Cropping terminology has me confused.

[JBW]

Hi all, I am as is usual for me, confused over terminology. I crop all of my shots. I do this to re-position, highlight, or must get rid of what I don't like. I have my camera set to take the largest shot possible. This means that even after some serious cropping I usually have a larger shot than I post. Naturally I downsize to 1600px on the long side.

I often end up with good detail and sharpness.

Why then do I constantly read about cropping being a bad thing for detail / sharpness?
Brian

[James G]

Brian, I think this may be a syntax related issue in part.
Cropping or resizing?

If you simply crop but do not resize (upwards?) then no detail/ sharpness in the basic image is affected. BUT if after cropping you resize upwards to 'display' on full screen/print you effectively have to 'stretch' the reduced number of pixels/points.
This is done by filling the gaps as the expansion progresses with similar pixels, but that inevitably results in less sharpness/softer contrast transitions.

If you imagine your image as a flat rubber sheet, think about cutting about 50% of it away (ie cropping). Now imagine pulling the sheet back to its original size and the way the its clarity/sharpness will be affected. (You can 'simulate' this for real by drawing a cartoon on a deflated balloon and then inflating it!)

Hope this helps

James

[JBW]

Brian, I think this may be a syntax related issue in part.
Cropping or resizing?

If you simply crop but do not resize (upwards?) then no detail/ sharpness in the basic image is affected. BUT if after cropping you resize upwards to 'display' on full screen/print you effectively have to 'stretch' the reduced number of pixels/points.
This is done by filling the gaps as the expansion progresses with similar pixels, but that inevitably results in less sharpness/softer contrast transitions.

If you imagine your image as a flat rubber sheet, think about cutting about 50% of it away (ie cropping). Now imagine pulling the sheet back to its original size and the way the its clarity/sharpness will be affected. (You can 'simulate' this for real by drawing a cartoon on a deflated balloon and then inflating it!)

Hope this helps

James

Yes it does. Cropping isn't the villain, up sizing is!

[xpatUSA]

Yes it does. Cropping isn't the villain, up sizing is!

Absolutely, Brian. In fact, down-sizing an image increases sharpness and local contrast just as you have found, unless the down-sizing method compensates for the effect, e.g. Adobe's 'bicubic smoother'.

The effect can be seen clearly in images with a sharpish high-contrast edge, which gets the "jaggies" after down-sizing. Or in images with repetitive patterns like roof-tiles or fence-posts where, after down-sizing, moiré appears where there was none before.

[NorthernFocus]

All else being equal, more pixels equals more detail. The point is that for a given scene, you're better off shooting it full frame than if the scene only fills part of the frame and you crop. It is more important if you are making large prints where the detail will be evident. Not as big of a deal when downsizing for web.

[Thlayle]

Hi Brian,

When I first jumped into digital photography, I was very lax about how I framed my shots. I was quite taken with the ability to control the post-processing myself, being "free at last" from the constraints of the darkrooms managed by the labs I previously sent ALL of my shots to. I was also impressed with the resolution of the camera I was using (a 6 megapixel --woo hoo!) so I cared little about my exact framing.

It really wasn't until I began a more serious interest in completing the process, that is...taking the shot all the way to a printed result, that I saw how important it was to do my capture of the shot as close as possible to my vision of what the final image should be.

So I will say that in addition to the great responses already posted, part of the trick is 'having the eye' -- a better sense of knowing what you want from your image before pressing the button. Thus, cropping or rescaling an image becomes less of a concern.

And here I throw in a great quote from an article I read recently:

"To expose film is not quite to photograph, and the photographer who does not consider his finished pictures is like a pianist who plays only on a silent keyboard. In the absence of proof, mistakes multiply, craft becomes theory, and good thinking passes for art."

John Szarkowski, “Winogrand: Figments from the Real World” MoMA, 1988

[James G]

Brian +1 to Randy's comment about developing

-- a better sense of knowing what you want from your image before pressing the button

We all developed shooting habits that helped build our confidence when starting out in photography. The problem is that we often become somewhat locked into them and fail to recognise when it's time to either abandon, or modify them.

It probably means moving outside your current comfort zone, but I'd suggest you try changing your current captureprocess slightly and focus on maximising the subject in the viewfinder before capture. Framing will follow on naturally. This will result (hopefully) in an optimally large image (re pixels), which in turn will limit what you might then lose when tweaking the final framing during post processing.

just a thought..... James

[Manfred M]

Brian - let me throw a slightly different spin on this for you.

I suspect that your laptop display is no better than about 1366 x 768 pixels (that is what the best resolution on mine is). That works out to about 1MP. So in theory, you could post a 1MP image and you could crop everything else out. You could probably get away with a bit less, as upsampling (which is adding data to make a small image scale up) won't have a lot of visual impact.

On the other hand, I have often written that the main reason for large MP cameras are really important to people who make prints. The native resolution of my Epson 3880 printer is 360 dpi which means I can print a 20.4" x 13.6" image without upsampling. So to post on the internet, my camera is total overkill and I can crop to my heart's content without anyone being any the wiser. For printing, I can still crop. In fact I have to, as the sensor format and paper formats are not identical, so I either have to trim the full-sized print to size of have some margins on the long side to crop off to get the image to fit.

So you can crop all you want, Brian and as long as you post to the web, this is unlikely to make much difference to you.

There is a big BUT, though. If you shoot the way you are right now and get that killer image that you (or someone else) wants on their wall, you could have a serious problem with image quality and you might not be able to produce an acceptable print. If you adopt good shooting practices and do your composition in camera rather at edit, they you won't ever run into this issue. In my work flow, I assume any shot I take might be printed to a large size. Most of the work I post here are either uncropped or cropped to standard print size (depending on how I was feeling that day).

Adopting good shooting practices is never a bad thing.

[jcuknz]

There is another point in all this and that is what and how you save. If I re-size for web-use the result is saved in a folder under the site's name [ CiC likes 1600 pixel files while Photonet has a limit of 700 if you want the image to show with the text written].
One of my reasons for never 'save'ing but always 'save as' so I have control of the process and do not over-write where the file came from. destroying the original, in case I have a great shot somebody wants on their wall
While the original is available un touched and a work-up elsewhere... so I could have four versions of a photo available to me depending on their purpose.
Storage is cheap these days unlike when I started and compressed files because storage was so costly

[rpcrowe]

Brian,

I suggest that you save an uncropped and unsharpened (except for entry sharpening) image as a "MASTER FILE" as a Photoshop Document (PSD). That way you will have the best of all worlds...

1. The original RAW images with no adjustments.

2. The Master image which you have adjusted contrast, color and density as well as doing whatever spotting and or other manipulations except: no cropping and no output sharpening.

3. Then you can crop, resize and output sharpen the master image for whatever use you wish to make of the image and save it as a JPEG or a TIFF (whichever you need for printing or submitting for the final use). You will always have the RAW image and the Master File Image to go back to.

I have also found that having a RAW image that has been saved is great when a new editing program or editing style comes down the road. You can work on the old images just the way you would be able to if they had just come out of your camera...

I standardized on this workflow after I needed a number of dog rescue images for a rescue calendar. I had originally shot the images in JPEG (this was years ago, before I began shooting RAW) and I cropped these images and resized them to the dimensions I needed. The cropping and resizing was destructive and my images were not of the correct format use in making a calendar. Naturally, the shots I had previously worked on and destroyed were my BEST shots. So I needed to use second best shots for the calendar...

Saving the partially edited images as Master Files is a little bit of extra work but, I have reaped great benefits from this workflow...

[JBW]

Hi all, I am as is usual for me, confused over terminology. I crop all of my shots. I do this to re-position, highlight, or must get rid of what I don't like. I have my camera set to take the largest shot possible. This means that even after some serious cropping I usually have a larger shot than I post. Naturally I downsize to 1600px on the long side.

I often end up with good detail and sharpness.

Why then do I constantly read about cropping being a bad thing for detail / sharpness?
Brian

I appreciate all of the answers and will answer all later. But for now... I appreciate the theory of filling the screen with the main object but it simply isn't possible for now. I shoot with an 18 - 55 lens and I can only get so close and so big. As an example: This shot started off more than 5000 px on the long side. I was at 55mm and as close as I could get and focus. You can see how much I needed to crop to get a good close up.

[NorthernFocus]

Nothing wrong with doing what you're doing, Brian. It meets your needs right now. And you are seeking to understand which is good. As needs/desires change and within limits of your kit, you'll have the knowledge to change technique if/as necessary.

I view a lot more of your images than I comment on. Often because by the time I look, others have already offered opinions/advice similar to what I'd contribute. But it has been interesting observing your growth and seemingly insatiable interest in photography over the past months. Good on you, brother.

[James G]

Hi Brian, I take your point about being limited in your ability to fill the screen with the 18-55 lens there is not much you can do except crop to get a 'better' composition.

That said, and in the context of your original question, it is worth while pointing out that although upsizing can be detrimental to an image, there are a number of sharpening filters/techniques available which go a long way to mitigate/offset the softening effect of upsizing.

I don't know if you have investigated sharpening in any detail yet, but I seem to remember you have been experimenting with GIMP recently.

You might find the attached link of interest..http://www.gimp.org/tutorials/Smart_Sharpening/

[xpatUSA]

. . . it is worth while pointing out that although upsizing can be detrimental to an image, there are a number of sharpening filters/techniques available which go a long way to mitigate/offset the softening effect of upsizing.

I don't know if you have investigated sharpening in any detail yet, but I seem to remember you have been experimenting with GIMP recently.

You might find the attached link of interest..http://www.gimp.org/tutorials/Smart_Sharpening/

I for one found it very interesting - that kind of processing promises much for difficult images.

It could wean me off RawTherapee which only has global adjustments, good as they are.

[JBW]

Hi Brian, I take your point about being limited in your ability to fill the screen with the 18-55 lens there is not much you can do except crop to get a 'better' composition.

That said, and in the context of your original question, it is worth while pointing out that although upsizing can be detrimental to an image, there are a number of sharpening filters/techniques available which go a long way to mitigate/offset the softening effect of upsizing.

I don't know if you have investigated sharpening in any detail yet, but I seem to remember you have been experimenting with GIMP recently.

You might find the attached link of interest..http://www.gimp.org/tutorials/Smart_Sharpening/

I have been using Gimp for a while and it does have the best sharpen, unsharpen and scaling that I have found so far. If I couold figure out how to get it to edit RAW and particularly Sony RAW it would become the kleft hand to the right hand Capture 1 Sony

[JBW]

I for one found it very interesting - that kind of processing promises much for difficult images.

It could wean me off RawTherapee which only has global adjustments, good as they are.

Ted for what it's worth I have tried, though never mastered RawTherapee and I always went to Gimp for sharpening.

[xpatUSA]

Ted for what it's worth I have tried, though never mastered RawTherapee and I always went to Gimp for sharpening.

I understand. RawTherapee has some awesome sharpening options along with detail contrast controls but sometimes when an entire image is sharpened to perfection it ain't quite what you wanted

I'll be trying out the GIMP shortly but I see a bit of a learning curve ahead . . .

[xpatUSA]

I have been using Gimp for a while and it does have the best sharpen, unsharpen and scaling that I have found so far. If I could figure out how to get it to edit RAW and particularly Sony RAW it would become the left hand to the right hand Capture 1 Sony

I know how you feel about that. I have sought the one-app solution for Sigma X3F files, for literally years. Sigma doesn't help by changing their raw file format for each new type of model - Adobe gave up long ago and I don't blame them!

With the GIMP, I understand that downloading UFRaw as a plug-in for opening raw files lets you open Sony raws.

However, I have tried UFRaw as a stand-alone for Sigma raw files and have not been impressed - based as it is on Dave Coffin's DCraw - it is nowhere near as "good" as Sigma's Photo Pro (SPP) proprietary converter.

So, I'll still be converting with SPP and using other applications such as the GIMP or RawTherapee for final editing.

[James G]

Brian, I use Photoshop and it is the same story I'm afraid. I can do a lot with ACR and can apply a blanket level sharpening and noise reduction to the raw capture, but have to use Photoshop or 3rd party plugins for the final stages.

I suppose I do not mind too much, I've learnt to use layers and luminosity masks along with the selection tool, and can target fine/medium/large detail within the image for attention. It also seems to be the 'right' way, having read around the subject quite a lot.
ie sharpening is best? tackled in 3 stages.
a) initial sharpening of the raw capture
b) sharpening of of specific content for detail/content
c) sharpening for final presentation, Monitor,Web,Print

Since every image presents it's own particular processing issues this 3 stage approach does work best for me.

I'll be trying out the GIMP shortly but I see a bit of a learning curve ahead . . .

Ted, twas ever so.....

[JBW]

Brian, I use Photoshop and it is the same story I'm afraid. I can do a lot with ACR and can apply a blanket level sharpening and noise reduction to the raw capture, but have to use Photoshop or 3rd party plugins for the final stages.

I suppose I do not mind too much, I've learnt to use layers and luminosity masks along with the selection tool, and can target fine/medium/large detail within the image for attention. It also seems to be the 'right' way, having read around the subject quite a lot.
ie sharpening is best? tackled in 3 stages.
a) initial sharpening of the raw capture
b) sharpening of of specific content for detail/content
c) sharpening for final presentation, Monitor,Web,Print

Since every image presents it's own particular processing issues this 3 stage approach does work best for me.

I'll be trying out the GIMP shortly but I see a bit of a learning curve ahead . . .

Ted, twas ever so.....

and the congregation said AMEN