Pages: 1 2 [3] 4 5 ... 8   Go Down

Author Topic: comparining micro four thirds to full frame  (Read 82766 times)

Emmett

  • Newbie
  • *
  • Offline Offline
  • Posts: 24
Re: comparining micro four thirds to full frame
« Reply #40 on: February 18, 2018, 12:37:36 pm »

Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: comparing micro four thirds to full frame
« Reply #41 on: February 18, 2018, 05:18:54 pm »

...  a D800 with the same 50mm prime but is not capable of capturing the same subject matter from the same position. If you don't have the time to move further back, or as is often the case, don't have the physical possibility of moving back, then you've lost the shot.

To give the smaller format that capability, you'd need a very high quality, and very expensive 33-50mm/F1.4 zoom.
Ray, you have in a rather quirky way illustrated my second point, on which we all seem to agree: if and when the best results in the larger format involve using a sufficiently low aperture ratio, then it is harder or impossible to get those results in a smaller format. That is, whenever you choose to be at or near the low DOF limit of the larger format.

Given the extra cost and weight involved in using a larger format, especially when equipped with the lenses that offer those large apertures and related advantages in low light/high shutter speed situations, some questions to ask are:
Would some of the images from one's current format, when displayed and viewed as one wishes, benefit from
- larger apertures than the current format allows? (For more OOF blurring, less noise at high enough shutter speeds, etc.)
- less visible shadow noise and finer tonal gradations in the deep shadows? (I agree mostly with Guillermo Luijk that a larger sensor does not give finer tonal gradations other than in the deep shadows, because elsewhere, the gradations and local SNR ratio in the smaller ILC formats are already far finer than the eye can distinguish.)
- more visible fine detail, in the sense of more "lines per picture height" of resolution? (Achieved if the sensor and the lens and aperture choice combine provide that — not true when the current format is already constrained by DOF and diffraction trade-offs.)


Of course, if the answer for the user of 4/3" format [22mm diagonal] is "yes, often enough", the next question is whether the step to a good 24x16mm [28mm diagonal] format system is enough to scratch those itches, or if instead the jump to 36x24mm [42mm diagonal] or 44x33mm [55mm diagonal] or even beyond is worthwhile.
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: comparining micro four thirds to full frame
« Reply #42 on: February 18, 2018, 05:32:53 pm »

The Nikon D800 pixel is approximately the same size as the Olympus EM5 pixel, but is slightly better quality over all, and noticeably better quality with regard to DR at base ISO.
The D800 pixels are of just under twice the area, and so about 1.4x greater linear size compared to EM5 pixels. The same is roughly true with the newer 45MP D850 and 20MP MFT sensors. In fact, this pattern has been surprising steady, with 36x24mm format having about twice the pixel count, twice the pixel area and so a 1.4x higher image resolution ("lines per picture height") and 1.4x lower sensor resolution ("lines per mm") — even since it was 5MP in 4/3", 11 to 14MP in 36x24mm.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: comparing micro four thirds to full frame
« Reply #43 on: February 18, 2018, 09:18:20 pm »

Ray, you have in a rather quirky way illustrated my second point, on which we all seem to agree: if and when the best results in the larger format involve using a sufficiently low aperture ratio, then it is harder or impossible to get those results in a smaller format. That is, whenever you choose to be at or near the low DOF limit of the larger format.

Given the extra cost and weight involved in using a larger format, especially when equipped with the lenses that offer those large apertures and related advantages in low light/high shutter speed situations, some questions to ask are:
Would some of the images from one's current format, when displayed and viewed as one wishes, benefit from
- larger apertures than the current format allows? (For more OOF blurring, less noise at high enough shutter speeds, etc.)
- less visible shadow noise and finer tonal gradations in the deep shadows? (I agree mostly with Guillermo Luijk that a larger sensor does not give finer tonal gradations other than in the deep shadows, because elsewhere, the gradations and local SNR ratio in the smaller ILC formats are already far finer than the eye can distinguish.)
- more visible fine detail, in the sense of more "lines per picture height" of resolution? (Achieved if the sensor and the lens and aperture choice combine provide that — not true when the current format is already constrained by DOF and diffraction trade-offs.)


Of course, if the answer for the user of 4/3" format [22mm diagonal] is "yes, often enough", the next question is whether the step to a good 24x16mm [28mm diagonal] format system is enough to scratch those itches, or if instead the jump to 36x24mm [42mm diagonal] or 44x33mm [55mm diagonal] or even beyond is worthwhile.

Hi BJL,
It's important to know what the differences actually are, and be aware of the visible significance of the differences in specific situations that are relevant to the nature of your photographic activities. That's why I find DXOmark a very useful site.

For me, the issues of price, weight and general flexibility of the system are very relevant. I'm always prepared to sacrifice a certain degree of technical image quality for the sake of a lighter, less expensive and more flexible system. That's why I've never been able to justify the purchase of those ridiculously expensive and heavy MF digital cameras and lenses.

However, the extra weight of a D800 body, compared with a D7000 body which has pixels of equal quality and size, is a mere 240 gms, which is also the same weight difference between the new D850 and the D7200. When I consider the 'effective' extended range of all lenses attached to the full frame body (compared with the same lens on the cropped-format body), then the slight extra weight of the full-frame body, plus the modest increase in the initial purchase price of the full-frame body, seem well-justified to me.

However, I realize that this situation of being able to use the same lenses on the two different formats does not apply to the Olympus 4/3rds system. A lens which is specifically designed for the smaller format can be lighter, so there can be a combined weight-saving of both lens and body, except when one chooses a lens which can mimic the effective focal length range, and wide-aperture DoF effects, of the equivalent lens on a full-frame.

For example, if you consider that 16mp is your minimum standard for all your purposes of future printing, cropping and display, then to get the same flexibility on an Olympus E-M5 as you could get on a Nikon D800, in relation to that required minimum pixel count of 16mp, the lens on the Olympus 4/3rds that would be fully equivalent to a standard 50mm/F2.8 lens on the D800, would need to be a 25-50mm/F1.4 zoom.

I wonder what the weight of a 25-50/F1.4 zoom designed for the 4/3rds format would be?  ;)

However, I agree we haven't really got full equivalence in this comparison because there are other advantages to a wider aperture, apart from a shallower DoF. One can use a faster shutter speed to freeze movement, and/or use a lower ISO setting in poor light, which can improve image quality.

On the other hand, the full D800 image, when downsized to the 16mp of the E-M5, will tend to have improved noise and resolution which will at least partially offset the lower noise of the Olympus shot at 25mm/F1.4, as opposed to 50mm/F2.8 on the D800. Agreed?

Also, if this is an issue, then using a Nikkor 50/F1.4 instead of 50/F2.8 removes that advantage of a 25-50mm/F1.4 zoom on the 4/3rds format. I don't think it would be realistic to expect anyone to manufacture a 20-50mm/F0.7 zoom.  :D
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: comparining micro four thirds to full frame
« Reply #44 on: February 18, 2018, 09:22:57 pm »

The D800 pixels are of just under twice the area, and so about 1.4x greater linear size compared to EM5 pixels. The same is roughly true with the newer 45MP D850 and 20MP MFT sensors. In fact, this pattern has been surprising steady, with 36x24mm format having about twice the pixel count, twice the pixel area and so a 1.4x higher image resolution ("lines per picture height") and 1.4x lower sensor resolution ("lines per mm") — even since it was 5MP in 4/3", 11 to 14MP in 36x24mm.

Yes. I was confusing the comparison between the D7000 and the D800. The 4/3rds sensor is smaller than the DX D7000 sensor, yet they both have the same number of pixels, so the MFT pixels must be smaller. The full frame Nikon FX cameras have 2.25x the area of the cropped format DX, so the 36mp of the D800 sensor divided by 2.25 = 16, the pixel count of the D7000.

Thanks for pointing that out, because there are other ramifications which I've got wrong. 50mm on the E-M5 produces an effective FF focal length of 100mm, with the full 16mp. However, 50mm on the D800, cropped to the same angle of view, results in a 9mp image, so the image from the E-M5 in these circumstances should be at least slightly better, at least in the centre of the image.

In the corners and edges, the heavily cropped D800 image might be better.

However, this discrepancy of pixel count works both ways. If one has a 25mm prime on the E-M5, and one want's the equivalent angle of view of a 100mm lens, full frame, then one has to crop the 16mp E-M5 image to just 4mp.

On the D800, to get an equivalent 100mm angle of view with a 50mm lens, one needs to crop to 9mp.
If Hugh thinks there is no quality difference between cropped 4mp and cropped 9mp images on an A2 size print, then there's even more reason to think there would be no difference between a cropped 9mp image and the full E-M5 16mp image. Agreed?




« Last Edit: February 19, 2018, 01:32:15 am by Ray »
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: comparing micro four thirds to full frame
« Reply #45 on: February 18, 2018, 10:29:01 pm »

Ray,

However, the extra weight of a D800 body, compared with a D7000 body ...
     Your comparisons of the 2010 Nikon D7000 to the 2012 Nikon D800, with roughly equal pixel size, might be relevant to your particular gear ownership, but not to this thread's main topic of "comparing micro four thirds to full frame" or to the far more common situation of the larger format having larger pixels, and this requiring longer focal lengths to get an equally detailed image of a subject. (Typically about 1.4x longer in the Four Thirds to 36x24mm comparison.) So I will leave that one and return to the main topic.

... if you consider that 16mp is your minimum standard for all your purposes of future printing, cropping and display, then to get the same flexibility on an Olympus E-M5 as you could get on a Nikon D800, in relation to that required minimum pixel count of 16mp, the lens on the Olympus 4/3rds that would be fully equivalent to a standard 50mm/F2.8 lens on the D800, would need to be a 25-50mm/F1.4 zoom.
You have a strange fascination with comparing that case of a 50mm prime and use of larger apertures (low f-stops) to what would be needed to reproduce it in a smaller format.  I instead am more interested in a portable kit offering a wide range of focal lengths well into the telephoto range, and so I ask with what I would replace my current light-weight walk-about wildlife lens, the 423g Olympus 75-300? Allowing for a 1.4x crop possible while matching pixel count, a lens for a Nikon FX body would need to cover about 150-420mm; there are lenses that roughly do so, but all are substantially heavier and more expensive. They are faster too of course, but so long as the 75-300 gets the job done for me, the extra bulk and cost does not attract me. If I did get more ambitious, I would look at the 985g Panasonic 100-400mm, covering roughly what would require 200-560mm in Nikon DX, so the 2,300g Nikon 200-500 is the closest I see.

Your comparison and mine both make sense, according to different photographic goals. The difference illustrates a general trend: that larger format advantages are most often realized at shorter focal lengths, while smaller format advantages are most noticeable towards the "long and narrow" end (also including macro, where I love my 60mm/2.8, always used at high f-stops except for focusing of course).

However, I agree we haven't really got full equivalence in this comparison because there are other advantages to a wider aperture, apart from a shallower DoF. One can use a faster shutter speed to freeze movement, and/or use a lower ISO setting in poor light, which can improve image quality.
Indeed, and that was all included in my two previous discussions of the low f-stop scenario, which includes using larger aperture sizes (larger entrance pupil diameter: focal length divide by aperture ratio) to get higher shutter speeds or less noise at the same shutter speed — this being inexorably tied (for better or for worse) to less DOF. Indeed, it seems to me that the main historic reason for using large apertures was to get adequate shutter speed with moving subjects (or shaky hand-held cameras!), with reduced DOF more often a price to be paid for that speed rather than an advantage in itself.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: comparing micro four thirds to full frame
« Reply #46 on: February 19, 2018, 02:09:53 am »

Ray,
You have a strange fascination with comparing that case of a 50mm prime and use of larger apertures (low f-stops) to what would be needed to reproduce it in a smaller format.  I instead am more interested in a portable kit offering a wide range of focal lengths well into the telephoto range, and so I ask with what I would replace my current light-weight walk-about wildlife lens, the 423g Olympus 75-300? Allowing for a 1.4x crop possible while matching pixel count, a lens for a Nikon FX body would need to cover about 150-420mm; there are lenses that roughly do so, but all are substantially heavier and more expensive. They are faster too of course, but so long as the 75-300 gets the job done for me, the extra bulk and cost does not attract me. If I did get more ambitious, I would look at the 985g Panasonic 100-400mm, covering roughly what would require 200-560mm in Nikon DX, so the 2,300g Nikon 200-500 is the closest I see.


BJL,
I've merely used the example of the 50mm lens in order to make the comparison simple. In fact, I have a general preference for zoom lenses, because of their flexibility, and a general preference for high pixel-count cameras because the lens quality of the zooms is effectively upgraded as one downsamples the images in accordance with whatever purpose one requires for the image.

Let's upgrade to the Nikon D850, which is 45mp. To get a 16mp image would require no more than a 1.7x crop. In terms of a minimum 16mp image size, the popular Nikkor 80-400/F4.5-F5.6 zoom, which I happen to use a lot, becomes the equivalent of an 47-680 zoom, in relation to your E-M5 with 75-300 lens (which is equivalent to a 150-600 zoom on full frame).

To get a Nikkor zoom lens for the D850, equivalent to the Olympus 75-300 with no pixel-count disadvantage, it would need to be a 127-353 lens. There's no such lens. The closest is the Sigma AF 70-300mm f/4-5.6 DG OS (FX), which weighs 610 grams, or just 187 grams heavier than the Olympus 75-300. Not a big deal for me.

One of the benefits of that extra weight is a much wider lens at one end. However, the slightly shorter focal length at the long end, 510mm instead of 600, can be compensated through a reduction of pixel count, from 16mp to 11 mp. No big deal.
The less wide angle of the Olympus zoom (150mm FF equivalent, compared with the 70mm of the Sigma zoom) cannot always be compensated. When needing to capture the moment, when the subject is moving, stitching might not work. You might lose the best shot you ever had the opportunity to take.  :D

In addition to those advantages of the Sigma zoom at the wide end, there are other significant advantages. At certain specific focal lengths of both lenses, from 70-300mm on the D850, and from 75-176mm on the E-M5, the D850 will produce a significantly better image in all respects. Even if you consider 45mp to be overkill and unnecessary for the prints you might want to make, when those 45mp images are downsampled to 16mp, the resolution and low noise will beat any results you could get from the finest Olympus prime lenses available.

That more expensive and slightly heavier D850 body has effectively converted a $400 Sigma zoom, and all of one's other zoom lenses, into a huge range of top notch prime lenses (or at least decent quality prime lenses), in relation to the smaller 16mp sensor of the E-M5. Wow!  ;D

And all this for a total weight increase of just 677 grams, including body and zoom lens.

Hope I've got my calculations right.  ;)
« Last Edit: February 19, 2018, 02:43:09 am by Ray »
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: comparining micro four thirds to full frame
« Reply #47 on: February 19, 2018, 07:19:49 am »

Try again with one of the 20 MP MFT bodies, restoring the 1.4x focal length factor. Actually, a better comparison is an entire kit, so how about my two lens dream of Olympus 12-100 and Panasonic 100-400. There is probably about a 500g difference in camera body weights, and I am for now ignoring the price gap.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: comparining micro four thirds to full frame
« Reply #48 on: February 19, 2018, 08:36:10 am »

Try again with one of the 20 MP MFT bodies, restoring the 1.4x focal length factor. Actually, a better comparison is an entire kit, so how about my two lens dream of Olympus 12-100 and Panasonic 100-400. There is probably about a 500g difference in camera body weights, and I am for now ignoring the price gap.

Which camera are you referring to? The 20mp Olympus Pen-F? And why use a 1.4x focal length factor?

The Nikon DX format has a 1.5x focal length factor; the Canon cropped format a 1.6x focal length factor, and the Micro 4/3rds a 2x focal length factor, comparing the long side of the frame. Comparing the height of the 4/3rd's frame with the height of the full frame 35mm format, the focal length factor is 1.8x.

I just checked the DXOmark graphs, comparing the 20mp Pen-F with the D850 at the pixel level. At the lower base ISO of the D850, the D850 pixel has almost 4dB better SNR at 18%, which translates to smoother skin tones in portraits.

DR at the pixel level is almost a whopping 2 stops better at the lower base ISO. At ISO 200 the DR advantage drops to about 1.5 stops, which is still very significant. At the nominal 400 ISO for the Pen-F, and at all higher ISOs, the DR of the D850 pixel is still one full stop better, after adjusting for differences in the actual ISO sensitivities. Refer attached image.

Need I go on?  ;D


Logged

SrMi

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 298
Re: comparining micro four thirds to full frame
« Reply #49 on: February 19, 2018, 11:52:00 am »

Which camera are you referring to? The 20mp Olympus Pen-F? And why use a 1.4x focal length factor?
 <snip>
Need I go on?  ;D

The best 20mp m43 sensor is in OM-D E-M1mII not in Pen-F (and probably also in Lumix G9, but there is no data about it yet). Interestingly, if you compare dynamic range of X1D, D850, and E-M1mII on DXOmark, you'll see that there is more difference between X1D and D850 than between D850 and M1mII. Of course, that graph is not the only relevant one.
Logged

Telecaster

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3686
Re: comparining micro four thirds to full frame
« Reply #50 on: February 19, 2018, 04:56:23 pm »

I use my m43 gear mostly for travel, where its small size and weight are big advantages. A small shoulder bag carries the lot of it, and sails easily through airport checkpoints. In use the camera and lenses attract no notice. Real-world image quality differences between a Panasonic GX8 and a Sony A7rii are, for my travel needs and purposes (YMMV), small enough to not matter. Really!

I use my larger format gear mostly with fast lenses shot wide open or nearly so. This is where it has an advantage for me. If for some reason I felt the urge to use 35mm format gear with lenses at ~f/8, after a second thought I'd likely opt for m43 at ~f/4 instead.

-Dave-
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: comparining micro four thirds to full frame
« Reply #51 on: February 19, 2018, 08:25:05 pm »

Ray,
    I was referring to any of the six MFT bodies with 20MP sensors, from either Olympus or Panasonic.

The 1.4x factor is how much longer lenses on the D850 need to be in order for a 20MP crop to cover the same field of view as on such a MFT body. The same as for the previous 16MP vs 36MP comparisons.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: comparining micro four thirds to full frame
« Reply #52 on: February 19, 2018, 08:31:11 pm »

I use my m43 gear mostly for travel, where its small size and weight are big advantages. A small shoulder bag carries the lot of it, and sails easily through airport checkpoints. In use the camera and lenses attract no notice. Real-world image quality differences between a Panasonic GX8 and a Sony A7rii are, for my travel needs and purposes (YMMV), small enough to not matter. Really!

I use my larger format gear mostly with fast lenses shot wide open or nearly so. This is where it has an advantage for me. If for some reason I felt the urge to use 35mm format gear with lenses at ~f/8, after a second thought I'd likely opt for m43 at ~f/4 instead.

-Dave-

Fair enough! Whatever satifies you and suits your purpose is the best option.  Most of my photography is done during my travels, so I'm reluctant to compromise the degree of quality that I'm used to, for the sake of a kilogram, or so, less weight. I actually go through the baggage check with my D810 and 14-24 zoom slung around my neck so it's not included in the weight of my carry-on luggage. I've never been asked to weigh the camera that's around my neck.  ;D

However, I can see the advantages of a lighter system, obviously. A few years ago I bought the Panasonic Lumix FZ200, which has a very impressive, fixed, 25-600mm/F2.8 lens attached. I wondered if this camera would satisfy me, regards image quality. I spent a lot of time comparing the image quality with equivalent focal lengths from my Canon and Nikon DSLRs, and it became clear that the Panasonic was no match, so I sold it to a neighbour.

There is also another issue which might tend to be glossed over. Whilst the image quality at the long end of an MFT zoom might compare very favourably with the significantly cropped image from a full-frame lens, the purpose of a zoom is to provide a number of different focal lengths to use, without the need to change lenses. The image quality from those wider focal lengths, from the D850 for example, will tend to be significantly better.

Let's take the example of one of BJL's dream lenses, the Panasonic 100-400, which becomes a very impressive 200-800 in full-frame equivalence. The closest equivalent zoom on the D850, to get the same reach, might be the Sigma 150-500, about 800 grams heavier.

A 1.6x crop at 500mm results in a 17.5mp image at 800mm equivalence. The resolution from the 20mp Micro 4/3rds should be better, because of the significant crop of the Sigma lens at 500mm. The difference in pixel count is insignificant, but the difference in DR will still favour the Sigma/D850 image.

The over all impression might be that the Panasonic image at 400mm is at least slightly better. It would be interesting to see real world comparisons.

However, what are the quality differences at wider focal lengths with each zoom? At all focal lengths between 100 and 250mm with the Panasonic lens, the comparison will be between 20mp and 45mp. Downsampling the 45mp image to 20mp, for the sake of a sensible comparison, the image from the D850 becomes significantly better in all respects.

The DR advantage at the lower base ISO becomes 2.5 EV. Even at ISO 200, the DR is a full 2 stops better. At all other ISO's the D850 image is around 1.5 stops better (compared with the Pen-F).

At the pixel level, the greater SNR (at 18%) is only noticeable at the lower ISO of the D850. However, when the image is downsampled to 20mp, that SNR advantage becomes much more significant. At the lower base ISO of the D850, it's about 7.1dB better, and about 5dB better at all other ISOs.

https://www.dxomark.com/Cameras/Compare/Side-by-side/Nikon-D850-versus-Olympus-PEN-F___1177_1070

Color sensitivity and tonal range are also noticeably better. I would expect resolution also to be better, even if the Sigma zoom is of lower quality than the Panasonic.


Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: comparining micro four thirds to full frame
« Reply #53 on: February 19, 2018, 08:36:14 pm »

Ray,
    I was referring to any of the six MFT bodies with 20MP sensors, from either Olympus or Panasonic.

The 1.4x factor is how much longer lenses on the D850 need to be in order for a 20MP crop to cover the same field of view as on such a MFT body. The same as for the previous 16MP vs 36MP comparisons.

BJL,
The angle of view has nothing to do with pixel count. If the sensors have a different pixel pitch, equal angles of view, or equal fields of view, will unavoidably consist of a different number of pixels, in all circumstances, at the same focal length equivalence.
« Last Edit: February 19, 2018, 08:43:12 pm by Ray »
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: comparining micro four thirds to full frame
« Reply #54 on: February 19, 2018, 09:02:08 pm »

The best 20mp m43 sensor is in OM-D E-M1mII not in Pen-F (and probably also in Lumix G9, but there is no data about it yet). Interestingly, if you compare dynamic range of X1D, D850, and E-M1mII on DXOmark, you'll see that there is more difference between X1D and D850 than between D850 and M1mII. Of course, that graph is not the only relevant one.

Thanks,
I don't keep track of all the developments in Micro 4/3rds format, so I just did a search on 20mp MFT cameras and the Pen-F came up first. I'll check the DXOMark results for the OM-D E-M1 mll.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: comparining micro four thirds to full frame
« Reply #55 on: February 19, 2018, 09:34:32 pm »

Okay! I've chacked the DXO results for the OM-D E-M1 MkII, and the DR and SNR are better than the Pen-F. However, the advantages of the Sigma 150-500 zoom on the D850 are still significant, and I presume still very noticeable.

Between the focal lengths of 100 to 250 on the Panasonic zoom lens, equivalent to 200 to 500 on the Sigma/D850 without cropping, the increased DR at the lower base ISO of the D850, has fallen to a mere 2 stop advantage. At other ISO's the advantage is sometimes as little as 1 full stop.  ;)

SNR is only around 6.3dB better, at base ISO. At all other 'equal' ISOs it's between 3 and 4dB better. Still noticeable. Resolution of the downsampled image has to be better, unless that Sigma zoom is a crap lens.
Logged

SrMi

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 298
Re: comparining micro four thirds to full frame
« Reply #56 on: February 19, 2018, 09:43:21 pm »

Thanks,
I don't keep track of all the developments in Micro 4/3rds format, so I just did a search on 20mp MFT cameras and the Pen-F came up first. I'll check the DXOMark results for the OM-D E-M1 mll.

Yes, m43 has made quite a jump forward with M1m2 and G9. I prefer using photonstophotos to DXOMark:

http://www.photonstophotos.net/Charts/DXOPDR.htm#Nikon%20D850,Olympus%20OM-D%20E-M1%20Mark%20II

Of interest may be also the dynamic range graph for the high res mode of M1m2, which shows similar dynamic range as D810/D850:

http://www.photonstophotos.net/Charts/PDR_HighResShotMode.htm#Olympus%20OM-D%20E-M1%20Mark%20II(HR)

Knowing the limitations and possibilities of m43 is crucial to extracting best image quality from the camera.

« Last Edit: February 19, 2018, 09:50:52 pm by SrMi »
Logged

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Re: comparining micro four thirds to full frame
« Reply #57 on: February 19, 2018, 10:08:32 pm »

The angle of view has nothing to do with pixel count. If the sensors have a different pixel pitch, equal angles of view, or equal fields of view, will unavoidably consist of a different number of pixels, in all circumstances, at the same focal length equivalence.
Ray, I think I do not explain myself clearly enough. I am talking about the fact that you have referred to in other ways: when a 36x24mm format body has pixel pitch 1.4x larger than a MFT body, matching both the FOV and pixel count in the larger body is possible with only 1.4x times the focal length if one is willing to crop to the pixel count of the smaller format sensor, not the full 2x times the focal length. For example, to match the FOV options that I get at 16MP with my diminutive 211g 12-50 lens and 425g EM5 body, I might not need a 24-100mm lens on a D800, but only a 24-70. Though in this case, the 485g 24-85/3.5-4.5 would be a closer match than the 1070g 24-70/2.8 in comparison the far smaller and slower 12-50/3.5-6.3. But even then, with the D8xx bodies all weighing 980g or up (slightly less than my standard kit of EM5 + 12-50 + 75-300) the Nikon kit would be far heavier, so some very big trade-offs in both directions.

By the way, you never got back to me about a proposed Nikon kit to match my dream kit of 20MP MFT body with 12-100 and 100-400 lenses. (I am guessing that there will be a 20MP EM5 Mk III this year, but as I mentioned above, there are six other 20MP MFT options; for those somewhat bigger lenses, the EM1 Mk II with its deeper handgrip might be the best fit, despite its whopping 574g.)

P. S. Please remember that I am not aspiring to match what your kit can do; in the spirit of the OP, the question is whether the IQ advantages of a larger format kit are, for my purposes, sufficient to justify that bulkier and more expensive kit. For example, I have not been bothered by insufficient DR when using the EM5 at base ISO speed, so the attraction of even more base ISO speed DR is a bit like the attraction of replacing a 200 kph car by a 300 kph one with far worse fuel economy.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Re: comparining micro four thirds to full frame
« Reply #58 on: February 20, 2018, 12:27:49 am »

Ray, I think I do not explain myself clearly enough.

BJL,
You are still not explaining yourself clearly enough. If one is using equivalent prime lenses with two different formats which have a different pixel pitch, the pixel count will inevitably be different. Adjusting the crop factor in order to equalize the pixel count causes the 'field of view' to be different. Downsizing or upsizing doesn't change the field of view.

Field of view always takes precedence over pixel count. All of my hundreds of thousands of images taken over the years, are all cropped to a specific angle of view, without exception. I would never allow pixel count to influence the composition of my images.
My first DSLR was the 6mp Canon D60. I would never think, "Oh! That fuzzy foreground needs cropping, but Hey! this will reduce the pixel count from 6mp to 4mp. Maybe I shouldn't crop."  :D

Quote
By the way, you never got back to me about a proposed Nikon kit to match my dream kit of 20MP MFT body with 12-100 and 100-400 lenses.

Easy. The Nikkor 24-120/F4 plus the Sigma 150-500. The gap between 120mm and 150mm is easily fixed with just a slight cropping of the D850 image at 120mm. The resulting pixel count and image quality would still be better than the equivalent 20mp MFT format quality between 120 and 150mm.
« Last Edit: February 20, 2018, 12:36:14 am by Ray »
Logged

Guillermo Luijk

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2005
    • http://www.guillermoluijk.com
Re: comparining micro four thirds to full frame
« Reply #59 on: February 20, 2018, 03:31:40 am »

The DR advantage at the lower base ISO becomes 2.5 EV. Even at ISO 200, the DR is a full 2 stops better. At all other ISO's the D850 image is around 1.5 stops better (compared with the Pen-F).
(...)

Ray, are you aware that DR or ISO (noise in general) differences, are becoming negligible with time? do you understand that a difference of 2,5 stops in DR between cameras with 11 and 13,5 stops of DR has not the same real world effect as the same 2,5 stops between those cameras with 8 and 10,5 stops of DR we used to have years ago?. Real scenes are not increasing their dynamic range, typical poorly lit scenes are not becoming more poorly lit with time. This means that while technology improves the advantages in noise of larger sensors are vanishing in practical terms.

I am an engineer and I love numbers and measurements, but I think you are losing the real world usage perspective. I have a FF camera with the only purpose of easily obtaining shallow DOF when I want to (something that can be enjoyed even at web size), and to adapt legacy 35mm lenses at cheap prices. Noise and DR stopped being a problem with my M4/3 years ago, by the time Olympus began to use Sony sensors.

Regards
« Last Edit: February 20, 2018, 06:28:16 am by Guillermo Luijk »
Logged
Pages: 1 2 [3] 4 5 ... 8   Go Up