Month: May 2012
I remember when I first purchased my Canon 5D MKII. One of the drivers behind this decision was to have a 21MPixel sensor. This was partly due to the main stock library I supplied only accepting 50Mb files and the pixel count on the 5D making this easy to achieve without interpolation. The other factor was that I wanted to be able to print large; 30 inch, perhaps even 40 inch images with good quality. When I recently purchased my GX1 I was also keen to ensure it would allow me to print large.
The GX1 is a 16Mpixel camera but it’s only the 4×3 format that gives an image this size. The 3×2 (similar format to my 5D) gives a smaller size and the 16×9 a smaller image still. To put this into context the 4×3 image would produce a 45Mb TIFF image, which is just short of the size required by the stock library but it’s not too far off and easily achieved with some interpolation. Print size was however the more important to me and out the camera the image is around 15” on its longest side when printed at 300dpi.
Now you are possibly reading this and thinking that the resolution is more than enough to create a 30” print given the viewing distance should be at least 3 feet and I would agree with you. I am however quite fussy (as are most photographers) and I want to look more closely at areas of my photographs and feel happy and confident that the image stands close scrutiny. Now you might think I am ignoring lens quality (which is true) but I know the 14-45 standard lens I use is more than capable of resolving sharp detail. No, what matters to me is the question am I satisfied when I look at the image closely when I print at 30 or even 40 inches.
To test this I resized my starting image (above) to 30 and 40 inches at 300 dpi using Genuine Fractals. I then extracted A4 sized sections from each image, sharpened and printed these on A4 Gloss paper ready to examine the results. If you want to know why I picked this image it’s because it has lots of fine detail. The glass as you can see from the images below has lots of fine straight, parallel lines on it that will show up any problems.
Image sample when viewed at 100%.
So the results? I’m not happy with the 40 inch print when viewed closely but it will be fine at normal viewing distance. The 30 inch print as show by the sample below is however is very good and you need to look very closely indeed to see the problem areas.
Image sample from 30 inch print scanned at 75ppi.
I do however know that I could achieve a good 40 inch print as I would simply reduce the resolution of the 30 inch print to 200dpi rather than 300dpi. I am also questioning the results of the Genuine Fractals software as I find it produces quite blocky results and this is what I can see in the 40inch print, but that’s another issue for another day.
Have I proved anything? Only that I am now satisfied that the GX1 is a really credible performer and fast becoming my camera of choice.
In a previous blog (Don’t Let Noise Kill Your Images) I wrote about the steps you could take to minimise noise in your images. It’s long been recognised that cameras with small sensors have higher levels of noise than they would if they had a larger sensor as a result of packing more pixels into less space. Whilst there have been great advances in this area, it can still be a problem.
If I look at the images produced by my GX1 (16Mpixels) and compare these to images shot on my GF1 (12Mpixels) both of which have the same sized sensor, I can see real improvements in the GX1 both at base and higher ISO levels. I would say that I am examining the images in great detail for any trace of noise in case you think these cameras are poor performers. To give you something to compare against, when I do this for check with my 5D MkII I can also pick out noise at ISO100 in dark areas and in the Blue channel. So what do you do if you find you can’t avoid capturing noisy images?
In the past I have used a noise reduction tool called Neat Image but to be honest it’s quite a lengthy process to get good results and whilst it has a batch mode, I prefer to fine tune the software to each image (hardly a lightweight processing workflow). This weekend however I decided to download a trial of DeNoise from Topaz labs and the Noise reduction plug in from PhotoWiz (I already use Contrast Master, B&W Styler and Focal Blade plug-in and rate these highly). The results were a bit of a surprise.
I found both solutions did a better job than Neat Image however the PhotoWiz product took some time to process my sample image, something I want to avoid. Comparing this with the performance of Topaz Labs DeNoise solution I found a huge difference. DeNoise was incredibly fast to process my image but it also gave the cleanest and most lifelike results. Finding the right level of reduction was as simple as moving one slider but it was then possible to further fine tune the results. It gave me lots of control in an interface that was very easy and fast to use.
I want to experiment further before purchasing, but early results look very promising.
The Lightweight Portfolio Challenge
New York, March 2011
Copyright: Robin Whalley 2011
I was recently reading about Haiku which is a form of Japanese poetry. The objective is to describe an idea in just 17 syllables in three lines (there is a bit more to it but this is the key point for me). As you can imagine, this is a huge undertaking if you are to convey the idea effectively with accuracy, beauty and completeness. It got me wondering if there were any parallels that we could draw as Lightweight photographers.
The idea I came up with is the Lightweight Portfolio Challenge.
The idea is simple; you have to create a portfolio of 5 images that completely express a subject or an idea:
- The portfolio must contain 5 images
- The images must explore the idea as fully and completely as possible
- All the images in the portfolio should work together with no image standing out from the others
You might also want to consider setting yourself some additional rules around this to make it Lightweight and prevent it from expanding into a huge task:
- Are you going to set a timeframe on how long you have to complete this exercise e.g. 1 month from the start?
- Will you include images from your archives or must the work be new?
- Consider restricting yourself to just one camera or even just one lens.
- Only have one of these projects on the go at any one time. It’s much better to have the focus.
Initially you might want to try to create such a body of work from images you already have. This will help you understand just how difficult the exercise can be. Once you are happy with this approach you can move on to shoot new material.
Now, once you are proficient in producing these mini portfolios consider broadening either the theme you are trying to represent in 5 images. What would you say if I asked you to produce such a set of 5 images which were to represent your work for the current year and that it would be only these 5 images that you could publish or share?
How about the 5 images that would represent your life’s work and would be how you want to be remembered as a Photographer?
Copyright: Robin Whalley 2012 www.thelightweightphotographer.com
See more of my work at www.lenscraft.co.uk
Creating a Lightweight Photoshop Workflow
Rough seas, Norway, March 2012
Copyright: Robin Whalley 2012
Last night I was the guest speaker at a Camera Club; something that I like to do a number of times each year in order to share my knowledge, but also to pick up information from others. The event went well and the presentation I delivered was entirely new, not only in content but in approach. The real innovation for the session was that I demonstrated how I use Photoshop to craft my images.
What became apparent from talking to people after the event was that my workflow in Photoshop was quite minimal yet achieved a lot. You see I have standardised my approach to be adequate for around 95% of my images and most of my changes can be applied in just a few minutes. Sure some of the better images I will spend longer on, but only if the additional effort is justified.
Here then is my approach:
- Clean up the image by removing dust spots etc on a new empty layer. I do however try to keep my sensor clean by using an Arctic Butterfly brush so that any cleanup work is minimal.
- Use a Curves adjustment layer to adjust brightness
- Use a curves adjustment layer to adjust contract
- Use a saturation layer to adjust saturation
- Add a new empty layer and stroke the image to create a frame
Whilst I didn’t demonstrate it last night I have created an action in Photoshop to automate the addition of the above layers so that all I need do is activate the action using Ctrl-F3 (which is the key combination I have assigned).
Consider also that all my adjustments are on layers which give rise to a number of benefits:
- I can turn the layer on or off to hide the effect
- I can adjust the opacity of the layer to control how strong its effect is on the image
- Each layer has a blending mode that you can use to make further enhancements
- Each layer has a layer mask attached to it which I can use to target the adjustment onto a specific area of the image
Once I am happy with the image I create a new consolidated version from all the layers by pressing Shift-Ctrl-Alt-E. I then sharpen this for output e.g. for paper or for the screen. My finished image is then saved as a PSD file with all the layers remaining in place. Again this gives rise to timesaving benefits:
- Each time I output the image to new media I can create a version of the image with the correct level of sharpening and simply turn off the other layers
- I can create modified versions of the same image e.g. versions with and without the frame
- If in a couple of weeks I decide I was a little heavy handed with one of the adjustments e.g. saturation, I can simply adjust the opacity of the relevant layer to control the effect
- If I later decide to work on the image further I have all my earlier layers already in place and ready to enhance further
By following my approach I find that I minimise the total time I spend on each image whilst achieving good results.
Don’t Let Noise Kill Your Images
Cornwall, October 2011
Copyright: Robin Whalley 2011
One of the complaints that has dogged small and compact cameras in the past is that they are prone to noise and noise can kill images dead. Unfortunately, whilst advances have been made, these cameras by virtue of their smaller sensors can still have problems. So what to do?
The first thing to realise is that not all image noise is the same or has the same negative impact on the final image. Colour noise appears as random colour speckles in the image and can result in images that appear “muddy” and ultimately lose definition and detail. This noise is quite ugly and should be removed whilst trying to do as little damage as possible to the image.
Luminosity noise appears as speckles of light and dark areas on the image and resembles film grain. If you remove all of this noise your images can look a little “plastic” like. If you have too much of this noise in your images you can lose the definition and they don’t appear sharp. But if you have a small amount of this noise it can actual make your images appear more detailed and sharper when printed.
The best approach is not to try to remove the noise from your images but to avoid introducing the noise in the first place. And as smaller sensor lightweight cameras have smaller sensors that are more susceptible to noise, you need to consider some of these strategies:
- Keep ISO’s low. Don’t rely on having your camera ISO set to auto. Yes it will increase the ISO when light is low to avoid slow shutter speeds but it will probably do this much too early.
- Use your camera in Aperture Priority mode where you select the aperture to determine the depth of field and the camera calculates the shutter speed. See next point for why.
- Consider how much depth of field you really need and what is the widest aperture to allow this. With small sensor cameras you will find you can use a much wider aperture to achieve a given depth of field. A wider aperture means more light reaching the sensor which means faster shutter speeds so you can use lower ISO settings.
- Shoot in RAW. You can then take control of the level of noise reduction applied to the final image. With most RAW converters you can even apply different levels of colour and luminance noise reductions to suit your image.
- Use a monopod in low light. This allows you to use slower shutter speeds without the need to increase the ISO.
- Expose the image to the right in the histogram. To use this approach you really need to be shooting RAW format but the technique involves slightly overexposing your images by perhaps 2/3 of a stop.
- Try to avoid long exposures if possible. Over about a second the sensor will warm up and create noise.
Whilst cameras have improved greatly in their handling of noise the above techniques are still valid and will help you create a quality image.
Are Compact Cameras up to the Job
Wasdale Head, January 2011
Copyright: Robin Whalley 2011
I was having a discussion with an acquaintance the other day who was arguing with me that Compact Cameras are just not up to the job of taking good quality images. His argument revolved around his own experience with a compact digital, where the results were “mushy” and much of the detail had been lost. Before we get into why his argument is flawed, let’s take a moment to look at what being “up to the job” actually means.
If we shoot an image with a digital compact camera and shoot a second image with comparable settings on a DSLR, then the DLSR image will almost certainly be the better quality. This does not however mean the image from the compact camera isn’t up to the job as this would depend on the intended use of the image. If the image will only be used to produce a 6×4 inch print or perhaps a low resolution image for a web site, then there is no advantage to be had with the DSLR. If however you want to produce a 30×20 inch fine art print then the compact camera is unlikely to be up to the job.
But what about a 35mm film compact camera? It’s quite possible you would be able to take a great image which could produce an image of this size. The person I was discussing this with had somehow muddled up his experience of digital photography with compact cameras in general and had forgotten entirely about film. Looking further at his argument I was also able to identify the camera he had been using was a cheap consumer compact and not one aimed at the photo pro/enthusiast. This is likely to make a big difference in that the lens might be of a lower quality, the camera sensor is likely to be smaller and the image is probably stored as a JPG rather than RAW. Let’s explore each of these in turn.
Lens quality, even on a compact camera is very important. Lenses must be sharp across the entire zoom range and display little diffraction. The ability of the lens to resolve detail is also an important factor. If the lens won’t resolve sufficient detail or isn’t sharp, the images produced will seem to be “mushy”. Now an expensive compact camera isn’t a guarantee of a good lens but a quality lens costs money and you can be sure no camera manufacturer will sell a camera with an expensive lens cheaply.
Next we have the subject of sensor size and whilst even high end compact cameras have small sensors in comparison to DSLR’s or Micro 4/3 cameras, they are larger than the sensors in cheap compact cameras. Add to this the tendency for cheaper compacts to be of a higher resolution e.g. 14Mpixels rather than the 10Mpixel you find on most high end compacts, you end up with a camera that produces images with a lot of noise. The solution most of the manufacturers have to this problem is to apply a lot of noise reduction, which is simply the blurring of the image to remove the appearance of noise. The result is “mushy” images.
Finally, the images tend to be recorded in JPG once the noise reduction has been applied. Unfortunately not every image needs the same level of noise reduction and often too much is applied. A much better approach is to capture your images in RAW format and then apply the noise reduction in software as part of the conversion process. This allows you to make judgements about how much noise should be left in the image with a trade off against sharpness.
So, to prove my point, look at the above image. This was captured on my LX5 compact camera in RAW format before process in Lightroom. The resulting image was then resized in order to produce a A3+ print at 300dpi. Is it sharp? Yes. Does it contain lots of crisp detail? You bet it does. In fact, here is a small section of the image shown at 100% so you can judge for yourself, and keep in mind a print from this will be even more impressive once sharpened.
Unfortunately I still don’t think I convinced my acquaintance.