comparining micro four thirds to full frame

[BJL]

There is also the point to consider that Guillermo made. As DR improves generally, as camera technology progresses, DR becomes less of an issue. It certainly has for Guillermo, but he used to be a fanatic about DR ...

Completely agreed! That is what I am actually more interested in measures of noise at various exposure levels than various attempts to define and then measure a photographically relevant flavor of "dynamic range". DXO's SNR 18% is a good start, if the horizontal axis is reinterpreted correctly.

[Ray]

While people are gram-counting, it might be worth thinking about how you carry that equipment. Gram counting may be essential for peak-bagging or through-hiking, less essential for day hikes and weekend hikes. A good pack and carriage system makes a huge difference in how the weight feels. Photo-specific packs don't fit me, even the high quality brands such as F stop. I have a short torso length (15") and need a women's extra-small or small size pack with a properly fitting shoulder harness and properly fitting hip belt. I use an ordinary panel-loading hiker's pack with a camera insert, and may also use the Cotton Carrier vest with it for instant camera access. 20 pounds of gear feels like "nothing" when not dealing with significant elevations.

Good points, Nancy. I don't have a problem with a short torso length, being 6ft tall, yet I still don't bother using my 'photo-specific' back pack with various cushioned compartments for camera bodies and lenses.

I sometimes tend to buy things because they seem a good idea at the time, then later discover there are certain inconveniences or awkwardness that make it not such a good idea. My photo-specific back pack has remained in the storage shed, unused for the past 15 years or so.

I always hike with a small and ordinary, lightweight, back pack, such as a cyclist might use on a push bike. I place some bubble-wrap material in the base of the back pack in case the bag is dropped to the ground too suddenly, perhaps damaging the equipment. The weight on my back tends to counter balance the weight of one of my cameras and lenses on my front.

Nevertheless, I can see the advantages of light equipment. I have a Nikon D5300 with Nikkor 18-140 zoom, which weighs in total just over 1Kg, roughly equivalent to an E-M1 MkII with 12-100 zoom.

[Ray]

Ray, can you define EXPOSE To The Right [ETTR] and explain why — with your definition — an ETTR exposure is relevant in "low light" situations where the exposure index needed well above the "base ISO speed"? (Which by the way is a nickname for what ISO 12232 defines as SSat, the basis for DXO's "true ISO".)

BJL,
As I've mentioned before, whatever the ISO setting, one can overexpose or underexpose. At higher than base ISO, the internal amplification of the signal is fixed in accordance with the ISO setting and the camera design. The photographer doesn't have any control over that degree of amplification in relation to a specific ISO setting.

Most cameras are not ISO invariant. An underexposure at any ISO results in an image with more noise than an ETTR shot at the same ISO. In other words, an ETTR at ISO 1600 will usually produce a better image than the same shutter speed used at ISO 800, except for truly ISO invariant cameras.

An ETTR shot is the maximum exposure, or the slowest shutter speed, that does not result in a loss of detail in the brightest part of the scene that one considers relevant to the composition.

That will generally exclude areas such as spectral highlights and small patches of sky visible through the leaves of a tree, or any bright patches which are devoid of detail or which one might consider to be less significant than the detail in the shadows of the scene.

There are various techniques of achieving an ETTR exposure. One method, which I tried years ago when using my Canon 5D, was to use a single focusing square in 'spot meter' mode, direct the square at the brightest part of the scene, such as the brightest part of the sky, take note of the shutter speed in the viewfinder, then manually reduce the shutter speed by 3 stops, or maybe it was 2.5 stops, can't remember. It's different for different models of camera.

Unfortunately, I found such a process  rather slow and cumbersome. It was quicker and easier to simply bracket exposures for all shots, +/- at least 1 EV, in auto-exposure mode, then choose the most appropriate shot in Photoshop's Camera RAW, which would usually be the shot which looked slightly overexposed before adjustments were made to retrieve highlight detail.

Now that I'm using Nikon equipment, I find it quicker and easier to use the AF-On button which separates both focusing and exposure from the shutter button. All I have to do is observe the exposure meter in the viewfinder as I swing the single focusing square to the brightest part of the scene, and adjust the shutter speed with my thumb.

The exposure meter has a zero in the centre, a minus sign at the extreme left, and a plus sign at the extreme right. With the focusing square covering the brightest part of the scene, I adjust the shutter speed so that the meter reading in the viewfinder is at the far right. Literally, an exposure to the right. No need to fuss around examining the histogram.

[NancyP]

One other slightly off-topic item to consider is whether you tend to concentrate on one genre per photography outing, or several, and whether you are willing to forego some equipment options to save weight (or brain-space  ). At this point, I am day hiking mostly, and choose between birding-specific and landscape / macro options. I can get by with a traditional DSLR. The smaller format weight advantage shines when one is doing long-distance hiking or traveling, and one wants to have all options on hand.

One of my favorites for lightweight day hiking is the Canon 6D plus pancake 40mm f/2.8 - a PITA to manually focus (it is focus by wire), but this is really a nice landscape lens stopped down. I like the 35mm to 40mm angle of view in local wooded landscapes (hint, Canon or other - 35 mm tilt/shift?). I throw in a light macro lens (Voigtlander 125 mm) if I want to capture any incidental flora, fungi, or insects as well.

Light but good tripods are another issue. I have been happy with my full-size Feisol carbon fiber 4-section tripod CT-3442 plus an Arca-Swiss p0 head with  Sunwayfoto screw Arca-style QR clamp, total circa 2.8 pounds / 1.3 kg, for landscape / macro. However, with a smaller-format kit, I suppose that the really tiny "travel tripods" like the Sirui T025X (2 pounds with head) would be more than adequate.

[Ray]

One other slightly off-topic item to consider is whether you tend to concentrate on one genre per photography outing, or several, and whether you are willing to forego some equipment options to save weight (or brain-space  ). At this point, I am day hiking mostly, and choose between birding-specific and landscape / macro options. I can get by with a traditional DSLR. The smaller format weight advantage shines when one is doing long-distance hiking or traveling, and one wants to have all options on hand.

True. I tend to be a peripatetic photographer, wandering from place to place where the unexpected scene can appear at any moment, whether the appearance of a beautiful landscape as the track bends around the corner, a Tibetan in traditional costume racing around the bend on a horse, or a flock of vultures in a nearby field feeding on some dead animal.

My ideal camera for such circumstances would be something like the DX Nikon D7200 with a really high quality DX 18-300mm F3.5-F6.3 zoom (which of course doesn't exist). That would result in an effective 27-450mm range in full-frame terms. Cropping the 24mp sensor to just 6mp would double the range to 900mm which would be sufficient, if the lens were of high quality, for most bird and wildlife shots, prints up to A3 size, and HDTV display.

Unfortunately, such lightweight, DX zoom lenses tend not to be sharp at the long end and generally show poor resolution at the edges and corners, so there is always a significant compromise in image quality for the convenience of that lighter weight and more useful FL range.

I can't help wondering if there is a technical obstacle to producing a lightweight zoom lens with such a range, which is really high quality. It would obviously have to be more expensive than the current Nikkor 18-300mm DX zooms, but I guess such a lens, if possible to design, would undermine the sales of existing heavier lenses.

[BJL]

BJL,
As I've mentioned before, whatever the ISO setting, one can overexpose or underexpose. At higher than base ISO, the internal amplification of the signal is fixed in accordance with the ISO setting and the camera design. The photographer doesn't have any control over that degree of amplification in relation to a specific ISO setting.

I agree that with a particular camera at a particular EI setting, that IF you can increase exposure to have the raw histogram at the right edge, it is better than giving less exposure. But that fact is of little use when judging or predicting IQ differences between different cameras that might apply a different amount of amplification at the same EI setting, or even comparing at different EI settings on the same camera. It is also a bit irrelevant in low light situations where one cannot vary the exposure because it is fixed (at well below full well capacity) by limits of minimum usable shutter speed and largest usable aperture. Then the question instead is something like deciding the best EI setting (analog amplification level) — and also deciding how much changes in EI setting matter: often very little in "near ISO-less" cameras.

However, we have dragged this thread far from its title, and I am not sure that anyone else cares! So maybe I will start another thread to discuss the use and abuse of "EXPOSE To The Right" and its evil twin "Raw Histogram After Amplification To The Right".

[Ray]

I agree that with a particular camera at a particular EI setting, that IF you can increase exposure to have the raw histogram at the right edge, it is better than giving less exposure. But that fact is of little use when judging or predicting IQ differences between different cameras that might apply a different amount of amplification at the same EI setting, or even comparing at different EI settings on the same camera. It is also a bit irrelevant in low light situations where one cannot vary the exposure because it is fixed (at well below full well capacity) by limits of minimum usable shutter speed and largest usable aperture. Then the question instead is something like deciding the best EI setting (analog amplification level) — and also deciding how much changes in EI setting matter: often very little in "near ISO-less" cameras.

BJL,
Sorry! Can't follow your logic. The purpose of the DXO measurements is to enable you to predict the IQ differences between different cameras that do in fact apply different amounts and different qualities of amplification at the same ISO setting.

How a camera behaves at a particular ISO setting is something that is useful to know in order to make the best choice of camera purchase, and also the best choices of ISO setting and F/stop, in relation to the desired shutter speed, when using the camera.

For example, it is well known that Canon DSLRs at base ISO have much worse DR than Nikon DSLRs. The Nikon D810 has a full 2 stops better DR than the Canon 5DSR at their nominated ISOs of 100, which are measured as being almost identical at ISO 77 for the 5DSR and ISO 75 for the D810.

However, as ISO settings are increased, that DR advantage of the D810 decreases. From ISO 800 onward, the DR is basically the same for both cameras, as well as the SNR. Isn't that useful to know?

Supposing, for example, that you are shooting a landscape where you consider that both DR and DoF is important, so you want to use your 5DSR at base ISO and and an f/stop of 16. Unfortunately, you haven't got your tripod with you and realize that at F16 and base ISO the shutter speed will be too slow for a sharp, hand-held shot, so an ISO setting of 200 might be more appropriate. This might result in a period of indecision if you don't know how your camera behaves at different ISOs. Is it better to sacrifice a bit of DoF and use F11, or better to sacrifice perhaps a full stop of DR for the sake of more DoF?

However, if you have examined DXO's test results for the Canon 5DSR, you will understand that the DR at ISO 200 is almost identical to the DR at the base ISO of 100. The difference is a mere 0.06 EV, of no consequence and probably within the margin of testing error.
If one is using the Nikon D810 in the same circumstances, increasing ISO to 200 will result in a drop of almost a full stop of DR, but that reduced DR is still more than a full stop better than the Canon 5DSR at ISO 200. Is that not useful to know?

Another example of the practical use of being aware of the different qualities of signal amplification in different cameras, relates to my first full-frame DSLR, the Canon 5D. It had impressively low noise at high ISOs, such as ISO 1600 and 3200 which I often used in low-light situations

However, it became widely known through various tests, before DXO began publishing its results, that using ISO 3200 with the Canon 5D served no purpose for those shooting in RAW mode. The camera merely amplified the signal without any further reduction in noise. An underexposure at ISO 1600 resulted in the same quality of image when the same exposure was used at ISO 3200, after appropriate adjustments in Photoshop's Camera Raw.

The advantage of underexposing at ISO 1600, instead of attempting an ETTR shot at ISO 3200, was better detail in the highlights, and/or lower risk of completely blowing some highlights. Is that sort of information not useful to know?

[Guillermo Luijk]

It certainly has for Guillermo, but he used to be a fanatic about DR, even promoting a procedure known as UniWB to extract slightly more DR from an image by avoiding the clipping of perhaps just one color that could result from a camera-selected white balance.

I was never that fanatical. 

I have never been a fanatical of anything but Alfa Romeo. If I made up a procedure for UniWB is just because at that time it was useful for my camera's poor DR and mostly because doing it was a beautiful engineering exercise. Using UniWB today is of very limited use for the reasons I gave here.

You should improve your deduction skills. Probably you consider David Coffin, a guy with 'DCRAW' on his car's plate a RAW fanatic.



Dave shoots JPEG.

Looking at your endless arguments on this thread makes me think you don't have a very reallistic vision of your own fanatism either

Regards

[DP]

btw what coffin is up to ? dcraw updates are long gone, no ?

[Guillermo Luijk]

He's not anymore interested in maintaining the code, but alive and living happily. DNG is very useful in this case.

[Ray]

You should improve your deduction skills.

How are your own deduction skills, Guillermo? Why do you appear to have ignored the 'smiley' in my comment about your fanaticism? 

I understand quite well that DR becomes less of an issue as sensor DR improves generally. However, when comparing Micro Four Thirds to a Full Frame format which usually has a much higher pixel count, there are many other performance advantages of the full-frame format, which should be considered in total.

One major advantage which I've already addressed, is the greater range of 'equivalent' focal lengths with any given lens on full-frame, in relation to the minimum standard image quality and image size of the MFT format under comparison.

In other words, a 50mm prime lens on the full-frame format becomes an effective 50-100 mm zoom, equivalent to a 25-50 mm zoom on the MFT format, with possibly slightly worse image quality towards the 100mm end, depending on which camera models are compared, but almost certainly better image quality towards the short end when the larger full-frame image is downsized to the MFT size.

[HSakols]

Nancy,

When it comes to hiking more than a couple of miles, I don't use a photo backpack.  Instead, I put my camera and lenses into smaller cases and put those in one of my Ospray backpacks.  This is much more comfortable.  Yes, my wife is only 5frt so I understand the difficulty of finding the right fit. 

Ray,

You have discussed dynamic range but what about comparing a high frequency images (eg a bush with lots of branches) to say an image of a mountain?

[Ray]

Ray,

You have discussed dynamic range but what about comparing a high frequency images (eg a bush with lots of branches) to say an image of a mountain?

Hugh,
I'd be very interested to see such comparisons. At the wider end of any lens, prime or zoom, I'd expect the full-frame 35mm format to show a clearly sharper and less noisy image.

However, at the long end of the zoom effect, comparing the full frame of an MFT shot using a 50mm lens, with the cropped image of a D850 shot using a 50mm lens, (to make it an effective 100mm shot), the MFT shot would probably show greater  resolution, at least in the centre, depending on the quality of the lenses used and which cameras are being compared.

At the edges and corners, the cropped D850 image would probably be sharper than an E-M1 MkII shot at 50mm.

If I had the time and the motivation, I'd try to hire an E-M1 MkII with lenses, in order to do my own comparisons, to see how significant any differences are, not only between MFT and my Nikon FX cameras but also my DX cameras.

[Frodo]

Lots of theoretical analysis.
Or you could look here: https://www.dpreview.com/reviews/nikon-d850/8
The D850 beats the Olympus hands down.
As for the theoretical argument about equivalent DoF, get a cellphone. I find the shallow depth of field of FF one of the advantages for subject isolation, e.g. for weddings.
But I walked 450km through the Swiss Alps in 2016. I took my Canon M3 and two lenses and saved over a kilogram.
Horses for courses.

[BJL]

Lots of theoretical analysis.
Or you could look here: https://www.dpreview.com/reviews/nikon-d850/8
The D850 beats the Olympus hands down.

Of course it does — at least in situations where very high resolution or very shallow depth of field or handling scenes of very great subject brightness range without exposure bracketing are sufficiently important: the question here is how big the differences are in various situations, and if and when they are worth the extra cost/weight/bulk involved. As you conclude,

... But I walked 450km through the Swiss Alps in 2016. I took my Canon M3 and two lenses and saved over a kilogram.
Horses for courses.

As for the theoretical argument about equivalent DoF, get a cellphone. I find the shallow depth of field of FF one of the advantages for subject isolation, e.g. for weddings.

Let's not play this game of "people who are not as demanding as me about some aspect of performance will therefore be satisfied with the very low end of the performance spectrum". There is a huge difference between a cell phone's "f/16 equivalent DOF all the time, and the limited low light abilities that go with it" and what is available in 4/3" format — or APS-C, or even 1" format.

[Frodo]

Hi BJL
Totally agree with all your points.
We are on the same page.

[HSakols]

OK here are my rough comparisons between a Nikon D800 160mm f16 at iso 100 and an Olympus EM5 I 75mm f8 at iso 200.  I used a Nikon 70-200 f4 and the Olympus 75mm f1.8

[Ray]

OK here are my rough comparisons between a Nikon D800 160mm f16 at iso 100 and an Olympus EM5 I 75mm f8 at iso 200.  I used a Nikon 70-200 f4 and the Olympus 75mm f1.8

Thanks for that, Hugh. The EM5 I image definitely looks better. It's just a pity the comparison is so rough. Perhaps we could say, for those who shoot rough, the EM5 I is a better camera. 

For a start, it's not clear if these images are out-of-camera jpegs. They look like it. The D800 shot look underexposed.

When comparing the image quality from different cameras, especially when comparing resolution and noise in the shadows, it's essential to start off with RAW images that are fully exposed (ETTR), and ensure that the shots of the same scene have the same lighting, otherwise the comparisons are not valid.

Having download your images into ACR, I am very puzzled as to the exposures. According to DXOMark, the ISO sensitivities of both cameras are approximately equal. The EM5 I is a mere 1/4th of a stop less sensitive.

The metadata shows that the shutter speed for the Nikon shot at ISO 100 and F16 is 1.6 seconds. For the EM5 I at ISO 200 and F8, the shutter speed is 0.6 seconds.

If we equalize the ISO and the f/stop for both cameras, the shutter speed should be approximately the same, but it's not. There's a significant difference. What's going on?

At ISO 200 and F8, the shutter speed for the Nikon shot should be 1/8th of 1.6 seconds, which is 0.2 seconds, about 1.5 stops less exposure that the 0.6 second of the EM5.

Nevertheless, when an extensive DoF is required, the ability to use F8 instead of F16, or F11 instead of F22, with the MFT format, is a significant advantage, which I don't deny, all else being equal.

[HSakols]

Ray,
I shot both of these using raw and didn't adjust the images.  But of course I converted to JPEG to post.  Yes the D800 image is underexposed.  Still I see better details in the D800 image - look at the glasses on the table.  Also there may be more noise in the the Olympus image.  But at a printed size of 16x20?   Yes, eventually I'm getting my hands on a D850, but I'll have to leave it home when I go on backpacking adventures because good lenses are just too heavy.   

[Ray]

Ray,
I shot both of these using raw and didn't adjust the images.  But of course I converted to JPEG to post.  Yes the D800 image is underexposed.  Still I see better details in the D800 image - look at the glasses on the table.  Also there may be more noise in the the Olympus image.  But at a printed size of 16x20?   Yes, eventually I'm getting my hands on a D850, but I'll have to leave it home when I go on backpacking adventures because good lenses are just too heavy.

Hugh,
If the 12mp of the Em5 Mk I is sufficient for your usual print size, then the advantages of the cropping options offered by the 45mp D850 should be taken into consideration when choosing lenses.

For example, the very light AF-S Nikkor 85mm F1.8G prime becomes effectively an 85-170mm/F1.8 zoom, with a maximum aperture of F1.8 across the entire zoom range, in relation to the 12mp of the OM5 I (approximately).

How much would such a zoom lens designed for the MFT format weigh and cost, with a maximum aperture of F1.8 at all focal lengths?  The Nikkor 85/F1.8 prime weighs only 350 grams and costs less than US$500.