likely to add to light falloff at corners on a digital lens which has a smaller light ring.

 

But I am not buying it. Well, I am buying it (I bought a filter marked 'pro1 digital' but I don't believe it has any powers over a filter that isn't marked digital.

 

The camera equipment industry has gone through a revolution in the last decade with so many new technologies. Even new lenses are coming out for digital. The filter makers were probably just feeling disappointed that nobody had a reason to buy new filters.

 

So they 'invented' something to change about the filters.

 

Get a Pro1 Digital or a Pro1 without digital, whatever they have in stock. It doesn't matter.

I have an SB-400 and it is pretty handy.

 

Some positives:

 

- Contrary to one comment above, it does do balanced fill flash. It is a camera dependent feature, and my D40 allows you to dial in a separate exposure compensation and flash compensation mode allowing you to set flash amount independent to exposure. I can also put it in manual mode and choose any power setting if I wanted to.

 

- It allows bounce flash, which the built in flash doesn't. That alone makes it a huge bonus, for me.

 

- It is small enough to fit in the pouch of my camera bag that's intended for spare cf cards, lens wipes, or whatever.

 

Drawbacks:

 

- It doesn't swivel, so bounce flash is upwards, or to the side in portrait mode. You can't swivel it backwards to bounce it onto a wall or screen behind you.

 

- It doesn't do wireless or trigger other units, you can only use it remotely with an iTTL extension cord. You'd probably buy one or more SB-800 if you wanted a more complex setup anyway.

Moving from 6 MP to 12 MP sounds like you are doubling the resolution, but you're not. You are only multiplying the resolution by 1.414.

 

So, even stating pixel counts on cameras is misleading, as for example it may lead people to believe that an 8 MP camera is twice the resolution as a 4 MP camera. So consumers deciding between 6 and 10 or 5 and 8 are really only talking about a very minor increase in resolution and it would easily be outweighed by other factors of the camera.

 

It would be more appropriate to state the resolution like in video - by quoting the horizontal and/or vertical resolution or "lines of horizontal resolution". For instance manufacturers could say "This camera has a resolution of 3008x2000 compared to the A95's resolution of 2592x1944". It's a more realistic idea. Or, assuming that you'll always have square pixels manufacturers could just quote the vertical resolution. "This camera has a massive 2000 lines of resolution".

 

Not to mention that sensor size is also a factor - a 6 MP sensor that is 4mmx3mm will give very different results to a 6 MP sensor that is 24mmx16mm. Not that this is something you need to worry about with DSLRs because the sensor sizes are all good, but with compact digital it just seems ridiculous to me that people still judge between cameras based on pixel counts more than sensor size.

 

---

 

With regard to that above posted link, there was a serious flaw in that guy's methodology in that he took a 13 megapixel image and downsampled it in Photoshop for the 8 and 5 megapixel versions.

 

The Bayer sampling and interpolation that happens in camera will limit the effective full colour resolution, because each pixel needs information from its surrounding four pixels in order to arrive at a value. This means that an image taken on a 5 MP sensor won't look nearly as good as an image taken on a 13 MP sensor scaled down in photoshop, due to the limitations of the Bayer interpolation when taking a full res shot out of camera.

 

Even the "zooming and cropping" method shown on page 2 has its faults, because the sharpness of a lens (and other visible factors) varies at different zoom levels. To do the test you would need to use the same lens at the same setting, and just change the body, ensuring that all bodies have the same size sensor, but just different pixel counts.

 

Bottom line is that there is a difference between a 6 and 10 MP camera, but the difference is way less than the megapixel counts would have you believe. In practical terms, other than viewing zoomed 1:1 on your monitor, the difference is always meaningless other than when you are printing it out poster size at high resolution and even then most people won't be able to tell, even when the difference is more than a factor of 2 megapixels.

 

On the other hand, if you take a shot with two different lenses, the differences in quality may be immediately obvious to most people, even if printed really small or viewed at a small resolution on screen. This is because factors like light falloff, distortion and contrast/flare are immediately apparent and if you are pushing the lens to its limits, will be easily visible.

Note: more care could be taken with the coloured stains on the bottom - a number of layers with different adjustments could be used rather than one.

Remove dust and scratches with clone tool. Apply slight sharpening.

 

Make a feathered mask around the pink stain and apply levels to match the rest of the picture.

 

Adjust overall levels of RGB channels separately to account for different fading on different channels.

 

Apply overall curves to account for fading removing some of the contrast.

 

Increase saturation (not sure why saturation is so low - old film?)

 

Correct colour cast using warm filter (Photo Filter control).

 

Apply overall curves to taste.<div></div>

The aperture you select will influence the depth of field of your camera. For example a wide open aperture will lead to narrow depth of field, or in other words, things that are out of focus will look most blurred. A narrow aperture will lead to wide depth of field where things that are out of focus are not nearly as blurry.

 

When you look through the viewfinder you are always seeing it with the aperture wide open, regardless of the aperture setting on your camera. This is because the aperture remains wide open while it's not taking a shot, and closes down to whatever setting while the shot is taken.

 

Depth of field preview will close down the aperture as desired without taking a photo, so you can see the effect in the viewfinder.

 

It will also make it darker in the viewfinder and will make increasingly out of focus things look sharper. Never focus while in depth of field preview mode - focus in normal viewfinder mode.

 

DoF preview is not nearly as useful these days with digital - because you can see the result on the LCD as soon as you take a photo anyway and it's better than pressing the depth of field preview because

- your finger is already on the shutter button

- it will be just as bright

- it is the actual photo taken, ie any slight differences (in alignment, dust, etc) between the viewfinder and the sensor won't matter

- you can zoom right up to get a closer look at the focus and out of focus areas (bokeh)

Essentially you are looking at resolution of 3872 x 2592 vs a resolution of 3008x2000. That means the 10 MP camera has 29% more resolution as the 6 MP camera. That's a tiny amount - to get a doubling of resolution you would need to multiply the pixels by four. So in order to get double the resolution of a 6 megapixel you need 24 megapixel. 6016x4000. And if you got that 24 megapixel sensor any fuzziness with your lens will annoy you so much more.

 

On the other hand a lens can make all the difference, depending on how you use it. If you always use the sweet spot of the lens (f/5.6-f/8.0) then it'll be sharp whether it's a toy lens or top grade one. Professional lenses give you luxuries like faster apertures, and sharper images when wide open. As well as build quality. It translates into better performance in low light, both in terms of sensitivity and sharpness.

 

The 18-70 Nikkor is a nice sharp lens though at 18mm at wide open it's sharpness is only medium and it has a lot of light falloff. And it's 'only' f/3.5 to f/4.5.

 

The downside to a better lens though is the weight. You may find it more painful to carry around, if you like to take lots of photos on the move.

 

So to summarise:

The 6 megapixel to 10 megapixel upgrade is puny - you'll only get 29% more resolution. Go with a lens upgrade instead - get one with a wider aperture and which is nice and sharp and well built.

 

If a lens that is twice as heavy and four times as expensive doesn't appeal to you, take advantage of what you already have and try to make better photos.

I kind of like that DX is becoming a standard sort of size now.

 

Of course, I'm in a position where it is a good thing to me - I don't own a whole bunch of expensive old lenses that cover the 36x24mm frame size - which puts me in a great position to immediately benefit from the smaller physical size and cost of the DX lenses.

 

I understand that those who have expensive lenses designed for the 36x24mm format would want a digital camera that will act the same way with the same lens.

My camera usually won't show any values below around 1/4 second and it then goes to "lo". It will still let me take a picture, and I'd guess the exposure is about half a second.

 

So the speed at which it will go "lo" will depend on how fast your aperture is and your ISO.

 

As pointed out above it is the camera indicating that there is not enough light for its meter to perform accurately, regardless of the settings you're using.

When you look at the histogram on your camera this is not related to white balance. White balance is about the colour of light, not its intensity. If objects that are white in real life look too blue or too orange in your pictures, then your white balance is wrong.

 

When you drag the white point to the left in the levels control of photoshop you are effectively increasing the exposure - this would indicate that your exposure was too low to begin with. If you only need to do this a little bit, then that is fine. It's better to underexpose a little than overexpose a little in digital.

A rule of thumb is that in film you should err on the side of overexposure and on digital you should err on the side of underexposure - if you can fit everything in the middle of the histogram for example. Not sure what your skill level is - if you know all this already just skip over my post ;)

 

In digital you should avoid overexposing your subject, and always expose for highlighted areas of your picture. The exception is if there is strong backlighting and you don't mind blowing out the background - but even then I would consider some fill flash or bouncing some light on their face with a white sheet.

 

So you look at the histogram of the photo after you take it, and if there is a tall thin white line on the right hand side, then areas are overexposed. If there is no tall thin white line on the right hand side, and the graph tails out a little bit before the right hand edge (up to say one sixth of the width), then you have successfully avoided

overexposure. If the graph tails out way before that, say only halfway from left to right and there is nothing further right, then you have underexposed and you can stand to increase the exposure.

 

If you underexpose in digital by only a stop or two then it is not a huge problem - boost it later. You'll just get a fair bit more noise. But if you overexpose you can't recover the lost detail and you will get 'hot' edges that won't go away no matter what you do. Shoot RAW and you get a little bit more latitude in this regard.

 

Note that if the histogram has only one graph, rather than separate graphs for red, green, and blue, then it won't be as accurate for determining if you have overexposed. This is because you can overexpose either the red, green, or blue, in any one place, and while it will look fine on a histogram that shows luminance only, it will still lead to 'hot' areas where the colour shifts and which have unpleasant borders. This is because the total luminance of that part will be normal, but the red, green or blue component will still be too bright.

 

In a general sense, a histogram can be used to give you an idea about the contrast of your image. If it is bunched up on the left with only tiny areas on the right then it is a low key image, whereas if it is bunched up on the right (but not overexposed) then it is a high key image (like a polar bear in the snow). And you might get used to recognising what a normal mid tone such as a person's skin might look like on a histogram.

To each their own - I guess you are free to choose your own compromise between righteously defending your rights to photograph as you choose, and trying to avoid a confrontation in the first place. If like me you'd lean towards the latter, perhaps you could try being a little more inconspicuous.

 

This does NOT mean sneaking up on people or acting weird. In fact, part of being inconspicuous is being confident in yourself. If you act timid as if you feel you are doing something wrong, you're more likely to evoke a 'creepy' response from others. So be confident.

 

Some people assume that if there is no flash then you didn't take a photo - particularly indoors or when it isn't sunny. Even when people ask me to take their photo, if it doesn't flash they often ask if I really took it. If you are using flash, then, just be aware that you're more likely to draw attention to yourself.

 

Try not to concentrate on just one person. If there is only one jogger in the park and you are taking a photo of them, it is going to be more uncomfortable for her. If there were several people then she wouldn't feel 'singled out' as it were. Perhaps if there is nobody else then take some photos of trees or animals in addition to taking her photo. If you are taking photos and you see someone coming, avoid walking up to them to take a photo. Walking up to them is a confrontation, and can make them feel uncomfortable. If you are confident enough you can strike up a conversation and then take a photo but you may not be. So I guess just try not to make it obvious.

 

People are also less likely to think 'potential pervert' when they see you if you dress nicely and are reasonably well groomed. I don't mean dress formally. Just if wearing casual clothes, don't wear the ones that might make you look like a creep.

 

Other people's suggestions like using a telephoto (though that will rarely give the desired effect) or using a waist level viewfinder are good ones - as is having a smaller camera in general.