Relationship between cropping and effective focal length

[splproductions]

I've been doing some digging around the internet and I think I now know the answers to these questions. Someone chime in and let me know if I've come to the right conclusions.

I shoot with a 6D (20.2MP). Lets say I have the 24-105 mounted. Then I upgrade to the 5D Mark IV (30.1MP). That's a 49% increase in pixels. Are the following assumptions correct?

1) I could take a shot at 105mm with the 5D Mark IV, crop it to a size that would be equivalent to 20.2MP (still plenty MPs to print large), and I essentially now have an effective "reach" of 156mm. (105mm x 1.49).

2) Assuming I keep that new photo at 20MP (I don't enlarge the photo), the image quality will be the same as if I would have had a telephoto zoomed to 156mm.

3) In terms of perspective and image compression, the cropped photo will have the same amount of compression as if I would have used a telephoto lens at 156mm.

4) Using a high-resolution sensor is a reasonable way to extend the reach of whatever lens you have mounted (assuming you retain enough MPs for the size print you are planning on).

Thanks!


Sent from my iPhone using ThePhotoForum.com mobile app

 

[Ysarex]

I've been doing some digging around the internet and I think I now know the answers to these questions. Someone chime in and let me know if I've come to the right conclusions.

I shoot with a 6D (20.2MP). Lets say I have the 24-105 mounted. Then I upgrade to the 5D Mark IV (30.1MP). That's a 49% increase in pixels. Are the following assumptions correct?

1) I could take a shot at 105mm with the 5D Mark IV, crop it to a size that would be equivalent to 20.2MP (still plenty MPs to print large), and I essentially now have an effective "reach" of 156mm. (105mm x 1.49).

Yes.

2) Assuming I keep that new photo at 20MP (I don't enlarge the photo), the image quality will be the same as if I would have had a telephoto zoomed to 156mm.

Do you mean the cropped (20mp) 5D photo (105mm) will have the same image quality as an uncropped 5D photo (156mm)? No.

3) In terms of perspective and image compression, the cropped photo will have the same amount of compression as if I would have used a telephoto lens at 156mm.

Yes, note: perspective is not a function of lens focal length.

4) Using a high-resolution sensor is a reasonable way to extend the reach of whatever lens you have mounted (assuming you retain enough MPs for the size print you are planning on).

Yes, reasonable. Or you could just say a high resolution sensor is more forgiving of an image crop and not have to do all that math.

Joe

Thanks!


Sent from my iPhone using ThePhotoForum.com mobile app

 

[splproductions]

Ysarex - thanks for the replies.

You said "no" to #2 - image quality. I believe you are understanding my question correctly. I'm wondering if the 5D cropped photo taken with a 105mm lens will have the same quality (sharpness, clarity, contrast, etc) as using a 5D with a 156mm lens and not cropping anything.

What is your reasoning for saying that the image quality won't be the same?


Sent from my iPhone using ThePhotoForum.com mobile app

 

[Ysarex]

Ysarex - thanks for the replies.

You said "no" to #2 - image quality. I believe you are understanding my question correctly. I'm wondering if the 5D cropped photo taken with a 105mm lens will have the same quality (sharpness, clarity, contrast, etc) as using a 5D with a 156mm lens and not cropping anything.

What is your reasoning for saying that the image quality won't be the same?

Digital photos are data. They begin as an electrical signal from the sensor and wind up as RGB pixels and maybe dots of ink. (All else being equal -- how the data is managed), more data = better quality and less data = lesser quality. Cropping forces less data.

Now you can play with that and make a case that within certain limits (I just look at pictures on my phone, or I don't post process and camera JPEGs are fine) the difference may be slight enough not to mater. OK -- that works. But all else being equal and you decide you want the very best quality then more data = better and less data = worse.

Joe

Sent from my iPhone using ThePhotoForum.com mobile app

 

[Derrel]

If you need to do very heavy cropping, you might find it helpful to "up-rezz" the original file,perhaps by quite a large margin, such as to at least twice the original size, before beginning any cropping. For example, back in the day, when the Nikon D1 digital SLR was king, with its impressive 2.7 megapixel captures (lol!), I would at times open a .NEF file, and create a 16-bit image file, and up-rezz it in small steps, over and over and over and over and over and over (using an automated software routine), and only then would I work on the large, 16-bit file, process it, adjust it, etc, and then I would down-size the image, in steps, with sharpening done in steps (again, using an automated action) which seems to make images look better in many cases.

Genuine Fractals was one software application program that was very often used back then, along with others, as a way to "up-rezz" images; back in those days, many digital cameras were putting out 2.7 to 4.0 megapixel-sized original images.

If you want to get into cropping and such as a regular part of your workflow, maytbe look into some of the ways you can up-rezz images, as well as performing image sharpening in multi-step processes. For example...instead of ONE, single unsharp mask step, say 100% at a threshold of .7 pixels, maybe do 100% at .3 pixels, twice, meaning applying about the same amount of total sharpening but done in two steps; or do the 100% at .7 pixel radius, but go to Filter>Fade USM, and fade the first application a bit, and then repeat the unsharp masking step once again, and fade that a bit until it looks correct.

As Ysarex has mentioned, working with MORE DATA usually really helps, and this is for example, why I think it is best to take a large image, say 5,000-pixels tall, and shrink it down, in STEPS, with very slight USM passes done at the intermediary sizes, to get down to say, a 1,200-pixel tall image; making the down-sized images using as much data as possible. If you need to add sharpening to compensate for an unsharp image, it seems to me to work better when the sharpening is applied to a lot of data; results are often better when the brunt of heavy Photoshop work is applied to a MASSIVE amount of data, rather than when applied to say, a 1,200 pixel-wide image. Of course, this was when an "original capture" was 2.7 to 4.2 megapixels; the 5D Classic is as I recall, 12.8 megapixels.

 

[splproductions]

My entire reason for this post was because I am trying to come up with my ideal "travel kit". I did Europe a couple months ago and I'm planning Hawaii and China for next year. I like to do a mix of landscape and travel/street photography on trips like this.

My thinking was... I could bring a 16-35, a 24-105, and then upgrade to a higher resolution sensor, and effectively have a "24-156mm", which would be awesome because I wouldn't miss leaving my 70-200 at home.

Derrel - after reading your post, I think carrying the 70-200 would be easier than learning everything you just said. Haha!

It sounds like from what you two are saying, I could use the high-res sensor to be able to crop in a pinch, but that it's not something I should rely on - I shouldn't pretend I have a 24-156mm lens (if that makes sense).


Sent from my iPhone using ThePhotoForum.com mobile app

 

[Derrel]

If you really want effective long reach, maybe consider an inexpensive camera body that has a high-density, APS-C sized sensor...the best sensor you can obtain in your brand, which I guess would be one of the newest Canon Rebels. High pixel density and a very good lens makes for a good image. I would ALSO consider the very inexpensive 24mm and 40mm pancake lenses for APS-C Canon cameras, which are pretty low-cost yet ultra-compact lenses, and the 50mm "brand-new" model Canon prime lens. My ideal travel kit fits into a large waist-pack, and has a body, 70-300 VR lens, a 60mm macro, and 85mm f/1.8 ultra-sharp lens, and a 24mm wide-angle, and a 35mm f/2.

 

[splproductions]

If you really want effective long reach, maybe consider an inexpensive camera body that has a high-density, APS-C sized sensor...

The thing is... I want the longer reach while still retaining a "true" 24mm. If I go to a crop sensor I lose that wide-angle.




Sent from my iPhone using ThePhotoForum.com mobile app

 

[Ysarex]

If you need to do very heavy cropping, you might find it helpful to "up-rezz" the original file,perhaps by quite a large margin, such as to at least twice the original size, before beginning any cropping. For example, back in the day, when the Nikon D1 digital SLR was king, with its impressive 2.7 megapixel captures (lol!), I would at times open a .NEF file, and create a 16-bit image file, and up-rezz it in small steps, over and over and over and over and over and over (using an automated software routine), and only then would I work on the large, 16-bit file, process it, adjust it, etc, and then I would down-size the image, in steps, with sharpening done in steps (again, using an automated action) which seems to make images look better in many cases.

Genuine Fractals was one software application program that was very often used back then, along with others, as a way to "up-rezz" images; back in those days, many digital cameras were putting out 2.7 to 4.0 megapixel-sized original images.

If you want to get into cropping and such as a regular part of your workflow, maytbe look into some of the ways you can up-rezz images, as well as performing image sharpening in multi-step processes. For example...instead of ONE, single unsharp mask step, say 100% at a threshold of .7 pixels, maybe do 100% at .3 pixels, twice, meaning applying about the same amount of total sharpening but done in two steps; or do the 100% at .7 pixel radius, but go to Filter>Fade USM, and fade the first application a bit, and then repeat the unsharp masking step once again, and fade that a bit until it looks correct.

As Ysarex has mentioned, working with MORE DATA usually really helps, and this is for example, why I think it is best to take a large image, say 5,000-pixels tall, and shrink it down, in STEPS, with very slight USM passes done at the intermediary sizes, to get down to say, a 1,200-pixel tall image; making the down-sized images using as much data as possible. If you need to add sharpening to compensate for an unsharp image, it seems to me to work better when the sharpening is applied to a lot of data; results are often better when the brunt of heavy Photoshop work is applied to a MASSIVE amount of data, rather than when applied to say, a 1,200 pixel-wide image. Of course, this was when an "original capture" was 2.7 to 4.2 megapixels; the 5D Classic is as I recall, 12.8 megapixels.

Wow! Bringing back memories there -- I remember Genuine Fractals and 4 megapixel cameras. Seems so far away now. Might be fun to see what an old copy of GF could do with modern 24-36 megapixel files.

Joe