DPChallenge: A Digital Photography Contest You are not logged in. (log in or register) 
 

DPChallenge Forums >> Photography Discussion >> Sharpness / Crispness / Focus
Pages:  
Showing posts 51 - 65 of 65, (reverse)
AuthorThread
 11/06/2006 03:22:32 PM · #51
Remember, there are two aspects of sharpness, call them "resolution" and "acutance".

Resolution is given by lines-per-inch or some similar measure and is a product of the lens and sensor system. Once the picture is taken, this cannot be made better.

Acutance is the "edge contrast", the perceived sharpness between two different colors. For example, a black line on white may look very sharp, but a red line on black may not look sharp at all. Some people talk about one lens being more "contrasty" than another, so the lens may matter. The type of lighting (front, back, raking) may affect this also.

Basically, USM increases the Acutance. It can't increase the Resolution. It may be better to think of USM as improving the "local contrast" rather than sharpening the image.

11/17/2006 04:08:48 PM · #52




Both of these images received comments that indicated viewers were not pleased with the "sharpness" or possibly the focus. While I'm embarrassed as heck by the score of BOTH of these, I would like constructive advice on what to do to prevent such "bad results" in the future. Obviously a lot of things about my photography SUCK, but these little things, well, maybe I can do SOMETHING about them and help a little.

Then again, I know it's pretty bad :)

11/17/2006 05:29:16 PM · #53
Originally posted by nards656:


One possibility on this image (besides simply needing an additional application of USM) would be the original focus of the image. Here's what I'm thinking:

Your EXIF data says you were using a Canon 300D, with a 135mm lens, at f/5.6. What it doesn't say is the subject distance. So let's make a guess that you were only 6 feet away. If you plug those values into this online depth of field calculator you will find that your total depth of field (distance that is "in focus") would be 0.12 ft. That's not a lot! If you were standing there hand holding the camera and wobbled back and forth ... or if there was a gentle breeze moving the objects back and forth, you could easily move outside of the DOF range in between the time that the focus is locked and when the shutter is opened.

There is also the possibility that f/5.6 is on the soft end of your lens. Usually a lens is at its sharpest when it is stopped down a couple of stops from wide open. But I see that your lens is wide open at f/5.6 when you're at 135mm. So the two thoughts here would be: don't go fully telephoto (100mm instead of 135mm) or stop down the lens to f/8 and bump the ISO if you need to keep the shutter speed up.

And again... back to the USM thing. If you are shooting raw, then the image doesn't have the built-in sharpening that you get on jpegs. So you have to more aggressively sharpen a raw file than you would a jpeg.


11/17/2006 05:32:22 PM · #54
I often will comment on sharpness if feel it benefits/hampers the overall feel of a shot for me. Not sure how this is perceived but it comes from a place of wanting to express how the shot is coming thru to me. But, while I agree that sharpness & focus has much to do with the quality of lenses, it also has MUCH to do with lighting & other factors that are well withint he control of the fotog.
 11/21/2006 12:03:51 PM · #55
Originally posted by Rooster:
...while I agree that sharpness & focus has much to do with the quality of lenses, it also has MUCH to do with lighting & other factors that are well withint he control of the fotog.


Those other factors are what I'm trying to get a solid grip on here. Can you (or anyone else) elaborate?
 11/21/2006 12:26:37 PM · #56
Originally posted by nards656:




Both of these images received comments that indicated viewers were not pleased with the "sharpness" or possibly the focus. While I'm embarrassed as heck by the score of BOTH of these, I would like constructive advice on what to do to prevent such "bad results" in the future. Obviously a lot of things about my photography SUCK, but these little things, well, maybe I can do SOMETHING about them and help a little.

Then again, I know it's pretty bad :)


Looking at the glass one, it doesn't strike me as a 'post processing' issue. It doesn't look like it was sharp when you captured it.

Let's start with some basics.

Was it shot on a tripod ?
Where/ how did you focus on the crystals ?
Did you focus then recompose?

Message edited by author 2006-11-21 12:27:37.
 11/21/2006 12:36:33 PM · #57
Originally posted by Gordon:

Looking at the glass one, it doesn't strike me as a 'post processing' issue. It doesn't look like it was sharp when you captured it.

Let's start with some basics.

Was it shot on a tripod ?
Where/ how did you focus on the crystals ?
Did you focus then recompose?


No tripod, bounced flash against wall to the left.

I believe the focus was AF on the center crystal. That's a memory issue at this point and it's been about 3 weeks, I think.

I don't believe I FTRed, but again I could be mistaken.

Am I trusting the flash too much? Is this a motion problem?

I just studied the EXIF and I've got to do some investigation. It almost looks like I shot this in Manual Focus, and that's a pretty stupid thing to do in this case. If that's true, I have no further questions regarding THIS image. It would appear that I'm a doofus moron :)

The other image, I still don't quite know about. I think perhaps I just focused on the wrong place in the picture????
 11/21/2006 12:45:04 PM · #58
Originally posted by nards656:
Originally posted by Gordon:

Looking at the glass one, it doesn't strike me as a 'post processing' issue. It doesn't look like it was sharp when you captured it.

Let's start with some basics.

Was it shot on a tripod ?
Where/ how did you focus on the crystals ?
Did you focus then recompose?


No tripod, bounced flash against wall to the left.

I believe the focus was AF on the center crystal. That's a memory issue at this point and it's been about 3 weeks, I think.

I don't believe I FTRed, but again I could be mistaken.

Am I trusting the flash too much? Is this a motion problem?

I just studied the EXIF and I've got to do some investigation. It almost looks like I shot this in Manual Focus, and that's a pretty stupid thing to do in this case. If that's true, I have no further questions regarding THIS image. It would appear that I'm a doofus moron :)

The other image, I still don't quite know about. I think perhaps I just focused on the wrong place in the picture????


So a couple of things.
No tripod. 135mm (175mm ish effective focal length)
1/125s. Too slow to handhold effectively at that focal length.
Some people claim to be able to handhold a lot slower than 1(2*focal length) my opinion is, they don't look closely enough at the end results.

For a predictably sharp handheld shot, you'd need to be at least shooting at 1/250s. As you are working very close in to the subject, 1/500s would have been pushing your luck for a sharp result. Use a tripod for this sort of effectively macro work. It is that simple.

To convince yourself about this is pretty straightforward. Get a small laser pointer. Hold it in your hand. Point it at a spot on the wall about 3-10 feet away. Now do the same thing but put the laser pointer on a tripod/ rest it on a table. Think about the difference.

Manual focus may or may not also be the problem, as you mentioned.
 11/21/2006 12:47:52 PM · #59
The landscape/ juxtaposition one doesn't particularly look like it has a sharpness problem. It is sharp, where you focused. But you've shot it at a fairly close to wide open aperture, so you don't have much depth of field. The foreground is in focus, the background mountain/ trees fall out of the depth of field at f6.3.

You'd need to stop down/ use a tripod to get more depth. While you are at it, I'd consider shooting at a time of day when the light is more interesting, creating lower side light, to build texture (depth) and as a result, the appearance of sharpness.
 11/21/2006 12:53:04 PM · #60
Originally posted by Gordon:
The landscape/ juxtaposition one doesn't particularly look like it has a sharpness problem. It is sharp, where you focused. But you've shot it at a fairly close to wide open aperture, so you don't have much depth of field. The foreground is in focus, the background mountain/ trees fall out of the depth of field at f6.3.

You'd need to stop down/ use a tripod to get more depth. While you are at it, I'd consider shooting at a time of day when the light is more interesting, creating lower side light, to build texture (depth) and as a result, the appearance of sharpness.


Thanks. While I already know all that, I somehow had convinced myself that I was "getting by" without always using the tripod. 'Bout time to be kicked back to reality and ALWAYS use the darn thing unless I can't.

As to lighting that particular landscape shot... Even though it's shot from my driveway, it's rarely lit well when I'm there, so my choice of lighting is more driven by opportunity than by good light. Good point, though. Thanks again.
 12/31/2006 01:41:18 PM · #61
Originally posted by hankk:
Remember, there are two aspects of sharpness, call them "resolution" and "acutance".

Resolution is given by lines-per-inch or some similar measure and is a product of the lens and sensor system. Once the picture is taken, this cannot be made better.

Acutance is the "edge contrast", the perceived sharpness between two different colors. For example, a black line on white may look very sharp, but a red line on black may not look sharp at all. Some people talk about one lens being more "contrasty" than another, so the lens may matter. The type of lighting (front, back, raking) may affect this also.

Basically, USM increases the Acutance. It can't increase the Resolution. It may be better to think of USM as improving the "local contrast" rather than sharpening the image.


I just notice when I shoot raw my dimensions are 3072 X 2048, 240 DPI at 16 bit, when that shot is converted to tiff it is 3072 X 2048, 350 DPI (converted by DPP) at 16 bit. Saving to jpeg (after switching to 8 bit mode) the image is 3072 X 2048 , 350 DPI at 16 bit .

If I shoot in jpeg mode to begin with I get 3072 X 2048 , 180 DPI at 8 bit. Now most folks say you dont need more that 200 DPI for nice prints and therfore shoot in JPEG. Shooting JPEG I am right off giving up the 20 DPI I don't need more than.

So.. I wonder, is it a 20 DPI trade off for time when folks recommend shooting jpegs (180 DPI for my cam) for weddings vx raw (240 DPI) ? I can see the process of converting 800 raw shots to tiff or psd then to jpeg would be a good chunk of time for the 20 dpi improvement "if" it is not noticed. My wordy post is to question wheather that 20 dpi loss is noticible and is the 350 DPI jpeg (a result of raw>tiff>jpeg) a better product for labs than a jpeg straight out of camera? This does relate to sharpness for me.

Message edited by author 2006-12-31 13:44:11.
 12/31/2006 01:54:42 PM · #62
Ignore the DPI number, it means nothing for web display. Literally nothing. The shot is 3072x2048, that's all you need to know.
Now for printing, different story, but still, the pixel dimensions determine how much information (detail) can be there, not the DPI. All the DPI does is determine what size the final image will be. For instance, a 2000px by 3000px image at 100 dpi will print at 20x30 inches, whereas at 200dpi it will print at 10x15 inches. Still the same image, with the same information content. Make sense?
 12/31/2006 02:16:06 PM · #63
Originally posted by stdavidson:
Originally posted by Tlemetry:
Does anyone here use selective sharpening techniques?

I do. It will come in one of two forms:

1-Add a layer mask to a sharpened layer and paint with a black or grey brush to reduce sharpening to specific areas.

2-Use the sharpen tool (very gingerly) to increase local sharpening in selected areas.

For DPC purposes these techniques are allowed only in 'advanced' editing.


This is the technique I use as well although I do it just a bit differently.
I'll duplicate the layer and add the unsharp mask and watch the point of interest until it is sharpened to an acceptable level. I will then gradually erase at different strengths the sharpening effect. The important thing to note is that most times you get a comment about oversharpening, it is because the outer edge of any subject shows the sharpening done with UM. Either by a halo or by jagged lines etc. What I'll do is take my soft round eraser brush while still in the layer, and at a fairly low strength using a fairly large brush, I'll go around the outer edge of the obvious oversharpened parts and tone them down. You'll notice that the eraser brush is gradual so you don't need to run the brush over the edges, merely bring them close and you'll see the changes as you go. Hope that made sense.
 12/31/2006 02:31:06 PM · #64
Originally posted by kirbic:
Ignore the DPI number, it means nothing for web display. Literally nothing. The shot is 3072x2048, that's all you need to know.
Now for printing, different story, but still, the pixel dimensions determine how much information (detail) can be there, not the DPI. All the DPI does is determine what size the final image will be. For instance, a 2000px by 3000px image at 100 dpi will print at 20x30 inches, whereas at 200dpi it will print at 10x15 inches. Still the same image, with the same information content. Make sense?


Yes..no..ah.. I feel like I am trying to learn binary math again. I think I get the concept but I have to screw the lid down on it or it get's away from me. 100 to 200 DPI results in size reduction so if I get the math.. a 2000px by 3000px image at 240 DPI would be an 8.3x12.5 inch print and at 250 DPI an 8x12 print.
 12/31/2006 02:44:59 PM · #65
Yup, you got it! Screw the lid on tight!
Pages:  
Current Server Time: 08/02/2025 01:44:47 PM

Please log in or register to post to the forums.


Home - Challenges - Community - League - Photos - Cameras - Lenses - Learn - Help - Terms of Use - Privacy - Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 08/02/2025 01:44:47 PM EDT.