Still haven’t seen enough GH3 image samples? Here we have three new links:
ePhotozine posted a new set of full size (and ISO test) image samples. More GH3 image samples have been posted at ePrice. But the record goes to Photographyblog that posted 85 Jpeg full size samples in one time! And thanks to PB pictures we can make a short E-M5 vs GH3 ISO JPG comparison. The GH3 ISO images are here while the E-M5 ISO images are here.
So what do you think? Who wins?
—
US deals: The 12mm Black and Limited and Expensive(!) Olympus 12mm lens is now in Stock a Amazon (Click here) and Bhphoto (Click here). The Olympus E-PL5 is now in Stock at Digitalrev (worlwide shipment) and should be in Stock next week at BHphoto too. There is also a two lens kit deal on the E-PM1 at Adorama (Click here).
GH3 preorders at Amazon (Click here), Adorama (Click here), Bhphoto (Click here), Panasonic US and in Europe at Wexphotographic UK, Photogalerie FR and Technikdirekt.
60 Responses
links are broken D:
Uhm, the image links seem to be forbidden, throwing a 403 error 😛 I guess they don’t want hotlinking.
ISO links are error 403
Your links are “forbidden.” But it hardly matters as I’m sure there is no winner, only minor differences and interpretations. Video a different matter.
I was about to order the e-pl5 from digitalrev until I noticed that they are asking for $50 more than anybody else. No thanks.
What does your post have to do with this thread?
Maybe it has to do with the news that the e-pl5 is in stock?
Cork is the Blog Patrol…oooooops.
hmmm….would an EPL-5 and a viewfinder be nice back up for an EM-5…. Think I will wait a little while for the price to drop…
Depends much on where you live. I’ve pre-ordered the E-Pl5 from digitalrev. Here in Germany it would be more expensive than from digitalrev (not much, but still a bit) and local shops still don’t have a date.
Hmm. Links are fine with me! No Error 403. 🙂
Yes, links are fine.
E-M5 and GH3 almost the same IQ up to 3200. above 6400, the Sony sensor has proven its might.
Is it the moire we see just below the “Geek” of the hi-res version of Laura the second?
doesn’t matter the links are forbidden, unless they ‘recalibrate’ test, e.g. real RAW ISO of EM-5 is 100 instead of 200, 200 instead of 400 or 1600 instead of 3200 – the test is pointless.
I do need certain shutter speed at certain aperture – so I set ISO to 3200 on OM-D and voila – getting lower shutter speed than necessary. Why? Because in reality, it’s only 1600!
Or they should set shutter speed and aperture to fixed values, switch off IBIS and try to see the exposure. And ISO.
Cheating is not good for you my friend, it’s excellent for marketing!
Eh, these are JPEGs not RAWs so the ISO values are correct. For any JPEG shooter, this is an accurate comparison.
Nonsense, its the image brightness that counts and in that case the E-M5 pushes almost 1 stop after underexposing by one stop.
Some ppl however do not want to get it because they rather like carping about the cheating with the ISO.
Well, whether they like it and care about it – i do not ..
according with DP Review, the E-M5 is about one stop worst, with same lenses and same aperture value, compared, for example with GH2 (I don’t remember now about G3 and G5). Indeed E-M5 keeps more time open the shutter. This means more light with same ISO this means that the ISO are biased.
Either you look at an alternative dpreview version, which says “1 stop” or you are trying to convince me that I cannot read:
“By our tests, the E-M5’s measured sensitivities are about 1/3 stop lower than indicated across the ISO range (i.e. images are fractionally darker than expected for any given set of exposure values). A discrepancy this small has little practical impact in real world use but means the camera will be slightly flattered in our tests.
“
I said comparison between other recent m43. Your statement, is referred maybe with oter things, if you can send me the page link I will looka at it.
What I mean is that you look at DxO mark and DPR EXIF data you can readely notice that the descrepancy in ISO performance has been “equilibrated” by taking a lower shutter speed. Thats my opinion, as well as both the things makes the bias between these same “sensor-sized” more than 1/3 stop(also, maybe less then 1 stop… let’s say 2/3?)
well – I do care about it the most!
what is the crucial feature of a camera is metering! Look on Nikon 1!
On the shutter half-pressed I want to see in the VF the EXACT same scene I see in reality – assuming I am useing the proper exp. mode (say I don’t have too much dynamic scene so it can be all captured with evaluative)
Then from time to time I like to expose to the right by +1/3 EV (2/3) – a habit from old Canon days, and ‘return back to normal’ in LR and retain some shadows detail (much more info there). The highlights are saved 95% of the time because of RAW 12 bit colors range. THEN apply any post-processing.
But what OM-D is doing – is ‘thinking for you’. It applies post processing in background. Means it tries to do everything for its JPEGs looks good and not to overblown highlight – which I COULDNT CARE LESS! I love the highlights in my JPEGs to be cut – means nothing in 8bit file!
But screwing the reality (and altering RAW) just for JPEG doesn’t feel right!
You know you can set an exposure shift to match your personal exposure index?
No, not talking about exposure compensation. Go into the menu, goto the custom menu, then to utility and then to exposure shift.
A personal exposure index is something you need to develop for any new medium you deal with, regardless of it being film or a digital sensor. The reason is that ISO values as used by the camera are based on subjective judgment and not on some measurement. This has always been the case even in the days of film and even when it was called ASA (or DIN if you prefer that scale).
yes but in this case camera requires either longer shutter speeds or higher ISO – and thus comparing it with standard values against any other camera is no fair (e.g. they cheat)
and proper exposure is indeed defined – look (again) at Nikon 1 and its almost flawless metering. Any pro camera needs to be properly calibrated – to have proper ‘feel’ of the world according to exposure mode.
As I said, I want reality in my liveview – imagine if camera constantly underexposes (this case – RAW) and you would only have optical (not electronic) viewfinder. If you wouldn’t know that you are screwed.
Another thing is, that due to digital nature of RAW and how photons are captured into information stored by bits (and how they are distributed from shadows (low bits) to highlights (top bits)) it is better for camera to overexpose (there is surplus of information in highlights), than to underexpose (there is not enough bits to store detail).
It latter case – LOTS of information is lost as the ‘world’ has been stored to lower bits – offering not many combinations in lower tones and leaving a LOT OF for just the highlights (and zillion unused).
You can make the underexposed image brighter, but there wasn’t enough bits to store it in the first place (at least the more important area) so you are brightening ‘losy’ information (something like mp3 -> wav conversion stupidity). You can find more about it in RAW definition by ADOBE.
No energy to answer to ISO stupidity below as there’s no reason to argue with someone offensive – of course you can measure it.
You change exposure shift and the camera will behave more like you want it to, that is the entire point of such a feature.
And no, perfect exposure does not exist, it is always a matter of intentions and a call of judgment.
Ansel Adams’ The Negative is mandatory reading for people who believe perfect exposure exists.
Thank you for the tip – I haven’t read it. What I meant with perfect exposure – is the world I am actually perceiving.
I do get that human eye is more than 14EV of a current best sensors, but that’s what I said – if I don’t have too much dynamic scene to look at.
Yes, I’ll adjust exposure to my liking, but the STARTING POINT should be well balanced. Then I can play with shadow silhouettes or highlights ‘cutting the boobs’ of my girlfriend 😉 but the problem described here was – that the starting point is screwed and Olympus is cheating.
So to close it – you can have ISO comparisons, but is pointless unless you take all into account.
hi Charles,
(sorry, I was also the ‘anonymous’ from the previous post)
But what makes for a ‘well balanced’ exposure? This seems to be about preserving both enough shadow detail and enough highlights, but where exactly should the balance be? I think you’ll get as many slightly differing opinions on that as the number of photographers you ask.
Anyway, I agree with what you say about exposure compensation and the ‘starting point’, but the exposure shift option I was referring to changes this starting point, exposure compensation is independent of this.
With regards to ISO and ‘cheating’, read
http://theonlinephotographer.typepad.com/the_online_photographer/2012/09/why-iso-isnt-iso.html
It’s your bone head which needs percussive recalibration.
Only real world usefull definition of ISO is certain standardized brightness of picture taken in standard illumination and exposure parameters.
Photosite’s RAW signal doesn’t even have any real ISO and every different sensor surely give different signal for same exposure parameters.
For getting any usefull image with standardized brightness for those exposure parameters that signal needs to be “calibrated”. That applying of tone curve could happen before ADC or after that in numeric RAW to RGB image conversion process but it has to happen in some point.
Just like film needed correct processing to produce correct brightness picture for that sensitivity.
And actually when looking PB’s numbers it’s actually GH3 which uses longer shutter speeds for same sensitivity!
But guess that means Panasonic is even worser…
Looks like in your film era Canikonist’s 4/3 hate you forgot to check what you were complaining.
And different cameras/makers have different auto exposure emphasis on shadows/highlight with automatically applied chosen tone curves so in scene with brightness variation there’s always some variation in exposure parameters if camera is allowed to select them.
Imaging-resource’s low light image quality test serie with varying shutter speed/ISO combinations is basically done with manual settings for every camera exactly because because of that.
Miraculously every camera outputs image of same brightness at same settings. (while sensitivity standard don’t even have single unambiguous way for defining sensitivity in all situations)
sorry to spam this thread more – but If the situation hasn’t changed dramatically and from what I remember from Canon 5D days- it’s better if starting point is a little bit overexposed (balance is shifted towards ‘cutting’ the highlights) if you have good sensor and do not fear of ‘lost sky’ in JPEG.
It’s because in original unprocessed RAW image you’ll see only ‘white’ in highlights – but that doesn’t mean RAW hasn’t stored hundreds of different ‘highlights tones’ but they look washed originally (in RAW, they are true white in 8bit JPEG) due to human eye perception of bright colors.
You’ll correct the exposure first (shifting it down with -2/3 EV in LR) to get good ‘well balanced’ starting point, and immediately the ‘whites’ becomes more detailed and saved (again, 12bit per channel in RAW and top bits for highlights + highl. recovery), but also the dark areas are now darker – but still contain more nuances.
They are a bit ‘lost’ (mostly for human eye) – as you shifted the exposure so not big variations in shadows first – but that’s why you have ‘save shadows’ option in LR (still using 3.x) that brings those details out (put brightness to them) to define them more.
But that information that you are working with is there, unlike with underexposed image, where true blacks will remain pitch black with no definition and brightening them later only increase their luminance, not definition. Again – if situation of storing RAW data hasn’t changed somewhat.
Indeed your aim is to get image right first (not overdo it and get washed-out HDR image) and apply any (more serious) post-processing later on…that’s why I love digital – getting a ‘proper exposure’ is then done in LR presents (many interesting) instead of on-the-spot as with analog…
Again, I need to try it on OM-D and put a blog post out – if I’m right with it, as I might not be.
Hi Charles,
First of all, expose to the right is indeed going to do what you expect from it with an E-M5, tho keep in mind that you have a bit less headroom for highlight recovery.
This said, keep in mind that all manufacturers use ‘camera iso’, which is based on jpeg output, not on any ‘inbetween’ step. This is no different from how things used to be in ‘film days’ where ISO/ASA/DIN were never about the latent image, but about the developed image. If you decide to divert from the manufacturer designated development process then you will have to use a ‘personal exposure index’ to deal with how you divert from the designated development process.
You describe one of the very interesting sides of digital, based on raw processing, but there are many different sides to this. For things like ‘fine arts’ and portraiture for example, your approach is totally valid and probably gives the best results regardless of what system you use. However, for things like ‘news’, event photography, sports etc, optimizing for good jpeg output directly from the camera is really very important, as in those cases very short turnaround times are key, and there simply is no time to do raw development, rather, you want to be able to make the resulting images available as quickly as you can. For a weekly magazine, better quality helps compete, but for a daily newspaper, the fastest results of sufficient quality is all that counts to compete, and for ‘online news’ that is even more true.
So, jpeg output on high-end cameras does matter, and producing good jpeg output, and exposing for it, also makes a lot of sense.
Such high-end cameras can almost always be configured to expose differently. In case of Olympus, the ‘exposure shift’ I have been pointing at has the same effect as changing exposure compensation when looking at the picture, but in operation it really is something totally different. Exposure shift (in the utility menu) changes the starting point, setting it to +4/6 for example will cause the starting point to be 2/3 stop brighter, but exposure compensation will still be 0 and you have the full range of exposure compensation at your disposal. So, exposure shift really does the thing that you want, change the balance between highlights and shadows.
Additionally, make sure you set ‘gradation’ to ‘none’. Auto gradation will let the camera try to judge the scene and pick what it thinks is a good tone-curve, and adjust the exposure based on that. This can make things much much worse for what you want because it is trying to automate the kind of thing you want to control yourself.
FYI – B&H is now showing an expected avail date of Dec 5 for the GH3.
just compared the iso 3200 bookshelf images; the GH3 jpg is much cleaner, more NR or not?
broken links
Comparing ISO 6400 here (What I find to be entirely usable on the E-M5) and I’m seriously hoping the GH3’s JPEGs are not final and will change dramatically. The colors are muted and the noise looks very odd.
Bahhh jpegs…..I want raw comparisons >:( These cameras are catering to a slightly more advanced than a beginner point-n-shoot market, right???? Give us raw samples to compare, gad dummit.
And what RAW converter are you going to use to view the GH3 images?
My kick-ass TomRAW converter, what do you think? >:(
The EM5 at ISO 1600 has the same shutter speed as the GH3 at iso3200, something fishy going on here – are the EM5 iso ratings that exaggerated?
I think it’s the other way around. The E-M5 has been accused of showing a higher ISO value than it really is using (compared to other manufacturers). In this case, what the E-M5 calls ISO 1600 is what GH3 calls ISO 3200 (assuming the lighting in consistent in this scene). In other words, if the E-M5 is cheating by 1/3rd of a stop, the GH3 is exaggerating even more and cheating by 1 and 1/3rd of a stop.
Even with this possible “cheating” by Panasonic the E-M5 looks better.
Use this tool : http://www.dpreview.com/reviews/panasonicdmcgx1/20 and choos e-m5. DPReview has consistent shooting conditions. The settings are equal. If Oly lies so Panasonic lies too.
Imaging resource shoots their low light studio scene serie with everything at manual settings because of whole exposure thing and ISOs being not completely unambiguous to define.
But still every camera manages to output same brightness image with same exposure parameters.
(with differences in shadows/highlights depending on sensor’s DR/noise performance)
DpReview do not have controlled anything, I expect nothing but whining from the Olympus fanboys , the biggest bunch of BS posters on the net lol .one decent camera ( compared to rivals) since digital started and they think it is king. If it wasn’t so damn annoying it would be funny
One? Would that be my E-400, still working since 2007 and >35,000 shutter count?
One good camera since digital began? Seriously? I’m annoyed F/T seems to be dead, but Olympus are still bringing out innovative cameras. Shame they’re not for me. The GH3 looks more my size, but I’m open to whatever I can afford and I think will see me for the next 5 years.
It’s just a camera.
Is it the moire we see just below the “Geek” of the hi-res version of Laura the second?
The links are working fine for me. The winner is the E-M5, because that’s the camera I have, and I don’t plan on buying a new camera anytime soon. I justify my purchase by not looking at the GH3 pictures, rather assuming the E-M5 are superior.
If and when I buy a GH3, it will be superior to the E-M5.
E-M5 definitely better
I don’t know IF the settings and IF the lens used are the same, but to me the OM-D files looks cleaner and sharper.. plus, pushing the ISO, the Oly’s noise is much more pleasant to my eyes (almost remind me film grain, well… almost..) for 300 euro less… i would take the GH3 only if i was a video addict, but i’m not..
They look very similar except the E-M5 has more NR and contract which helps hide the noise. The E-M5 also has a lot more sharpening applied. I expect the RAW files to be very close.
OMD for me.
The two portraits shots with the gh3 are ugly ( at full res).
It is very thrifty and frugal to shoot with M4/3 lenses now, it seems. First, you pay MORE for a given M4/3 lens than you could get a comparable APS-C coverage lens for. Next, you pay anywhere from $300 to $500 MORE to have the lens colored black instead of silver. What’s not to like here, after all?
Boy, oh boy, is everyone shooting w. M4/3 lenes a blue-blooded richy-rich, or what?
Wow, buying these M4/3 lenses, one definitely must be well heeled. First, the M4/3 lens is going to cost MORE than a comparable APS-C sensor coverage goes out the door for. And then, if you want the lens in black — heck, expect to cough-up another $300 or $400 at the cash register.
Boy, oh boy, is everyone shooting with M4/3 lenses a blue-blooded richy-rich, or what?
It´s the question if it´s good to compare jpg´s in this early phase of the GH3.
Looking at the bookshelfs with iso 1600, the GH3 has a little more noise, but : the border lines between colours are much cleaner! You can see this also on the photos of the colour checker.
With cleaner I mean : It looks like ” digital dirt ” on the OM-D photos, at least at the edges.
If you want to pixel peep to excess, compare the photos with Fast Stone Viewer at 200 %, then it´s getting very clear …
From my own experience with my OM-D I know, that there is little difference when shooting raw if you only look at the ” dirty ” edges ..
PS : you can see this also at ISO 800..
Both E-M5 & GH-3 has one thing in common, and the CLEAR winner: Sony.
Yup, pretty sad they’ve monopolized the sensor market. I still hope Panny could make a comeback (and with the multi-aspect thing) in GH4.
As far as jpeg goes, GH3 looks cleaner @iso6400 but the OM-D keeps more details.
All I can tell now is GH3 got stronger NR. Waiting to see the raw though.
well the evidence that the gh3 uses the same sony sensor is getting stronger. I hope they saw the om-d and thought they had to respond but their global shutter (multi aspect!) sensor wasn’t finised so they had to use the sony sensor. which means the om-d could maybe be as good for video as the gh3 with only a firmware update?
The EM-5 seems to be holding back, not quite reaching the full potential of it’s Sony-provided sensor. Looks like the GH3 has about 1 stop greater dynamic range over the Oly.
None of these help me in regards to what the camera will do. When I shoot high ISO it is almost always wide open and lower shutter speeds in dark conditions. It’s a different animal. Show me some night time street photography.
About m4/3 VS APS-C/H, well :
m4/3 cameras AND lenses are much smaller. That counts when you travel a lot or have your camera all the time with you.
You don’t have to stop 3 or 4 times with m4/3 lenses to get something very sharp : one or two stops are enough.
With m4/3 you get less DOF than with APS-C, and that goes the same for APS-C vs Full Frame.
m4/3 are 25% more expensive in general than APS-C DSLR and lenses. Smaller is more expensive I guess.
My good friend “M” who owns both 5D Mark2 and big Hasselblad loves my OMD-EM5. If he didn’t have to buy his 7000€ used Hasselblad for work, he told me he would have bought and OMD EM5 as a serious complement to his 5D.
But what we all want is a digital Mamiya 6-7, no? 😀 Maybe something in the like will show up in 5 years. After all, sony is ALREADY releasing a FF compact body! Oooh a 6×8 “compact” camera would be so nice. Come on Sony! Do it! You have Olympus lenses and stabilisation technology! Go!
My money goes to GH3. I want a well-rounded camera for everything(still, video, wifi for smartphones and tablets, timelapse(fav), heat dispersion, controls, multi-aspect sensor,power input to replace the battery using dc couplers, and etc). OMD is an awesome camera for STILL ONLY, but video? mehh. Dont get me wrong, Im not a fan of Oly or Panny, my previous camera was a canon s100 and a gopro hero2. In my opinion, As an undergraduate student GH3 is the “New Generation Camera”. The blooming of tablets and smartphones are parallel to the wifi feature of this camera.
I knew I had spotted a heavy ‘Orange’ cast in the GH3 footage that we’ve seen so far. All I can say is Panasonic had better fix that csst in shipping cameras. No one has time to be dealing with it.