My E-M1 Mark II Sensor Measurements

Started Jan 23, 2017 | Discussions
bclaff Forum Pro • Posts: 11,230
My E-M1 Mark II Sensor Measurements
39

Thanks to raw files contributed by various people I have E-M1 Mark II results which are posted at my site. (Still interested in testing other camera models if this type of analysis interests you.)

There were some interesting surprises.
Photograph Dynamic Range (PDR) looks like this :

Note that the PDR at ISO 250 is the same as that at ISO 200 and ISO 64.

This is confirmed by Read Noise in DNs:

Note the lowest value at ISO 250.
It's a mystery why ISO 64 and ISO 200 are above and not level, as if they are based on about ISO 320 rather than ISO 250.

Also, exciting news, the drop in read noise above ISO 1600 indicates the use of dual gain conversion (first seen in a high end camera with the Sony A7).
This explains the big improvement at high ISO settings.
If you're curious about the technology read Aptina DR-Pix Technology Whitepaper.

I confirmed with 2D Fourier Transforms (FTs) that there is little noise reduction apparent in the raw files.

I'll post a separate thread on my Fixed Pattern Noise (FPN) results which were also quite interesting.

-- hide signature --

Bill ( Your trusted source for independent sensor data at http://www.PhotonsToPhotos.net )

Olympus E-M1 Olympus E-M1 II
If you believe there are incorrect tags, please send us this post using our feedback form.
boxerman Senior Member • Posts: 1,808
Re: My E-M1 Mark II Sensor Measurements
1

Thanks for this kind of work. It helps a lot understanding the capacities of the sensor. The difference from Mark I to Mark II seems VERY large. Around 2 stops. Can this be real? Is it naive to expect this should be obvious in real-world shooting?

Oh, and from the linked description, it appears the señor has quite specific and significant per-pixel circuitry to implement dual gain conversion. Puts the thinking that it's a slightly revamped Pen F sensor to bed. I always thought that was naive, on the surface.

Thanks.

-- hide signature --

The BoxerMan

 boxerman's gear list:boxerman's gear list
Olympus E-M1 III Olympus M.Zuiko Digital ED 9-18mm F4.0-5.6 Panasonic Lumix G 20mm F1.7 ASPH Olympus M.Zuiko Digital ED 40-150mm F4-5.6 R Olympus 12-40mm F2.8 Pro +3 more
OP bclaff Forum Pro • Posts: 11,230
Re: My E-M1 Mark II Sensor Measurements
7

boxerman wrote:

Thanks for this kind of work. It helps a lot understanding the capacities of the sensor.

Thanks. And this is why I'm a broken record about getting people to contribute

The difference from Mark I to Mark II seems VERY large. Around 2 stops. Can this be real? Is it naive to expect this should be obvious in real-world shooting?

Looks like about 1 1/3 stops.
No improvement is obvious unless it is needed.
In other words, I would expect the Mark II to be able to handle more DR; but that would only show in a situation where that additional DR is needed.

Oh, and from the linked description, it appears the sensor has quite specific and significant per-pixel circuitry to implement dual gain conversion. Puts the thinking that it's a slightly revamped Pen F sensor to bed. I always thought that was naive, on the surface.

I generally avoid speculation and stick to what I can measure

-- hide signature --

Bill ( Your trusted source for independent sensor data at http://www.PhotonsToPhotos.net )

boxerman Senior Member • Posts: 1,808
Re: My E-M1 Mark II Sensor Measurements
1

bclaff wrote:

boxerman wrote:

...

The difference from Mark I to Mark II seems VERY large. Around 2 stops. Can this be real? Is it naive to expect this should be obvious in real-world shooting?

Looks like about 1 1/3 stops.

Oops, looks like I read wrong. Sorry.

No improvement is obvious unless it is needed.
In other words, I would expect the Mark II to be able to handle more DR; but that would only show in a situation where that additional DR is needed.

Yes, of course, but who hasn't run across situations where increased DR is a help? Well, maybe portrait photographers, where they can control the light. I'm out in the woods (so to speak) so high DR shots are almost always a challenge. Sunsets almost always stretch DR. Photography on the water (with reflections). And, I've fairly often had small areas of high interest, at high brightness, which get killed in contrast, unless you're very, very careful with exposure and willing and skillful with post processing.

Oh, and from the linked description, it appears the sensor has quite specific and significant per-pixel circuitry to implement dual gain conversion. Puts the thinking that it's a slightly revamped Pen F sensor to bed. I always thought that was naive, on the surface.

I generally avoid speculation and stick to what I can measure

Ah, but I think the measurement nailed something that others might only speculate about!

-- hide signature --

The BoxerMan

 boxerman's gear list:boxerman's gear list
Olympus E-M1 III Olympus M.Zuiko Digital ED 9-18mm F4.0-5.6 Panasonic Lumix G 20mm F1.7 ASPH Olympus M.Zuiko Digital ED 40-150mm F4-5.6 R Olympus 12-40mm F2.8 Pro +3 more
photofisher Senior Member • Posts: 1,567
Re: My E-M1 Mark II Sensor Measurements
1

bclaff wrote:

Thanks to raw files contributed by various people I have E-M1 Mark II results which are posted at my site. (Still interested in testing other camera models if this type of analysis interests you.)

There were some interesting surprises.
Photograph Dynamic Range (PDR) looks like this :

Note that the PDR at ISO 250 is the same as that at ISO 200 and ISO 64.

This is confirmed by Read Noise in DNs:

Note the lowest value at ISO 250.
It's a mystery why ISO 64 and ISO 200 are above and not level, as if they are based on about ISO 320 rather than ISO 250.

Also, exciting news, the drop in read noise above ISO 1600 indicates the use of dual gain conversion (first seen in a high end camera with the Sony A7).
This explains the big improvement at high ISO settings.
If you're curious about the technology read Aptina DR-Pix Technology Whitepaper.

I confirmed with 2D Fourier Transforms (FTs) that there is little noise reduction apparent in the raw files.

I'll post a separate thread on my Fixed Pattern Noise (FPN) results which were also quite interesting.

How do you explain DXO not finding an increase in base ISO dynamic range, where you did?

Just curious .

-- hide signature --

Amateur Photographer

 photofisher's gear list:photofisher's gear list
Olympus E-M1 Olympus Zuiko Digital ED 50mm 1:2.0 Macro Olympus Zuiko Digital ED 9-18mm 1:4.0-5.6 Panasonic Leica Summilux DG 25mm F1.4 Olympus M.Zuiko Digital ED 75mm F1.8 +3 more
OP bclaff Forum Pro • Posts: 11,230
Re: My E-M1 Mark II Sensor Measurements
2

photofisher wrote:

bclaff wrote:

Photograph Dynamic Range (PDR) looks like this :

How do you explain DXO not finding an increase in base ISO dynamic range, where you did?

I assume you mean as opposed to this:

A small part of the answer is that their x-axis is "Measured" ISO and mine is the ISO setting.

But mostly DxoMark calculated their Dynamic Range (Landscape score) differently than I calculate Photographic Dynamic Range (PDR).
Very technically theirs is based on pixel read noise while mine is based on a Signal to Noise Ratio (SNR) criteria.

DxOMark measured more cameras than I do but most people find my PDR lines up better with actual experience.

-- hide signature --

Bill ( Your trusted source for independent sensor data at http://www.PhotonsToPhotos.net )

photofisher Senior Member • Posts: 1,567
Re: My E-M1 Mark II Sensor Measurements
4

Thanks.....I think your results match this review nicely:

http://mirrorlesscomparison.com/olympus-vs-olympus/omd-em1-vs-omd-em1-ii/#ISO-performance

-- hide signature --

Amateur Photographer

 photofisher's gear list:photofisher's gear list
Olympus E-M1 Olympus Zuiko Digital ED 50mm 1:2.0 Macro Olympus Zuiko Digital ED 9-18mm 1:4.0-5.6 Panasonic Leica Summilux DG 25mm F1.4 Olympus M.Zuiko Digital ED 75mm F1.8 +3 more
PurpleFringe Contributing Member • Posts: 663
Re: My E-M1 Mark II Sensor Measurements

So PDR of the E-M1 mk II is better than 'Ideal 4/3' at ISO 250 and above?

http://www.photonstophotos.net/Charts/PDR.htm#Ideal%204/3,Olympus%20OM-D%20E-M1,Olympus%20OM-D%20E-M1%20Mark%20II

Inquiring minds would like to know more...

 PurpleFringe's gear list:PurpleFringe's gear list
Fujifilm FinePix X100 Olympus PEN E-P1 Olympus E-M1 Olympus E-M1 II Panasonic Lumix G 20mm F1.7 ASPH +7 more
OP bclaff Forum Pro • Posts: 11,230
Re: My E-M1 Mark II Sensor Measurements
2

PurpleFringe wrote:

So PDR of the E-M1 mk II is better than 'Ideal 4/3' at ISO 250 and above?

http://www.photonstophotos.net/Charts/PDR.htm#Ideal%204/3,Olympus%20OM-D%20E-M1,Olympus%20OM-D%20E-M1%20Mark%20II

Inquiring minds would like to know more...

Once again, that's because of how aggressively Olympus shifts their true raw ISO.
Mentally move the curve left about 1.23 stops (!)

Regards,

-- hide signature --

Bill ( Your trusted source for independent sensor data at http://www.PhotonsToPhotos.net )

David
David Regular Member • Posts: 343
Re: My E-M1 Mark II Sensor Measurements
1

bclaff wrote:

Thanks to raw files contributed by various people I have E-M1 Mark II results which are posted at my site. (Still interested in testing other camera models if this type of analysis interests you.)

There were some interesting surprises.
Photograph Dynamic Range (PDR) looks like this :

Note that the PDR at ISO 250 is the same as that at ISO 200 and ISO 64.

This is confirmed by Read Noise in DNs:

Note the lowest value at ISO 250.
It's a mystery why ISO 64 and ISO 200 are above and not level, as if they are based on about ISO 320 rather than ISO 250.

Also, exciting news, the drop in read noise above ISO 1600 indicates the use of dual gain conversion (first seen in a high end camera with the Sony A7).
This explains the big improvement at high ISO settings.
If you're curious about the technology read Aptina DR-Pix Technology Whitepaper.

I confirmed with 2D Fourier Transforms (FTs) that there is little noise reduction apparent in the raw files.

I'll post a separate thread on my Fixed Pattern Noise (FPN) results which were also quite interesting.

Hi Bill,

Can you elaborate more about the use of dual gain conversion with the E-M1 Mk II vs Mk 1?  Specially, how does color saturation be effected?  Will it be the same as the Mk1 but with less noise?

In my observation helping a pro photographer of mine who shoots with both a Nikon D4 and D5, he and I noticed that the D5 loses color saturation faster than the D4 from 12800, 25600 and up with the same ISO, shutter speed and lenses despite the D5 having more DR than the D4.

I wonder if you have measurements like that in your website and if so, what should I be looking for?

Anyhow, I truly find your work to be "MORE" accurate representation of shooting results than DXO does.

Thank you..

 David's gear list:David's gear list
Nikon Coolpix AW110 Panasonic ZS100 Olympus PEN E-P5 Panasonic Lumix G X Vario PZ 14-42mm F3.5-5.6 ASPH OIS Samyang 7.5mm F3.5 Fisheye +4 more
JaKing
JaKing Senior Member • Posts: 6,299
Re: My E-M1 Mark II Sensor Measurements

Thanks for your hard work, Bill. Shows at a technical level what I am seeing in my files.

The Aptina white paper was very interesting, as far as my limited understanding allowed. Is the DR technology Aptina's patent? From your analysis it appears that the Olympus claim that this is a brand new sensor is true, and it may not even be a Sony sensor. Your thoughts?

-- hide signature --

br, john, from you know where
My gear list and sordid past are here: https://www.dpreview.com/members/1558378718/overview
Gallery: http://canopuscomputing.com.au/gallery2/v/main-page/

 JaKing's gear list:JaKing's gear list
Olympus E-M1 Olympus E-M1 II Olympus M.Zuiko Digital ED 14-42mm F3.5-5.6 EZ Olympus 12-100mm F4.0 Olympus E-1 +28 more
OP bclaff Forum Pro • Posts: 11,230
Re: My E-M1 Mark II Sensor Measurements
1

David wrote:

bclaff wrote:

... the drop in read noise above ISO 1600 indicates the use of dual gain conversion (first seen in a high end camera with the Sony A7).
This explains the big improvement at high ISO settings.
If you're curious about the technology read Aptina DR-Pix Technology Whitepaper.

...

Hi Bill,

Can you elaborate more about the use of dual gain conversion with the E-M1 Mk II vs Mk 1?

To better understand dual conversion gain consult the Whitepaper link above.

Specially, how does color saturation be effected? Will it be the same as the Mk1 but with less noise?

In my observation helping a pro photographer of mine who shoots with both a Nikon D4 and D5, he and I noticed that the D5 loses color saturation faster than the D4 from 12800, 25600 and up with the same ISO, shutter speed and lenses despite the D5 having more DR than the D4.

Color differences have nothing to do with the signal processing chain.
The most significant factor is the Color Filter Array (CFA, often mistakenly called "Bayer").
The strength of the filters is a trade-off between color fidelity and Quantum Efficiency (QE).

I wonder if you have measurements like that in your website and if so, what should I be looking for?

I don't measure color, DxOMark does but I'm not too familiar with the interpretation of their Color Sensitivity (CS) and other color measures.

Anyhow, I truly find your work to be "MORE" accurate representation of shooting results than DXO does.

Thanks for the feedback.

Thank you..

You are welcome.

-- hide signature --

Bill ( Your trusted source for independent sensor data at http://www.PhotonsToPhotos.net )

OP bclaff Forum Pro • Posts: 11,230
Re: My E-M1 Mark II Sensor Measurements
1

JaKing wrote:

Thanks for your hard work, Bill. Shows at a technical level what I am seeing in my files.

You are welcome.

The Aptina white paper was very interesting, as far as my limited understanding allowed. Is the DR technology Aptina's patent?

I'm sure there is a patent but I don't know if the use is by a licensing agreement or if the implementation is sufficiently novel to not infringe on the Aptina (now On Semicondutor).
Regardless, this type of technology is becoming more and more common.

From your analysis it appears that the Olympus claim that this is a brand new sensor is true, and it may not even be a Sony sensor. Your thoughts?

No thoughts. I don't care about who designed or what foundry was used.
I just live by the numbers

Regards,

-- hide signature --

Bill ( Your trusted source for independent sensor data at http://www.PhotonsToPhotos.net )

SHood Veteran Member • Posts: 5,731
Re: My E-M1 Mark II Sensor Measurements
2

bclaff wrote:

JaKing wrote:

Thanks for your hard work, Bill. Shows at a technical level what I am seeing in my files.

You are welcome.

The Aptina white paper was very interesting, as far as my limited understanding allowed. Is the DR technology Aptina's patent?

I'm sure there is a patent but I don't know if the use is by a licensing agreement or if the implementation is sufficiently novel to not infringe on the Aptina (now On Semicondutor).
Regardless, this type of technology is becoming more and more common.

From your analysis it appears that the Olympus claim that this is a brand new sensor is true, and it may not even be a Sony sensor. Your thoughts?

No thoughts. I don't care about who designed or what foundry was used.
I just live by the numbers

Regards,

Sony and Aptina signed a cross-license for their patent porfolio 4 years ago. This was likely included in it.

http://image-sensors-world.blogspot.ca/2013/02/aptina-and-sony-cross-license-their.html

-- hide signature --
 SHood's gear list:SHood's gear list
Olympus E-M1 II Olympus OM-D E-M1X Olympus M.Zuiko 300mm F4 IS Pro Olympus 12-100mm F4.0 Olympus 17mm F1.2 Pro
John Brawley Contributing Member • Posts: 712
Re: My E-M1 Mark II Sensor Measurements
1

bclaff wrote:

I'm sure there is a patent but I don't know if the use is by a licensing agreement or if the implementation is sufficiently novel to not infringe on the Aptina (now On Semicondutor).

It's used by others.

In motion imaging most of Blackmagic cameras are dual gain sensors and the Arri Alexa's are too.

http://www.arri.com/camera/alexa_mini/technology/arri_imaging_technology/alexa_mini_sensor/

I believe Panasonic do something similar in their Varicam / Varicam LT cameras but instead of summing for more DR, have a dual ISO of 800 / 5000.

JB

-- hide signature --

John Brawley
Cinematographer
Sydney Australia
www.johnbrawley.com
http://johnbrawley.wordpress.com/

 John Brawley's gear list:John Brawley's gear list
Leica M8 Leica M-E Typ 220 Olympus E-M5 II Olympus E-M1 II Leica M10 +37 more
SHood Veteran Member • Posts: 5,731
Re: My E-M1 Mark II Sensor Measurements
2

Call me crazy but I have set the ISO Default to 250 for when I use Auto ISO.

-- hide signature --
 SHood's gear list:SHood's gear list
Olympus E-M1 II Olympus OM-D E-M1X Olympus M.Zuiko 300mm F4 IS Pro Olympus 12-100mm F4.0 Olympus 17mm F1.2 Pro
OP bclaff Forum Pro • Posts: 11,230
Re: My E-M1 Mark II Sensor Measurements
3

John Brawley wrote:

bclaff wrote:

I'm sure there is a patent but I don't know if the use is by a licensing agreement or if the implementation is sufficiently novel to not infringe on the Aptina (now On Semicondutor).

It's used by others.

In motion imaging most of Blackmagic cameras are dual gain sensors and the Arri Alexa's are too.

http://www.arri.com/camera/alexa_mini/technology/arri_imaging_technology/alexa_mini_sensor/

Note that Dual Gain Architecture (DGA) is not the same a Dual Conversion Gain.
DGA reads out both high and low gain from a pixel.
("DGA simultaneously provides two separate read-out paths from each pixel with different amplification.")
Dual Conversion Gain changes the conversion gain in the pixel as a function of the ISO setting. Dual Conversion Gain lowers the read noise (and FWC) at higher ISO settings.

Regards,

-- hide signature --

Bill ( Your trusted source for independent sensor data at http://www.PhotonsToPhotos.net )

dkyl Contributing Member • Posts: 892
Re: My E-M1 Mark II Sensor Measurements

bclaff wrote:

Once again, that's because of how aggressively Olympus shifts their true raw ISO.
Mentally move the curve left about 1.23 stops (!)

Regards,

-- hide signature --

Bill ( Your trusted source for independent sensor data at http://www.PhotonsToPhotos.net )

Do you by any chance know how much Olympus shifts their true RAW ISO for the EM1 MKI?

Perhaps it would be beneficial to be able to implement a way to offset your graphs on the x-axis by a certain number of stops.

 dkyl's gear list:dkyl's gear list
Olympus Tough TG-4 Nikon D700 Panasonic Lumix DC-G9 Olympus E-PL9 Nikon Z6 II +6 more
brunobarolo Senior Member • Posts: 1,118
Re: My E-M1 Mark II Sensor Measurements

dkyl wrote:

bclaff wrote:

Once again, that's because of how aggressively Olympus shifts their true raw ISO.
Mentally move the curve left about 1.23 stops (!)

Regards,

Do you by any chance know how much Olympus shifts their true RAW ISO for the EM1 MKI?

Perhaps it would be beneficial to be able to implement a way to offset your graphs on the x-axis by a certain number of stops.

I'm not Bill, but I suppose you'll find your answer in the DxO graphs:

https://www.dxomark.com/Cameras/Compare/Side-by-side/Olympus-OM-D-E-M1-Mark-II-versus-Olympus-OM-D-E-M1___1136_909

Jorginho Forum Pro • Posts: 15,267
Thx for a nice explanation!! Good job.
1

No text.

 Jorginho's gear list:Jorginho's gear list
Olympus PEN E-PL5 Panasonic Lumix DMC-GH4 Olympus E-M1 II Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G 20mm F1.7 ASPH +8 more
Keyboard shortcuts:
FForum MMy threads