Olympus versus Full Frame

[Mark Johnson]

Olympus and all the other smaller sensor cameras................

Is the size of the sensor very relevant??

Or is a matter of pixels, ie 16 or 20 or 28..............

Without wanting to blame tools )) I wonder whether a full frame camera / sensor is able to record more detail than M43 cameras.

I don't think the lens is a factor, but suspect the sensor size might be?

Mj

 

[Mark twiglet]

A top of the range camera body with a milk bottle on....... Verses a starter slr with a decent lens on.......

What's better?

 

[john.margetts]

The limits of a smaller sensor will produce less detail than the limits of a larger sensor (all else being equal). It is a simple matter of how much magnification is required to produce the final viewed image.

 

[Mark Johnson]

A top of the range camera body with a milk bottle on....... Verses a starter slr with a decent lens on.......

What's better?

Yea but...........

Not quite sure what you are getting at. Are you saying the lens is all important?

I accept the lens is very important, but the q was about the sensor.....

 

[Mark Johnson]

The limits of a smaller sensor will produce less detail than the limits of a larger sensor (all else being equal). It is a simple matter of how much magnification is required to produce the final viewed image.

Sounds though you are saying the sensor size does matter

I am comparing my astro shots taken with aEM10 and Em1 against some really superb shots taken with =Canons and Nikons. I am assuming=g they are full frame.

My photos are just not as sharp and don't have the detail of some of these bigger cameras.

Could be technique I guess, though I don't see how to improve that aspect.

Could be that I live in area with a lot of light pollution................

 

[damianmkv]

I don't think that the early OMDs are fantastic for Astro / LE, certainly the em1 is considerably worse than the em10 and em5 from what I've seen. Both the below have had PP and an amount of NR

Em1

nightime picnic by damianmkv, on Flickr

Rx100

the milky way by damianmkv, on Flickr

So, you should be able to get sharp pictures with Olympus but the noise will be nothing like FF - take a look at d750s at iso12800 for example, and the images are still very clean.

So, you want a wide quick lens so that you can shoot and keep the ISO down with m43.

And as said above, if you need to crop, the FF will retain more detail. I'm considering dumping everything for a d750 but it would cost me £2000 to change and I'd end up with a body and 2 lenses ( 24 to 200mm ) instead of what I have now ( 7.5-210mm or 15-420mm in "35mm equivalent" )

 

[snerkler]

Bigger sensors will give better IQ with equal lenses and similar aged tech. How much better will vary, and will depend on the viewer.

I ran two systems for a while, the D750 and EM5-II. In certain scenarios the difference was pretty imperceivable, in some situations the difference was quite considerable.

Astro is one of the scenarios where you're likely to see more difference.

 

[woof woof]

Olympus and all the other smaller sensor cameras................

Is the size of the sensor very relevant??

Or is a matter of pixels, ie 16 or 20 or 28..............

Without wanting to blame tools )) I wonder whether a full frame camera / sensor is able to record more detail than M43 cameras.

I don't think the lens is a factor, but suspect the sensor size might be?

Mj

Sounds though you are saying the sensor size does matter

I am comparing my astro shots taken with aEM10 and Em1 against some really superb shots taken with =Canons and Nikons. I am assuming=g they are full frame.

My photos are just not as sharp and don't have the detail of some of these bigger cameras.

Could be technique I guess, though I don't see how to improve that aspect.

Could be that I live in area with a lot of light pollution................

I have Panasonic 16mp cameras which should be more or less the same as an Olympus camera and I also have a 24mp Sony A7 and these are my thoughts after comparing the two...

Assuming you use a good lens on each and account for the crop factor by using lenses of the same field of view and adjust the MFT camera aperture accordingly (for example use a 35mm lens at f8 on FF and a 17mm lens at f4 on MFT) the FF camera will be sharper if you go pixel peeping. If noise isn't going to be a problem I'd expect more detail to be recorded as the mp count rises assuming the lens is up to the job.

As John says, the amount of magnification will hurt the MFT setup so you'd need to start with a very good starting point to match FF for close inspection and you may therefore want to be careful with lens selection as with any system some are going to be better than others.

 

[HoppyUK]

If it's image quality you're after (all aspects of) then the order of merit runs: sensor size, lens quality, pixels. In terms of noise, low light capability and high ISO performance, I'd also add that some of the latest generation sensors ('ISO-invariant' for want of a better description) are astonishingly good.

 

[woof woof]

If it's image quality you're after (all aspects of) then the order of merit runs: sensor size, lens quality, pixels. In terms of noise, low light capability and high ISO performance, I'd also add that some of the latest generation sensors ('ISO-invariant' for want of a better description) are astonishingly good.

Similar is said of audio... source, amp, speakers... but of course each link in the chain does matter and just as there's little point buying a top end audio source and playing it through your transistor radio there's little point expecting top end FF image quality if the shot is spoilt by noise banding due to a big but behind the times sensor or a poor lens.

I know that's obvious but I felt someone had to say it

 

[HoppyUK]

Similar is said of audio... source, amp, speakers... but of course each link in the chain does matter and just as there's little point buying a top end audio source and playing it through your transistor radio there's little point expecting top end FF image quality if the shot is spoilt by noise banding due to a big but behind the times sensor or a poor lens.

I know that's obvious but I felt someone had to say it

A chain is only as strong as its weakest link.

 

[frod]

With regards to audio, your money spent on speakers will go a lot further, once you've reached a certain acceptable minimum on the other two links. I don't like chain discussions because of this kind of thing. Nobody spends enough on speakers.

 

[gad-westy]

Agree with a above. Like for like (lens quality etc), a larger sensor will record more detail than a smaller sensor. The only question after that is whether you need all that detail. The answer is usually 'sometimes'. Depends what you shoot

 

[snerkler]

Agree with a above. Like for like (lens quality etc), a larger sensor will record more detail than a smaller sensor. The only question after that is whether you need all that detail. The answer is usually 'sometimes'. Depends what you shoot

I don't think it is just detail that's better though. FF seems to have more 'depth' to the image to my eyes. Also FF files are more malleable in PP and they crop better too.

 

[Mark twiglet]

Yea but...........

Not quite sure what you are getting at. Are you saying the lens is all important?

I accept the lens is very important, but the q was about the sensor.....

You said the lens was not a factor!

M4/3 is definately catching up with fx! The new Olympus is great! You can get a full frame Sony mirror less system that is bloody fantastic! I nearly chopped in all my fx nikon stuff in last year for the Sony a7r

 

[snerkler]

You said the lens was not a factor!

M4/3 is definately catching up with fx! The new Olympus is great! You can get a full frame Sony bridge system that is bloody fantastic! I nearly chopped in all my fx nikon stuff in last year for the Sony a7r

A7 is a compact system camera (aka mirrorless), bridge cameras are very different beasts

 

[Mark twiglet]

A7 is a compact system camera (aka mirrorless), bridge cameras are very different beasts

Ok all I meant was it was smaller but with big sensor

 

[woof woof]

With regards to audio, your money spent on speakers will go a lot further, once you've reached a certain acceptable minimum on the other two links. I don't like chain discussions because of this kind of thing. Nobody spends enough on speakers.

I know we're getting off track here but I have some relatively cheap Monitor Audio speakers and they're very characterful. They have a sound I've not heard from any other and with the right track they're stunning. Just like an old lens like my Minolta Rokkor 50mm f1.2 when you want that look. I also have some more expensive Tannoy's that are more like my Sony 55mm f1.8, incredibly accurate, almost technically faultless and stunning for that.

Every component matters

 

[Mark Johnson]

Speakers!!! Audio......

I think the drift is that FF is probably better for Astro......... where minute detail is important.

For most of my efforts in photo's the EM1 and Em10 are more than adequate for my skill level. But just looking at moon and star shots, the camera and lens are not as good as the bigger cameras.

When I use my 8" scope, I het good detailed results as the magnification or image size is so much more..

Thanks all

Just occurred to me that If I want high detailed shots, the scope is the way to go!!

 

[Phil V]

Whilst quality continues to improve...

...
M4/3 is definately catching up with fx!

Clearly this is nonsense, because quality improves across the board, and the inherent advantages of a larger system are a constant (what with them being dependant on well understood mathematical principals and all).
So in 10 years time when M4/3 systems are better than current FF offerings, the contemporary FF systems will still be better. What's more I'd predict ever larger and cheaper MF systems which will raise the bar even further, so M4/3 systems will be even further behind 'the best'.

 

[woof woof]

I think you need to look at the whole thing and not just the maximum achievable image quality and if you use kit to it's strengths the playing field can be levelled just a bit. Get a nice compact MFT camera and lens and shoot considering the crop factor and MFT can look a lot like FF until the limits are pushed and you look closely and go looking for differences.

For example my MFT GX7+17mm at f4 gives a look very similar to my FF Sony A7 with 35mm at f8. The Sony gives the ultimately better image quality if you go looking for it so the question for people considering the MFT system is what advantages does it offer. I expect the same to be true when/if MF drops into the more affordable price range and who knows, when/if that happens MFT may still make sense for some people and some uses.

 

[odd jim]

You said the lens was not a factor!

M4/3 is definately catching up with fx! The new Olympus is great! You can get a full frame Sony mirror less system that is bloody fantastic! I nearly chopped in all my fx nikon stuff in last year for the Sony a7r

It's not catching up - it's all relative. As sensor tech for the Olympus / m43 etc sensors improves, so does that of full frame. All sensor tech improves. FF will always have the edge generation for generation simply due to the laws of physics and the advantages that entails (sharper, more dynamic range, contrast, low levels of noise / cleaner images, etc etc) and of course shallower dof, etc etc, and it holds that advantage for longer, relatively speaking...

I liken it to performance cars. You can buy hot hatches today with the same power super cars had 25 years ago. But super cars have also moved on, and now have triple the power of yesteryears supercars, so of course they're well ahead

 

[snerkler]

Speakers!!! Audio......

I think the drift is that FF is probably better for Astro......... where minute detail is important.

It's not just the detail due to the physics of the sensor size and demands on the lenses, it's the noise handling at high ISO and low light. FF will give you cleaner images than m4/3.

 

[Mark Johnson]

Is the pixel in my MFT better or worse than the pixel in a FF xyz?

If not it must just a factor of cropping/magnification. The FF has a bigger sensor, but not many more pixels proportionally to the MFT.

If the pixel quality is the same, and the quality should be much the same proportionally to the lens size?

Any one know what I am talking about

 

[woof woof]

Is the pixel in my MFT better or worse than the pixel in a FF xyz?

If not it must just a factor of cropping/magnification. The FF has a bigger sensor, but not many more pixels proportionally to the MFT.

If the pixel quality is the same, and the quality should be much the same proportionally to the lens size?

Any one know what I am talking about

I think it's more than just simply the pixels, I think it's more the wider technology. On chip A/D conversion, backlighting, microlenses and all that Voodoo... etc. It's difficult to make like for like comparisons as you can rarely if ever find cameras with different sensor sizes and other than that with exactly the same technology.

The easiest thing to do is to take a FF camera, take a picture and crop it. What you'll then have is a FF image and a MFT image taken with the same camera and the only difference in the final image quality will be due solely to the magnification of the image to produce the final picture. The FF image will be sharper. That will IMO be the most obvious difference.

 

[Phil V]

Is the pixel in my MFT better or worse than the pixel in a FF xyz?

If not it must just a factor of cropping/magnification. The FF has a bigger sensor, but not many more pixels proportionally to the MFT.

If the pixel quality is the same, and the quality should be much the same proportionally to the lens size?

Any one know what I am talking about

We know what you're talking about, but you're missing about 90% of the story.

In photography, size really does matter, a 16 megapixel sensor in a camera phone can't compete with an 8mp crop camera, which is also inferior to a similar MP FF camera, it's about resolving detail, signal to noise and some other optical phenomena.

 

[Mark Johnson]

Yea but.......

mines a pint, lets get on with life

 

[ancient_mariner]

We have an EM10 and a D610 (and Sony A58) here. All things being equal in terms of lens quality, FF (and good APS-C) is better than M43 for noise, dynamic range and resolution. M43 will still give very usable images from a small and light camera system, but cannot compete in extreme situations where noise DR and resolution are factors determining whether the image is acceptable or not. Buy your tools according to what you need.

 

[Mark twiglet]

If we were projecting images onto the moon I'd get a 36mp sensor! But bit pointless otherwise!

 

[odd jim]

If we were projecting images onto the moon I'd get a 36mp sensor! But bit pointless otherwise!

This isn't about MP. The advantages of FF don't have anything to do with image size (Some compact 1" sensors produce images bigger than APS-C sensors) and the advantages of FF can easily be seen at screen level, especially when processing raws.

 

[snerkler]

We have an EM10 and a D610 (and Sony A58) here. All things being equal in terms of lens quality, FF (and good APS-C) is better than M43 for noise, dynamic range and resolution. M43 will still give very usable images from a small and light camera system, but cannot compete in extreme situations where noise DR and resolution are factors determining whether the image is acceptable or not. Buy your tools according to what you need.

TBH DR seems to have as much to do with tech as sensor size. For example the Olympus EM5-II has more DR at base than the Canon 5D3 which is one of the best regarded FF cameras (the 5D3 soon overtakes as ISO increases). Yes they've improved the DR with the 5D4 but this still falls short of others such as the D7200 which is 'only' crop sensor. I used to think sensor size was the determining factor of base DR but my guess is that the processor plays a large part nowadays. Perhaps someone who knows can enlighten me?

 

[odd jim]

TBH DR seems to have as much to do with tech as sensor size. For example the Olympus EM5-II has more DR at base than the Canon 5D3 which is one of the best regarded FF cameras (the 5D3 soon overtakes as ISO increases). Yes they've improved the DR with the 5D4 but this still falls short of others such as the D7200 which is 'only' crop sensor. I used to think sensor size was the determining factor of base DR but my guess is that the processor plays a large part nowadays. Perhaps someone who knows can enlighten me?

...but only if you look at DXO which needs to be taken with a pinch of salt. They say my mk1 RX100 has more DR than my FF 6d - absolutely no chance (as good as the rx100 clearly is), the 6d nukes it for DR (and the resolved detail therein).

 

[Swanseajack]

@Mark Johnson ........Mark I'm of to Cornwall for a week this coming Monday, but when I'm back I'll drop you a PM or do you still have my number? As you may know, I have the D810 now and you're more than welcome to have a play to see if it's what you need. I have a long lens @ 600mm and a 20mm f1.8 lens plus others, not sure if this would be suitable for your Astro photography needs, but we can try them out. Bring your 300mm lens and you can compare with my longer lenses if you need.

Bring plenty of cards as the D810 files are big.

ETA: I also recently bought a Panasonic GX80 for a smaller system which has some of the latest tech, (I then sold after a couple of weeks) as there was just no comparison to the D810, as expected, and I decided I would rather take that out with some lightweight primes for my needs.

 

[Mark Johnson]

Thanks for that Simon. Be good to have that comparison directly....

Hope the trip to Sunny Cornwall goes well, used to live in Pendeen for a while.

Its always a toss up between size and capability. M43 has suited me for some while now, but I can see here the "quality" of astro shots do not match up to the bigger cameras.

May just be I use the scope for astro, ........................

Mj

 

[ianpiper]

Olympus and all the other smaller sensor cameras................

Is the size of the sensor very relevant??

Or is a matter of pixels, ie 16 or 20 or 28..............

Without wanting to blame tools )) I wonder whether a full frame camera / sensor is able to record more detail than M43 cameras.

I don't think the lens is a factor, but suspect the sensor size might be?

Mj

The sensor size and resolution are both factors, as is the glass you stick in front of it. Zack Arias has a great video where he talks about the different sensor sizes and formats, and his view is that the differences between successive sizes are not that important (or as Zack says, neg-li-gible). I can't post the link because I'm new here, but if you google zack arias sensor size it's the first hit. It's quite amusing too.

There's another, slightly counter-intuitive factor that confused me for a long time; the size of the individual image-gathering cells on a senosr. Think of these like buckets for gathering photons. The bigger, the better, with lower noise in the final image and maybe better tonal range too. The same number of MP in a bigger sensor gives bigger cells. Just packing more MP into the same sensor size will result in smaller cells.

So here are some examples.

First, my X-T1 is an APS-C with 16 MP. That means it has 4896 cells across its width (23.7mm) and 3264 cells across its height (15.7mm). The individual light-gathering cells are thus 0.00483mm in width and height.

Now take a Nikon D810 sensor. It offers 36.3 MP, with a sensor that has 7360 x 4912 pixels in 36mm x 24mm. The individual image gathering cells are 0.00489mm, only a tiny fraction larger than on the X-T1. In other words, the image-gathering cells in a 16 MP APS-C sensor are about the same size as those in a 36 MP full-frame sensor - there are just fewer of them. The full-frame sensor is not giving you more resolution, it is giving you bigger pictures at the same resolution. I guess this explains why the flagship full-frame cameras have a smaller number of pixels in the sensor; each pixel is capable of gathering more light.

So I've stopped worrying about just increasing the number of megapixels in the sensor alone to improve the image. An MFT sensor won't necessarily give worse results than an APS-C sensor - it depends on the size of the individual pixels (do the same calculations with a 12 MP MFT camera and you get cell sizes of 0.00446mm, again not much different from the X-T1 or D810).

 

[woof woof]

ETA: I also recently bought a Panasonic GX80 for a smaller system which has some of the latest tech, (I then sold after a couple of weeks) as there was just no comparison to the D810, as expected, and I decided I would rather take that out with some lightweight primes for my needs.

I have to wonder about this as the GX80 should be a good little camera. I have slightly older cameras with afaik the same chip and they're quite good performers. FF will give you better image quality but imo you must have gone through a critical testing regime or be looking at exacting situations and requirements to give up and dismiss the camera after two weeks.

What lenses and settings did you use on the GX80 and what lead you to sell it so quickly?

 

[Swanseajack]

I have to wonder about this as the GX80 should be a good little camera. I have slightly older cameras with afaik the same chip and they're quite good performers. FF will give you better image quality but imo you must have gone through a critical testing regime or be looking at exacting situations and requirements to give up and dismiss the camera after two weeks.

What lenses and settings did you use on the GX80 and what lead you to sell it so quickly?

There were a couple of issues where the camera would over expose and the files were mushy although intermittent. I was using the 12-32mm kit lens and usually shooting in manual and RAW. There was no data in the files to pull things around if needed in LR, but perhaps their LR profiles were still work in progress as it was a new camera?

To be honest I think I had a faulty camera and never experienced this when using the GX1 a few years ago.


An example where this was OE and the next ones were fine.

P1000109 by Swansea Jack, on Flickr



Look at the green mush on the top of the conifer left side.

P1000090-4 by Swansea Jack, on Flickr

Full file.

P1000090 by Swansea Jack, on Flickr

 

[HoppyUK]

The larger the sensor, the more light falls on it per square mm and regardless of the pixel count it's there to be captured. It used to be true that in practise higher pixel density effectively reduced the light gathering capability, but that's not really true now and gapless microlenses, off-set microlenses and tricks like Back-Side Illuminated BSI designs that remove the clutter from the front of the sensor mean much more of the light is actually captured and used. Over and above that, improved electronics have dramatically reduced the amount of noise generated. The net result is that pixel count and pixel density are not an accurate guide to noise levels and dynamic range capability, and indeed some of the best sensors on that score have the highest pixel count/density.

The other area where sheer sensor size impacts image quality is lens performance. Smaller formats need greater enlargement and therefore demand higher resolution from the lens. Fact of physics is that as lens resolution goes up, so image contrast goes down, and it's actually contrast that has most impact on our visual impression of sharpness. Hence bigger formats produce sharper images.

 

[woof woof]

There were a couple of issues where the camera would over expose and the files were mushy although intermittent. I was using the 12-32mm kit lens and usually shooting in manual and RAW. There was no data in the files to pull things around if needed in LR, but perhaps their LR profiles were still work in progress as it was a new camera?

To be honest I think I had a faulty camera and never experienced this when using the GX1 a few years ago.

I'm not going to comment on pictures posted here or on lr as I've never used it so I'll stick to sweeping generalities

If you use good lenses and are sensible with your settings (like not shooting everything at f16 and ISO 6400) and try to get the best from the kit you should get good and consistant results from mft and you certainly shouldn't be getting repeated instances of picture ruining over exposure, that's something I haven't seen. Yet

In my little tests which I've long since stopped doing no one I've roped in could consistantly and accuratetly pick mft prints from a pile of ff and aps-c prints. I could on screen and really looking, often, sometimes, but even now when processing ff and mft shots at the same time I often have to look at the settings to tell which is which. ff is better but I personally have to look closely or push the envelope to see it.

Sorry it didn't work well enough for you.

 

[Riz_Guru]

I believe you'll always have to accept that smaller sensors just physically can't gather the same amount of light compared to larger format sensor.
But as long as your happy with the outputs who cares? Plenty of people out there are perfectly happy with iPhone snaps etc.
I think at present Sony are the ones to watch in terms of where the sensor market is going.... I think it's going to aim for high megapixel, sensitive and speed.