PDA

View Full Version : E-M1 and Sony A7R comparison (warning - lots of images)


pdk42
6th February 2016, 05:40 PM
As per my post on "Buying into another system", here are some comparison shots I've made with my E-M1 and a Sony A7R. The E-M1 had the Oly 17mm f1.8 fitted, the Sony the Zeiss 35mm f2.8.

All the shots are from raw and processed with LR 5.7.1. Except where noted (and some slight WB tweaks), I have used default LR settings for all images (sharpening at 25, no NR).

The 100% crops are from images converted to jpgs at 4600 pixels long side. This is the native resolution on the E-M1. I did this to get comparable magnification and in any case it's representative of a print 20" wide (at 240 dpi) or displayed on a 4k monitor. It means that the Sony shots are down-resed, but this will make the Sony shots apparently sharper and less noisy even if it robs it of resolution.

1) Bookshelf. A test of the cameras at base ISO. Shot on a tripod. ISO 200. Both at f5.6.

Both shots:
http://www.famillekaye.com/EM1vsA7R/BookshelfFull.jpg

Sony at 100% (of 4600 wide image)
http://www.famillekaye.com/EM1vsA7R/BookshelfSony_100.jpg

E-M1 at 100%
http://www.famillekaye.com/EM1vsA7R/BookshelfEM1_100.jpg

My view - there's clearly more noise and grittiness on the E-M1 and the Sony lens is a better performer than the Oly, but the differencea aren't enormous. Interestingly (but not shown here), the blown-out area of the window recovers to some degree on both cameras. The Sony manages slightly better, but it's not night and day.

2) Wash basket. A test at higher ISO. Handheld at ISO 6400. Both shot at f2.8 (so Sony wide open, but Oly 1.5 stops down). Since I was testing high ISO, I added some NR on both shots (since that's how we'd do it in the real world isn't it!). I added 30 NR on the E-M1 and 20 NR on the Sony:

Both shots:
http://www.famillekaye.com/EM1vsA7R/BasketFull.jpg

Sony at 100% (of 4600-wide image):
http://www.famillekaye.com/EM1vsA7R/BasketSony_100.jpg

E-M1 100%:
http://www.famillekaye.com/EM1vsA7R/BasketEM1_100.jpg

My view - the Sony is clearly doing a lot better here. Especially in terms of sharpness on the towels and the lack of colour bleeding of the cyan towel. There's also much more detail in the fabrics. If shooting at this sort of ISO is something you need, then the Sony will do it a lot better.

3) Outdoor shot. A test of the sort of thing I use the camera for mainly. It was a dull day so I shot both at ISO 400 - mainly to keep the shutter speed high enough on the Sony ! This time at f4.

Both shots:
http://www.famillekaye.com/EM1vsA7R/ShedFull.jpg

Sony 100%:
http://www.famillekaye.com/EM1vsA7R/ShedSony_100.jpg

E-M1 100%:
http://www.famillekaye.com/EM1vsA7R/ShedEM1_100.jpg

My view - The Sony clearly shows more detail in the trees. I'm not sure whether this is the lens or the sensor, but either way it's better. It's not a major difference, but it's there nonetheless. Would it make a difference on an otherwise great shot with good composition and visual impact? - I doubt it.

4) "Pushed" shot at ISO 800. This shot was underexposed due to poor metering on my part, but it's useful to see how the files cope with exposure pushing. Both at ISO 800 with the Sony pushed 1.5 stops and the Oly 0.5 stop. Not symmetrical I know, but there it is!

Both shots:
http://www.famillekaye.com/EM1vsA7R/TomFull.jpg

Sony 100%:
http://www.famillekaye.com/EM1vsA7R/TomSony_100.jpg

E-M1 100%:
http://www.famillekaye.com/EM1vsA7R/TomEM1_100.jpg

My view - well, another win for the Sony. Would it matter on a print? You decide!

5) Portrait in natural light. Not the same ISO here - but close. Sony at 800, Oly at 640.

Both shots:
http://www.famillekaye.com/EM1vsA7R/CatherineFull.jpg

Sony:
http://www.famillekaye.com/EM1vsA7R/CatherineSony_100.jpg

E-M1:
http://www.famillekaye.com/EM1vsA7R/CatherineEM1_100.jpg

My view - well, more detail in the Sony image again - but the Oly is more than good enough.


In summary, it's clear that the Sony is better; but that's hardly surprising given its sensor is 4x the size and there are more than twice the number of pixels. However, the E-M1 gets pretty close at lower ISO. All other things equal you might conclude that the Sony is the better option - but all other things aren't equal:

- The E-M1 is faster, better handling and with the IBIS allows shots that the Sony would probably miss.

- The size/weight of a bag of Sony lenses would be a lot bigger.

- The cost of Sony lenses is pretty high.

I don't think I'll be staying with the Sony, but it's a closer call than I thought it might be!

Zuiko
6th February 2016, 06:03 PM
Thanks for this, Paul, it's a very interesting comparison and you have clearly put a lot of work into it. The Sony wins in every situation, but not by a huge margin. Given the difference in surface area of the sensors it's rather like comparing an Olympus OM 35mm SLR with a Mamiya RB67. I suspect that with such a comparison you would see a truly enormous difference between the two formats. It would be interesting to see a similar comparison between the EM-1 and an APS-C camera; considering the results of this test I doubt there would be much difference at all.

Harold Gough
6th February 2016, 06:05 PM
Paul,

I don't like to accuse anyone of bias but you have a definite leaning tendency.

Allowing for a bit of vertical offset alignment, you have the best cross-eye stereo pair of a basket of washing I have seen. OK, there is no competition. However, the last pair are spot on, all the more surprising as they are at wide aperture.

Harold

Kiwi Paul
6th February 2016, 06:18 PM
Paul,

I don't like to accuse anyone of bias but you have a definite leaning tendency.

Allowing for a bit of vertical offset alignment, you have the best cross-eye stereo pair of a basket of washing I have seen. OK, there is no competition. However, the last pair are spot on, all the more surprising as they are at wide aperture.

Harold

I was impressed how the woman and boy didn't seem to move at all between changing cameras :D

Paul

pdk42
6th February 2016, 06:20 PM
I was impressed how the woman and boy didn't seem to move at all between changing cameras :D

Paul

I've got them well trained! Although TBH, my son was playing on his phone so was oblivious to what I was doing and my wife, as you can probably see, had that "not more photographs" look!

birdboy
6th February 2016, 07:14 PM
As John has said you have put a lot of work into this comparison but I do question if it is truly representative, the lenses are so different in FL and aperture for a start (35mm f2.8 = true dia 12.5 mm; 17mm f 1.8= true dia 9.4mm so the 35mm will let considerably more (43%) light onto the sensor). As far as I could see they are all well defined pictures as one has become accustom to with modern DSLR's Did you use the anti-shake mode on the Olympus, how does IS system compare? Does the A7R do composite or focusing stacking in camera? Yes I am probably guilty of being bias towards the EM1 because I have one and this is a Olympus forum. Making true comparisons between cameras and lenses is so difficult with everything being so well made these days. At the end of the day it is what you are happy with and at least you have been able to make your own subjective comparison. Good luck if this is the system that gets you the pictures you crave for.

Kiwi Paul
6th February 2016, 07:26 PM
The way I compare is to go out and take real world photos of things I take photos of, landscapes, abstract stuff etc, process them so they look as good as I can get them independently, then look at the photo as a whole, if there's is no perceivable difference then it doesn't matter which camera, lens or whatever, I use, if the final picture has a definite difference to the other then I may consider using that camera, lens etc. But I think you have to evaluate equipment over a period of time, allowing for different lighting, different scenes and genres etc to get a truly balanced idea, there's the "novelty factor" too that you have to get over to remove any bias, and only after extended evaluation can you really ascertain which you prefer.

Paul

birdboy
6th February 2016, 07:40 PM
I have just checked the sensor details on sensorgen. The EM1 has lower read noise so I suspect that the noise difference you are seeing is down to the lens not the camera.

http://www.sensorgen.info/SonyA7R.html

http://www.sensorgen.info/OlympusOM-D-E-M1.html

pdk42
6th February 2016, 07:46 PM
I have just checked the sensor details on sensorgen. The EM1 has lower read noise so I suspect that the noise difference you are seeing is down to the lens not the camera.

http://www.sensorgen.info/SonyA7R.html

http://www.sensorgen.info/OlympusOM-D-E-M1.html

Well, I can certainly say that the Sony lens is a sharper lens than the Olympus, but I can also say without any shadow of a doubt that the Sony delivers lower noise. Not only is this what I observe, it's what DxO have measured, what every other review on the internet says and what is intuitively right given the difference in sensor size.

pdk42
6th February 2016, 07:52 PM
As John has said you have put a lot of work into this comparison but I do question if it is truly representative, the lenses are so different in FL and aperture for a start (35mm f2.8 = true dia 12.5 mm; 17mm f 1.8= true dia 9.4mm so the 35mm will let considerably more (43%) light onto the sensor). As far as I could see they are all well defined pictures as one has become accustom to with modern DSLR's Did you use the anti-shake mode on the Olympus, how does IS system compare? Does the A7R do composite or focusing stacking in camera? Yes I am probably guilty of being bias towards the EM1 because I have one and this is a Olympus forum. Making true comparisons between cameras and lenses is so difficult with everything being so well made these days. At the end of the day it is what you are happy with and at least you have been able to make your own subjective comparison. Good luck if this is the system that gets you the pictures you crave for.

Ermmm - I'm pretty sure that the amount of light exiting a lens is not related to its absolute aperture size, but its ratio with the focal length (aka the f-stop). This must be right since otherwise my 50-200 would be delivering several times the light that my 17mm does since the aperture sizes are significantly different. But of course it doesn't since it's f2.8-f3.5 and the 17mm is f1.8. I can readily see this by looking at the shutter speeds that each lens uses when shot wide open.

I know it's not an uber scientific test, but I think it's pretty representative of what you'd get using both cameras in real life.

As for bias etc - I hope you read the end of my post where I said that on balance I think I prefer the E-M1. The Sony's IQ is definitely better, but it's more than a question of IQ.

Ricoh
6th February 2016, 07:57 PM
Ive compared the files from my Leica 'system', ie the camera and the all important lens (I have a Summicron 35 and a more recent summilux 50) against the U4/3 EM5 with a PL 25/f1.4, the files show quite a bit of difference, especially in the maluability if I want to raise shadows or recover highlights - but the main reason for preferring the Leica system is the fully manual camera, almost like the cameras I grew up with (my next door neighbour gifted a Kodak brownie 120 film). I'm not a big fan of computer controlled cameras, and would convert to film if I had the patience to develop and lash out on a reasonable scanner. I could have purchased a little used MP for almost half the purchase price of the M240, but hey ho I didn't. Anyway, M lenses work equally well on analogue and digital cameras.

There's a plus point on side of FF, and that's the DoF compared with FF/4 - just need an appropriate lens. But Paul (PDK42, as a landscape photographer, I understand limited DoF isn't your main objective.

birdboy
6th February 2016, 07:58 PM
I thought you were comparing cameras not lenses as per your heading. The information I provided is clear that the EM1 sensor out performs the Sony Sensor regarding read noise and I am assuming that is the dominant noise for the pictures you were taking. To get a truer picture and get the same light gathering for both lenes you would need to take a picture with the 17mm at f1.8 and the 35mm at f 3.7.

Nawty
6th February 2016, 08:44 PM
You had an itch, you scratched it and found out that there isn't thaaaat much difference in terms of real word usage. I did the same (but in the other direction) and now have no FF kit and a whole stack load of Olympus.

I would also say that, in my experience, some of the differences you are seeing is down to the Oly lens which isn't as brilliant as some of their other primes, it lacks contrast wide open, isn't as sharp as others and suffers CA when wide open. You can see this in the shot of your wife where there is a noticeable loss of contrast in the frame (could be because it was wide open, the flare or a bit of both).

Anyway, I love these kind of posts as they show the differences in real world shooting to achieve a similar shot, not having everything equalised to a non-sensical absolute.

birdboy
6th February 2016, 08:50 PM
Ermmm - I'm pretty sure that the amount of light exiting a lens is not related to its absolute aperture size, but its ratio with the focal length (aka the f-stop). This must be right since otherwise my 50-200 would be delivering several times the light that my 17mm does since the aperture sizes are significantly different. But of course it doesn't since it's f2.8-f3.5 and the 17mm is f1.8. I can readily see this by looking at the shutter speeds that each lens uses when shot wide open.



My understanding is the bigger the lens the more light it lets in. I have been trying to properly understand this light issue because I want to do astrophotography. I have come across a very good web site that properly explains this complex issue. The guy is a Dr Roger Clark who has spent years in imaging some including with NASA. I think he makes a lot of sense but his explanations seem to contradict a lot of what is published on the web. Here is his explanation about light gathering and f/ratios.

http://www.clarkvision.com/articles/low.light.photography.and.f-ratios/

Naughty Nigel
6th February 2016, 09:16 PM
You have a fine triple dunny there Paul!

There is a difference, but not as much as I would have expected given the difference in sensor size.

I would be interested to see a comparison with the EM-5 as, from what I have seen, it seems to provide a marginally cleaner image with better colour. I certainly have no complains with mine.

pvasc
6th February 2016, 09:31 PM
I think the picture of the tree top is a DOF issue or a missed focus issue. The top of the shed looks sharper in on crop than the other, or do I need to clean my glasses?

Good comparison though, I would not have guessed it would be this close.

Nawty
6th February 2016, 09:42 PM
My understanding is the bigger the lens the more light it lets in. I have been trying to properly understand this light issue because I want to do astrophotography. I have come across a very good web site that properly explains this complex issue. The guy is a Dr Roger Clark who has spent years in imaging some including with NASA. I think he makes a lot of sense but his explanations seem to contradict a lot of what is published on the web. Here is his explanation about light gathering and f/ratios.

http://www.clarkvision.com/articles/low.light.photography.and.f-ratios/

As far as I can tell, what he is talking about is crop factor wrapped up in an over elaborate explanation and a number of equations I haven't seen before.

I could be wrong as I'm not an expert and I have had a few glasses of vino but as far as I can tell he is saying:

1. Real aperture size is the real measure of light let in (obvious no?)
2. The fact that the less light you have the more noise you will have (correct assuming a base noise floor).

And that's pretty much it. He does include some examples from different lenses but they are different focal length so comparing them at the same framing (either shrinking one or enlarging the other - he doesn't say which but certainly you can't frame the milky way the same with a 15 and 35mm lens) isn't a level playing field.

What I surmise is that basically these are all things we intuitively understand as crop factor, without the need for pages of explanation and misdirection. He basically says as much when he resamples the 7D image to the same field of view as the 5d2.

Nawty
6th February 2016, 09:43 PM
I think the picture of the tree top is a DOF issue or a missed focus issue. The top of the shed looks sharper in on crop than the other, or do I need to clean my glasses?

Good comparison though, I would not have guessed it would be this close.

I think some of that is the lens - it's Oly's least sharp 'enthusiast' lens up against a Zeiss - not surprising really.

pdk42
6th February 2016, 10:03 PM
On this issue of aperture size and f-stop I can assure you all that the light intensity reaching the sensor is proportional to the f-stop, not the absolute aperture size. A moments thought makes this obvious - if it were just the size of the aperture alone, then why do longer focal lengths have bigger glass for the same f-stop? Look at your 17mm f1.8 lens and the aperture size is about 1.25cm. Now look at your 75mm - here it's more like 5cm and yet both are f1.8 and both will give the same exposure when used for a photo at the same shutter speed and ISO.

The logic of it in more mathematical terms is due to the inverse square law. The inverse sq law states : "The intensity of the light is inversely proportional to the Sq of the distance it travels."

That is I is proportional to 1/(D * D)
where
D is the distance and
I is the Intensity

For a 32mm lens at f/8
Diameter is 32/8 = 4 mm, therefore radius = 2mm
The area of the aperture is then 2 * 2 * pi = 4pi
The focal length of the lens is 32mm so
the intensity of the light = 4pi/(32 * 32) = 0.01227 .............. A

For a 80mm lens at f/8
Diameter is 80/8 = 10 mm, therefore radius = 5mm
The area of the aperture is then 5 * 5 * pi = 25pi
The focal length of the lens is 80mm so
the intensity of the light = 25pi/(80 * 80) = 0.01227.............. B

A and B are the same - the same intensity at the same f-stop, but the aperture size are 4mm and 10mm respectively.

Absolute aperture size matters for the amount of blur for out of focus parts - hence the reason why FF cameras show more blur than u43 cameras.

birdboy
6th February 2016, 10:07 PM
As far as I can tell, what he is talking about is crop factor wrapped up in an over elaborate explanation and a number of equations I haven't seen before.



He is using the same camera body so its not crop factor or a sensor difference issue. His explanation do seem very academic and his whole web site takes a considerable amount of time to digest and understand (I'm not sure I properly understand it yet, hence the opening it up to this forum) but I think he makes some very valid points which are backed up with amazing images he produces. His theme throughout his web site is that true aperture area is the key to reduce noise more than any other element of a image system.

pdk42
6th February 2016, 10:12 PM
1. Real aperture size is the real measure of light let in (obvious no?)

No, that's not right. As per my post above, the amount of light let in is the ratio of aperture size to focal length - i.e. the f-stop.


2. The fact that the less light you have the more noise you will have (correct assuming a base noise floor).

Yes, because you need to increase the amplification on the pixels to get a high enough output in to the A to D converter.


And that's pretty much it. He does include some examples from different lenses but they are different focal length so comparing them at the same framing (either shrinking one or enlarging the other - he doesn't say which but certainly you can't frame the milky way the same with a 15 and 35mm lens) isn't a level playing field.

The field of view of a 17mm lens on a u43 sensor and a 35mm lens on an FF sensor is the same so you can absolutely compare them and it IS a level playing field. At the same f-stop, the light intensity will be the same, the exposure will be the same and the difference in IQ will be down to sensor performance and then magnification effects as you turn them into a viewable image.

The "shrinking" I did is nothing to do with the lens - it's simply to compensate from one sensor having 4600 pixels on its longest edge while the other is 7200 (ish). If I hadn't done that, 100% views from the Sony would have been larger and therefore harder to correct. Reducing the image size will however reduce noise and also increase sharpness - which is exactly one of the reasons why larger sensors deliver better image quality!
[/QUOTE]

pdk42
6th February 2016, 10:14 PM
I think the picture of the tree top is a DOF issue or a missed focus issue. The top of the shed looks sharper in on crop than the other, or do I need to clean my glasses?


That's possible, but I did try to focus on the trees. In any case at f4 I think both the shed and the trees are within the DOF for a u43 sensor. The fact that the shed and the trees are in focus for the E-M1 but not the Sony is due to the narrower DOF at the same aperture on FF.

pdk42
6th February 2016, 10:29 PM
I have just checked the sensor details on sensorgen. The EM1 has lower read noise so I suspect that the noise difference you are seeing is down to the lens not the camera.

http://www.sensorgen.info/SonyA7R.html

http://www.sensorgen.info/OlympusOM-D-E-M1.html

I think you need to look at the read noise in combination with the max saturation (which is shown in the last column in the links you provide). You see these (at IẞO 200):

E:M1: read noise: 3.0 saturation: 15716
A7R: read noise: 4.3 saturation: 49714

Nawty
6th February 2016, 10:36 PM
The "shrinking" I did is nothing to do with the lens - it's simply to compensate from one sensor having 4600 pixels on its longest edge while the other is 7200 (ish). If I hadn't done that, 100% views from the Sony would have been larger and therefore harder to correct. Reducing the image size will however reduce noise and also increase sharpness - which is exactly one of the reasons why larger sensors deliver better image quality!


The "shrinking" was nothing to do with your post but the link to Clarkvision website where he (for example) took a milky-way photo with a 15mm and 35mm lens and then showed them at the same framing and then compared the noise between them, which to me seems rather disingenuous.

:)

Also, (a few glasses later) I guess the absolute size of the aperture is only as important as the amount of total light gathered, for example a wide angle will take in the whole sky and a telephoto will take in a frame of a bird - the total light form the bird is less than the total light from the sky but it is the aperture that manages that difference. As your calc's show.

birdboy
6th February 2016, 10:59 PM
"Light gathering" is one of my pet hates as it is usually overused by FF fanboys. In this situation I think it is a nonsense point as light gathered is not being projected onto the same area as the M4/3 lens is projecting it onto its sensor size. In other words the light is more concentrated on the E-M1 sensor than the Sony gets, hence needing that larger diameter lens for that larger format. The main difference between these two cameras (apart from two different manufacture's sensor) is the pixel size & the larger pixel size is going to produce less noise & more of those larger pixels in a larger format is going to store more detail.

Ross How would you explain the lower read noise figures for the EM1 shown on the sensorgen web pages? I am beginning to think that the generally accepted WWW view that FF must be less noisy than cropped sensors is suspect. I accept that the final image does appear clearer in FF but that image has been produced by a sensor and lens and I currently support the view that the lens plays a bigger part in this than the sensor.

Nawty
6th February 2016, 11:09 PM
Ross How would you explain the lower read noise figures for the EM1 shown on the sensorgen web pages? I am beginning to think that the generally accepted WWW view that FF must be less noisy than cropped sensors is suspect. I accept that the final image does appear clearer in FF but that image has been produced by a sensor and lens and I currently support the view that the lens plays a bigger part in this than the sensor.

I read somewhere recently something along the lines of:

Take a photo. Fold it in half and then fold it in half again so you have a quarter of the original, blow them up to the same size (say 8x10, A4, whatever) and which do you think will have the most noise?

That's exactly what m4/3 does.

Basic lens MFT theory says that sharpness and contrast are directly in opposition to each other, i.e. if sharpness goes up then contrast goes down. For smaller sensor sizes you need sharper lenses to resolve the detail and this means that contrast must go down and it is actually contrast, rather than sharpness, that affects perceived sharpness.

Or, to put it another way, the less you have to magnify (i.e. the bigger the sensor) then the less 'hard' the lens has to work, the less sharp it needs to be and the higher contrast it has, which is actually what looks good to our eyes.

yorky
6th February 2016, 11:19 PM
Very interesting conclusions, I can see a smallish difference and the Sony does seem to have a slightish benifit But! What about the cost and availability!

Greytop
6th February 2016, 11:41 PM
I think some of that is the lens - it's Oly's least sharp 'enthusiast' lens up against a Zeiss - not surprising really.

I would agree, the 17mm f/1.8 is good little lens but not the greatest small prime Oly manufacture. I suspect the Sony 35mm is a step up in quality.

Very interesting thread Paul :)

birdboy
6th February 2016, 11:50 PM
I read somewhere recently something along the lines of:

Take a photo. Fold it in half and then fold it in half again so you have a quarter of the original, blow them up to the same size (say 8x10, A4, whatever) and which do you think will have the most noise?

That's exactly what m4/3 does.

Basic lens MFT theory says that sharpness and contrast are directly in opposition to each other, i.e. if sharpness goes up then contrast goes down. For smaller sensor sizes you need sharper lenses to resolve the detail and this means that contrast must go down and it is actually contrast, rather than sharpness, that affects perceived sharpness.

Or, to put it another way, the less you have to magnify (i.e. the bigger the sensor) then the less 'hard' the lens has to work, the less sharp it needs to be and the higher contrast it has, which is actually what looks good to our eyes.
It not about sensor size per se or pixel density. The A7R has a pixel density of 4.8 micron /pixel whereas the EM1 has 3.75 the E5 has 4.3 and the E3 4.7 so the E3 is close to the pixel density of the A7R! When one looks at the images then it is not FF or pixel density that is determining the noise the trick lies in the electronics that measure the photons in the photsites and the Olympus sensor and electronics is doing an admirable job at reducing noise and I would say better than the Sony A7R.

Ricoh
7th February 2016, 12:03 AM
Very interesting conclusions, I can see a smallish difference and the Sony does seem to have a slightish benifit But! What about the cost and availability!

I'm not sure. I'm viewing images presented for web using an iPad, very difficult to make a judgement. If people do no more than present web-grade images, most cameras are capable, the iPhone 6 being more than capable.

The dominant distinction for me is DR and the ability to manipulate the files.

I agree with you on cost, it's like making a choice between an Aygo and a BMW, both get you from a-b, and if you obey the speed limits you'll arrive at more or less the same time.

Greytop
7th February 2016, 12:11 AM
I'm not sure. I'm viewing images presented for web using an iPad, very difficult to make a judgement. If people do no more than present web-grade images, most cameras are capable, the iPhone 6 being more than capable.

Paul has presented 100% crops albeit with the Sony resized to match the image area of the E-M1, which does make sense given that's how you would view a comparative print (for example).
I wonder how popping the 12-40 on the E-M1 and zooming to 17mm would stack up. I'm pretty sure it's a better performer than the worst of the Oly primes at their respective focal lengths.
As an alternative the 25 (Oly or Panny) verses the 55 would be interesting.

Kiwi Paul
7th February 2016, 12:21 AM
Paul has presented 100% crops albeit with the Sony resized to match the image area of the E-M1, which does make sense given that's how you would view a comparative print (for example).
I wonder how popping the 12-40 on the E-M1 and zooming to 17mm would stack up. I'm pretty sure it's a better performer the worst of the Oly primes at their respective focal lengths.
As an alternative the 25 (Oly or Panny) verses the 55 would be interesting.

The 2 images I posted earlier today compared the Pany 25 f1.4 and Ziess 55 f1.8.
Here, I've resized them in Photoshop so they are the same size then taken a crop of each.

Oly + 25 F1.4

https://farm2.staticflickr.com/1504/24862309265_f67e405aeb_b.jpg (https://flic.kr/p/DSZQmz)

A7ii + 55 f1.8

https://farm2.staticflickr.com/1489/24836037436_3a87a88811_b.jpg (https://flic.kr/p/DQFbE1)

Paul

Nawty
7th February 2016, 12:48 AM
It not about sensor size per se or pixel density. The A7R has a pixel density of 4.8 micron /pixel whereas the EM1 has 3.75 the E5 has 4.3 and the E3 4.7 so the E3 is close to the pixel density of the A7R! When one looks at the images then it is not FF or pixel density that is determining the noise the trick lies in the electronics that measure the photons in the photsites and the Olympus sensor and electronics is doing an admirable job at reducing noise and I would say better than the Sony A7R.

Sorry, that is wrong. All other things being equal, a larger sensor will have less noise as it captures more light.

Greytop
7th February 2016, 12:56 AM
The 2 images I posted earlier today compared the Pany 25 f1.4 and Ziess 55 f1.8.
Here, I've resized them in Photoshop so they are the same size then taken a crop of each.

Paul

Thanks post these Paul *chr
Are these 100% or larger? They look too sharp to be 100% for either lens camera combination to me :)
Oh and this was with your A7 II 24MP (now returned) and not the A7R II correct?

Nawty
7th February 2016, 01:01 AM
The 2 images I posted earlier today compared the Pany 25 f1.4 and Ziess 55 f1.8.
Here, I've resized them in Photoshop so they are the same size then taken a crop of each.


Paul

The thing that strikes me most about those is the difference in processing. Don't know if they are jpgs or raw but there ar some massive differences in the curves in them. Sony is slightly warmer too, or the sun came out a bit.

Ross the fiddler
7th February 2016, 03:28 AM
Ross How would you explain the lower read noise figures for the EM1 shown on the sensorgen web pages? I am beginning to think that the generally accepted WWW view that FF must be less noisy than cropped sensors is suspect. I accept that the final image does appear clearer in FF but that image has been produced by a sensor and lens and I currently support the view that the lens plays a bigger part in this than the sensor.

Thanks, but I had deleted my post, but obviously not soon enough. :rolleyes: ;)

I love my E-M1 & am not (can't afford to if I wanted to) going to get into anything different anyhow. *yes

*chr

Zuiko
7th February 2016, 04:55 AM
It seems that Paul's comparison is being questioned for several reasons but I think it is important to realise that it was never intended to be a scientific test, merely an observation based on several real life situations.

It may well be true that the lenses used are not an equal match in terms of quality, but who's fault is that? If the 17mm f1.8 is the best prime of that focal length Olympus can produce they must expect to be judged upon it. Actually, I perceive that although perhaps not quite top draw it's still a more than capable lens.

As for resizing the Sony images before cropping giving that camera an advantage, Paul could have done it the other way - upsizing the Olympus files to match the Sony before taking 100% crops. I expect that would have been far less kind to the E-M1.

Regarding the confusion over f-stops, I think we can accept that, regardless of the physical diameter of the aperture, in terms of light gathering f2.8 is the same for any lens on any format. We don't need to involve ourselves in complicated physics equations, the lens designers have already done that!

It was no surprise to me that the full frame Sony beat the MFT Olympus in all of Paul's examples; what did surprise me was how little the differences were in terms of real world usage.

pdk42
7th February 2016, 08:04 AM
Sorry, that is wrong. All other things being equal, a larger sensor will have less noise as it captures more light.

I agree - all other things equal, that's exactly true. However, whether that's because it "captures more light" or whether it's because it needs less magnification is definitely up for debate.

Try this thought experiment:

- Imagine two sensors made with the same pixels (same technology, pitch, size), but one u43 and the other FF

- Put a lens in front of each with the same field of view and set the same f-stop

- Set them both to the same ISO and take a shot with the same exposure time

Now, each pixel would receive exactly the same light intensity on both sensors and would therefore yield the same noise.

However, the u43 image would need more magnification so whatever noise was there would be magnified. That's got nothing to do with the sensor itself capturing more light per se has it?

pdk42
7th February 2016, 09:08 AM
The "shrinking" was nothing to do with your post but the link to Clarkvision website where he (for example) took a milky-way photo with a 15mm and 35mm lens and then showed them at the same framing and then compared the noise between them, which to me seems rather disingenuous.


Ah, apologies - I thought you were referring to my post! :)

Otto
7th February 2016, 10:08 AM
The larger lens will indeed capture more light but when that light is focussed on the larger sensor it will be spread over a larger area, so all other things being equal (aperture, euqivalent focal length etc) surely the light intensity per unit area at the sensor will be the same?

Anyway, on the evidence of this thread I won't be changing to a Sony system as the differences seem very small. Thanks for the demo Paul :).

Graham_of_Rainham
7th February 2016, 10:09 AM
What if all the sensor data from both cameras was printed such that each pixel was produced by one "dot" of the printer head.

4800 x 3200 @ 300dpi = 16 x 10.67

4608 x 3456 @ 300dpi = 15.36 x 11.52

Could you see a difference? Perhaps the colours would be slightly different, any specular highlights may look dissimilar, whites whiter, blacks blacker, etc., etc., et...

Hang on a minute, we had all this with film choice, grain size, speed, backing colour, and all the wonderful variations that came from chemistry :eek:

We still discus the colour richness of the E-1 & E-500 sensor and all the various stages that have brought us to where we are today with the "mysterious" unknown PEN-F sensor.;)

I do enjoy a good technical discussion, and this forum seems to be able to do it so well, with hardly any Teddies being harmed.

*chr

pdk42
7th February 2016, 11:13 AM
Ross How would you explain the lower read noise figures for the EM1 shown on the sensorgen web pages? I am beginning to think that the generally accepted WWW view that FF must be less noisy than cropped sensors is suspect. I accept that the final image does appear clearer in FF but that image has been produced by a sensor and lens and I currently support the view that the lens plays a bigger part in this than the sensor.

Read noise is only half the story. What's important is the signal to noise ratio so you need to look at the max saturation (also called the pixel well capacity) as well. Both are measured in electrons (created by photon interaction) and the ratio gives you the s/n. As the sensor size (area) increases so does the well capacity. Although the read noise is lower on the E-M1, so is its well capacity. The Sony has a well capacity three times bigger than the E-M1 but the read noise is only about 50% higher. This means the Sony has a better s/n ratio. In heavily filled pixel wells the s/n will be low for either camera, but as the well gets more lightly filled (in low light) the s/n increases. Since the larger area of the FF pixel and its larger well capacity are able to capture more photons, more electrons accumulate in the well leading to lower s/n ratios.

And in fact, there are other sources of noise than read noise - "shot noise" being the main one which is random quantum fluctuation in the capture and conversion of photons into electrons in the well. Shot noise becomes the biggest contributor to noise in the darker parts of an image.

As regards lenses, since we're magnifying a 4/3 image twice as much as FF (in linear dimensions, not area), any lack of sharpness is exaggerated. Lenses for smaller formats need to have higher acuity to deliver the same perceived sharpness in the final image. I reckon the Oly 12mm f2 lens is one of the sharpest in the u43 system, but at the same final viewing size, an old OM 24mm f2.8 (when stopped down to f8) delivers an apparently sharper image from the Sony.

Kiwi Paul
7th February 2016, 11:53 AM
Thanks post these Paul *chr
Are these 100% or larger? They look too sharp to be 100% for either lens camera combination to me :)
Oh and this was with your A7 II 24MP (now returned) and not the A7R II correct?

They are 100% crops or there abouts, and yes with the A7ii.

Paul

Kiwi Paul
7th February 2016, 12:02 PM
The thing that strikes me most about those is the difference in processing. Don't know if they are jpgs or raw but there ar some massive differences in the curves in them. Sony is slightly warmer too, or the sun came out a bit.

Shot as RAW and processed in LR CC, shot on a different day, so the colour temp wouldn't be the same, the compositions aren't exactly the same as I weren't standing in exactly the same place either. The processing was done independently but I used the same process and the difference in the LR sliders isn't that great, even with the sliders set the same there is still a difference in the histogram, different day, different composition, different cameras and lenses, there will be a difference in the curves / histogram not necessarily attributable to processing.

Paul

birdboy
7th February 2016, 01:46 PM
On this issue of aperture size and f-stop I can assure you all that the light intensity reaching the sensor is proportional to the f-stop, not the absolute aperture size. A moments thought makes this obvious - if it were just the size of the aperture alone, then why do longer focal lengths have bigger glass for the same f-stop? Look at your 17mm f1.8 lens and the aperture size is about 1.25cm. Now look at your 75mm - here it's more like 5cm and yet both are f1.8 and both will give the same exposure when used for a photo at the same shutter speed and ISO.

The logic of it in more mathematical terms is due to the inverse square law. The inverse sq law states : "The intensity of the light is inversely proportional to the Sq of the distance it travels."

That is I is proportional to 1/(D * D)
where
D is the distance and
I is the Intensity

For a 32mm lens at f/8
Diameter is 32/8 = 4 mm, therefore radius = 2mm
The area of the aperture is then 2 * 2 * pi = 4pi
The focal length of the lens is 32mm so
the intensity of the light = 4pi/(32 * 32) = 0.01227 .............. A

For a 80mm lens at f/8
Diameter is 80/8 = 10 mm, therefore radius = 5mm
The area of the aperture is then 5 * 5 * pi = 25pi
The focal length of the lens is 80mm so
the intensity of the light = 25pi/(80 * 80) = 0.01227.............. B

A and B are the same - the same intensity at the same f-stop, but the aperture size are 4mm and 10mm respectively.

Absolute aperture size matters for the amount of blur for out of focus parts - hence the reason why FF cameras show more blur than u43 cameras.

Paul my reasons for questioning what you did because I am trying to fully understand this FF versus cropped sensor issue because I have thought about going FF. I wonder where you got your suspect formula from maybe http://www.photographycorner.com? I was not going to do this but I feel it is so wrong that it needs to be corrected before more folk believe in your formula. I say its suspect because you state "The intensity of the light is inversely proportional to the Sq of the distance it travels." and that the intensity = 0.01227 yet your formula can be simplified to pi/4/(fstop)^2 which contradicts your statement as there is no distance in the equation! Intensity is a measure of watts per area.

I would guess that few on here have read the clarkvision article and properly digested it. But I would like to highlight some of the important statements made in it. For those who still think he is wrong would do well to look at his background.

http://clarkvision.com/rnc/

“The prevailing view by photographers is that faster f/ratios delivers more light. I will show you in this article that is not necessarily the case. This bucks the prevailing view in photography, but is not the prevailing view in imaging science. This article will explain why.”

The clarkvision article explains all those issues that folk have raised on here. He uses the same exposure settings (ISO aperture and shutter speed) but the conclusions are that f/stop does not tell the whole story and that a truer exposure is described in his Etendue equation.

“There are three factors that determine the true exposure in a camera +lens. 1) The lens area, or more accurately, the lens entrance pupil, which is the effective light collection area of a complex lens. The area determines how much light the lens collects to deliver to the sensor. 2) The angular area of the subject. The product of these two values is called Etendue, or A*Ω (A*Omega) product. (A= the lens entrance pupil area, and Ω, omega = the angular area of subject). The third value is 3) exposure time, the length of time the sensor is exposed to light.”

“We have shown 3 examples where f/ratio does not indicate the true amount of light in a recorded image. The true exposure is better described by Etendue, the A*omega product, or EtS. This is the case for all photography.

But one need not do EtS calculations. Given a choice for imaging a subject, use the lens that has the largest aperture diameter, not the fastest f/ratio. Of course this applies only if the selected lens will fit the subject in the frame, or else you may need to do a mosaic.”

Graham_of_Rainham
7th February 2016, 02:31 PM
I'm so pleased that this debate is in the Lounge area :)

"The concept of ´etendue is applied to the propagation of luminescent radiation, and to the transformation of such radiation in absorbing and luminescent media. Central to this analysis is the notion of ´etendue as a measure of the number of rays in the beam which permits the definition of entropy and transition to the formalism of statistical mechanics. When considered from the statistical viewpoint, ´etendue conservation along the path of a beam in clear and transparent media then implies the conservation of entropy. The changes in thermodynamic parameters of a beam upon absorption and re-emission can then be determined in terms of the corresponding changes resulting from the addition or removal of photons from the incident and emitted beam. The thermodynamic theory which follows gives the rate of entropy generation in this process. At moderate light intensities, the results resemble the thermodynamics of a two-dimensional gas. The formalism allows an extension to absorption/emission processes"

Which having read and completely understood, leads me to take up charcoal sketching.

I will just add that when I'm in a studio, the only thing I trust to get the exposure right is my Sekonic hand held flash meter. :cool:

Kiwi Paul
7th February 2016, 03:20 PM
But do make sure, the media you sketch on has sufficient reflective properties to anable the sketch to be seen at its best, i I'll save the lecture on the properties of charcoal reflectance for another time :D

Paul

Graham_of_Rainham
7th February 2016, 03:39 PM
I hope there is someone out there with enough disposable income to follow up this comparison, with one between the PEN-F and the a6300.

pdk42
7th February 2016, 04:01 PM
Paul my reasons for questioning what you did because I am trying to fully understand this FF versus cropped sensor issue because I have thought about going FF. I wonder where you got your suspect formula from maybe http://www.photographycorner.com? I was not going to do this but I feel it is so wrong that it needs to be corrected before more folk believe in your formula. I say its suspect because you state "The intensity of the light is inversely proportional to the Sq of the distance it travels." and that the intensity = 0.01227 yet your formula can be simplified to pi/4/(fstop)^2 which contradicts your statement as there is no distance in the equation! Intensity is a measure of watts per area.

I would guess that few on here have read the clarkvision article and properly digested it. But I would like to highlight some of the important statements made in it. For those who still think he is wrong would do well to look at his background.

http://clarkvision.com/rnc/

“The prevailing view by photographers is that faster f/ratios delivers more light. I will show you in this article that is not necessarily the case. This bucks the prevailing view in photography, but is not the prevailing view in imaging science. This article will explain why.”

The clarkvision article explains all those issues that folk have raised on here. He uses the same exposure settings (ISO aperture and shutter speed) but the conclusions are that f/stop does not tell the whole story and that a truer exposure is described in his Etendue equation.

“There are three factors that determine the true exposure in a camera +lens. 1) The lens area, or more accurately, the lens entrance pupil, which is the effective light collection area of a complex lens. The area determines how much light the lens collects to deliver to the sensor. 2) The angular area of the subject. The product of these two values is called Etendue, or A*Ω (A*Omega) product. (A= the lens entrance pupil area, and Ω, omega = the angular area of subject). The third value is 3) exposure time, the length of time the sensor is exposed to light.”

“We have shown 3 examples where f/ratio does not indicate the true amount of light in a recorded image. The true exposure is better described by Etendue, the A*omega product, or EtS. This is the case for all photography.

But one need not do EtS calculations. Given a choice for imaging a subject, use the lens that has the largest aperture diameter, not the fastest f/ratio. Of course this applies only if the selected lens will fit the subject in the frame, or else you may need to do a mosaic.”

For sure the f-stop and illumination relationship is an approximation. The physical design of a lens alters the transmission of light such that the f-stop might not be a faithful measure of the final light intensity. For this reason, there is also the t-stop which is an actual measure of light that passes through the lens. The deviation of t-stop to f-stop is sometimes given in lens tests (e.g. on DxO). Videographers seem to use t-stops more than photographers for some reason.

Etendue is a technical term used to express essentially the same concept. It expresses the efficiency of the overall optical system and can also be called "light gathering power". In this sense it is a measure of a lens's transmission. It's important in fibre optic systems, microscopes, telescopes etc. Whilst in edge cases it matters, for most photographic purposes we can just use f-stop and not get into additional complexity.

From a physics point of view, what's important is the luminance, measured in candelas per square metre.

My formula is correct and it does have the distance in it - it's the 32*32 and 80*80 in the equation.

Let's try it without formulas:

We can consider a lens to be projecting an image from the lens's centre point (for a simple thin lens) onto the image plane. When focused, the distance from this centre point to the image is the focal length. The image circle cast by the lens will widen with distance and the light intensity will drop off as per the inverse square law. So, if we double the focal length we reduce the light intensity by four times. That's the inverse square law. To compensate we need an aperture that lets in 4 times the light, or one that's double in radius. So, by doubling the focal length we must double the aperture, but since f-stop is aperture / focal length we're back to the same f-stop. So, if we double the focal length we must double the aperture size to get the same illumination and that means the same f-stop.

pdk42
7th February 2016, 04:12 PM
So far as the FF vs crop argument goes, it's actually quite simple. Ignoring diffraction effects, we can in principle design a sensor and lens system of any size. As we get smaller though, three things happen-

- The area of a pixel reduces. This means that its charge well gets smaller which means that the signal to noise ratio deteriorates, That is, we get more noise in the image. Of course, we can improve noise handling and charge well efficiency by using better technology, but for a given technology state, a smaller pixel will ALWAYS produce more noise.

- The focal length of the lens reduces in order to keep the same field of view. In turn this means that the lens's absolute aperture size reduces and this means that blur from out of focus areas reduces. Voila - wider DOF.

- The image will need more magnification to produce something for viewing. That means that noise is amplified and sharpness is smeared. That means that lenses must resolve better and we've got a double whammy on noise.

That's it. No more need for anything else really!

mstphoto
7th February 2016, 04:17 PM
And here's me thinking that f4 is f4 regardless of what lens I'm using without all this physics parafenalia
:D :D :D

birdboy
7th February 2016, 04:30 PM
My formula is correct and it does have the distance in it - it's the 32*32 and 80*80 in the equation.



Sorry Paul but I beg to differ it is not correct the 32*32 and 80*80 you refer to cancels out the 32*32 and 80*80 you used in your first equation.

I only found one reference to that formula on google and it was on a very suspect page. Just because its on the internet does not mean it is correct.

Graham_of_Rainham
7th February 2016, 04:36 PM
And here's me thinking that f4 is f4 regardless of what lens I'm using without all this physics parafenalia
:D :D :D

I wrote an extensive paper on this but the spell checker turned it into a "double entendre" :D

I'm also quite concerned that I used to know and understand all sort of things like this, but as the science and technology evolves, my mind is going the other way :o

Thankfully Olympus gave us iAuto :D

Harold Gough
7th February 2016, 04:38 PM
When it comes to showing differences, or lack of them, between cameras?sensors/lenses, it may well be that any differences are lost during image downsizing for the internet.

Harold

OM USer
7th February 2016, 04:42 PM
I must thank Paul for the time and effort in producing the images for comparison and the discussion that followed.

...I think it is important to realise that it was never intended to be a scientific test, merely an observation based on several real life situations....
I agree wholeheartedly. For me, I want the sharpest pictures that I can take with equipment that is within my budget, that I can carry, that I can handhold, and that I can operate.

I think some of that is the lens - it's Oly's least sharp 'enthusiast' lens up against a Zeiss - not surprising really.

I would agree, the 17mm f/1.8 is good little lens but not the greatest small prime Oly manufacture. I suspect the Sony 35mm is a step up in quality

In choosing my equipment I take account of availability so I am quite happy to see comparisons between lenses (in fact whole systems) that I can purchase even if they are not exactly equivalent.

...I wonder how popping the 12-40 on the E-M1 and zooming to 17mm would stack up. I'm pretty sure it's a better performer than the worst of the Oly primes at their respective focal lengths.

Ah, so that's why I have not put my 17mm on my camera since I got the 12-40mm?

...I do enjoy a good technical discussion, and this forum seems to be able to do it so well, with hardly any Teddies being harmed

This is why we continue to be members of this fantastic forum even if our photographic habits stray into the dark side.

As a final point, along with sensor performance surely the main impact on signal to noise ratios is the cross sectional area of the photocell (or perhaps the microlens above it) rather than the pixels per inch measurement which makes no allowance for the gap between the sensor sites.

pdk42
7th February 2016, 04:56 PM
Sorry Paul but I beg to differ it is not correct the 32*32 and 80*80 you refer to cancels out the 32*32 and 80*80 you used in your first equation.

I only found one reference to that formula on google and it was on a very suspect page. Just because its on the internet does not mean it is correct.

John - no, there is no cancelling out. Let's look at it again:

That is I is proportional to 1/(D * D)
where
D is the distance and
I is the Intensity

So, if we double the distance, D, the intensity reduces by 4 (1/2^2). All the above is stating is the inverse square law. There is nothing controversial in that!

For a 32mm lens at f/8
Diameter is 32/8 = 4 mm, therefore radius = 2mm
The area of the aperture is then 2 * 2 * pi = 4pi
The focal length of the lens is 32mm so
the intensity of the light = 4pi/(32 * 32) = 0.01227 .............. A

I guess I didn't explicitly state that the intensity is proportional to the area of the aperture as well as inversely proportional to the distance squared, but that's what the above equation is saying. That the light transmitted varies by area of the aperture (or exit pupil or entrance pupil or whatever) is entirely accepted and in any case is part of the definition of luminance - candelas per sq m.

In other words: I = a / D^2 (ignoring the constant of proportionality)

"A" in the quoted section above is being calculated from the area of the aperture (2^2 * pi) and D is being used - it's being divided into the area, a.

For a 80mm lens at f/8
Diameter is 80/8 = 10 mm, therefore radius = 5mm
The area of the aperture is then 5 * 5 * pi = 25pi
The focal length of the lens is 80mm so
the intensity of the light = 25pi/(80 * 80) = 0.01227.............. B

This is the same equation with different values for a and D.

Why do you think that D is being cancelled out? Where is that happening?

timboo
7th February 2016, 05:01 PM
I like this thread as it shows what different peoples priorities are. You have the analysts and the artists on here.
It goes to show how good the em-1 is I for one dont shoot with really poor lighting so dont have need for high iso generally I also dont need higher resolution for billboard prints.
Paul has produced some good images comparison wise (i personally dont care about equations too much)
If the end result to the human eye is as good who cares.
I love my em-1 and Pauls approach in my opinion is similar to typical type of shot. This re enforces the capabilities of the em-1 :-)
Also there is a good comparison em-1 v d750 for those techno folks out there, where they have tried to make the test as true as possible.

I see pauls comparison more of an artists approach.

birdboy
7th February 2016, 05:22 PM
Lets take the 2*2*pi =4 pi in your equation where does the 2 come from it comes from the 32/8/2 then you divide by 32 cancelling it out.

pdk42
7th February 2016, 05:36 PM
Lets take the 2*2*pi =4 pi in your equation where does the 2 come from it comes from the 32/8/2 then you divide by 32 cancelling it out.

OK, let's try this a step at a time.

I contend that light intensity, I, is related by the following formula:

I = k a / d^2

where:

k = some arbitrary constant
a = area of aperture
d = distance from lens centre (simple thin lens) to image plane


We can ignore k since all we want to do is prove that the results are the same for a 32mm and an 80mm lens.

For a 32mm lens at f8, the radius of the aperture will be 2mm (32/8). So, the area of the aperture will be:

a = 2 * 2 * pi

Hence the intensity I will be:

I = 2 *2 *pi / (32 * 32)
I = 0.01227

If I do the same for a 80mm lens, I get the same answer.

Hence, same f-stop means same illumination means different aperture absolute size.

birdboy
7th February 2016, 06:33 PM
OK, let's try this a step at a time.

I contend that light intensity, I, is related by the following formula:

I = k a / d^2

where:

k = some arbitrary constant
a = area of aperture
d = distance from lens centre (simple thin lens) to image plane


We can ignore k since all we want to do is prove that the results are the same for a 32mm and an 80mm lens.

For a 32mm lens at f8, the radius of the aperture will be 2mm (32/8). So, the area of the aperture will be:

a = 2 * 2 * pi

Hence the intensity I will be:

I = 2 *2 *pi / (32 * 32)
I = 0.01227

If I do the same for a 80mm lens, I get the same answer.

Hence, same f-stop means same illumination means different aperture absolute size.

check your units mm^2/MM*MM= dimensionless

Nawty
7th February 2016, 06:53 PM
check your units mm^2/MM*MM= dimensionless

Presumably the forgotten constant has units?

pdk42
7th February 2016, 07:16 PM
Presumably the forgotten constant has units?

Precisely! I started with a proportionality statement and only introduced the constant k because I can't type the proportionality character (Greek alpha) on the keyboard!!

Zuiko
7th February 2016, 07:30 PM
Does anyone other than Sheldon Cooper and Leonard Hofstadter actually understand all this? :confused:

Think I'll just take another look at the pictures...... :)

Phill D
7th February 2016, 07:56 PM
Wow 5 pages in one day that's some interest. I'm with you though John I got lost after the pictures :confused:

timboo
7th February 2016, 07:59 PM
does anyone other than sheldon cooper and leonard hofstadter actually understand all this? :confused:

Think i'll just take another look at the pictures...... :)

thank you the artists approach:-)

mstphoto
7th February 2016, 08:16 PM
IMHO, who cares!!
Just go out, enjoy yourself taking pictures and forget all this palony :D

Mike

pdk42
7th February 2016, 08:55 PM
IMHO, who cares!!
Just go out, enjoy yourself taking pictures and forget all this palony :D

Mike

Aye - I guess you're right. But the tech interests me too. We're living in photographic utopia these days - and it's all down to technology.

Nawty
7th February 2016, 09:02 PM
To be fair, it is rare that there's anything to discuss nowadays as normally "Google Knows". I would draw my conclusions from the photos (and as said previously shifted my FF kit after doing my own comparisons) but its been good to have a bit of discussion that flexes the old grey cells - it's been far too long and mine didn't really jump into action but fun none-the-less, and I think I've learned something (or at least a new way of looking at something I understood in a different way) :)

mstphoto
7th February 2016, 09:09 PM
Good onya Paul *chrs
I'll leave that to the geeks in white coats that make our hobby the great hobby it is so that i can go out to take photos without worrying about the scientific aspects :D

Thanks for your comparison images Paul.
I came to the same conclusion last week that at the end of the day there isn't much in it.
I'll probably end up with one of the Sonys - eventually ;)

pdk42
7th February 2016, 09:20 PM
I'll probably end up with one of the Sonys - eventually ;)

You know, I sort of feel the same. Whilst the A7R isn't there yet, the improvement in image quality is definitely there to see and it's hard to go back when you've seen it. The A7Rii seems a much improved camera but the price is truly eye-watering. However, having read Ming Thein's review of it, it's clear that issues remain. Probably the mark iii will nail it it completely (probably next year knowing Sony's rate of releasing new gear).

My sincere hope is that Olympus are able to up the IQ with a new u43 sensor. It seems the 20Mp one in the Pen F doesn't raise the bar much though, which is a real disappointment. There are some big step changes in the research stage (organic sensors, stacked sensors, global shutter), but they all seem 2-3 years out.

Ross the fiddler
7th February 2016, 09:49 PM
You know, I sort of feel the same. Whilst the A7R isn't there yet, the improvement in image quality is definitely there to see and it's hard to go back when you've seen it. The A7Rii seems a much improved camera but the price is truly eye-watering. However, having read Ming Thein's review of it, it's clear that issues remain. Probably the mark iii will nail it it completely (probably next year knowing Sony's rate of releasing new gear).

My sincere hope is that Olympus are able to up the IQ with a new u43 sensor. It seems the 20Mp one in the Pen F doesn't raise the bar much though, which is a real disappointment. There are some big step changes in the research stage (organic sensors, stacked sensors, global shutter), but they all seem 2-3 years out.

It seems the reliance is on Sony anyhow since both cameras use a Sony sensor. :rolleyes:

PeterBirder
7th February 2016, 10:05 PM
It seems the reliance is on Sony anyhow since both cameras use a Sony sensor. :rolleyes:

And Sony, I believe still have a significant share holding in Olympus (even though they are no longer the largest single stock holder) and the two companies are working together in the medical imaging business (Olympus' main business). .;)

Graham_of_Rainham
7th February 2016, 10:28 PM
Reading this thread between episodes of FRINGE makes it easy to use ∝ in equality equations, so that the decision between tea or coffee become easier ;)

pdk42
7th February 2016, 11:00 PM
makes it easy to use ∝ in equality equations

How did you do that?

Ricoh
8th February 2016, 12:31 AM
Hmm, I'll have to consult Erwin Schrodinger about these 'impressive' equations and solutions provided here - I'm assuming of course he's both alive and dead, or am I too late already... perhaps someone has captured him already, with an EM1, a Sony A7R, or more likely a Leica III, waving then immediately collapsing !!

Ross the fiddler
8th February 2016, 12:41 AM
How did you do that?

Probably copy & paste (off google). At least that's what I've resorted to when desperate. :rolleyes:

Zuiko
8th February 2016, 01:13 AM
How did you do that?

Type your post in Microsoft Word using the Cambria Math font. To get the ∝ symbol hold down the Alt Key whilst typing +8733. When your post is complete, copy and paste it into your forum reply. There may be an easier way to do this!

I'm leaning towards a coffee as I'm expecting another sleepless night anyway! :D

Ross the fiddler
8th February 2016, 02:08 AM
And Sony, I believe still have a significant share holding in Olympus (even though they are no longer the largest single stock holder) and the two companies are working together in the medical imaging business (Olympus' main business). .;)

I think that 'significant share' might be about 5% perhaps. :rolleyes:

Ross the fiddler
8th February 2016, 02:13 AM
Type your post in Microsoft Word using the Cambria Math font. To get the ∝ symbol hold down the Alt Key whilst typing +8733. When your post is complete, copy and paste it into your forum reply. There may be an easier way to do this!

I'm leaning towards a coffee as I'm expecting another sleepless night anyway! :D

Or just go to this site (or others). http://math.typeit.org/

BTW, I guess the coffee will help ensure it will be a sleepless night though. :rolleyes:

Zuiko
8th February 2016, 07:38 AM
BTW, I guess the coffee will help ensure it will be a sleepless night though. :rolleyes:

I didn't have a coffee after all, but it was still a sleepless night! :(

Ross the fiddler
8th February 2016, 08:38 AM
I didn't have a coffee after all, but it was still a sleepless night! :(

Sorry to hear that. :(

pdk42
8th February 2016, 09:14 AM
Type your post in Microsoft Word using the Cambria Math font. To get the ∝ symbol hold down the Alt Key whilst typing +8733. When your post is complete, copy and paste it into your forum reply. There may be an easier way to do this!

I'm leaning towards a coffee as I'm expecting another sleepless night anyway! :D

Well, I typed my responses on a tablet so didn't have a real keyboard. I'd have spent all night trying to do that on the Windows 10 on-screen keyboard!!

birdboy
8th February 2016, 10:46 AM
In windows there is program called Character Map (type char in program search). The default font for this forum seems to be Verdana so select Verdana from the font drop down and scroll down to end of the char map the infinity symbol is 3rd row up from bottom. Select it copy it and then paste it into your response. vol la ∞ you could have ∏ as well.

Olybirder
8th February 2016, 11:08 AM
Nice and easy on my Mac, using Keyboard Viewer. ∞§ś∑Ąřπß∂∆Ω√∫µ etc.

Ron

birdboy
8th February 2016, 11:20 AM
I did try an copy and paste from equation editor in word but it did not work so had to PrtScn and save as image.
http://e-group.uk.net/gallery/data/500/Intensity_BS.jpg (http://e-group.uk.net/gallery/showphoto.php/photo/88273)
:D

Graham_of_Rainham
8th February 2016, 12:11 PM
Hmm, I'll have to consult Erwin Schrodinger !!

The biggest problem with Schrodinger jokes is you never know if they will be funny until you tell them...

*chr

Zuiko
8th February 2016, 12:56 PM
Nice and easy on my Mac, using Keyboard Viewer. ∞§ś∑Ąřπß∂∆Ω√∫µ etc.

Ron

Now that's just showing off! :D

pdk42
8th February 2016, 01:03 PM
The biggest problem with Schrodinger jokes is you never know if they will be funny until you tell them...

*chr

All my jokes proceed as a wave function with the reality of them being amusing collapsing when they're observed.

Ricoh
9th February 2016, 01:03 AM
The biggest problem with Schrodinger jokes is you never know if they will be funny until you tell them...

*chr
There's a 'principle of uncertainty' in joke telling - like god playing roulette without knowing the outcome before placing the bet.

Zuiko
9th February 2016, 09:12 AM
There's a 'principle of uncertainty' in joke telling - like god playing roulette without knowing the outcome before placing the bet.

The funniest joke I ever heard was when a friend retold a mildly amusing joke that I had told him to a mutual group of friends, but he got the punch line spectacularly wrong. It was at the time of the Chernobyl disaster (yes, that long ago) and it should have gone like this:-

"Why shouldn't you wear Russian underpants? Because cher knob'll fall out."

Ken actually said "Because cher balls will fall out." This resulted in an audience of very confused people, quizzically looking at each other as if they had somehow missed the point and feeling even more puzzled because I was rendered speechless with hysterical laughter at Ken's faux pas.

RogerMac
9th February 2016, 11:40 AM
To go back to the original subject of the comparison

I use a mixture of Canon FF and Oly E5. If I am anywhere near base ISO it is very difficult to tell the difference between the two formats and in a significant proportion of cases the Oly output will be superior - especially as I own the ZD 50mm f2 lens. The difference comes when one gets away from base ISO and in my case case the Canon is perfectly happy at ISO10,000 and not upset at 25,000. That's the main reason I bought the FF and a surprisingly large proportion of my shots do use these very high settings, for example a bride cutting a cake by the light of a single candle. In the OP's comparison no shots went above 800 which is just not stretching the FF at all but higher ISOs might begin to embarrass the smaller format.

There was an article in AP a few weeks ago making just this point - that the larger formats have a wider envelope of good exposures but within the comfort zone of the smaller formats there is little real world difference.


Roger

pdk42
9th February 2016, 11:53 AM
In the OP's comparison no shots went above 800

Except for the ones at ISO 6400 I suppose ;)

RogerMac
9th February 2016, 12:10 PM
Except for the ones at ISO 6400 I suppose ;)

Oops I missed those , sorry and the Oly ones are perfectly good. I do not think it invalidates my point that, that at very high ISOs the FF really does come into its own.

Zuiko
9th February 2016, 12:23 PM
Oops I missed those , sorry and the Oly ones are perfectly good. I do not think it invalidates my point that, that at very high ISOs the FF really does come into its own.

Well of course that's true, in much the same way as 5x4 Large Format cameras really come into their own for huge, finely detailed landscapes with enormous depth of field by virtue of their shift and tilt movements. Each format has its advantages and disadvantages - we just need to pick the tool that is most appropriate to our own needs.

pdk42
9th February 2016, 03:23 PM
Oops I missed those , sorry and the Oly ones are perfectly good. I do not think it invalidates my point that, that at very high ISOs the FF really does come into its own.

What I also noticed is that the A7R delivered nearly noise free images at between ISO 800 and 1600. In good lighting and with careful PP you can get pretty close with u43 at the same ISOs - but if you need to push shadows or correct under-exposure or sharpen then it's a different story altogether. That for me, rather than increased resolution or DR, is what's appealing about FF. Since I came to u43 from Canon FF (5dii), it's interesting to see how much better the A7R is to the Canon.

Kiwi Paul
9th February 2016, 03:30 PM
I think there's always the caveat that it's just fun to try different systems out of interest, we each have different ideas of an ideal camera and format, and unless you give other formats a try you will never know and I prefer to make my own judgement from my own experiences where I can.

Paul

Harold Gough
9th February 2016, 03:33 PM
Well of course that's true, in much the same way as 5x4 Large Format cameras really come into their own for huge, finely detailed landscapes with enormous depth of field by virtue of their shift and tilt movements. Each format has its advantages and disadvantages - we just need to pick the tool that is most appropriate to our own needs.

John,

I'm not entirely sure that that is right. Shift (maybe) and tilt (definitely)can chose another principal plane of focus but I don't think they can increase overall DOF.

Harold

Harold Gough
9th February 2016, 03:38 PM
What I also noticed is that the A7R delivered nearly noise free images at between ISO 800 and 1600.

Paul,

This makes me very happy.

I notice that a lot of contributors have a strong interest in street photography and/or low light applications. I have neither. My default ISO is 400, using 200 or lower if I can, and 800 if I can't avoid it. Venturing to 1000 or 1600 is an adventure for desperate circumstances.

Harold

pdk42
9th February 2016, 05:18 PM
Paul,

This makes me very happy.

I notice that a lot of contributors have a strong interest in street photography and/or low light applications. I have neither. My default ISO is 400, using 200 or lower if I can, and 800 if I can't avoid it. Venturing to 1000 or 1600 is an adventure for desperate circumstances.

Harold

Since I do a lot of landscape and cityscape photography, most of my shooting is at ISO 200 so I'm pretty much in the same boat. For that kind of shooting IBIS is a massive advantage too - hence why I decided not to keep the A7R.

Ricoh
9th February 2016, 05:30 PM
Since I do a lot of landscape and cityscape photography, most of my shooting is at ISO 200 so I'm pretty much in the same boat. For that kind of shooting IBIS is a massive advantage too - hence why I decided not to keep the A7R.

Or the lens, I suppose. That's a very short 'rental', Paul, hope you recover a reasonable proportion of your outlay!

Ricoh
9th February 2016, 06:36 PM
Developments are on the horizon... improved DR will be welcome.

http://thephoblographer.us3.list-manage2.com/track/click?u=c936e6d50a1b870cefefb7762&id=bb12a3b211&e=b8a5b56ba0

pdk42
9th February 2016, 09:39 PM
Developments are on the horizon... improved DR will be welcome.

http://thephoblographer.us3.list-manage2.com/track/click?u=c936e6d50a1b870cefefb7762&id=bb12a3b211&e=b8a5b56ba0

Yes, I read that too Steve. Realistically though, I think a real camera with that technology is 2-3 years away. However, it'll revolutionise digital imaging for sure. I suspect it'll remove many of the advantages of FF - aside from subject isolation.

Kiwi Paul
9th February 2016, 09:53 PM
Developments are on the horizon... improved DR will be welcome.

http://thephoblographer.us3.list-manage2.com/track/click?u=c936e6d50a1b870cefefb7762&id=bb12a3b211&e=b8a5b56ba0

There ain't half some clever folk out there.

That really will change the game, in lots of ways, the mind boggles.

Paul

Ricoh
9th February 2016, 10:48 PM
Yes, I read that too Steve. Realistically though, I think a real camera with that technology is 2-3 years away. However, it'll revolutionise digital imaging for sure. I suspect it'll remove many of the advantages of FF - aside from subject isolation.

I was out earlier practising with the 50 lux f1.4. At near distance the DoF is seriously shallow, just breathing in and out could alters the focus plane. I've tried to imagine what life would be like with the Noctilux f 0.95; it must take some phenomenal skills to use correctly - just a heart beat and focus has vanished.

Kiwi Paul
9th February 2016, 10:52 PM
I was out earlier practising with the 50 lux f1.4. At near distance the DoF is seriously shallow, just breathing in and out could alters the focus plane. I've tried to imagine what life would be like with the Noctilux f 0.95; it must take some phenomenal skills to use correctly - just a heart beat and focus has vanished.

Yes the m43 PL 25 f1.4 gets shallow DOF but that's only dof equiv of 50mm f2.8 on FF, the 55 f1.8 on the A7ii was really shallow DOF, I was really surprised just how shallow.

Paul

Ricoh
9th February 2016, 10:57 PM
Yes, at 1m we're talking 4 cms. Looks damn impressive if performed accurately - and I'm using a RF!!

Ralph Harwood
12th February 2016, 09:16 PM
John,

I'm not entirely sure that that is right. Shift (maybe) and tilt (definitely)can chose another principal plane of focus but I don't think they can increase overall DOF.

Harold

Hi Harold!

I think that the trick with the tilt lens is to angle it so that the plane of focus follows the contours of the landscape, as it isn't vital that the sky is perfectly focussed but you can get all of the ground in front in fairly good focus if there is a fairly constant angle through the picture.

I have used a tilt adapter on my OM lenses to get a good depth of field on the side of a boat which was at a 45 degree angle to me such that all of the boat was in focus at f2.5 at 28mm - the water in front was oof and the background behind was oof too, but that was all to the good as they were not the focus of the picture.

Cheers,

Ralph.

Ian
12th February 2016, 09:57 PM
My understanding is the bigger the lens the more light it lets in. I have been trying to properly understand this light issue because I want to do astrophotography. I have come across a very good web site that properly explains this complex issue. The guy is a Dr Roger Clark who has spent years in imaging some including with NASA. I think he makes a lot of sense but his explanations seem to contradict a lot of what is published on the web. Here is his explanation about light gathering and f/ratios.

http://www.clarkvision.com/articles/low.light.photography.and.f-ratios/

A bigger entrance pupil gathers more light but the selected aperture should limit the light to a given brightness for a comparable are of illumination on the film/sensor plane (assuming the light source is the same brightness). So a larger lens does gather more light but its potential is limited by the aperture in the optical centre of the lens. This should be regardless of the lens focal length. With astrophotography involving telescopes there is no variable aperture - the optics work differently so the size of the entrance pupil is effectively the limiting factor - hence the larger the lens the more light you transmit.

Ian

Ian
12th February 2016, 10:09 PM
I think there is a flaw in the cropped sample images comparison. Do correct me if my assumption is wrong but if you are using the same field of view lens (35mm FF and near enough 17mm with the E-M1) then while the framing will be comparable between the two cameras if positioned at the same distance from the subject, the '100%' crops are misleading at best. They look to me to be the same size so the crops are the same fraction of the frame but of course the A7R has a 36 megapixel sensor so to get crops the same size on the screen here either the E-M1 has to be interpolated up to A7R size or the A7R has to be down-sized to the E-M1 size. The former does the E-M1 no favours while the latter can mitigate noise.

In the past to do a comparison like this I have made prints at a decent size and then scanned them.

While the A7R has four times the sensor area of the E-M1 it also only has barely more than twice the number of pixels so its pixel pitch is massively greater than the E-M1.

I'd like to know what you thought of the A7R in terms of usability - especially shutter sound/loudness and image stabilisation effectiveness. How did the Sony 35mm perform at f/1.8? :)

Ian

pdk42
12th February 2016, 11:14 PM
I think there is a flaw in the cropped sample images comparison. Do correct me if my assumption is wrong but if you are using the same field of view lens (35mm FF and near enough 17mm with the E-M1) then while the framing will be comparable between the two cameras if positioned at the same distance from the subject, the '100%' crops are misleading at best. They look to me to be the same size so the crops are the same fraction of the frame but of course the A7R has a 36 megapixel sensor so to get crops the same size on the screen here either the E-M1 has to be interpolated up to A7R size or the A7R has to be down-sized to the E-M1 size. The former does the E-M1 no favours while the latter can mitigate noise.

In the past to do a comparison like this I have made prints at a decent size and then scanned them.

While the A7R has four times the sensor area of the E-M1 it also only has barely more than twice the number of pixels so its pixel pitch is massively greater than the E-M1.

I'd like to know what you thought of the A7R in terms of usability - especially shutter sound/loudness and image stabilisation effectiveness. How did the Sony 35mm perform at f/1.8? :)

Ian
Ian - I downsized the Sony images to match the E-M1's size. As I said in the original post, I know this improves the Sony's noise and sharpness but OTOH, an image at about 4600 pixels would print to 16x20 at 240dpi and would present well on a 4k monitor so I thought it a reasonable compromise. As you say, the only way to truly level the field is to do real prints.

As to usability - well the issue isn't really usability per se but the overall package.

The Sony is like an American muscle car - massive V8 under the hood but the chassis is made from prams springs and the steering as vague as a Tony Blair speech. The E-M1 though is like a Golf GTI - small turbo-charged engine and a supremely engineered body around it.

If ripping up straights is what you want, and you can stomach the old-school handing, then the muscle car is the thing to go for; but if you want to make progress on a twisty road and get a smile from using a finely-tuned suspension and steering setup, then you'll want to be in the Golf. Oh, and dont even think about the differences parking them!

It's the same with the Sony and the E-M1. If the very best IQ in terms of noise and sharpness is your aim then you can't beat the Sony - just make sure you keep the shutter speed up and take care with focussing and camera stability. But if you want to take shots with a high degree of confidence in getting a usable result then the Oly beats it hands down. The IBIS and AF are extremely reliable and together return consistent results.

And just like the car analogy, where chances are that you'll arrive at your destination no faster in the V8 than the Golf, so it is with the cameras. Yes, the Sony's IQ is better, esp if you go pixel peeking, but I doubt on a print it would matter at all.

As to shutter noise - well the A7R's shutter is not only loud but it seems to take forever doing it - no matter what the shutter speed.

Maybe the A7Rii fixes much of the issues with the A7R - but at Ł2500 it's way beyond what I'd be prepared to pay.

Kiwi Paul
12th February 2016, 11:54 PM
In my short term experience with the A7ii I see is as a very similar camera to the original EM5, a lovely camera, great to use, no real flaws, a few things you wish it had, but a solid performer and a camera to like.
The trouble for Sony is Olympus now have the EM5ii, a much improved design with most of those "wish for" things accommodated. It has better ergonomics, great tactile feel to the buttons and dials and a solid build with a professional feel. The Sony A7ii feels a bit cheap in comparison, lacks the ergonomics, the tactile feels of the buttons and dials and just doesn't have the same completeness about it, not to mention all the additional features the EM5ii has, silent shutter, keystone compensation, touch screen, etc etc. I think the A7Rii is a more complete camera and is more like the EM5ii in feature set but will still lack the ergonomics and tactile feel of the EM5ii. I'll let you know next week ;)
The 2 prime lenses I have for the A7 are very nice lenses, the Sony/Zeiss 55 1.8 really is a delight to use and a very good lens, good build quality, sharp, detailed, good performance (focus speed and feel) and at f1.8 has such a shallow DOF, nice if you want that, the Sony 28 f2 is similar certainly no complaints.

Paul

drmarkf
13th February 2016, 11:39 PM
Fascinating thread.
It's also been fascinating to follow your thread on the Sony dpreview forum, although not solely for the photographic aspects!

Greytop
13th February 2016, 11:40 PM
Fascinating thread.
It's also been fascinating to follow your thread on the Sony dpreview forum, although not solely for the photographic aspects!

How about a link? ;)

drmarkf
14th February 2016, 09:27 AM
How about a link? ;)

I've just tried to find it again and have failed: that might just be my incompetence, but one or two of the 'difficult' dpreview forum members had joined in and at least one moderator had tried to calm things down. It might have been temporarily pulled. Maybe Paul can elucidate.

Dpreview has some very fine members, but thre is a very prominent Aspergoid Tendency as well...

Kiwi Paul
16th February 2016, 02:30 PM
I got my A7Rii yesterday, very nice camera, excellent detail and resolution with the Zeiss 55 and Sony 28 lenses.
It has the same EVF flicker the A7ii had but I've tried a few different A7 series cameras and they all have it, I've reduced the effect by turning down the EVF brightness, otherwise a very nice EVF.
I'll post more once I've had a chance to properly evaluate it and the photos I've taken so far.

Paul

Ivor
17th February 2016, 04:17 PM
I like the work that went on here.

I often wonder whether the minute differences that can be seen when 'pixel peeping' are that important. Compare a 100% view of any good-quality SLR or CSC with models from five years ago and the current ones will win hands down and the difference will be greater than comparing two contemporary cameras. But, people were taking fantastic, printable images five to ten years back, images that still hold there own.

The variation in image quality between different cameras now is so small that in almost any circumstance it makes no difference. Don't other factors like ergonomics, highest and lowest usable ISOs and fastest shutter speeds become more important?

Kiwi Paul
18th February 2016, 12:43 AM
I like the work that went on here.

I often wonder whether the minute differences that can be seen when 'pixel peeping' are that important. Compare a 100% view of any good-quality SLR or CSC with models from five years ago and the current ones will win hands down and the difference will be greater than comparing two contemporary cameras. But, people were taking fantastic, printable images five to ten years back, images that still hold there own.

The variation in image quality between different cameras now is so small that in almost any circumstance it makes no difference. Don't other factors like ergonomics, highest and lowest usable ISOs and fastest shutter speeds become more important?

That is true what you have said. But each format has it's advantages and disadvantages, full frame can provide higher usable ISO's, more dynamic range and shallower dof. On the contrary m43 (and ASP-C) can provide greater dof so it really depends what you are shooting as to which has the advantage from that perspective.
There are also other factors, particularly with lenses where FF lenses can suffer from more vignetting, edge softness and distortion than m43 lenses and are larger and heavier.

I do agree 100% though that the only way to judge a photo is by looking at the whole picture standing back at the distance it's meant to be viewed at, if it looks better using one system over another then that's definitive, pixel peeping I only ever use to see the resolution of a camera or lens, not to compare different cameras and lenses.

Paul