Beginner Dx to Fx?

But youre not enlarging a digital file physically like a negative where you talk e.g. 3x or 8x enlargements from LF or 35mm negatives. youre printing a file of a certain size dependin go the MP count, not? What I see matters is whether the lenses are capable of resolving to the pixelsize whether it is a 24mp apsc or 35mm or a 50mp 35mm sensor.
Im stupid I dont get it (n)
OK so if you start with an image that is 1cm squared and enlarge it to one that is 100cm squared you have enlarged the area by 100. If you start with an image that is 10cm squared and enlarge it to one that is 100cm squared you have enlarged it by 10 times (in simple terms). So yes you have the same pixels to start with but with a smaller sensor you are enlarging each MP more and therefore you are going to see more 'flaws'/degradation in IQ.
 
OK so if you start with an image that is 1cm squared and enlarge it to one that is 100cm squared you have enlarged the area by 100. If you start with an image that is 10cm squared and enlarge it to one that is 100cm squared you have enlarged it by 10 times (in simple terms). So yes you have the same pixels to start with but with a smaller sensor you are enlarging each MP more and therefore you are going to see more 'flaws'/degradation in IQ.
But thing is your file is not size dependable its XX MB of 1's and 0's only thing thats size dependable is how many pixels it takes to cover the lens' resolution power in lp/mm so as far as I can get my brain working the size thing is all about the optimization of the lens. But maybe that should be the "mystery" that keeps me awake at night :)
Maybe its the limit in resolution of the lens that will show with the smaller pixels size thus giving less detailed images.( it takes more pixels to resolve the linepairs on apsc than on fullframe. ????
 
Last edited by a moderator:
But thing is your file is not size dependable its XX MB of 1's and 0's only thing thats size dependable is how many pixels it takes to cover the lens' resolution power in lp/mm so as far as I can get my brain working the size thing is all about the optimization of the lens. But maybe that should be the "mystery" that keeps me awake at night :)
Maybe its the limit in resolution of the lens that will show with the smaller pixels size thus giving less detailed images.( it takes more pixels to resolve the linepairs on apsc than on fullframe. ????
Whilst it's capturing (in theory) the same amount of data in terms of file size, the physical size of the image captured on APS-C is smaller than that of FF. You are talking about file size which is different to image size. Would you agree that if you magnify an image it doesn't look as sharp? How many times have you looked at the image on the back of your camera and thought it's nice and sharp, only to see it on your computer and it's not as sharp as you thought? The more you enlarge the image the more that you see the flaws. MP is not the only thing that helps resolve detail, size also matters. Which has more of an impact I can't say for sure. For example you can resolve more detail from the 24mp D3400 using the 24-70mm lens than you can using the D700, I would therefore expect to be able to get larger prints from the D3400. However, the 24mp D750 can resolve more detail than the 24mp D3400 and 24mp D7200. Obviously same MP count but the D750 has less pixel density, larger pixels and larger physical image size at the point of capture.
 
Whilst it's capturing (in theory) the same amount of data in terms of file size, the physical size of the image captured on APS-C is smaller than that of FF. You are talking about file size which is different to image size. Would you agree that if you magnify an image it doesn't look as sharp? How many times have you looked at the image on the back of your camera and thought it's nice and sharp, only to see it on your computer and it's not as sharp as you thought? The more you enlarge the image the more that you see the flaws. MP is not the only thing that helps resolve detail, size also matters. Which has more of an impact I can't say for sure. For example you can resolve more detail from the 24mp D3400 using the 24-70mm lens than you can using the D700, I would therefore expect to be able to get larger prints from the D3400. However, the 24mp D750 can resolve more detail than the 24mp D3400 and 24mp D7200. Obviously same MP count but the D750 has less pixel density, larger pixels and larger physical image size at the point of capture.

I tried the A7II vs my A6000 using among some other lenses a 105mm AFD micro nikkor and found hat there was a difference in detail but it was small. mounting the 60mm f/2,8 sigma on the A6000 gives me very detailed images comparable to what I saw on the A7II with the Nikkor. Couldnt make a really good comparison because I didnt have lenses that would give me good comparable pictures

Yep, it's one of very few ways we can get and idea. I used this when I upgraded from the D750 to D850 to decide whether to keep the 24-120mm f4 or swap to the 24-70mm f2.8 (my decision wasn't solely based on this result, but it was part of it). Now unfortunately DXO haven't tested any lenses with the D850 yet (not sure why, it's been out long enough yet) so I had to use the D810.

Anyway on the D750 the resolution that the 24-120mm can resolve is 14mpix, and with the D810 it only increases to 15mpix (a hike of approx 8% in resolution). With the 24-70mm the results are 17mpix on the D750 and 21 on the D810 (a hike of 25% resolution) which suggested to me that at 15mpix you're probably reaching the limits of what the 24-120mm can resolve, whereas the 24-70mm makes a more significant jump and may possibly even resolve more with the D850.

Could be fun to see the numbers form D7200 vs D750. Seems to me more a resolution on sensor than a sensor size thing.
Well I think I have to settle on "I dont understand this one"
 
Last edited by a moderator:
Seems to me more a resolution on sensor than a sensor size thing.
Yes, you wind up with a digital file of 1's and 0's with nothing to differentiate them other than how many there are (pixel dimensions). But it's not that simple, and in the end it comes down to sensor size much more than MP's for many reasons... just as it comes down to negative size.
The main reason that it still primarily comes down to the physical size of the sensor is because you're still collecting/recording light on a light sensitive medium. It makes little difference that the sensor is using silicon pixels in a fixed pattern rather than the negative's more random silver halide crystals. That aspect, and everything else about the process is essentially the same. And the fact is a larger sensor receives more total light for any captured image... In order for the image/exposure to be the same; the larger sensor requires a longer FL which makes the details larger on the sensor (same % coverage, easier to resolve), and it requires a larger entrance pupil (same f-ratio).

There was never a concept of "crystals/inch" with photographic printing per se, but it was/is a factor... i.e. when we think about modern ISO performance with higher resolution sensors a realistic correlation would be that Kodak invented an ISO100 film with the crystal grain size/structure of ISO64 film (and then it would still come down to the size of negative it's used for).

That said, when we are talking about the difference between 24mm DX and 35mm FF the effects/results are not significant enough to be a big factor for most users/applications.... it really is more like your correlation of 110 and 35mm being fine for general "every day" use. But if you had to push the boundaries, the larger format would be the better choice.

FWIW, the PPI/DPI printing thing is almost complete garbage... it is based upon the fact that you are reproducing dots (pixels) with dots (ink). However, a person with 20/20 vision can only see ~ 260DPI at a distance of 12"... and a person with perfect vision can only see ~320DPI. And as distance increases the minimum size of what you can see also increases (same relative size; reduced DPI). By keeping the DPI requirement constant you are basically saying the images will always be viewed at a distance of 12".
 
I tried the A7II vs my A6000 using among some other lenses a 105mm AFD micro nikkor and found hat there was a difference in detail but it was small. mounting the 60mm f/2,8 sigma on the A6000 gives me very detailed images comparable to what I saw on the A7II with the Nikkor. Couldnt make a really good comparison because I didnt have lenses that would give me good comparable pictures

Could be fun to see the numbers form D7200 vs D750. Seems to me more a resolution on sensor than a sensor size thing.
Well I think I have to settle on "I dont understand this one"

That kind of comparison is difficult, where both sensor format and pixel count change - hard to calculate, it's not just straight mathematics. But if you compare different formats with similar pixel counts (eg Canon 5D2 vs 7D, Nikon D700 vs D300, Nikon D750 vs D7200 etc) then the larger full-frame sensor always wins. The benefits of FF are almost all derived from the physically larger sensor area - better sharpness, less noise and improved dynamic range (assuming same generation sensors), and more depth-of-field control - than from any increase in pixels.

Basic lens MTF sharpness theory is that as resolution demands go up, so image contrast goes down, and it's contrast that contributes most to perceived sharpness. The smaller the sensor format, the more it has to be enlarged, which requires higher resolution and so sharpness drops. By how much it reduces depends on the lens, but as a rough guide the MTF graphs published by LensRentals are useful. Looking here for example https://www.lensrentals.com/blog/2018/08/mtf-tests-for-the-canon-70-200mm-f4-is-ii/ there are sharpness traces at different resolution levels. A good comparison is the green 30-lpmm against purple 50-lpmm, given that 30-lpmm x 1.6x crop factor would be 48-lpmm, so pretty close. They show a drop of 12-15% MTF contrast which ties in pretty well with my own MTF tests as a fairly representative average difference between the two formats.
 
But youre not enlarging a digital file physically like a negative where you talk e.g. 3x or 8x enlargements from LF or 35mm negatives. youre printing a file of a certain size dependin go the MP count, not? What I see matters is whether the lenses are capable of resolving to the pixelsize whether it is a 24mp apsc or 35mm or a 50mp 35mm sensor.
Im stupid I dont get it (n)

It's just an analogy. sk66 is using an old fashion enlarger to print from negatives as an analogy, because printing from a digital image files does sometimes feels somewhat akin to printing from negatives.

A 2MP camera will result in an image size of something like 1600 x 1200 pixels. A 14MP camera will result in an image size of something like 6000 x 4000. (According to what I read somewhere on the Internet, but let's just use them as an example.)

Print each on a big paper. The 2MP will look small in the middle of the paper. The 14MP will look a little bigger, but still surrounded by white space.

If you want the picture to fill the whole paper, you tell the printer to fit to page or something like that, different printers and different software vary, but they usually give you an option to fit to page.

This is the computer and printer's way of stretch out the image to fill the whole page. It feels like similar to moving the enlarger up so you see the negative image fill up the paper.

It do not physically move an actual enlarger inside the computer. It is when you print them out and look at the image, it reminds you of the old fashion darkroom printing with an enlarger.

When printing a negative, you will find a tiny image on a big paper, so you move the enlarger up to fill the paper. When printing from a printer, you will find a tiny image on a big paper, so you tell the computer to fit to page.

With a computer software, you zoom in or fit to page, but it feels like you enlarge a negative. Different technology but somewhat same job.
 
Different technology but somewhat same job.
If you send your files out to have photographic prints made the image is projected onto the paper with LEDs, this is very much like the photo enlarger scenario.
Most of the prints I have made are photographic and not inkjet... I don't do any serious printing at home.
 
The difference, to me at least, isn't so abstract as even needing huge enlargements. When I followed Snerklers link to his 2 images, one from FX and 1 from M43 I could see a difference even at the small scale displayed on TP, let alone on Flickr.
 
The difference, to me at least, isn't so abstract as even needing huge enlargements. When I followed Snerklers link to his 2 images, one from FX and 1 from M43 I could see a difference even at the small scale displayed on TP, let alone on Flickr.
Yeah, you have superhuman vision ;)
 
I chose correct too, but it was a 50/50, c'mon ... and many on here couldn't tell the difference. So you guessed right, nothing more.
 
Yeah, you have superhuman vision ;)

I've also been told I have golden ears by other people making guitar amplifiers because I can pick out cryogenically treated valves in a blind test. :D

I've just been back to look at the pictures again. It's hard to say what's different, but looking closely at small details like the fern sticking up on the RHS about halfway down the image, even on the small version it looks like it has more depth and detail in the top image than the bottom. Generally fine detail is subtly picked out, rather than on the lower image, with a wider range of small differences in tones and shades. I think this is what contributes to the perception of greater depth. There was someone who posted comparative images of a flower shot with a Lumix superzoom camera and a canon DSLR (IIRC - think it was @GardenersHelper ) with the DSLR showing a greater range of fine and subtle tones & shades that contributed to make the image less flat - although that's much less obvious here, I think the situation is similar.
 
I've also been told I have golden ears by other people making guitar amplifiers because I can pick out cryogenically treated valves in a blind test. :D

I've just been back to look at the pictures again. It's hard to say what's different, but looking closely at small details like the fern sticking up on the RHS about halfway down the image, even on the small version it looks like it has more depth and detail in the top image than the bottom. Generally fine detail is subtly picked out, rather than on the lower image, with a wider range of small differences in tones and shades. I think this is what contributes to the perception of greater depth. There was someone who posted comparative images of a flower shot with a Lumix superzoom camera and a canon DSLR (IIRC - think it was @GardenersHelper ) with the DSLR showing a greater range of fine and subtle tones & shades that contributed to make the image less flat - although that's much less obvious here, I think the situation is similar.

That sounds like the sort of thing I would do, with one of my two main types of subject, and it sounds like the sort of conclusion I would have come to, which is why I don't use my bridge cameras much for botanical imaging (unlike for invertebrates, for which I almost always use a bridge camera). I have to say though (in case someone asks me to produce the comparison) that I don't recall the particular example. (But my memory is embarrassingly terrible so it might well have been me.) In fact, I'm not aware that I have any sufficiently like for like botanical examples which definitively illustrate to my satisfaction the benefits of a larger sensor for post processed images of the sort of size that I produce (1400 pixels high for viewing on screen). I believe that "on average" I get more (to my eye) appealing images with a larger sensor, but I don't know that I could convince anyone else of that.

Actually, these days I use a micro four thirds Panasonic G80 for botanical work in preference to my slightly larger sensor Canon 70D, but that is do with usability rather than image quality considerations. That said, and somewhat relevant for this thread, I am continuing to wonder about using full frame for botanical imaging for (disputed) potential improvements in the rendering of the subtleties of colours, texture, light, clarity and/or "dimensionality". Reading this thread, like others at dpreview including one where I specifically asked some (Sony as it happens) full frame users about the (alleged) "special" qualities of full frame images, reinforces my feeling that the only way I am going to find out whether there is a difference that matters to me is to try it. But it would be a rather expensive experiment, and that holds me back (as well, now, the possibility of a Panasonic full frame camera, which could have two very specific, and significant, advantage for my particular approach).

Very interesting thread.
 
Last edited:
The other point to bear in mind is the age of sensor technology. I get the impression that a crop sensor camera of the same MP as a full frame will have better IQ etc. than the full frame if the sensor is 3-6 younger in technology. In other words the extra image quality you buy with full frame can be had with crop sensor if you just wait for a few to several years. Crop sensors were invented because they're a particular sweet spot in digital camera technology, rather good performance for the money. Full frame is pushing into diminishing returns territory.

Do I mind being a few to several years out of date in my camera technology? If I'm not forced into an early camera upgrade by breaking my camera I'll usually go on using it for several years. Looks like I'm not the kind of person likely to buy a full frame camera! Especially since the stuff I've got, mostly second hand, is way ahead of my own skills in image quality. It's not holding me back in developing my skills, even though it will always be true that spending thousands of pound on the latest gear would noticeably improve, at least to knowledgeable photographers, the quality of some of my photographs.
 
Especially since the stuff I've got, mostly second hand, is way ahead of my own skills in image quality. It's not holding me back in developing my skills, even though it will always be true that spending thousands of pound on the latest gear would noticeably improve, at least to knowledgeable photographers, the quality of some of my photographs.

Absolutely cant argue with that.

Enlargements - I have some A2 prints from a 6mp Fuji S5 pro and there is no visible pixelation, and the colours are superb (I did make sure I filled the frame, but isn't that what we are supposed to do!!!)

I have a D90, D300s, and the S5 pro. I know I should sell them as I now have the D500, but they are worth so little, I keep them for dog walking where they may get knocked about. The D500 focusing is amazing and the IQ noticeably better, but the reality is my best photos were taken on the slow and clumsy S5 pro!

I too hanker after a D750, but know it wont take better pictures, it will just give me the satisfaction of ownership (OK I could justify it to take pictures of the wife doing indoor dressage where high ISO is required, but these days I dont need to justify anything I can afford!)

So to the OP if you want to upgrade by all means do so but dont underestimate how competent the D90 and other cameras of that era are.
 
The other point to bear in mind is the age of sensor technology. I get the impression that a crop sensor camera of the same MP as a full frame will have better IQ etc. than the full frame if the sensor is 3-6 younger in technology. In other words the extra image quality you buy with full frame can be had with crop sensor if you just wait for a few to several years. Crop sensors were invented because they're a particular sweet spot in digital camera technology, rather good performance for the money. Full frame is pushing into diminishing returns territory.

Do I mind being a few to several years out of date in my camera technology? If I'm not forced into an early camera upgrade by breaking my camera I'll usually go on using it for several years. Looks like I'm not the kind of person likely to buy a full frame camera! Especially since the stuff I've got, mostly second hand, is way ahead of my own skills in image quality. It's not holding me back in developing my skills, even though it will always be true that spending thousands of pound on the latest gear would noticeably improve, at least to knowledgeable photographers, the quality of some of my photographs.

Not quite true. Sensor technology plays a big part, but applies equally to all formats and in a like-for-like comparison full-frame will always be sharper and offer more DoF control. The question of diminishing returns is very valid though - just how sharp do we need and how much wafer-thin DoF can we take?!
 
Absolutely cant argue with that.

Enlargements - I have some A2 prints from a 6mp Fuji S5 pro and there is no visible pixelation, and the colours are superb (I did make sure I filled the frame, but isn't that what we are supposed to do!!!)

I have a D90, D300s, and the S5 pro. I know I should sell them as I now have the D500, but they are worth so little, I keep them for dog walking where they may get knocked about. The D500 focusing is amazing and the IQ noticeably better, but the reality is my best photos were taken on the slow and clumsy S5 pro!

I too hanker after a D750, but know it wont take better pictures, it will just give me the satisfaction of ownership (OK I could justify it to take pictures of the wife doing indoor dressage where high ISO is required, but these days I dont need to justify anything I can afford!)

So to the OP if you want to upgrade by all means do so but dont underestimate how competent the D90 and other cameras of that era are.

I know and love the D90 and am happy with it apart from few things like noise levels at iso Over 1600. It’s not the end of the world but in few instances it did annoy me especially when I needed fast enough shutter speed and aperture was at widest already.
This upgrade I was just asking as I wanted to know opinions and also I liked the specs of the D850 or D500 if sticking to dx cameras. Of course now some of you might make me look at mirrorless [emoji14]
But seriously though I may not even upgrade anyway.
The biggest barrier is justifying to the wife why I would want to upgrade (the camera that is not the wife)
 
But seriously though I may not even upgrade anyway.
The biggest barrier is justifying to the wife why I would want to upgrade (the camera that is not the wife)

Start complaining about the misbehaviour of your camera, occasional malfunctions which become more frequent...
 
I have shot with the D70s-D200 and the D300 all crop sensors and happy with them at the time. Then moved onto the D800 and then the D810 Full frame cameras. Why ? good question and I think the answer lies in getting frustrated with the camera limitations. One can study camera specifications until blue in the face, but won't give you the answers your looking for. There comes a time when one "out grows" a camera when skill levels increase beyond what a camera can do.

I went for the nikon D810 as I had damaged the D800 (now repaired), and it although has the same D8 designation is different in many ways without the AA filter and better. I did find I had to be more precise with the D810 in my handling technique to get the best from it and it made me concentrate more on what I was doing, Something I needed to do.

. I could have gone for the D850 but apart from the price difference I had to ask myself do I was a 40+mp camera without a flash and the answer for me is no. I do have the Nikon SB900 flash unit if one wonders

If upgrading I would seriously suggest looking at the Nikon D810 , I find it completely suits my needs.

Still have the D300 and D800 kinda got attached to them
 
Last edited:
Last edited by a moderator:
Back
Top