new Olympus EM1X - Everything and the kitchen sink - updated feature list rumor

AlanF

Canon 5DSR II
Aug 16, 2012
5,663
2,923
What I take away from this release are some of the things that can be done with a mirrorless camera if it has enough computing power. I would expect to see some of these features (but not all) on a higher end R camera. It will be very interesting to see what the R equivalent of the 1DX2 will be.
Agreed Don. As I wrote at the beginning; put an FF sensor in it and it will be a winner. But, Panasonic are coming out with an FF body and they are giving Olympus a good run in the MFT department. Also, Sigma will be supplying lenses for their new L mount, which will make up somewhat for the lack of Panasonic FF lenses.
 

AlanF

Canon 5DSR II
Aug 16, 2012
5,663
2,923
Price has just been released - £2799, presumably similar in USD.
 

AlanF

Canon 5DSR II
Aug 16, 2012
5,663
2,923
Wrote too soon about the lack of Olympus Telephoto lenses, they have just announced one https://www.olympus-global.com/news/2019/nr01035.html
M.ZUIKO DIGITAL ED 150-400mm F4.5 TC1.25x IS PRO with a built in 1.25xTC and an external 2xTC, available next year. But, to put it into perspective, the effective reach is only about the same as a Sigma or Tamron 150-600mm on a 5DSR (496mm without the TC, 620 with the TC), with half the field of view.
 

old-pr-pix

EOS RP
Dec 26, 2011
384
29
Some day I hope to get bird shots as good as yours, but for now, perhaps you can help me understand your math above. When you say "effective reach" with 2X TC is 620, what are you taking into account? To me, 400 mm w/1.25 TC is 500 mm; adding another 2X TC takes it to 1000mm... what am I overlooking?
 

Pape

EOS RP
Dec 31, 2018
381
225
he wasnt adding 2x cause its more future stuff i guess. It need to be really amazing otpical quality lens if allow 2 converter and give still more sharpness on suchs tiny pixel camera than 20mpixel mft . sounds very difficult + 2x tc makes it go under diffraction limit so i would think 2x is waste of money?
maybe that 2x is meaned for 300mm 4f zuiko?
or maybe not, doesnt diffraction start affect after 5.6 on MFTs? on aps-c lens tests give shapest pic on 5,6 and mft is even smaller.
i guess it depends lens sharpness ,if big white quality diffraction would drop sharpness on 4 or 2,8 already when 5,6 is for average lens ?
 
Last edited:
  • Like
Reactions: AlanF

AlanF

Canon 5DSR II
Aug 16, 2012
5,663
2,923
Some day I hope to get bird shots as good as yours, but for now, perhaps you can help me understand your math above. When you say "effective reach" with 2X TC is 620, what are you taking into account? To me, 400 mm w/1.25 TC is 500 mm; adding another 2X TC takes it to 1000mm... what am I overlooking?
Pape is correct, I was referring to the built in 1.25xTC. He is also absolutely right about diffraction. The diffraction limited aperture for the Olympus is f/5.3. The 1.25xTC on f/4.5 gives f/5.6 at 500mm, just breaching the DLA, and add 2x to that gives f/11, which is more than 2x the DLA. Adding the 2xTC will give no increase in resolution because the doubling of the focal length is negated by halving the resolution by diffraction and further so by the hit on IQ of a 2xTC (usually 20-30% loss).
 
  • Like
Reactions: Pape

Mikehit

EOS 5D MK IV
Jul 28, 2015
3,213
398
Pape is correct, I was referring to the built in 1.25xTC. He is also absolutely right about diffraction. The diffraction limited aperture for the Olympus is f/5.3. The 1.25xTC on f/4.5 gives f/5.6 at 500mm, just breaching the DLA, and add 2x to that gives f/11, which is more than 2x the DLA. Adding the 2xTC will give no increase in resolution because the doubling of the focal length is negated by halving the resolution by diffraction and further so by the hit on IQ of a 2xTC (usually 20-30% loss).
As far as I am aware diffraction is irrelevant in these circumstances.
For a given viewing size, the diffraction is identical. But given a choice between a lower resolution sensor and (theoretically less) diffraction versus a higher resolution sensor with (theoretically more) diffraction, the image from the higher resolution sensor will always win. Diffraction will not negate the higher resolution.
 

Don Haines

Beware of cats with laser eyes!
Jun 4, 2012
8,130
1,650
Canada
As far as I am aware diffraction is irrelevant in these circumstances.
For a given viewing size, the diffraction is identical. But given a choice between a lower resolution sensor and (theoretically less) diffraction versus a higher resolution sensor with (theoretically more) diffraction, the image from the higher resolution sensor will always win. Diffraction will not negate the higher resolution.
Exactly, the diffraction limit is not a brick wall!
 

Pape

EOS RP
Dec 31, 2018
381
225
nope it isnt but difference between using extender or not .may be so small that isnt really worth of 4x longer shutter speed
complex thingies to think ,should always test with real equipments :)
 

AlanF

Canon 5DSR II
Aug 16, 2012
5,663
2,923
Exactly, the diffraction limit is not a brick wall!
The MTF value (resolution) drops to zero at the diffraction limit, ie is a brick wall cut off, and above that increases progressively with increasingly widening aperture until eventually it is not diffraction limited. Here is a page from textbook with a computed graph https://spie.org/publications/tt52_151_diffraction_mtf?SSO=1 and the equations.
When you read articles that say the diffraction limit is not a brick wall they are oversimplifying and what they mean is that there is progressive loss of resolution as the aperture is decreased until you hit the diffraction limit. That is, you cannot resolve details that are smaller than the diffraction limited but larger ones get progressively sharper with increasing size. This has been a classic problem facing astronomers and microscopists throughout the generations. Once you hit the diffraction limited aperture, increasing the focal length of a telescope, for example, without increasing the size of the aperture does not increase the cut-off resolution dictated by the diffraction of light.

The above applies to a purely diffraction limited system. But, there are other factors such as aberration of the lens, Baeyer, AA-filter etc that will lower the diffraction-limited MTF still further. Oversampling, change of strength of filter and the effect of the TC on aberration might help to some extent.The take home message is that putting a 2xTC on a predominantly diffraction-limited system will double the size of an image but won't do much to the minimum size of an object that can be resolved. You would do just about as well by up-resolving in Photoshop.
 
  • Like
Reactions: mikekeck and Pape

Pape

EOS RP
Dec 31, 2018
381
225
ah yeah i see i got wrong about term diffraction limit .always thought its point where image quality start degrease . But it means point where image quality totally collapses. yep its brick wall there eventually. on f8 where it really start dive straight downward mtf curve .
 
  • Like
Reactions: AlanF

AlanF

Canon 5DSR II
Aug 16, 2012
5,663
2,923
ah yeah i see i got wrong about term diffraction limit .always thought its point where image quality start degrease . But it means point where image quality totally collapses. yep its brick wall there eventually. on f8 where it really start dive straight downward mtf curve .
That's right. Opticallimits in their testing with the 5DSR say that you see the effects of diffraction from f/4 upwards and you can see it in their charts.
 
  • Like
Reactions: Pape

Don Haines

Beware of cats with laser eyes!
Jun 4, 2012
8,130
1,650
Canada
ah yeah i see i got wrong about term diffraction limit .always thought its point where image quality start degrease . But it means point where image quality totally collapses. yep its brick wall there eventually. on f8 where it really start dive straight downward mtf curve .
Same here
 

Mikehit

EOS 5D MK IV
Jul 28, 2015
3,213
398
The MTF value (resolution) drops to zero at the diffraction limit, ie is a brick wall cut off, and above that increases progressively with increasingly widening aperture until eventually it is not diffraction limited. Here is a page from textbook with a computed graph https://spie.org/publications/tt52_151_diffraction_mtf?SSO=1 and the equations.
When you read articles that say the diffraction limit is not a brick wall they are oversimplifying and what they mean is that there is progressive loss of resolution as the aperture is decreased until you hit the diffraction limit. That is, you cannot resolve details that are smaller than the diffraction limited but larger ones get progressively sharper with increasing size. This has been a classic problem facing astronomers and microscopists throughout the generations. Once you hit the diffraction limited aperture, increasing the focal length of a telescope, for example, without increasing the size of the aperture does not increase the cut-off resolution dictated by the diffraction of light.

The above applies to a purely diffraction limited system. But, there are other factors such as aberration of the lens, Baeyer, AA-filter etc that will lower the diffraction-limited MTF still further. Oversampling, change of strength of filter and the effect of the TC on aberration might help to some extent.The take home message is that putting a 2xTC on a predominantly diffraction-limited system will double the size of an image but won't do much to the minimum size of an object that can be resolved. You would do just about as well by up-resolving in Photoshop.
I am not denying that diffraction exists. I was arguing the effect of diffraction between different sensor models.
Because what you seem to be saying is that if I put the 100mm f2.8 macro on the 30D (8MP) and on the 5DSR (50MP) and stop down to f16 for both cameras, then the image on the 5DSR will be worse than the 30D because the diffraction is 'higher' on the 5DSR.
The only reason you will not see diffraction on the 30D is because it has insufficient resolution to show the detail being affected by diffraction.

I agree totally that diffraction will cause an image to collapse, but the aperture at which this happens is irrelevant of the sensor resolution. And anyway, the aperture at which that happens is so tiny that it is in effect irrelevant to photographers.
You mention about astrophotography but that is because they are magnifying images of a tiny tiny fraction of the sky and are in effect doing extreme magnifications. Even doing wildlife needing heavy cropping you will hit atmospheric resolution issues way before diffraction becomes an issue.

Unless of course you can present actual pictorial evidence to show I am incorrect.
 

AlanF

Canon 5DSR II
Aug 16, 2012
5,663
2,923
I am not denying that diffraction exists. I was arguing the effect of diffraction between different sensor models.
Because what you seem to be saying is that if I put the 100mm f2.8 macro on the 30D (8MP) and on the 5DSR (50MP) and stop down to f16 for both cameras, then the image on the 5DSR will be worse than the 30D because the diffraction is 'higher' on the 5DSR.
The only reason you will not see diffraction on the 30D is because it has insufficient resolution to show the detail being affected by diffraction.

I agree totally that diffraction will cause an image to collapse, but the aperture at which this happens is irrelevant of the sensor resolution. And anyway, the aperture at which that happens is so tiny that it is in effect irrelevant to photographers.
You mention about astrophotography but that is because they are magnifying images of a tiny tiny fraction of the sky and are in effect doing extreme magnifications. Even doing wildlife needing heavy cropping you will hit atmospheric resolution issues way before diffraction becomes an issue.

Unless of course you can present actual pictorial evidence to show I am incorrect.
You are attributing to me ideas I have never had or promulgated and have distorted what I have written and what I am trying to get across. No way have I ever stated or implied that diffraction would cause a high density sensor to have lower resolution than a low density one. What diffraction does is to smear out points of light (by making Airy disks) so they become larger than the size of the pixels. The smaller the pixels, the sooner that happens with increasing f-number but, all things being equal, a high density sensor never has worse resolution than a lower density sensor. What I stated absolutely clearly and unambiguously is the take home message: "The take home message is that putting a 2xTC on a predominantly diffraction-limited system will double the size of an image but won't do much to the minimum size of an object that can be resolved. You would do just about as well by up-resolving in Photoshop."

I never even mentioned astrophotography - I wrote about telescopes, because the astronomers were aware of diffraction limitations way before sensors were invented. It was in 1835 that AIry published his work on the Airy disk, whose diameter is used to calculate diffraction limited apertures. The astrononomers knew nearly 200 years ago that their early telescopes with relatively narrow apertures were limited by diffraction. If you think that "Even doing wildlife needing heavy cropping you will hit atmospheric resolution issues way before diffraction becomes an issue." then I will give you the benefit of the doubt and assume that you must spend your time photographing wildlife in a heat haze.

An image doesn't collapse as f-number increases, it progressively loses the finer details as the size of the Airy disk increases and obscures the smaller elements.
https://en.wikipedia.org/wiki/Airy_disk
Edit: It's worth adding that when you double the focal length of the lens, you double the size of the image but if you are doing this with a 2xTC you double the size of the Airy disk because you double the f-number.
 
Last edited:
  • Like
Reactions: Pape

AlanF

Canon 5DSR II
Aug 16, 2012
5,663
2,923
Here is an amusing collage I made a little while ago that illustrates diffraction limitations with the P1000. I took a series of shots at different focal lengths with the diffraction limited Nikon, and compared them with some Canon gear. You can see that at a supposed 3000mm focal length the resolution is poorer than at 2000mm which is similar to 1600mm. Increasing the focal length without increasing the aperture didn't help the Nikon and its increased aberrations actually lowered resolution. (All taken at same distance and scaled to same size.)
NikonP1000_Collage.jpg
 
Last edited:

AlanF

Canon 5DSR II
Aug 16, 2012
5,663
2,923
And here is evidence about the loss of resolution with increasing f-number. ePhotozine has measured the MTFs of lenses on the 5DSR. I have plotted them for the best lenses under where they are diffraction limited. Under ideal conditions, the MTFs should tend to 5792 lines/picture height. You can see how the image loses resolution with increasing f-number
5DSR_ephotozine_New.jpg
.