Lens Quality vs User Error

68lbs

68lbs

Suspended / Banned
Messages
5,450
Name
April 2008
Edit My Images
No
Hi all...

Just wondering if there is a any easy way to test whether a lens is soft. I've only been using a dslr for a couple of months and don't really know what level of sharpness (is that a word?) to expect. And hence I don't know whether any lack of clarity is a lens issure, or user error on my part (incorrect aperture, missed focus point, camera shake, etc).

The main reason for me asking is that the two lenses we have are Sigma ones which I believe from what I have read, can have QC issues. Also, I've been looking into Alamy which requires images being upscaled to about 50Mb and I don't think ANY of my photos would be sharp enough to pass their QC - unless I am simply expecting too much.

I would like to get into stock basically because I am new to all this, could do with considerable practice shooting stuff, and kinda think the images are best sat on a few stock sites rather than hidden away on my HD. I am shooting in RAW, pre-processing in Photoshop CS3, making adjustments to exposure, wb, etc and then clicking the little linky under the image and choosing 16bit and upping the image resolution and dimensions, before importing properly into PS CS3. From here I can remove any dust, adjust levels, unsharp mask, etc. But I am still worried about IQ )mainly the sharpness) when viewed at 100%.

Any advice? Anyone submit to Alamy and willing to post a 100% cropped section so I can see what the sharpness 'should' be like?

Thanks in advance
 
Simple.

Get a tripod and set it up with the camera on at about 11am or 2pm in front of a brick wall. Shoot in JPEG to allow the camera to do in camera sharpening, or shoot in RAW and sharpen yourself.
Set the camera to self timer mode, and take a series of shots.

Preferably keep the shutter speed to about 1/50 and the ISO to 100.

Take pictures at different focal lengths and different apertures.
So, for the 10-20, you would ideally want to do the following:
10mm @ f/4
10mm @ f/7.1
10mm @ f/11
10mm @ f/16
10mm @ f/22
15mm @ f/4.5
15mm @ f/8
15mm @ f/11
15mm @ f/16
15mm @ f/22
20mm @ f/5.6
20mm @ f/8
20mm @ f/11
20mm @ f/16
20mm @ f/22

Then, put the photos into photoshop, add sharpening if shot in raw, leave be if shot in JPEG.
Zoom in to 100%, and look at the left, middle, and right. The shot should be sharp throughout, and is most likely to be sharpest at f/8 to f/11.

Do the same for the 24-70.
Alternatively if you don't have tripod, the same test can be achieved with the camera placed on a pillow on a table/chair, and if there isn't enough light you can use the flash/reflectors to fill the light in, bearing in mind that a lens like the 10-20 will have a spot of light where the flash is, with the edges dark.
 
foodpoison said:
Preferably keep the shutter speed to about 1/50 and the ISO to 100.
Click to expand...


Have I read this right?

If I keep the ISO and shutter speed the same the exposure will be all to cock on most.
 
I have never sent images to stock sites, but i would think if you post a 100%crop, then you will get some answers whether people think it is sharp enough or not.

i dont know if alamy charge for sending images, but if you submit a few and see what answers come back from them that might help as well
 
68lbs said:
Have I read this right?

If I keep the ISO and shutter speed the same the exposure will be all to cock on most.
Click to expand...

yes, but you are eliminating human error, this is to check the lens abilities. so using the same settings where possible would stop user error or better conditions
 
If I was a stock reviewer I would refuse this on technical grounds of noise and lack of focus.

This is 100% crop from one I shot for stock but not edited at all yet apart from cropping and changing to .jpg to upload here. To me this crop as it is would be borderline and possibly needs sharpening a fraction and the beak tidying up before submission.

As you used an image that was just about the same colour all over I tried to do the same

chickcrop.jpg
 
You must log in or register to reply here.
Back
Top