Micro 4/3 dynamic range

[ymc226]

In digital, I've only used full frame (Leica M9-P, MM and 240) and Nikon D800.

My usual places to take pictures are at the beach with salty air/sand blowing away and standing in the waves or deeper at times which has got me worried about sensor corrosion in the Leica's given the recent news.

How is the dynamic range performance in terms of printing on paper up to 16x20 of the OM-D EM5 (now Mk 2) compared to a full frame sensor? Any difference at this small size?

Advantages for me of an Olympus system are fast autofocus (for my aging eyes) and image stabilization good for up to 5 stops. Also if I happen to dunk my kit in salt water, the loss would be 10% of the Leica kit.

In terms of the PRO line of Olympus lenses, are they comparable to modern Leica ASPH lenses? I'm thinking about a OM-D EM-5 Mk II and M.Zuiko Digital ED 12-40 f/2,8 PRO lens to start with. I feel the Nikon D800 is too bulky to carry around. Not having a comparison, is the OM-D that much smaller?

 

[Godfrey]

I have a pretty complete set of FourThirds SLR/Micro-FourThirds bodies (E-1, E-M1, E-PL7) and lenses (from 11mm to 283mm). And I work with lots of Leica equipment (M4-2, CL, M9 becoming M-P, Leicaflex SL, Leica R8) along with a kaboodle of lenses, both R and M mount.

The E-M1 and E-PL7 sensors are 16 Mpixel sensors with, by my measurements, about 12 stops of dynamic range. They are about as clean at ISO 6400 as my Sony A7 (24 Mpixel), FAR cleaner than the Leica M9 (18Mpixel) at ISO 2500, and retain more dynamic range at the upper limit*of sensitivity than the Leica too. The Sony A7 sensor holds the performance another stop and a half, but their lossy raw format cuts back on the quality by a bit. I'm eager to see how the M-P sensor compares.

In fact, I have printed for and won photo exhibition contests with even the ancient E-1 (manufactured in 2003, 5Mpixel, heavy AA filter, smallest DR and least sensitivity of the bunch) with 20x24 inch sized prints. ALL of the modern cameras outperform it on DR, sensitivity, and resolution.

And with that said, the E-1 STILL produces some of the most beautiful photographs of all. Take away: beautiful photographs cannot be found in specifications alone.

...
In terms of the PRO line of Olympus lenses, are they comparable to modern Leica ASPH lenses? I'm thinking about a OM-D EM-5 Mk II and M.Zuiko Digital ED 12-40 f/2,8 PRO lens to start with. I feel the Nikon D800 is too bulky to carry around. Not having a comparison, is the OM-D that much smaller?

In my opinion, Olympus is a top-tier lens producer on par with Leica and Zeiss. Like with all the top-tier lens producers, the entire range is good but of course there are the exceptional ones in the range. And each of the top-tier producers' lens design foo is different from each other to some degree. I have no problem shooting with Leica and Olympus gear and just picking the best photos that work for a project. The results in print are indistinguishable.

The OM-D E-M1 like mine or E-M5 (+ Mk II) bodies are very compact; the E-PL7 body even moreso. They make a Nikon D800 body look like a behemoth. Easiest to see the difference from a top view:

The Olympus pro grade zoom lenses are not so compact: the 12-40/2.8 is large for an mFT lens, the 40-150/2.8 is too, and they're much larger than most Leica M-mount lenses. But they're much more compact than lenses of similar field of view for FF format SLR cameras.

So what suits what you want to carry around is a matter of what feels best in your hands, and how much you want to carry. The only way you're going to know what works for you is to go to a store that has the equipment and evaluate it in person, or buy it, use it, then return it if it isn't suitable. Far as I've been able to determine, there aren't many other ways to really get to know the gear other than to put it in your hands and use it...

G

 

[ymc226]

Thanks Godfrey. Just what I wanted to hear. The lossy compression of the Sony A7MkII was mentioned by Ming Thein and is reason why that is not an option.

I'm also missing a lot of shot opportunities (my kids on beaches) regarding the MF of the Leica so the Olympus is looking better and better. Can you just rinse off under the faucet after the beach and not worry about water damage.

 

[Godfrey]

Thanks Godfrey. Just what I wanted to hear. The lossy compression of the Sony A7MkII was mentioned by Ming Thein and is reason why that is not an option.

I'm also missing a lot of shot opportunities (my kids on beaches) regarding the MF of the Leica so the Olympus is looking better and better. Can you just rinse off under the faucet after the beach and not worry about water damage.

I don't know about the E-M5/E-M5 II, but with the E-M1 (fitted with MMF-3 and the high-grade ZD 11-22/2.8-3.5 SLR lens) I've had the moment of having it slip from my hands while shooting at the beach standing in the water. I caught the neckstrap as it went down so it didn't hit anything, and kinda "swished" through the water. I walked back to my spot on the beach, opened my water bottle, and rinsed it off. That was last Summer ... neither lens nor body have showed one iota of any effect from their salt water dunking.

My E-1 has seen a fair bit of duty in pretty wet circumstances with the same lens and has never shown the slightest issue from it.

I wouldn't recommend submerging them on a regular basis, or washing them off with too much water pressure, and of course to keep the camera and lens as water resistant as possible you leave all the port covers in place. But water-resistant works pretty darn well with this gear.

G

 

[Daniel Teoli Jr.]

M43 does good work. What I don't like is no shutter dial.

These were shot with a 12 mp M43

all nsfw

http://rangefindercamera.tumblr.com/image/108384024014

http://rangefindercamera.tumblr.com/image/108383797689

Compare them to the rest of the pix in the project shot with a M240 and Fuji X

http://bikermardigras.tumblr.com/

 

[ymc226]

Thanks Godfrey,

I pre-ordered the Oly EM-5 Mk II along with the 12-40 PRO lens directly from the Olympus Store and got $200 off the lens. Can't wait to try it out without worry of the elements hammering it.

 

[willie_901]

Here's a link with empirical dynamic range comparisons for many different cameras.

http://home.comcast.net/~nikond70/Charts/PDR.htm

 

[Godfrey]

Here's a link with empirical dynamic range comparisons for many different cameras.

http://home.comcast.net/~nikond70/Charts/PDR.htm

Don't see the E-M1, M9, E-M5, or MM there.
Test methodology?

G

 

[dfatty]

neat link, willie, never seen that site before.

E-M5 is there. as is M240. last update was 3-3-12.

i think the links on this page go into the methodology.
http://home.comcast.net/~nikond70/G...or_Analysis_Primer/Sensor_Analysis_Primer.htm

 

[Bingley]

Thanks Godfrey. Just what I wanted to hear. The lossy compression of the Sony A7MkII was mentioned by Ming Thein and is reason why that is not an option.

I'm also missing a lot of shot opportunities (my kids on beaches) regarding the MF of the Leica so the Olympus is looking better and better. Can you just rinse off under the faucet after the beach and not worry about water damage.

FWIW I have the Olympus E-P5, and its autofocus (same as on the E-M1) is blazingly fast with certain m4/3 lenses (I'm thinking of the Zuiko 25/1.8 and the Panasonic 14/2.5). And I second Godfrey's comments about high iso performance from the latest sensor in the top end Olympus m4/3 cameras.

 

[Godfrey]

neat link, willie, never seen that site before.

E-M5 is there. as is M240. last update was 3-3-12.

i think the links on this page go into the methodology.
http://home.comcast.net/~nikond70/G...or_Analysis_Primer/Sensor_Analysis_Primer.htm

Ah, I see the "OM-D E-M5" there now. Ordering the list by some parameter in the data makes it hard to find things! Thx

So, assuming the same measurement methodology, the E-M5 and M typ 250 sensors have fairly similar curves.

G

 

[dfatty]

yeah, the listing order isn't very intuitive, lol.

as expected the sony sensors do really well. i knew canon is a bit behind on sensor tech, but i'm a bit surprised that the e-m5 is right there with even the 5d3, at least in DR according to this methodology.

 

[YYV_146]

Thanks Godfrey. Just what I wanted to hear. The lossy compression of the Sony A7MkII was mentioned by Ming Thein and is reason why that is not an option.

I'm also missing a lot of shot opportunities (my kids on beaches) regarding the MF of the Leica so the Olympus is looking better and better. Can you just rinse off under the faucet after the beach and not worry about water damage.

In all honesty, I shoot both the A7 and A7S in semi-professional capacity. I also process files to punishing degrees. This is why I left Canon in the first place.

I can only find issues with compression after intense processing aimed specifically at uncovering the artifacts (shooting high contrast objects severely underexposed, then brightening by 2-3 stops while applying a sharpening mask). In real-life photography, including studio, landscapes and street work, I have never even seen the artifacts. Maybe Mr. Ming Thein's workflow isn't as good as mine, or maybe his work requires extreme post processing. But I doubt that many people will notice in reality.

The level of processing needed to break the A7 files at base iso will also break the loseless RAW files of the 5dmk3 or any other Canon FF body. The studio in Beijing I work for shoots mostly Canon and I cringe whenever I have to use one of their cameras. The difference in dynamic range between the Sony sensors (D600, A7) and Canon sensors is huge and, yes, in my experience many M43 bodies will readily outperform the 5dmk3 at low iso levels.

 

[Godfrey]

...
The level of processing needed to break the A7 files at base iso will also break the loseless RAW files of the 5dmk3 or any other Canon FF body. ...

Indeed.

I was referring to the high ISO DR limits. The Sony A7 would do even better if the raw files were encoded lossless. I shoot at the high limits mostly when doing pinhole and zone plate photography, where everything is shape and texture, not detail. 128000 is fairly good, 256000 gets a bit ugly without a bit of smoothing and cleanup.

G

 

[YYV_146]

Indeed.

I was referring to the high ISO DR limits. The Sony A7 would do even better if the raw files were encoded lossless. I shoot at the high limits mostly when doing pinhole and zone plate photography, where everything is shape and texture, not detail. 128000 is fairly good, 256000 gets a bit ugly without a bit of smoothing and cleanup.

G

I see. That is a fair point. I've also found the high iso performance on the A7 lacking. I only use the A7 if I can get away with 3200.

Beyond that I take the A7S, which IMO is usable up to 25600 and will work at 102400 with some processing.

 

[Godfrey]

LOL! Please pardon the extra zeros in my ISO numbers... 🙂

Yes, 3200 and occasionally 6400 for crisp rendering with the A7. That's still amazing... !

G

 

[willie_901]

First, the methodology is described in detail in the links at the bottom of the page. I guess Godfrey didn't take the time to see those either.

Second, Bill Claff computes Photographic Dynamic Range (his term) not engineering DR. The difference is a constant... 2EV (if I haven't mis-remembered). This means Claff's DR estimate will be lower than empirical estimates by other sources as everyone else uses the classic engineering definition of DR.

Third, some the data is generated by volunteers (usually multiple volunteers) using a method specified by Claff. This, and manufacturing variation accounts for small differences in plots from cameras with similar sensors and data streams.

Finally, DR is primarily dependent on the analog signal-to-noise ratio of the sensor at base ISO. Of course DR is also affected by other analog characteristics of the data stream including the read noise of the ADC. Differences in the DR plots for different camera data-stream designs are useful to consider exposure strategies and ISO choices for the camera(s) you own.

 

[Godfrey]

First, the methodology is described in detail in the links at the bottom of the page. I guess Godfrey didn't take the time to see those either.

Second, Bill Claff computes Photographic Dynamic Range (his term) not engineering DR. The difference is a constant... 2EV (if I haven't mis-remembered). This means Claff's DR estimate will be lower than empirical estimates by other sources as everyone else uses the classic engineering definition of DR.

Third, some the data is generated by volunteers (usually multiple volunteers) using a method specified by Claff. This, and manufacturing variation accounts for small differences in plots from cameras with similar sensors and data streams.

Finally, DR is primarily dependent on the analog signal-to-noise ratio of the sensor at base ISO. Of course DR is also affected by other analog characteristics of the data stream including the read noise of the ADC. Differences in the DR plots for different camera data-stream designs are useful to consider exposure strategies and ISO choices for the camera(s) you own.

No need to be snarky. I first looked at it using my iPad browser and the page didn't image fully. Poor page design, likely. Looking at it later with my desktop machine made a lot more apparent.

I do my own DR testing, treating all cameras with the same procedure, and derive useful values of DR in stops at the different ISO settings from that. It's not a very technical measurement, but it's something I've been using for half a dozen years because it models the behavior of a camera very predictably.

I'm always interested to see others' measuring procedures.

G

 

[willie_901]

People in glass homes...