The Phoblographer’s Autofocus Test: Olympus EP3 vs Canon 7D
There are many interesting E-P3 tests you absolutely have to see! On top you can see an [shoplink 23568]E-P3[/shoplink] versus Canon 7D autofocus test. Easy to see who is the winner! And here is the reviews list:
1) Photographyblog (Click here) made a very detailed Olympus E-P3 JPEG vs RAW comparison. And those are the most interesting notes:
– The camera-original JPEG shows considerably higher contrast, while the raw conversion has arguably better tonality.
– There’s a bit of luminance noise in the JPEG, while the raw conversion has some tightly-grained chroma noise.
– AT High ISO the JPEG has a monochromatic noise pattern, while the raw conversion has more chroma noise
– You can definitely achieve much crisper-looking results by shooting raw, but the actual amount of detail may not be that much higher, and you may also run into aliasing issues.
2) Micro Four Thirds high ISO noise compared – G1, E-P2, E-P3 at FourThirdsUser (Click here). Who wins? Easy…the E-P3 🙂
3) Focus Numerique (Click here) posted their well known ISO tests of the E-P3. At highest sensitivities (ISO 3200 and 6400) images of the E-P3 are better than the [shoplink 18844]E-PL2[/shoplink] but it gets beaten by the [shoplink 22077]G3[/shoplink]. Focus Numerique posted also some Image samples you can download as RAW!
And finally also Amazon.de is accepting E-P3 preorders. Here is the updated Hot preorder list:
Olympus E-P3 preorders at Amazon US, Adorama, BHphoto, Amazon UK and Amazon DE.
Olympus 12mm f/2.0 preorders at Amazon US, Adorama, B&H and [shoplink 23576 ebay]eBay[/shoplink].
Olympus 45mm f/1.8 preorders at Amazon US, Adorama, B&H.
76 Responses
Very, very cool! Good to know the raws have room to breathe. Hopefully this should go some way to stopping people complaining so much. Haha…yeah right…
Still no af-tracking test which is where the difference really matters, but the E-P3 is certainly really fast !
As for noise, well, I still prefer a 16Mp APS-C to 12Mp m43 for noise, but it is clearly better than previous PENs.
Good job Oly !
Wow, that E-P3 is expensive on amazon.de…
€1000 with 14-42!!!!!!!!!!!!!!!!!
Hope this is not the official price….
On amazon.co.uk you can get the kit with the two lenses for about the same price…
all is more euro and uk moneys than the NEX!
Comparing the images posted by focus-numerique at ISO 6400 I don’t know how they came to the conclusion that the E-P3 does better than the G3. It’s not even close.
However, the noise reduction is certainly better than in the E-PL2 in all the samples I’ve seen.
EDIT: now I read the page in French and obviously it says that the G3 is better. Wrong google translation.
Wait…wait….Photoblog is comparing raw files and jpegs from the same camera…and the raw file is better?
WOW….what incredible information. A jpeg is like an mp3 (tinny…no base sounds)..it is a compressed file with less information. I think it is embarrassing that someone is even doing this on a photo review website? LOL!!!!!!!!!!!!!!!!!!!!!!
How about the E-P3 raw file compared to the GH2 raw file?….and how about using the same lens…like the P/L 25mm f/1.4 or the Olympus 12mm f/2.0…on each camera so that the comparison is absolutely fair and using a high-quality prime. Then we would have some useful info!
mp3 is perfictly able to hold all the bass in the world…. its the top end that needs to be compressed…. 🙂
Jim, does that mean that jpegs can hold all of the shadow that we need…it’s the highlights that need to be compressed????? LOL!!!! YOU know what I mean!!!!!!! smartass….LOL! 🙂
LOL to you. It is not even possible to compare a RAW file with much anything else, because a RAW file is not an image. It needs to be converted into some kind of image first. And every JPEG image is in fact a converted RAW image — just converted by the camera itself.
Not so.
Faster is better, ok. But it’s worthless if I can’t track my two years old son with his bobby car. My 5DII is fast enough (= I never had an issue, not one, with focus on not moving things), but tracking is really something else. If the new Olympus also win this discipline I’m in.
Yep, plus change your summary, the G3 is obviously better and the google translation is wrong..
Exactly,
i dont know french but its clear google translation is at fault,
anyway the last sentence reveals the truth anyway
…”Even at ISO 6400, the compact Panasonic keeps a step ahead.”…
damn.. i advise everyone who lives in germany to preorder your E-P3 at amazon.uk. they offer it for 799 pounds, which is 895 €. shipping is 3.99 £. why pay 100€ more at amazon.de?
by the way, really interesting tests and comparisons
Some crazy sh*t !O
So much money for something with a two years old sensor :>
Pablum…did someone say Pablum?
take a chill pill bro! we, most user of pen camera, dont work for versace/LV/guess/etc taking pics for bill board ads.
“So much money for something with a two years old sensor”
Why bother trying to spread rubbish? It’s a new sensor.
http://fourthirds-user.com/sample_images/327/e-p2_iso-3200.JPG
http://fourthirds-user.com/sample_images/327/e-p3_iso-3200.JPG
Do they look the same to you?
RAW, or it didn’t happen
In your opinion, how often does a company have to change its sensor Pablo? My E-P1 took great photos, my E-PL2 takes great photos. It looks like the E-P3 is even better – what more do you want? You are sounding like a total measurebator.
I bet you would have lost your mind if you had been around in the film days – when a product like Fuji Velvia went 20 YEARS before being replaced by a newer film ( :
So dumb! it has same sensor and same plastic and same chips and i hate olympus’s camera the PEN because it is not a NEX which is just amazing. i masturbate thinking about the NEXXXXXXXXXXX!
LOL
because you’re gonna pay taxes along with the shipping
nope – no taxes going from UK to mainland europe.
they dont like to admit it themselves, but UK is part of the EU trade zone – not monetary union.
Amazon DE is just trying to take advantage.
Seems the Panasonic’s higher MP sensor is still a good bit better. The extra processing power and latest Raw images seem to suggest the sensor on the P3 isn’t really improved much other than maybe bandwidth to do the faster focus if you read the tea leaves. The dynamic range and better high ISO stuff seems a lot related to better JPEG processing more than better sensor output. The color output from the sensor may be improved a bit also, but I think we need more official Raw filters to really know just what all was improved on the sensor for certain. Seems quite clear though it isn’t an extreme low noise high ISO monster that a modern 12mp probably COULD be in 4/3rds size if done from the ground up as such by someone with vast experience designing them like Nikon, Panasonic, or Canon.
Still and awesome camera for Oly though! I am just one who wants a true low light not so high resolution monster of a sensor that we should have by now.
On the Focus Numerique comparison, the “A6” written on the green card in the sample image is still legible at ISO 6400 with the G3, but not with the E-P3.
The color is still mostly accurate too. Meanwhile the green card turned grey in the Olympus sample.
I’ll get mine in less than one week. MSRP here is HKD7,390 and Olympus HK confirmed that the camera will be delivered on July 13.
I think you misunderstood the traduction (I’m french) on Focus Numérique: they said that at 3200 and higher iso the G3 is better.
> Photographyblog (Click here) made a very detailed Olympus E-P3 JPEG vs RAW comparison.
Wow. Wow. Just wow.
No no – not E-P3.
The RawTherapee!
It made astounding progress in last two years I have touched it.
I am not too sure why all the whinning about ISO6400…even with my D7000, I dont even want to go pass ISO1600!!!
i take pictures of my butt with nex and not pen because it goes higher iso much better.
I think the true reason, whoever you are, is because the NEX has sweep panorama!
It is obvious that we now have two m43 “enthusiast cameras”, G3 and EP3, that are better than previous generation. That should be no surprise! Probably the IQ diffenences between them are more on the pixel-peeping level than anything else.
Concerning RAW comparison; Different RAW converters produce different results, some of the more obscure RAW converters are pretty lousy. So will rest myjudgement until the big guns have support for these cameras.
There is one difference though: Price. An EP1 with EVF costs twice as much as a G3! You can buy a very good lens for that difference.
Just tested my G3 with Zeiss 50/1,7: Excellent! Color, detail, contrast… Far better than 20/1,7 and all the zooms. (The Zeiss did very well on my EP1 too will do a compare when rain stops)
Previous Olympus E-P cameras had very good JPEG’s when NR was set to low or off if you want to preserve details and don’t mind some fine-grained luminance noise (the Panasonics are not bad too IMO.) Hopefully the E-P3 doesn’t mess with the formula, and only improves on it.
The Sony NEX applies too much NR in JPEG’s even when set to ‘weak’, forcing me to shoot RAW. What a shame, since the sensor is really great. Of course this is based on my subjective image preference, leaving a little luminance noise gives the impression of more detail. I would rather have a ‘grainy’ image than a ‘smudgy’ one.
For day to day use I prefer shooting JPEG’s if quality is good, to save space on my notebook HD.
NEX versus mFT: Here is my “milkmaid computation” comparing these two sensor sizes:
sqrt(23.4 * 15.6) / sqrt(17.3 * 13) = 1.27
The ratio of the square roots of that areas gives us the linear scale factor 1.27. Hence, the weight factor is 1.27^3 = 2.07.
Therefore, a lens just scaled up from mFT to NEX is expected to weigh more than twice as much. I would suggest to introduce a “glass factor” besides the well known crop factor. Is it worth to carry around 4 kg of lenses instead of 2 kg in order to achieve such minor improvements in image quality? For my needs, FT/mFT is fortunately good enough. Full frame compared to FT has a “glass factor” 7.54, i.e. 100 g of glass in a FT lens correspond to 754 g of glass in a comparably designed FF lens (or 1.64 kg in a Hasselblad lens). These considerations on glass volumes show us that there are three dimensions in engineering — only the photographic image is two dimensional 😉
Interesting post, although it does not correspond to reality,
for example take 2 similiar spec’d lens: APS-C 30mm f2 and m43 20mm f1.7, they are both equilent to 40-45mm ff and have similar apertures,
the m43 20mm weighs 100g and according to your methodology the samsung 30mm is supposed to weigh 100*2.07=207g, but it doesn’t, the Samsung weights 85g, 15% lower then the m43 counterpart, thus making the larger sensor look more attractive in weight/sensor size ratio
OK, lenses don’t consist of glass only, there is also metal and plastics. Usually, retrofocus type lenses have greater volume and less vignetting. Maybe this APS-C lens has been designed differently and optimized for lightweight such way that some lens errors like darkening to the edges have to be corrected by firmware/software. BTW, if we consider the difference between 3:2 at APS-C and 4:3 at mFT and take the radius of each circumcircle as the linear scale, the results are even worse, the “glass factor” is 2.2 (instead of 2.09). My impression of real-world Sony NEX lenses is that they have substantially more volume than their mFT counterparts, especially in the tele range.
Mountain walker, I’ve never actually come across a camera or lens designed by a milkmaid, and there seem to be some assumptions behind your post. There is no reason these numbers should match, you are assuming one format is in effect a scale model of another – absolutely not the case. There are 3 dimensions in engineering, but many other factors as well.
For example, by your reasoning a Canon 5D with lens should weigh almost 8 times as much as a Panasonic G3 with comparable lens. Does it?
G3 + std zoom = 583g ; 5D + old 28-90 zoom (nearest equivalent) = 1100g
So let’s compare some lenses:
Similar m43/APS-C (1.27 or 2.07 ratio?):
Panasonic 14mm pancake (6element/5group design) – 55grams, 55.5 x 20.5mm
Sony 16mm pancake (5 element/5 group design) – 67grams, 62 x 22.5mm
Similar m43/FF :
Olympus 45/1.8 (9 elements in 8 groups) – 116 grams, 56 x 46mm
Canon 85/1.8 (9 elements in 7 groups) – 425 grams, 75 x 72mm
……………Theory Busted!
Ok, I come from the ‘old days’ of 35mm film, haven’t caught on to digital for quite a while, and am reading/working my way back into the (enthusiast) game lately. One lingering question I have is: what is all that obsession with high ISO? With film, my preferred sensitivity was 25, or 64 – crisp as fresh toast. I would never have considered using anything higher than 400. If that wasn’t enough – well, you got to leave your shutter open longer. Or not take the picture at all (there’s a lot of pictures that are better not taken, but this would bring us to philosophical underpinnings…).
Isn’t this whole ‘high ISO’ thing only a distractor from lens quality? When I see prime lenses nowadays with f=2.8, or 3.6, my initial reaction is ‘whoa – how could you consider anything above 1.8 or 2.0?’
Curious about what people think about this.
@evenflo: +1
Even with my D300 setup, i used my wide aperture lenses and ISO200 (with couple of speedlights to help in low light indoor situation for beauty shot). I just love the high quality image and contrast with strobist technique.
High ISO performance is one of the properties that will always be tested by reviewers and it is where smaller sensors are likely to be compared unfavourably to larger ones. Even if it’s not important to some users, if performance at high ISO is comparatively poor, it will affect ratings in reviews and is thus likely to impact sales.
The same if true of AF speed: it is one factor that contributes to the overall score in reviews and thus Oly and Panasonic will always be seeking to improve it, because people are always comparing mirrorless to slrs.
Usually when I see landscape photographers with Canon 5d mark II or Nikon D3x cameras, they will have the cameras on a tripod, just as they would have done in the film era. Despite all the technological changes, sensors generally give lower better results at lower ISO, so it remains good practice to get the best lenses and use a tripod.
> how could you consider anything above 1.8 or 2.0?
And how much the glass along weighs?
> what is all that obsession with high ISO?
Can’t tell for everybody, but in my case, I hate shooting with flash. (No, you do not need to explain me why I should love shooting with flash – I have two colleagues who doing it every event.)
I shoot mostly people, candid if possible, and I attempt to preserve always the surrounding atmosphere. (Not only flash destroys the atmosphere, but it also takes space in my bag.) Now without flash, dim light, with f/1.4 lens at ISO 800 I can shoot at speeds 1/50-1/25. (At eq 50mm, without IS it is already literally impossible.) If I could shoot at ISO1600, then I would be able to raise shutter speed…
Though, again, in low light, first challenge is the focus. But if one managed to nail the focus, due to slow shutter speed, any movement result in blur. And at 1/25 there is quite a lot of the blur.
That type of shooting probably is more related to the fact that nowadays I simply have little time to go out while it is still a daylight.
i totally agree, fast lenses are about more options: no flash, if you want… discreet low light candids, more range of control on DoF, if you want… I too hate using flashes for the same reasons. I’ve shot available light theatre with film in the 1970s (oh for high iso and white balance control at least the lighting director built in colour correction into the lighting for me). In short there are so many ways we work and things we want to accomplish so please think a bit before parading your ignorance of the broad scope of photographic practice and people’s interests. Oh yeah and realy fast accurate autofocus would have been awesome to have back then as well.
Isn’t the G3 sensor quite a bit worse than the GH1 sensor?
only at base ISO. DXOMark ratings are heavily weighted towards based ISO which is why the GH1 scores higher.
Anonymous, where did you get that information?
i was speculating for quite some time as to why the is lower yet high-iso images looked better
I know they’re weighted towards ISO, but if DxOMark is to be trusted, the GH1 has a stop more dynamic range. That, in my book, is worth way more than better high ISO performance.
neither is as good as NEX!
LOL, Pablo, we get it, you really dont have to repeat yourself every other second, its becoming a joke 🙂
Haha, I don’t think that’s the real Pablo. It sounds exactly like him, but he never used an avatar before.
Unless I’m mistaken we now have TWO Pablos Frosti. The bitter, boring, original speaks English Pablo, and the stupid, foul mouthed copycat doesn’t speak English Pablo.
Wow. Lucky us.
@Admin:
Have you get any pic for new EVF? I think there has one at:
http://www.youtube.com/watch?v=cpoC2n11Bwk
check what mounted at E-P3 on 00:58.
I suppose the one on the left is what we’re looking at? Doesn’t quite look like the VF-2, you’re right…
but Pablo Numero Dos is funny!
Yeah, looks like the new viewfinder.
Yeah, looks like a new viewfinder.
Thanks Felix! Will post that soon!
Whats mounted on the black model? Some flash or GPS module or something?
1) Of course AF improvement is welcome but in my opinion it’s importance is over exaggerated. I don’t mind if my camera focuses in 0.35s or in 0.15s during day light as I won’t barely notice any difference. But I do mind if it’s unable to focus inside a café during the night and for what I’ve read it still happens sometimes.
2) What about shutter speeds? ISO 3200 on the G1 is actually 4055 while ISO 3200 on the EP-L2 is actually 2100. It’s not the same nominal ISO than real ISO and most people seems to ignore it! Still have no data on the EP-3 but we can assume no changes regarding this.
3) To me the G3 looks better at higher ISO even knowing that Panny’s jpgs algos are worse. EP3 looks really great at base ISO but I don’t see that “huge improvement” over the EP-L2. Seems that the sensor was only tweaked for video but not for stills. A real pity. The real advantage of the Olys in low light is the IBIS but IQ at the same parameters… no way.
from exif info –
g1 ISO3200, 1/60, f7.1
ep2 ISO3200, 1/60, f6.3 <— close!
ep3 ISO3200, 1/40, F7.1
ep3 ISO5000, 1/60, f7.1 <— ISO5000 to match GF1 ISO 3200 in exposure
“Still have no data on the EP-3 but we can assume no changes regarding this.”
Actually, no you can’t. Republicans. Meh!
To be honest, both G3 and E-P3 are pretty crappy beyond ISO1600 (or 3200 if you don’t print big).
For this, you’re best off with D3s or some other similar camera.
Sony 16mp APS does a bit better, but maaaaybe 1 stop over E-P3 at best, but it’s still got slower glass (at least on NEX).
On the other hand, E-P3 looks so much better than G3 or any other mirrorless out there, is smaller has much more compact lenses and 2 recent primes are yummy (as is third – PL25mm micro).
G3 is not significantly better at all compared to E-P3 in terms of high ISO (RAW), but it’s not worse.
For low light however, I’d give advantage to E-P3 because it’s faster focus and IBIS.
But those cameras can’t really match even 3 year old D3/D700 when it comes to low light (those FF models have 1.5-2 stops better high ISO), yet alone “Dark Knight” D3s which has over 3 stops of advantage (ISO 25600 looking like 3200 on m43).
New Pens are great cameras, new lenses are great too and probably “good enough” for 99.5% of photographers (who don’t need fast tele lenses).
lol “Dark Knight”
The E-P1 already matched the Canon 1ds though regarding the noise. I mean nobody complained about the noise from that camera like the majority does now about the mft models …. lets get over with the high iso sickness…
The Sony APS “Maaaaybe” over 1 stop at best? What a bunch of nonsense. Unless you mean at least the Nex model and not the D7000/K-5 models. At low ISO the difference it less but where it counts, at high ISO, the Sony sensor pulls ahead by a significant margin.
Since I am sure you don’t like dxo mark, here’s another data point
http://translate.google.com/translate?js=n&prev=_t&hl=en&ie=UTF-8&layout=2&eotf=1&sl=auto&tl=en&u=http%3A%2F%2Fwww.focus-numerique.com%2Ftest-1261%2Fcompact-olympus-e-p3-bruit-electronique-12.html
Why dont olympus just not make PENS when NEX is so good? i buy a NEX next (i made a funny!) and not a PEN.
Maybe you’ll get anNEXed by a NEX and so can live happily ever after inside it, without pestering this blog anymore. 😉
focus tracking test please, single shot it is pretty quick. But single shot does not really mean much for a sport shooter. If the AF is that fast then when focus tracking, then I will be impress.
can we get a sample of it in low contrast shots, and low light shooting. Contrast system seem not to AF as well in really low light.
I have important proposal to the author of the “test”.
Pal, if you do not know how to measure autofocus speed of the camera, do not try to do “test”. Instead, go back to school.
If you publish TEST you should also publish METHODOLOGY, RESULTS, and declare the winner. And you have no knowledge how to measure autofocusing speed, do not ask stupid questions to the public to get the clicks. It is SOOOOOOOO LOOOOOW.
DPReview has New high iso samples from the E-P3 that look really good. Take a look at the meal, shot at 3200 iso. That image should look totally ok printed at a4 size.
> Focus Numerique (Click here) posted their well known ISO tests
Well, I looked at their 1600 ISO raw files.
Relative to olympus-e-pl2-1600iso-nrstan-big, olympus-e-p3-1600iso-nrstan-big is 0.3 EV hotter while panasonic-g3-1600iso-nrstan-big is 1 EV hotter.
I still think the e-pl1 is the best JPEG Camera EVER