Olympus E-M5 Mark II - love and hate at first sight

[Bob Goldberg]

I've done another test and have come to a conclusion I think.  I think the EM5 mk ii settings in the camera are just a beginning and it shows it stuff when tweaked in grading.  The odd thing was I did my test and on looking what came out of the camera I considered the GH2 the winner.  When I was editing I could see the side by side the EM5 looked better than it was showing.  So I tried to simply grade them to make them more similar. After my grade I liked the EM5 better go figure.  I think it comes down to a flatter file on the EM5 where the GH2 has more contrast.  This all kinds of makes sense because sometimes what came out of the EM5 looked great to me and sometimes not!  Same with other peoples videos.

As far as the settings on the EM5 mk ii I was using natural -2 contrast +1 sharpness and -1 saturation.  As for the GH2 smooth -2 all except saturation -1.  Side note the EM5 files was 2,382,642 KB to the GH2 878,028 KB.  Both were auto WB and constant focus.  Both were set to 1/50th and auto exposure Iso was set to 200 on GH2 and the Em5 did what it thought best so no clue probably 200.  I ran them side by side the GH2 using the 14-140 and the Olympus the 12-40 which does give the EM5 a little help. I should also add I zoomed out to about 50 vs the EM5 40 to make up for the crop factor.

The very basic grade was for the EM5 -10 satuation -20 brightness 4 contrast WB and a slight hue adj.  The GH2 was saturation 5 brightness 19 contrast 7 WB and a slight hue adj.  I'm sure someone better versed in grading could get better on both cameras.  For me it was to see if they were in the neighborhood and after some tweaking I now do not share Mr Goldbergs opinion at least as far as the GH2 goes.  It's by no means perfect and I have big gripes manual control, moire, and crop that being said it is a great stills camera and not to embarrassing as a video camera I now discovered so it works for me I think??   A firmware update with a few tweaks could put it over the top.

Even at 480p, the one on the left just looks much sharper. Forget 1080p. The fact of the matter is you never labeled them, but if the one on the left isn't the GH2, you did something seriously, seriously wrong. Because I've never seen the GH2 look a bad as the one on the right.

And, forget the grading and YT posting. Most of us here are interested in seeing the unedited output from the cameras side by side. YT compression is known to destroy video quality. And someone else editing the video takes even more away from the comparison.

And finally, I think you should be using similar quality lenses on both cameras. But, it's understandable if you want to shoot side by side and don't have similar lenses.

[Flynn]

It would appear the EM5 ii is better than the GH2. I guess that amazing and incredibly useful 5 axis IBIS doesn't degrade the image after all.

[Bob Goldberg]

It would appear the EM5 ii is better than the GH2. I guess that amazing and incredibly useful 5 axis IBIS doesn't degrade the image after all.

​Based on what precisely?

You know, with communication skills like these, you're really going places.

[Flynn]

Ha! Based on what Inquisitive said about preferring the EM5 image. But in all honesty, was just needling you. I apologize. Gonna be hilarious if the image on the right is the GH2 though. 

[fuzzynormal]

my posts would decrease markedly if I didn't have to keep responding to your inane observations of me. 

​You don't, you know.

[Inquisitive]

Well the one on the left is the EM5 and the GH2 on the right.  I only did what I wrote which was not much.  Here is the file out of camera for both.  I did discover the EM5 is very flat and the GH2 has a lot of contrast in comparison check the tree trunk in the shadows towards the end and than look at it after my changes. You should be able to view it at 1080.

http://youtu.be/XUEjQOIHvIU

[fuzzynormal]

A firmware update with a few tweaks could put it over the top.

​I think that is possible, but not likely.  Moreover, I don't believe any upgrade will be forthcoming in a timely manner.  Even then I wouldn't expect it to change the IQ much.  Here's to hope though!

[Inquisitive]

It would appear the EM5 ii is better than the GH2. I guess that amazing and incredibly useful 5 axis IBIS doesn't degrade the image after all.

​Well actually stabilization was off on both as it was on a tripod. 

[Bob Goldberg]

Well the one on the left is the EM5 and the GH2 on the right.  I only did what I wrote which was not much.  Here is the file out of camera for both.  I did discover the EM5 is very flat and the GH2 has a lot of contrast in comparison check the tree trunk in the shadows towards the end and than look at it after my changes. You should be able to view it at 1080.

http://youtu.be/XUEjQOIHvIU

​I'm honestly at a loss for words for how poorly you've made the GH2 footage look. This is simply one of the most resolute 1080p cameras out there, and it looks like a toy the way you've shot it. Even with YT compression, I've seen this camera look vastly better.

For one thing, the fact that your graded original footage looks so markedly different, even at 480p, means that you've used some seriously different settings and different grading techniques. The contrast adjustments are probably hugely different for these two cameras, with the E-M5 II getting a lot more. And once you've fed these two completely differently contrasted videos into YT's sh1tty compression scheme, this is the final outcome.

The fact of the matter is, the GH2's contrast at baseline settings is far more understated than the E-M5 II. The E-M5 II is completely oversaturated and overcontrasted at all "0" settings, while the GH2 isn't. The adjustments you did for the GH2, you should have done for the E-M5 II, and vice versa. Or you could have left them both at all "0".

I'm not saying this is intentional, but it's completely botched.

I still believe that if you actually UPLOAD the two unedited files, we'll be able to see a lot more than what we're getting from YT.

[Inquisitive]

​I'm honestly at a loss for words for how poorly you've made the GH2 footage look. This is simply one of the most resolute 1080p cameras out there, and it looks like a toy the way you've shot it. Even with YT compression, I've seen this camera look vastly better.

​How poorly I made.... LOL well I told you how I set them up but I don't dismiss I may have screwed something up in post.  So I will post a video with the corrections I made to the EM5 only and leave the footage from the GH2 as it was out of camera.  I will even admit that out of the camera the GH2 looks better but we have had a long time to get it right and we all are getting are head around EM5 mk ii.  I won't agree as you stated in another post that we have to evaluate them strictly by out of camera results that's as silly as just comparing stills by a cameras jpgs alone.  If the camera needs some post to make it shine that's ok with me.  So the settings are posted and the cameras were on shutter priority auto so anything either camera did they did to themselves.  Also please keep in mind while evangelizing the Panasonic series it would be nothing without the hacking that was done to the GH1 and GH2 that motivated Panasonic to morph the GH series into what it has become.  If only the hackers could break this one down I'm sure we could be amazed.  As far as YT files go anything it's doing it's doing to both of them at the same time. I think it is a fair comparison.  

http://youtu.be/xc7VphzEpB8

 

[mat33]

​How poorly I made.... LOL well I told you how I set them up but I don't dismiss I may have screwed something up in post.  So I will post a video with the corrections I made to the EM5 only and leave the footage from the GH2 as it was out of camera.  I will even admit that out of the camera the GH2 looks better but we have had a long time to get it right and we all are getting are head around EM5 mk ii.  I won't agree as you stated in another post that we have to evaluate them strictly by out of camera results that's as silly as just comparing stills by a cameras jpgs alone.  If the camera needs some post to make it shine that's ok with me.  So the settings are posted and the cameras were on shutter priority auto so anything either camera did they did to themselves.  Also please keep in mind while evangelizing the Panasonic series it would be nothing without the hacking that was done to the GH1 and GH2 that motivated Panasonic to morph the GH series into what it has become.  If only the hackers could break this one down I'm sure we could be amazed.  As far as YT files go anything it's doing it's doing to both of them at the same time. I think it is a fair comparison.  

http://youtu.be/xc7VphzEpB8

 

​Ok Inquisitive, it time to fess up.  We know that you are the head of the global Olympus video fan boy club -the Olympinati, who is engaged in a global conspiracy to secretly convince the film making world that Olympus is the king of video with the same if not better resolution than the GH2/3/4 but with much superior stabilisation that they just can't live without.  If you succeed, Panasonic and other companies will have no choice but to put IBIS into all their cameras and in the process they will throw away their highly detailed, moire-free images and will plunge the world into the next video dark age.......

[Bob Goldberg]

​How poorly I made.... LOL well I told you how I set them up but I don't dismiss I may have screwed something up in post.

 

So I will post a video with the corrections I made to the EM5 only and leave the footage from the GH2 as it was out of camera.  I will even admit that out of the camera the GH2 looks better but we have had a long time to get it right and we all are getting are head around EM5 mk ii.  I won't agree as you stated in another post that we have to evaluate them strictly by out of camera results that's as silly as just comparing stills by a cameras jpgs alone.  If the camera needs some post to make it shine that's ok with me.  So the settings are posted and the cameras were on shutter priority auto so anything either camera did they did to themselves.  Also please keep in mind while evangelizing the Panasonic series it would be nothing without the hacking that was done to the GH1 and GH2 that motivated Panasonic to morph the GH series into what it has become.  If only the hackers could break this one down I'm sure we could be amazed.  As far as YT files go anything it's doing it's doing to both of them at the same time. I think it is a fair comparison.  

http://youtu.be/xc7VphzEpB8

 

​You screwed up a number of things, but, first and foremost, the biggest mistake you're making is posting to YT. I've seen this type of thing happen repeatedly with their compression. YT completely obliterates the fine detail. What they don't alter is the contrast you (or the camera) apply to the video.

What can (and often does) happen is that because much of the fine detail is gone, what we have left to compare cameras by is solely the contrast that's been applied. In that way, much less resolute cameras can look better on YT than actually good cameras. The better cameras often have contrast and saturation dialed down at default (while allowing you to adjust them up either in camera or in post).

BTW, if you want to compress the GH2 footage in some way, you MUST make all the proper contrast adjustments BEFORE you do so. There's no way to properly fix this after it's been compressed. If you want to upload the GH2 footage without doing anything in post, then bump up the contrast, saturation, and sharpness in camera.

But, this is all pointless anyway as the best way to compare the footage is OOC. I don't want to see your adjustments and they're of no use to me in evaluating a camera. Similarly, by the time it's been wrecked by YT, it's all pointless anyway.

If you want us to fairly evaluate the footage, upload them both unedited. It's as simple as that, and I've never said otherwise.

p.s. Judging things like AF is fine for YT, but judging the actual video quality is incredibly difficult there. If you do want to upload to YT, you absolutely must get contrast, saturation, and sharpness (either in camera or post) as close as possible before you upload. Even then, expect the fine detail to be whittled away. But you even failed to get the grading anywhere near the same for both cameras, so you did completely botch it. All you really have to do is upload the OOC footage. Nothing more, nothing less.

[Wulf]

the king is dead, long live the king

in other words: if you deliver mostly for YT, you have a marvelous tool by now - much smaller and cheaper as a 24-70/2,8, and at least as good as a GH2

 

[Bob Goldberg]

the king is dead, long live the king

in other words: if you deliver mostly for YT, you have a marvelous tool by now - much smaller and cheaper as a 24-70/2,8, and at least as good as a GH2

 

​You're joking?  Even most of the YT videos I've seen, including those done by professional review sites, have looked pretty poor. 

Look at the DPReview and Robin Wong videos shot with this camera. Both truly bad IMO.

Inquisitive's E-M5 II video looks terrible as well, except somehow he's managed to drive the GH2 even lower than the E-M5 II, making it look good by comparison.

It simply doesn't matter where you post, the GH2 is going to look better than the E-M5 II, unless you've done some intentional sleight of hand or you're so inept as to defy reason.

And why does he refuse to post the originals? He only gives us his hack jobs, but won't give us the OOC footage.

My guess is that in Inquisitive's zeal to justify the purchase of the E-M5 II, he completely forgot to try to make the GH2 footage look good. Why else would you set the GH2 for grading and then not grade it? He would have simply done better to leave it at default (all "0") and left his magic touch out of it.

[John Brawley]

​I have however a question: What is the native rated ISO on this sensor? I find more noise on 200 than 640, or is it me. 

Hello. 

Im lead to believe it's only ISO 200. I've certainly found it gets pretty noisy by ISO 1600. I haven't liked anything beyond that. 

Which codec are you shooting ?  I've only found noise presenting by 800+ ISO. 

 

JB

[noa]

@ Inquisitive While I always appreciate anyone taking the time to post comparison videos there are some I take more seriously then others, when you start using autowhitebalance and auto exposure, it's a test on how both camera's compare in automodus but anytime you will let the camera decide what to do you can't call it a test on IQ, I mean no disrespect but when I read "and the Em5 did what it thought best so no clue probably 200" says a lot about your experience level. When I shoot video with my dslr's I know exactly what the iso, shutter and f-stop is at any time because they do have an effect on the image, having no clue can mean you can get diffraction because your iris is almost completely closed resulting into a soft image, or having a way to high iso resulting in noise and again a softer image or not having a natural motion blur because your shutter is way too high. That's what you can get when you let your camera decide and that is what you see in many comparison videos where the viewer gets the wrong idea about a camera.

I feel that if you don't understand basic camera functionality any comparison video you put online will only confuse many people. Hope you won't be too offended by my comments, like I said, I do appreciate the time you put into this, I only find the test the way you intended it to be not of much value when you use auto functionality on both camera's.

 

[Wulf]

​You're joking?  Even most of the YT videos I've seen, including those done by professional review sites, have looked pretty poor. 

​me, joking?

the whole affair is a joke

it's all part of the big plan

btw, talking about pictures is the biggest joke; this is an amateur cam and unfortunely missing (edit:) _sharp_, moirefree  footage at least without IBIS, which nevertheless makes it quite perfect for simple family stuff in the hand of amateurs

or sometimes maybe even in productions, when ever camera movement is asked for. A stealthy B-cam in pocket size, marvelous. in the other pocket you could put your GH oder NX or whatever, perfect.

edit

 

Bob, did you have a photography background?

Compare for example the teaser for the new Canon C100 MKII and from the Oly and you will have a chance to understand that 'sharpness and detail' is only one asspect in story telling by motion pictures...

[John Brawley]

Thanks John Brawley for participating and giving hints on how to get the best from the camera.

I'm coming to the conclusion that the EM5 2 is similar to the GH4 in that you can seriously degrade the results by being too smart for your own good. Natural Picture Mode with everything set to 0 and standard noise reduction is giving me the best footage so far. I'm surprised at how good relatively high ISOs look. 

The moiré, on the other hand, is a major problem. I can't see how you could count on this camera to shoot paid video work in an urban environment. Far to risky. I can easily imagine situations where the moiré would make it impossible to deliver useable footage, it's that bad. No idea how Olympus could let a camera ship with that. It would be very interesting to hear from someone at the company. There appears to be a lot of goodwill around that Olympus is sacrificing due to what appears to be, at the very least, rather optimistic claims made in the marketing of the EM5 2.  

Stills are great. I did a job yesterday in awful light on my 5D MK III and the EM5 2. The Canon was at 3200 - 6400 f 5.6 most of the day, the Olympus at 800 ish and 2.8. The EM5 2 was a joy to work with, more stable and had better focus. The EVF made the Canon's optical viewfinder feel very old technology. Although the Canon files are bigger the Olympus seems to have better resolution. 

​Hello...

Moire...Well, I don't get offended by it as much as others... but for some reason, I don't tend to get it happening to me a lot....It's something I put down to my shooting style.....It's one of those things we all face as DOP's to varying degrees.

Right now I'm shooting principally with Sony F55's and have done for the last two TV series I've done ( Party Tricks and Hiding ) And despite the fact they have OLPF's they can moire pretty easily. More so when shooting 2K RAW but even in 4K RAW.  Sony actually have another OLPF they suggest you can use for 4K RAW for HD delivery, but I found it was wayyyyy to soft...  In fact, another peer DOP on a fairly mainstream show that shoots here had huge big problems with a particular set that they had on the series.  He had been shooting both F65 and F55 and had almost no problems at all in the same conditions with the F65.

We spoke at length about moiré minimisation strategies.  It's very hard to remove in post as I'm sure we all know.

The last two shows I did were with vintage / film lenses which are a bit softer and I tended to shoot pretty shallow so I only had a few shots across the whole eight episodes that were problems.  He had problem shots a lot more frequently...

I've found anecdotally it's moires less than a 5DMK2, which is my benchmark for borderline unusable for moire.

I do turn the sharpness down in the EM5MKII though so I think that probably helps a little.  I also tend to use older film era style cine lenses which again, help a little too I think. Having the sharpness turned down on the camera also gives, to me, a more pleasant less "processed" look.  Kind of like what happens when you over use the clarity slider in PS. A lot of consumer cameras like these tend to do a lot of processing under the hood (things like detail, coring, noise reduction etc).  There's also a lot of things that affect sharpness perception and I'm not much one for the hyper sharp 60 FPS look so I like my motion blur at 24 fps please and I think that helps too..I used to to exactly the same thing on 5DMK2's and turn the sharpness all the way down...

I've also been working through it on the Blackmagic cameras, none of which have OLPFs and all of which to some degree are fairly prone to moire.  Again, it hasn't stopped me form putting many many shots to air, and it rarely bites me in a way that makes something unusable.  Usually I can spot it when we're shooting and it's usually rpetty easy to make a small adjustment to avoid it....

I still get bitten though.  There are a couple of shots in Curiosity that are bad.  Personally, as I said, It has to be pretty obvious for me to find it offensive, but I have a higher threshold than others I think.  I know many many DP's hate seeing it at all.   I think if it's something you're really averse to seeing, then it's probably not for you.

JB

[noa]

I've found anecdotally it's moires less than a 5DMK2, which is my benchmark for borderline unusable for moire.

 

I have shot with a 550d in the past and the moire on that one could be pretty bad, maybe even worse then the 5DII, shooting with fast lenses wide open to blur out the background incase I had some fine textures like small bricks in the background did help as well though there have been times when shots, especially wide deep dof shots looked unusable. But from what I have seen in Gordon Laing's video, this one in particular: (vimeo.com/119316833) I would be certain that the same shot with a 550d the moire would have looked much worse.

 

[DPC]

Thanks John for adding that.

Last week I was shooting interviews in a laboratory environment. Two of the four walls were white with those tiny black holes in them (there must be a technical name for this that I don't know). I had a 5D MK III, GH4 and the EM5 2 with me. The 5D MK III allowed me to throw the background out of focus and kill the moiré that way but then I had all the limits of shooting a talking head with shallow depth of field. The GH4 4K wider angle ( for cropable cutaways) was entirely determined by having to eliminate the moiré on the background. So my options for the shoot were limited by the way the different cameras handled the moiré in that location. Kind of difficult to explain to my client...

I wonder if using a Pro-Mist filter would have helped?