Percentage crops definition

[LordV]

I understand what a 100% crop of an image is - ie it is a 1:1 pixel representation of part of the original image with no web size reduction and is often used to show the detail of the original photo, but is there an accepted definition of what for example a 50% crop would be?
eg is this half the picture by area or half the picture by linear dimension giving a quarter of the area of the original picture ?

This question came up in the macro section but would be interested in other people's views.

Brian v.

[C2S]

Interesting question. You already know the following, but:

If you want to halve an image by the linear dimensions, you divide the dimensions by 2, resulting in 25% area of original.
If you want to halve an image by the area, you divide the dimensions by the square root of 2, resulting in 50% area of original.

Personally, if "100% crop" means 1:1 crop, and I was asked to post a "50% crop", I doubt I would start messing with the square roots... I would just halve the pixel dimensions and post a 1:1 crop of the result. But that's my 0.02€.

[Wilt]

Using Paint Shop Pro X, I just looked at an image taken years ago with a 20D...a feature within the shot was displayed at 45mm when the image was at 25%. At 100% view, the same features where 180mm apart. Now you know the fractional value is a linear one.

[LordV]

Thanks for the comments.

Wilt although photosoftware may use a linear percentage scale, is that the accepted basis for people making percentage crops ?
I assume it probably is ?

Brian V.

[Wilt]

Thanks for the comments.

Wilt although photosoftware may use a linear percentage scale, is that the accepted basis for people making percentage crops ?
I assume it probably is ?

Brian V.

" the accepted basis" ??? 'acceptable' by what criteria? I'm failing to understand what your question is about. Earlier you stated, "is there an accepted definition of what for example a 50% crop would be?" and I came back with the response that the percentage is a linear measurement. So if you have a 1280x1024 monitor, at 50% you would see every other pixel for an area that occupies 2560 x 2048 of the original image.

[LordV]

" the accepted basis" ??? 'acceptable' by what criteria? I'm failing to understand what your question is about. Earlier you stated, "is there an accepted definition of what for example a 50% crop would be?" and I came back with the response that the percentage is a linear measurement. So if you have a 1280x1024 monitor, at 50% you would see every other pixel for an area that occupies 2560 x 2048 of the original image.


What I was after whether there was an officially accepted definition of percentage crops. There is certainly some confusion as depending on the camera a 100% crop in the normally used sense is actually somewhere between a 10% and 30% crop approx according to a "linear" definition.
Brian v.

[Radtech1]

Under what circumstance would you use a 50% crop?

In general a 100% crop is to show some technical aspect of the shot at full zoom, and not meant to be a general viewing crop. Otherwise, you would RESIZE to the media size, or CROP to the media aspect ratio. I am trying to think of why an exactly 50% crop would be needed.

Please elucidate.

Rad

[LordV]

Under what circumstance would you use a 50% crop?

In general a 100% crop is to show some technical aspect of the shot at full zoom, and not meant to be a general viewing crop. Otherwise, you would RESIZE to the media size, or CROP to the media aspect ratio. I am trying to think of why an exactly 50% crop would be needed.

Please elucidate.

Rad
It can be fairly common in macro shots to crop and sometimes you are asked how much of a crop it is - just wondered if there was a standard way of describing it. I was only using a 50% crop as an example.
Brian v.

[Radtech1]

It can be fairly common in macro shots to crop and sometimes you are asked how much of a crop it is - just wondered if there was a standard way of describing it. I was only using a 50% crop as an example.
Brian v.

Hmmm. Didn't know that. (Obviously :rolleyes: )

Thanks,

Rad

[tonylong]

I'd say there is no standard "50% crop". It would have no meaning. Like you've said, a 100% crop has a meaning that has nothing to do, actually, with a percentage of the image but rather a mapping to a display. If you try to apply that to smaller percentage numbers it gets pretty silly -- let's see, a 50% crop means 2 image pixels to one display pixel? What earthly reason would there be to think like that?

In practical terms, I either crop "a lot" (for things like wildlife) or "little if any, maybe a little straightening (for stuff I can compose and shoot) or crop to a 4:5 aspect ratio for a lot of things. I don't ever think of cropping at 50% or whatever.

[Wilt]

The crop percentage is simply the ratio of the camera pixels and the monitor pixels, as stated by TonyLong. So Yes, LordV, the percentage is certainly variable to the total pixel count of the camera taking that image!

1280 pixels across the screen (assuming I have a 1280x1024 monitor) is always '100%' whether I start with a 3504 horizontal pixel count image from a 30D, or I start with a 4752 horizontal pixel count image from a 50D

[SkipD]

I'd say there is no standard "50% crop". It would have no meaning. Like you've said, a 100% crop has a meaning that has nothing to do, actually, with a percentage of the image but rather a mapping to a display. If you try to apply that to smaller percentage numbers it gets pretty silly -- let's see, a 50% crop means 2 image pixels to one display pixel? What earthly reason would there be to think like that?

In practical terms, I either crop "a lot" (for things like wildlife) or "little if any, maybe a little straightening (for stuff I can compose and shoot) or crop to a 4:5 aspect ratio for a lot of things. I don't ever think of cropping at 50% or whatever.

The crop percentage is simply the ratio of the camera pixels and the monitor pixels, as stated by TonyLong. So Yes, LordV, the percentage is certainly variable to the total pixel count of the camera taking that image!

1280 pixels across the screen (assuming I have a 1280x1024 monitor) is always '100%' whether I start with a 3504 horizontal pixel count image from a 30D, or I start with a 4752 horizontal pixel count image from a 50DYup - in the context of the "100% crop" that you see in the forum, there is no other value of the number (such as 50% crop) that has any relevant meaning whatsoever.

[timbop]

geez guys, are you being thick or cranky? We throw terms around like "50% crop" all the time, without defining what we mean and expect everyone knows what we are thinking. He's saying if he takes a 3600x2400 pixel image and crops it by removing the pixels outside of an 1800x1200 area (i.e. leaving an image with 1800x1200 pixels), is that generally accepted as a "50% crop" or a 25% crop? Putting it another way, if you remove 6.48MP (75% of the image) is that generally considered a 50% crop.

That is a different issue than downrezing the image to fit on a typical monitor, or to meet the rules on most forums that restrict maximum image dimensions.

The corollary to that, is when you downrez a 3600x2400 image to 900x600 image is that generally considered a 25% reduction (aka crop)? IMHO I would call that a 25% "redux" (again commonly referred to as a "crop").

No need to be snippy, folks

[Wilt]

Cranky? your interpretation. I can assure you I was not cranky at the time of my post! Frankly I haven't heard '50% crop' used in the context of pixels downsampled from a full resolution image.

thickly yours,
--wilt

[LordV]

Thanks for the answers so far.
Timbop- glad you inderstand what I was asking.

It appears that a 100% crop is an accepted crop definition but may actually be a variable percentage of the original depending on sensor size .
The 100% crop definition then has nothing to do with any other cropping we may do to a picture for whatever reason prior to printing, for which there seems to be no accepted defintion.
However as commented near the top of the thread, a percentage based on linear dimension would seem to be a suitable definition for describing how much of the original picture is in a crop.
ie 75% crop of a 12X 8 photo would be 8X 6
50% would be 6X4
25% would be 3X2
or should it be the other way round ? :)
Brian v.

[tonydee]

Alternatively, you could avoid using "% crop" as your units, and phrase it ala "the crop left 50% of pixels", "cropped from 12MP to ~3", "cropped 30% off each dimension", "crop left 20% of original area", "cropped away 80% of the image" etc..

Cheers, Tony

[LordV]

Alternatively, you could avoid using "% crop" as your units, and phrase it ala "the crop left 50% of pixels", "cropped from 12MP to ~3", "cropped 30% off each dimension", "crop left 20% of original area", "cropped away 80% of the image" etc..

Cheers, Tony

That might avoid some of the confusion :)
Brian v.

[Wilt]

However as commented near the top of the thread, a percentage based on linear dimension would seem to be a suitable definition for describing how much of the original picture is in a crop.
ie 75% crop of a 12X 8 photo would be 8X 6
50% would be 6X4
25% would be 3X2
or should it be the other way round ? :)
Brian v.

The problem with that definition is that photos have no inherent size, apart from the original capture (film or sensor). No photo is inherently 12 x8 ...we might choose to make a 12 x8 but we equally as easily choose to make a 10 x 6.7 instead, and we might choose to employ 100% of the original image to make both sizes or we might choose to employ 27% of the original image or 68% or 67.8% or... (you get my drift)

I think the fundamental issue with many digital photographers today, is that they have no context like film shooters did. So they did not see a 24 x 36mm piece of film and choose a section of that image and choose to make some custom size print in the darkroom. With that context in their memory banks, digital as the analogy to that experience is much easier to grasp.

[LordV]

The problem with that definition is that photos have no inherent size, apart from the original capture (film or sensor). No photo is inherently 12 x8 ...we might choose to make a 12 x8 but we equally as easily choose to make a 10 x 6.7 instead, and we might choose to employ 100% of the original image to make both sizes or we might choose to employ 27% of the original image or 68% or 67.8% or... (you get my drift)

I think the fundamental issue with many digital photographers today, is that they have no context like film shooters did. So they did not see a 24 x 36mm piece of film and choose a section of that image and choose to make some custom size print in the darkroom. With that context in their memory banks, digital as the analogy to that experience is much easier to grasp.

Think you are taking the example too literally- it as just an example, but agree if you start making non standard format crops you would be all over the place :)

Brian V.

[Wilt]

The challenge in forum discussions is when figurative descriptions get read by novices and internalized wrongly as literal descriptions...which is why I can get so literal when it is not clear that a discussion thread is couched in figurative space, not literal space. :)

[Radtech1]

Here is what I am wondering - which is why the whole question makes no sense to me - In practice, a 100% crop is "A crop of any portion of a image intended to be viewed at 100% zoom." The primary purpose it to present a "loupe view" of a certain area to assess or demonstrate a characteristic that is too small to be seen at a "normal" viewing zoom, ie, sharpness, noise, c.a., etc.

A crop viewed at 100%.

A 100% crop.

So I STILL don't see what circumstances that you would want to show something at a 50% zoom. It doesn't make sense for two reasons - 1) If presenting an actual "crop", then a portion of the image seen at 50% would be useless because the noise, etc detail would be compromised by the interpolation, and 2) if presenting the whole image, 50% would be too large for most displays.

I think the term "50% crop" is meaningless. ???

Could someone please explain the purpose of a 50% crop to someone who does not shoot macros?

Rad

[Wilt]

Here is what I am wondering - which is why the whole question makes no sense to me - In practice, a 100% crop is "A crop of any portion of a image intended to be viewed at 100% zoom." The primary purpose it to present a "loupe view" of a certain area to assess or demonstrate a characteristic that is too small to be seen at a "normal" viewing zoom, ie, sharpness, noise, c.a., etc.

A crop viewed at 100%.

A 100% crop.

So I STILL don't see what circumstances that you would want to show something at a 50% zoom. It doesn't make sense for two reasons - 1) If presenting an actual "crop", then a portion of the image seen at 50% would be useless because the noise, etc detail would be compromised by the interpolation, and 2) if presenting the whole image, 50% would be too large for most displays.

I think the term "50% crop" is meaningless. ???

Could someone please explain the purpose of a 50% crop to someone who does not shoot macros?

Rad

Frankly I normally ignore the percentage value displayed at the edge of a window while working in post processing. Usually I think, 'I am viewing the whole image' or 'I am viewing part of the image' or 'I am zoomed absurdly close in the image (100%)'. And I zoom to the level required by the edit which I am performing (sometimes individual pixels or small zones, sometimes large sections of the image).

The only value I think there is in seeing '37%' or '50%' would be the awareness of how tightly I am zoomed (no real value to seeing the absolute value of the number per se).

I see no different value specific to doing macro work vs. to non-macro scenes

[Radtech1]

Frankly I normally ignore the percentage value displayed at the edge of a window while working in post processing. Usually I think, 'I am viewing the whole image' or 'I am viewing part of the image' or 'I am zoomed absurdly close in the image (100%)'.

That is how I work as well - um, actually either "Fit to Screen" or "300%", but I don't think the OP was talking about working zooms. The sense I got was that the original topic was regarding presentation. Showing the shot.

Which is why I can't quite wrap my mind around showing a crop of something at 50% zoom. As I said, the interpolation at 50% makes it useless for demonstrating something, and other than that, I don't know.

It sill makes no sense to me that you would zoom to 50%, crop, and present. I am hoping that someone could explain it.

Rad

[tonydee]

Hope Brian will chip in soon, but FWIW - my assumptions re where all this is coming from:

Someone posts a terrific looking macro shot to forum.
1) it looks really sharp, but may not be so at 100%, and/or
2) it look like there's a truckload of magnification, but that may be because it's been cropped heavily and there's not many more pixels than in the (possibly) resized image posted on POTN.

To get a feel for the real quality of the unresized image, forum members will want to know how many megapixels were left after any cropping. To get an idea how much the megapixel potential of the camera was compromised, a percentage may be useful. For example, getting a 12MP shot off a 1DsIII wouldn't be as effective a utilisation of hardware as getting the full 12MP off a 5D. The ratio gives an indication of how well they've framed and captured optically.

So, I think people saying it's a 50% crop probably mean one of:
a) from a PPed image (where one PP step happens to be cropping), a "Save for web..." version with 50% as many pixels was posted
b) ditto but 50% in each dimension (i.e. 25% overall)
c) the crop done during post-processing kept 50% of the pixels captured by the camera sensor (ignoring Bayer interpolation etc)
d) ditto but 50% in each dimension (i.e. 25% overall)
e) they're presenting a smaller area cropped from their work, but each pixel in that sample represents 2 from their work: this would be bizarre (why not a 100% crop?), but does at least provide some kind of loose "upper bound" for quality issues
f) ditto but 4 rather than 2...

Any of these would be useful insight for others contemplating the "success" of the macro shot: having used optical zoom to get a large-megapixel detailed result with good sharpness.

Unfortunately, it's anyone's guess which of these things is meant.

Cheers, Tony

[LordV]

The original question was to confirm what I thought a 100% crop meant. I'm happy I had got that correct and it is only used to show original picture quality.

It is fairly common in macroshots for someone to gain apparent magnification by cropping a shot both for web presentation and I think in some cases for print purposes. I suspect the same is often done by telephoto users. This may seem an odd thing to do but macro shooting does have some peculiarities all of it's own. For example when using high power macro lenses (eg MPE-65 or reversed lenses) the effect of magnification on apparent aperture and hence diffraction softening is so bad that you can actually get a sharper looking picture when viewed fullscreen on a cropped shot taken at half the magnification than an original shot a full magnification.

So many of the shots shown in the macro forum do have some cropping to increase the apparent magnification - nothing to do with showing the image quality of the original.
I was just trying to ascertain whether in this case where the picture has been cropped for magnification or even perhaps for compositional reasons whether there was a standard nomenclature for describing how much of the original picture the crop is showing.
I was also not refering to image sizes shown in PS when talking about percentages.

I was actually asking this on behalf of others, I actually rarely do anything more than minor compositional cropping.

Hope this makes sense

Brian V.

[Wilt]

I was just trying to ascertain whether in this case where the picture has been cropped for magnification or even perhaps for compositional reasons whether there was a standard nomenclature for describing how much of the original picture the crop is showing.
I was also not refering to image sizes shown in PS when talking about percentages.

In the macro world, things are normally stated in the context of on-film image vs. real life item, such as 0.5x. In the case of film, one did not question what fraction of the macro image was in a print, you only cared about the subject magnfication (0.5x). Similarly one would not care about what fraction of the total pixel count (either linear or area) was used, you only care about the subject magnification on the sensor.But you never hear any mention of taking that image and blowing it up to 11x14 vs. to 16x20 size, and expressing that

In the digital world the same applies...we talk about subject magnification on sensor, but never mention portraying that information on a monitor or on a final print.



You might be worrying about an aspect of presentation which never has been controlled, nor even stated as a 'size factor'.

[LordV]

In the macro world, things are normally stated in the context of on-film image vs. real life item, such as 0.5x. In the case of film, one did not question what fraction of the macro image was in a print, you only cared about the subject magnfication (0.5x). Similarly one would not care about what fraction of the total pixel count (either linear or area) was used, you only care about the subject magnification on the sensor.But you never hear any mention of taking that image and blowing it up to 11x14 vs. to 16x20 size, and expressing that

In the digital world the same applies...we talk about subject magnification on sensor, but never mention portraying that information on a monitor or on a final print.



You might be worrying about an aspect of presentation which never has been controlled, nor even stated as a 'size factor'.

I'm certainly not worrying - just wanted to know if there was a standard method of saying how much a shot had been cropped (apart from the 100% thing we all seem to agree about)

You are correct re the normal representation of magnification in macro but that only applies to the lens and takes no account of the sensor size. Agreed the different sensor sizes effectively mean images are being cropped anyway from the std 35mm size. This means though that at any standard print size, the print magnification of a shot taken taken at 1:1 on a 1.6crop DSLR is the same as the print magnification from a P&S at 0.3:1 magnification and both are larger than a FF at 1:1.

This is getting slightly off topic but I made the mistake I suspect of thinking there was a simple answer to the question of expressing how much a shot had been cropped (not including the 100% crop).

Brian V.

[SkipD]

I'm certainly not worrying - just wanted to know if there was a standard method of saying how much a shot had been cropped (apart from the 100% thing we all seem to agree about)Brian, in well over 40 years of being in photography, I have never heard of anyone referring to a percentage of cropping anything other than the "100% crop" thing we ask for on the forums here. Anybody cropping an image might talk about the format being used (8x10, for example) and might talk about "cropping heavily" or something like that, but I have never heard of a numeric value applied to the process and especially never a percentage.

[CyberDyneSystems]

100% crop as mentioned, means simply looking at the actual pixels with no reduction relevant to your screen real estate.. displaying each of the image files' pixels on each of your displays' pixels.. (or dots, or points as you see fit to call them)

A 50% crop means little to me, but it would arguably mean that one is showing a crop that has a 2:1 ratio of image file pixels to screen pixels. ie: the crop image being shown is at double the pixel density of the screen res, or an image sized to display 150 pixels of the original file per inch, on a screen showing 75 DPI.

I don't think this has any value to mention regarding image detail, it means like most images we see on screen that it too has been interpolated in the process of being displayed... it is just like all images that are "not" 100% crops. It seems more likely to be used simply to establish how much of the sensor the photog was able to fill with the subject, sort of an attempt to loosely describe how close the photog got, be it to a macro subject or a bird... etc..