What do Alamy and other libraries need in quality terms?

From RAW conversion to image editing and printing
David Kilpatrick
Site Admin
Posts: 5985
Joined: Sat May 19, 2007 1:14 pm
Location: Kelso, Scotland
Contact:

What do Alamy and other libraries need in quality terms?

Unread post by David Kilpatrick »

Today, a would-be Alamy image provider looked me up and made a telephone call to ask if I could help with her rejection from passing their initial quality control. She had made the mistake of trying to show them wildlife images which were not equivalent to normal photography, as far as assessing quality goes. I agreed to help and she sent me one image, which I will not re-post. This image was from a Canon 7D with 100-400mm lens, a recipe for poor quality if ever I knew one, but it was also very noisy at ISO 1000, soft and coarse with oversharpening.

Anyway, to respond I took a look at something comparable (meerkat not bird) and sent her this reply.

--------------------------

Dear XXXXX, the image you attached appears to be a small clip of a 200mm shot taken at ISO 1000 and f/8. The 7D really can't shoot a clean 1000, and the Canon 100-400mm is not particularly sharp at f/8, and there's also not a great deal of depth of field at f/8 and 200mm on 1.6X. But the processing has destroyed all the detail and produce very large grain noise - the whole image looks as if it has been unsharp masked, there is a white glowing line especially to the left (as we look) of the bird's neck and this sharpening halo will get you a rejection from Alamy even if the general noise, unsharpness, and lack of any fine detail do not.

Here's an image I rejected for Alamy last month, only because it has no point at all (in order to sell, animal and bird shots must either be of a rare species or have some great pictorial merit - humour, anthropomorphic behaviour, great light and setting, action, cuteness, expression, attitude - whatever). However having processed it I will probably upload it. It required better interaction and relationship between the two meerkats to have any kind of story for the viewer.

This is taken using the Sony NEX-7 with a fairly low-cost super zoom, the Tamron 18-200mm, at 200mm. The Sony is 24 megapixels and a slightly larger sensor than the Canon 7D, and like the 7D it needs careful processing to avoid noise at higher ISO settings. Generally, the Sony 24 megapixels have better fine detail and colour than the 7D but that is to be expected as the sensor is two or three years later in development. Exposure 1/320th at f/9 (an optimum aperture for just the right depth of field and soft background, with good sharpness) at ISO 800 (the fastest setting I am happy to use on the camera).
meerkatshotNEX7.jpg
meerkatshotNEX7.jpg (300.56 KiB) Viewed 6017 times
This is the full picture just at 600 x 900 pixels to show the field of view. I have since sold the NEX 7 as I was not happy with the accuracy and consistency of focusing when working fast. A static animal subject like this is just about its limit. I use the Alpha 77 which has much faster focusing a DSLR lenses, and I've gone back to using a NEX-5n which is only 16 megapixels but really great for low noise stuff like gigs, interiors, evening and night shots - the times I need a more compact pocketable sort of camera.

Below, is a 100% clip from the FULL size (6000 x 4000 pixel) file processed using Adobe Camera Raw. Parametric tone curve Strong Contrast; Sharpen 25, 0.5, 25, 0; Noise 50, 50, 0; Colour Noise 50, 50. No further noise reduction, sharpening or other processing applied.
meerkat-100%clip-clip.jpg
meerkat-100%clip-clip.jpg (304.34 KiB) Viewed 6017 times
Compare this to your bird. While I have not used any unsharp mask, I have used normal levels of capture sharpening, but very heavy noise reduction in raw conversion to counteract the ISO 800 noise levels in the blurred areas - this makes the fur look quite soft, but you can still distinguish individual single hairs. This would almost certainly pass Alamy QC, as they understand the difference between artificial sharpening and actual resolved detail in the image. What matters is stuff like being able to see, clearly, the small black single whiskers from the nose. This image will take sharpening and enhancement for printing without looking bad. Your image could not be processed any further, it already looks like a very unsharp shot which has been rescued to look OK for web use.

But I often will take an image like this, judge that it still looks slightly less sharp than my ideal because really I should have been using a top grade lens not a zoom, and working at ISO 100 not 800. So I will resize the image in Photoshop to 2400 x 3600, 24.7MB or thereabouts, for submission. A 100% clip from the resized image looks like this:
meerkat-24MB-100%-clip.jpg
meerkat-24MB-100%-clip.jpg (321.48 KiB) Viewed 6017 times
That is more like the level of sharp detail I expect to deliver for stock photography, no matter what the subject. I know that the reduction in file size means that those whiskers are now much less visible (technically the full size image is much better) but the overall impression of sharpness, and the degree of image noise, meets picture library standards.

When I asked if you could see the fronds of the feathers I did not mean the overall feather, I meant the tiny textural details of the individual strands of each feather. Unless you are shooting sharp enough to show that, you are only meeting internet bird photo standards, not the standards of current wildlife photography. Of course, if you get interesting behaviour or a unique shot in flight, fine detail sharpness is not so vital but for any static study of a bird you must meet the standard of 'does this look almost like I am looking at the real thing on the screen and not at a computer screen?'.

To pass Alamy QC, just photograph something other than wildlife. Shoot something like boats, buildings, objects. Here is a 100% clip from a 6000 x 4000 pixel image from the NEX-7, again using the Tamron 18-200mm, also at f/9 (optimum aperture for this lens) but at ISO 100 and with about half the overall level of noise reduction:
sharp-hardedgesubject.jpg
sharp-hardedgesubject.jpg (195.29 KiB) Viewed 6017 times
This is about the level of sharpness and detail your images should achieve at normal shooting size to be acceptable for publication and general stock sale, and it's what Alamy will routinely expect unless it is very clear from the shot that you were working at night, in unusual conditions etc. Anything well-lit, in sunshine, should be no worse than this.

David
User avatar
pakodominguez
Minister with Portfolio
Posts: 2306
Joined: Tue May 22, 2007 5:38 pm
Location: NYC
Contact:

Re: What do Alamy and other libraries need in quality terms?

Unread post by pakodominguez »

Hi David,
What did she answered?
BTW, Are you considering RX100 files for Alamy?
Pako
------------
http://www.pakodominguez.photo/blog" onclick="window.open(this.href);return false;
David Kilpatrick
Site Admin
Posts: 5985
Joined: Sat May 19, 2007 1:14 pm
Location: Kelso, Scotland
Contact:

Re: What do Alamy and other libraries need in quality terms?

Unread post by David Kilpatrick »

I'm sure some RX100 images will be perfect for Alamy but I'm concerned by levels of flare/glow from very brightly lit white objects in shot. It's not easy to control or retouch these, and I think that if Alamy see any of this, they will ban the camera.

No reply yet...

David
Post Reply

Who is online

Users browsing this forum: No registered users and 4 guests