« Vivian Maier at the Chicago History Museum | Main | OT: Monday Night Melee »

Monday, 24 September 2012

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

"An impressive step forward—much kudos to Olympus, especially the product designers and engineers who've kept plugging away despite the nonsense at the corporate level."

Keep in mind, the development cycle for this camera started long before the Olympus board became a public embarrassment, and the camera was in production before said board even imploded. It's the future models that will show what the product designers and engineers have been able to do while their stock value evaporated...

I shot my OM-D E-M5 pretty much in the dark at a Jazz festival Saturday night. I was able to get usable images from 800-3200 ISO. Granted the images were intended to be converted to B/W from the start. A couple of closeups were lit just with the reflected light from a cell phone. Pretty impressive by my standards. Not that I am a photographer is the same sense as some of the real photographers around here.

It was also my first outing using the OEM grip. It made handling much better. Looking forward to using the RRS grip and L plate when it get here.

Well, I think many thanks will go also to those corporates, who made the right decission for moving to Sony-made sensor. With previous Panasonic-sensors OMD would never have been such a succes.

Perhaps kudos also to Sony who supplied the sensor?

It's the first significant change since DXO started measuring Oly 4/3 cameras way back with the E510. Now if only they come out with the "made for Mike's hands" version....

From the DXO's measurements it seems that OM-D does have ISO 100. Apparently, the ISO 100 setting is called "ISO 200".

With E-M5 image quality stopped being relevant for me.

I reluctantly went to digital with Canon 30D, as I was satisfied with the IQ compared to 35mm, and the cost of developing film and ease of digital was attractive. I stayed with APS-C sensors for years. Then MFT came out, and I predicted it would catch up with APS-C sensors pretty quickly. My reasoning being that if 30D was good enough for me back then, any future camera would be good enough if it had the same IQ.

I sold my heavy and bulky APS-C kit, and now travel and ride my motorcycle with a full MFT kit: EM-5 and E-PL1 plus four primes and two zooms. The entire kit fits in a bag smaller than one APS-C body with a prime and L zoom, and weighs only slightly more. Now that EM-5 outperforms 30D in IQ, there's absolutely nothing that I miss in the bigger systems. For studio photography I still use 6x6 and 4x5 film cameras.

There are still some niches left to cover for MFT: fast telephoto primes and zooms are missing for birders, weather-sealed lenses for PJs, and ultra-wide primes and tilt-shift lenses for architectural and landscape.

DxO said a few months back that they're doublechecking some surprising results, which was the cause for the delayed analysis. I wonder what it was?

Didn't you read Ctein's post on DR?

Perhaps kudos also to Sony who supplied the sensor?

Yes!

I held this beauty at tempe camera. Liked it.
Sadly I am a film shooter.

Phew, that's a comfort to know. I'd hate to have discovered that all my enjoyment of this camera and it's apparently wonderful performance were entirely figments of my imagination.
Roy

yes, yes, whatever, but do they think the BLACK one is better or the SILVER ??? I really need some help here :-)

Dear Oren,

A small technical quibble-- an ISO 100 setting wouldn't kick the exposure range up beyond 13 stops; the system is already saturating at ISO 200. Note that there's only about a 0.3 stop improvement in range going from ISO 400 down to 200.

Dropping to ISO 100 might get you a smidge more, from improved S/N ratio in the shadows, but it wouldn't be much.

pax / Ctein

If you look closer, you can notice an interesting ISO strategy. The ISO 200 is really ISO ... But, this surely must have been taken into account by the DxO when comaring cameras. Which (as I understand) means that all images are pushed 1EV by the camera, but what happens with RAW ?

it looks like all major leaps on image quality these days (D800, D7000, K5, OM-D...) are when a brand adopts Sony's sensors... it makes you think!

I think a little more than kudos are due to Sony on this. It might be pointed out that this sensor is already in Sony Nex cameras where, according to the DxO info, it generally is performing marginally better.It might also be pointed out that Nikon is now using Sony sensors in their top dslrs.

My Canon 7D recently met a sad end, but happily the insurance paid out the replacement value. So my problem was what to buy? Although I've been a long time Canon user (and still use a 5Diii), I wasn't convinced that the 7D was the ideal backup camera.

After much thought, review reading and in store handling, I decided to opt for the OM-D.

First impressions - EVF does not replace an optical viewfinder, but I can certainly live with it. The tilting LCD is a big plus for me, so I think that's a fair trade off for the EVF.

AF - single shot is impressive. As far as I can see I'm not giving anything away compared to the 7D. Continuous focus, not so good - it works, but not up to the 7D's performance. I knew that in advance, but still a little disappointed (one always hopes for better than expected!). For now I guess this is the mirrorless camera's Achilles heel. Even if fast super-telephotos were available, I think it will be a while before these can compete head on with DLSRs for, say, sports or wildlife photography.

On the other hand, for those who care, having any form of continuous focus for video is of course a big plus.

Image quality is where this camera really scores for me. Comparing raw output from the OM-D at 200 to the 7D at 100, the OM-D is a little cleaner, but the big difference is how much detail you can pull from the shadows. I was stunned, as I found this a real weakness with the 7D. I now see why the Nikon people have been raving about the Sony sensors.

I did notice slightly less detail in the OM-D images, but perhaps this is more to do with the fact my shooting so far has been done with mid-range m43 zooms and compared to top end L glass on the Canon.

I was also surprised how well the OM-D IS works - certainly comparable to Canon's in lens IS.

Ultimately though its the sheer 'usability' of this camera that's won me over. Finally, a genuine take anywhere camera which doesn't leave me feeling I'm compromising on quality. I think OM-D experience must be similar to that of the early adopters of 35mm film.

Colin


Dear Marcin,

No, that's due to a fundamental misunderstanding of ISO by their readers. Industry-standard camera ISO and sensor ISO are determined in entirely different ways and mean different things. I've posted a lengthy comment to the thread there explaining this. Likely I will turn that into a column here, delaying my “RAW isn't raw” column, but it's an important lead-in to it.


pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

Still don't like the pointy top.

Oren said: "I'm always concerned about the DR score, and in that category the only reason it doesn't match the best APS-C cameras is that it doesn't have the ISO 100 setting that kicks the measured exposure range up beyond 13 stops."

That's not the reason.

As DxOmark says: "The pixel pitch of the OM-D E-M5 is the tightest [smallest] among the three [E-M5 and two Sony APS-C] sensors."

The difference in pixel size and a corresponding reduction in saturation level for a pixel defines the top of the sensor's dynamic range (given the same read noise defining the bottom). It's a pixel level property unlike, say, midtone (photon count dominated) noise.

If you have the same (sized) pixels from the same sensor generation and manufacturer you would have the same DR regardless of sensor size. A smaller pixel in the same generation will have lower saturation level.

That means a FourThirds user would want (roughly) 3/4 the number of pixels on the same generation APS-C sensor.

From the results this sensor seems to be the similar "generation" as the 16MPx NEX 5N rather than 24Mpx NEX 7 which seems to be "next generation" along. Those NEX 7 pixels in an (hypothetical) 18Mpx E-M6 should give the same DR as the NEX 7.

So this is the best micro 4/3 camera ever, with a sensor comparable to Canon's 60D that trounces the Panasonics and reduces the Nikon 1-series to 'expensive toy' status. I haven't had the opportunity to experience one, but I have no doubt it's a great ittle camera.
Then why don't I feel attached to it? Why don't I feel the urge to buy one, like I felt about the E-P1? Could it be that I feel the same Mike feels about the Leica M9? Its design just doesn't grip me; it's a pastiche, or a caricature, of the glorious OM-series (the one-digit models, that is). And I can't help thinking 'witch hat' everytime I look at its viewfinder hump. Labeling this camera 'OM' is almost insulting. Not that it is ugly - it isn't; if I called it so, what adjectives would I use to describe the Pentax K-01 or the Hasselblad Lunar? -, but it is Olympus imitating itself. Well, who cares - as long as it helps taking great pictures. At the end of the day that's what really counts, isn't it?
Now let's have an E-P4 with this sensor. THAT will be a riot!

I have an OMD and, for now, a Canon 7d for long lenses and fast moving aircraft. The Canon excels at continuous focus, the Olympus at everything else.

I'm hoping Canon buys a Sony sensor for the 7d Mark II.

The top isn't pointy enough.

@Manuel: +1

Still only good to max 800 iso with the small 4/3 sensor.

I agree, that the DxO mark results has to be taken with a grain of salt, but the real photos shows that the Olympus managed to design a serious step-forward camera, E-M5, in comparison with their MFT PEN series.

"On their measures it's a leap beyond any prior Micro 4/3 camera, including the 16MP Panasonic G3." Why would you compare the E-M5 with the G3, which is roughly 1/3 of the E-M5 price on the Amazon?
Even the more recently released Panasonic G5 is about half of the E-M5 price, while its sensor, according to the DxO marks has 1EV better Dynamic range than the G3.
Anyhow, the GH1 sensor is still DxO-marked as the best among those which are made by the Panasonic. The GH1 is also about half of the E-M5 price. Now let's wait for the GH3 sensor result ;-)

Manuel said:
Its design just doesn't grip me; it's a pastiche, or a caricature, of the glorious OM-series (the one-digit models, that is)

I must confess I'm not over keen on the looks - I'm not a fan of retro anything, it always seems a bit contrived to me.

But you do need to hold an OM-D to appreciate it. In the hand the design just works and feels like it was carved from a single lump of metal.

Colin

Well given that it's a recent Sony sensor, it seems to be about where I would expect given the same architecture shrunk slightly for MFT.

"I'd hate to have discovered that all my enjoyment of this camera and it's apparently wonderful performance were entirely figments of my imagination."

:)

I am certain that Ctein will explain precisely the differences between these three iso sensitivity curves and what conclusions can be drawn:

http://www.dxomark.com/index.php/Cameras/Camera-Sensor-Database/Pentax/K5 # tabs-2

http://www.dxomark.com/index.php/Cameras/Camera-Sensor-Database/Nikon/D800

http://www.dxomark.com/index.php/Cameras/Camera-Sensor-Database/Olympus/OM-DE-M5

Thank you in advance.

Today's news "Sony Corp will likely approve a plan this week to invest 50 billion yen ($642 million) in cash-strapped Olympus Corp, becoming its biggest shareholder with around a 10 percent stake, three sources familiar with the deal told Reuters."

I think there's a very good chance Olympus will stick with Sony sensors :-)

I still don´t know why this camera is not named as the Olympus "OM-G!".

"Still only good to max 800 iso with the small 4/3 sensor."

Terry,
No, the OM-D is fine to at least 1600. More about this tomorrow.

Mike

Dear Jean-Louis,

I did write the column explaining the difference between sensor ISO and camera ISO, so you will get your wish in the morning (or whenever Mike posts it tomorrow).

~~~~~~

Dear Terence,

This opinion is based on what photographs you've made with the camera?

I would agree that the previous generation, typified by the Olympus Pen series, was not good for serious work (my definition of serious being that I can make a portfolio-quality 17" x 22" print) above ISO 800.

Having now worked a fair amount with this camera under lighting conditions that would draw out any problems with noise or image uniformity, I am utterly confident of the image quality at ISO 1600; it is distinctly better than the previous generation was at ISO 800. The only evidence that one is photographing at that ISO (other than an inevitably-reduced exposure range) is a very fine and uniform “grain” pattern in the prints. There is no evidence of patterning whatsoever. What it most resembles, in magnitude and impact, is a print made from a fine-grained 6 x 7 cm negative.

The noise is so uniform and fine that a small amount of noise reduction, not enough to visibly degrade fine textures or subtle tonality, is sufficient to almost entirely eliminate it.

At ISO 2500, pattern noise does appear, in the form of fine horizontal lines. It's almost like every other row in the sensor is of slightly higher brightness. It is only visible in uniform areas in the deepest shadows, but it is present. It is entirely luminance noise, not chroma noise. It's far less annoying than the large-scale “plaid” chroma noise one saw at ISO 1600 with the Olympus Pen.

I must emphasize that I am NOT saying that ISOs above 1600 are bad, just that ISO 1600 is the upper limit for my serious portfolio work, unless I want to do a certain amount of cleanup in Photoshop.


pax \ Ctein
[ Please excuse any word-salad. MacSpeech in training! ]
======================================
-- Ctein's Online Gallery http://ctein.com 
-- Digital Restorations http://photo-repair.com 
======================================

"But you do need to hold an OM-D to appreciate it. In the hand the design just works and feels like it was carved from a single lump of metal."

I just wish they had used a bit more metal...

I love my OM D. After the initial frustration, it is a joy to use and the IQ is stellar. I just hope that Olympus sticks with it and brings out a pro level body soon rather than more art filters.

Mr. Ctein.
This opinion (max 800 iso) was based on the low light score of
826 iso from DXOMark.
Regards,
Terence

I agree with many here. I love my EM5 and today paired it with the new 75 1.8 which is just amazing. So glad I moved from Canon 5D. It was so worth it.

Dear Terence,

Ummm, there's no "Mr" in front of Ctein. No honorifics, here. [s]

That aside, you're seriously misunderstanding what DxOMark means by ISO-- see my column for this week.

Even if it meant what you thought it did (and it doesn't) it still wouldn't mean that a cameras good only up to that ISO.

pax / Ctein

On the "the low light score from DXOMark" from Terence Morrissey.

Lots of people misunderstand what this (and the other DxOmark "Use case scores") mean.

DXOMark explain it: the low-light ISO is the highest ISO setting for a camera that allows it to achieve an SNR of 30dB while keeping a good dynamic range of 9 EVs and a color depth of 18bits.

The camera has to meet all three of those criteria. And they're rather strict criteria to look very good to excellent in print.

If you have different criteria then that number will differ and Ctein does!

This is also the reason smaller sensor cameras (the Canon S95 and the like) have rather low low-light ISO values. Not because they don't deliver the SNR or the color depth but because they have smaller pixels that run out of dynamic range rapidly at higher ISOs. If you don't care as much about that DR then your low-light ISO score will vary.

The comments to this entry are closed.

Portals




Stats


Blog powered by Typepad
Member since 06/2007