I see, then the little 16 megapixel compact camera will beat all of them?
Pixels only attempt to reproduce the lens image, hopefully well. The way to bet is a large sharp lens image.
It's pixel pitch Wayne not Mpix. Pixel size/surface area.
About Crop Factors on lenses
- Thread starter Aleks
- Start date
- Status
It's pixel pitch Wayne not Mpix. Pixel size/surface area.
Yeah, so? What's the pixel pitch of 16 megapixels on a 1/2.3" 6.30 x 4.72 mm sensor?
(Nikon, Canon, etc). That's not much more than 1/4 the DX area, so it's really high resolution, right? At least pixel pitch implies it, so it ought to sample really well, right? But samples what? It samples a really small image, that has to be greatly enlarged.
Enlargement directly reduces resolution (among other issues).Yeah, so? What's the pixel pitch of 16 megapixels on a 1/2.3" 6.30 x 4.72 mm sensor?
That's not much more than 1/4 the DX area, so it's really high resolution, right? At least pixel pitch implies it, so it ought to sample really well, right? But samples what? It samples a really small image, that has to be greatly enlarged.
If Mike takes a shot with his P610 and I take one with my D810 and we both use a 300mm, do you think I can match his quality when having to crop/enlarge my shot to match his birdshot?
Overall my image quality will be better and if I have a decent lens, I can produce a sharp crop up to 100% cropped but once I'm there and have to go the rest of the distance, I will never match his version.
When we talk about digital images shown on a digital medium, the smallest dimension is one pixel, which is also the smallest size of the detail that can be displayed. Enlargement only occurs whenever we scale beyond that pixel size and "stretch" the pixels.
When taking the same shot with the same lens (able to resolve the detail) with both cams, at the same distance, it's about those pixels/surface area.
Last edited:
Come on J-see, I'm disappointed to be in this nonsense with you. All those buying FX sure are going to get a surprise to learn they should have bought a compact.
I don't know where this bird is, but you're saying that if you use the wrong choice lens for some cases, your camera may not do well. That is only evidence of your inability to deal with it, not evidence that a compact has more quality and capability than a D810. That's outright nonsense. I know you know that, so it is even worse nonsense. Why do you do that?
P610 is a 258 mm lens instead of 300mm.
The compact crop factor is 5.6, so are you actually claiming your D810 is useless without the equivalent 1440 mm lens?
Don't you have the 150-600mm? Try it, it should be hard to beat on a D810.
So for your example, but with the 600mm, that's 600/258 = 2.3x more focal length, compared to a little sensor (17% size) needing 5.6x more enlargement to view same size.
Some numbers... Say the bird is at 100 feet.
Field of view of FX at 100 feet with 600 mm is 6 x 4 feet.
Field of view of compact with 258mm is 2.44 x 1.83 feet.
So your bird is 41% size (but there is more to be understood). The compact has to enlarge it 560% more than you (to view it same size). Think again, who do you imagine wins that?
Pixels determine sampling resolution to reproduce the lens image well.
Pixels also determine printing resolution to make ink dots to reproduce the image well.
We do need sufficient pixels for the print size, but that is just a boundary condition (meaning more than enough won't help. And we can print smaller, or at less then 300 dpi.) Same as we need sufficient pixels to sample the lens image adequately.
But pixels do not create any image detail, the lens does all of that.
What all steps are reproducing digitally is the resolution of the lens image detail projected onto the sensor. And it requires enlargement to see it, which reduces that lens resolution when we see it (for example, 4x enlargement is 1/4 the detail level ... the resolved lines per inch are 4x wider apart.)
So starting with a large image needing less enlargement is the way to bet. It's always been true. Ansel Adams favored 8x10 inch film.
Factors like resolution in depth of field are about enlargement. Pixels are not mentioned.
Come on J-see, I'm disappointed to be in this nonsense with you. All those buying FX sure are going to get a surprise to learn they should have bought a compact.
I don't know where this bird is, but you're saying that if you use the wrong choice lens for some cases, your camera may not do well. That is only evidence of your inability to deal with it, not evidence that a compact has more quality and capability than a D810. That's outright nonsense. I know you know that, so it is even worse nonsense. Why do you do that?
P610 is a 258 mm lens instead of 300mm.
The compact crop factor is 5.6, so are you actually claiming your D810 is useless without the equivalent 1440 mm lens?
Don't you have the 150-600mm? Try it, it should be hard to beat on a D810.
So for your example, but with the 600mm, that's 600/258 = 2.3x more focal length, compared to a little sensor (17% size) needing 5.6x more enlargement to view same size.
Some numbers... Say the bird is at 100 feet.
Field of view of FX at 100 feet with 600 mm is 6 x 4 feet.
Field of view of compact with 258mm is 2.44 x 1.83 feet.
So your bird is 41% size (but there is more to be understood). The compact has to enlarge it 560% more than you (to view it same size). Think again, who do you imagine wins that?
Pixels determine sampling resolution to reproduce the lens image well.
Pixels also determine printing resolution to make ink dots to reproduce the image well.
We do need sufficient pixels for the print size, but that is just a boundary condition (meaning more than enough won't help. And we can print smaller, or at less then 300 dpi.) Same as we need sufficient pixels to sample the lens image adequately.
But pixels do not create any image detail, the lens does all of that.
What all steps are reproducing digitally is the resolution of the lens image detail projected onto the sensor. And it requires enlargement to see it, which reduces that lens resolution when we see it (for example, 4x enlargement is 1/4 the detail level ... the resolved lines per inch are 4x wider apart.)
So starting with a large image needing less enlargement is the way to bet. It's always been true. Ansel Adams favored 8x10 inch film.
Factors like resolution in depth of field are about enlargement. Pixels are not mentioned.
If I'm not mistaken I replied to the comment about FX cropping to DX size and the sharpness/detail difference between both. It is about pixel pitch/surface area since the analog signal is directly translated into pixels and when the ability to resolve detail is identical (same lens, same distance), the smaller pixels translate into more detail. When we start to crop and match the images, depending which cam we use, we will start to see differences in quality.
Enlargement does not occur until we stretch the pixels when taking the same shot. That's beyond 100%.
Why the nonsense about compacts and birds then?
Enlargement occurs anytime we want to look the image. The DX image is only 24x16 mm, not even quite small postage stamp size. We look at it larger than that, several times larger. Looking at DX necessarily requires 50% more enlargement than FX. So by your argument, I hope it started with at least 50% more detail.
If both DX and FX are 24 megapixels, then the DX has 50% more sampling resolution (of the image from the same lens). Assuming that the lens image of the scene has the resolution to give (depends on lens, on focus, on diffraction, how steady we hold it, and even on the scene). And DX is also necessarily enlarged 50% more to view it. So at best, this case is a wash (but still quite a feat indeed). But of course, the DX pixels are 2/3 size, another factor, so not actually a wash for quality, not in all cases. And if the lens image didn't quite have that much detail, then FX size wins over resolution.
But then a biggie, it's not the same scene of course, unless DX stands back 50% farther, which then is 50% less resolvable detail in the DX lens image. Enlargement is a rather large factor. It is a mistake to consider DX the equal of FX. Size matters.
Anyway, Depth of Field is not about pixels. The sensor aspect of DOF is about enlargement degree.
Last edited:
Maybe because that's the majority of cases when this is the subject and it becomes a factor on this site. I could have used macro too or anything else in which crop provides the advantage (depending the cam). Same shot, same distance, different sensor. The reason I mentioned microsensors is because it makes it a whole lot easier to see the difference. I'm used to conversing with people that are able to process information while they read but if I'm going too fast for you, say so and I'll break it down in steps.
On a computer screen the DX image has no metric dimensions, same with the FX. If we'd display it at its true dimensions on screen, the FX shot would be much larger since it has more irrelevant data for this comparison. But the image portion of the DX would contain more detail (the digital output being larger too) and if the FX would try to match it, it would only then require actual enlargement which comes at the expense of quality.
Last edited:
Say what? Do you not realize that you readily make up crap, and say it publicly as fact? Your notions do not make it be fact. We all do that in some degree, but we learn that it's good to keep quiet the first day or two that we think it up, while we think it over. We have to verify notions before it can thought to be fact.
Of course the screen shows dimensions. My 23 inch screen is 20x11.5 inches, which is dimension. And if an image fills any major portion of it, it's a rather large image - with corresponding reduced resolution due to enlargement (and resampling). The video system doesn't use inches or dpi, it just shows resampled pixels directly on what it's got, but don't be confused, of course we still certainly SEE inches and pixels per inch in the image result. My default browser view (Firefox with 110% Windows text size) on this forum sees your four pixel image as about 4.5 inches square. Which can vary with how I size my browser, but I see 360 pixels now, so that computes 80 pixels per inch. I think it was 400 pixels originally, but of course, browsers do CSS Pixel Ratio today, a long story, but image size depends on text zoom now. Browsers are NOT quite like other image apps.
But the 4.5 inch result I see certainly is a dimension.
Now ***THAT IMAGE*** is graphics, without any original source size, but DX camera images were all originally near 24x16 mm. That is very much a dimension. Which we enlarge to view. Which degrades resolution proportionately. DX 50% more than FX images, because their original dimension is cropped that much smaller.My raw editor typically shows my landscape images as about 12x8 inch dimension, or a bit more. From DX dimensions (24x16mm), that's the SAME 13x enlargement as DX to an 8x10 inch print (except the images have first been resampled to about 30% size at default view, so the pixels will fit the screen dimension). Video systems have different resolution needs than printing, able to show about 100 dpi vs 300 dpi. A 1920 pixel screen 20 inches wide is necessarily showing 1920/20 = 96 pixels per inch. The video system certainly did not figure it that way (totally ignores inches), but of course, it comes out that way and we see inches, and dimension. The large image we show has normally been greatly resampled first, to fit. But 100 dpi does look impressive on the monitor, because the color depth is so great (which is a form of resolution), and the dynamic range is so much greater than reflective prints.
Horoscope Fish
Senior Member
.....
.....
.....
Wayne, seeing your brain in action is like observing a fat man swim across the Channel. It's boring fast.
It's not personal attack; it's an observation.
Who talks about physical screen size as a measurement for image size when everyone knows the size of that image on your screen is defined by resolution. Pixels times pixels.
I am guilty of observations too.
The video system does draw images as pixels. It is defined as pixels. There is no concept of inches or dpi in the video system.
But we cannot stop there.
Our eyes can't see pixels (cannot resolve them). I don't know proper terminology, but our eyes see size as dimension, angular extent. An image is viewed at an enlarged size (larger than the camera sensor mm dimensions), even if we don't measure it in inches (we certainly can hold a ruler up to it though, the inches are there). We buy a 23 inch screen, and the image fills a good portion of the inches on that screen. It is a significant enlargement. If it came from a camera with an original size dimension in mm, and now it covers several inches, it is enlarged, and detail is now shown at wider spacing, i.e., any hypothetical lines per inch of detail have wider spacing, or much less resolution than the original sensor image. The larger we show it, the wider the image detail becomes. Size matters.
This is not hard.
Again, what is the relevance of this when we (at least I) talk about taking a shot with a smaller sensor but higher pixel density vs a bigger sensor with lower pixel density cropped to the same level as the smaller?
It doesn't matter at all how big we display them or what the initial sensor dimensions were when we compare both shots.
It doesn't matter at all how big we display them or what the initial sensor dimensions were when we compare both shots.
You don't listen or grasp anything?
I already covered that, #29. Many things can prevent that.
- Status