Top 6 videos on EOS R with combined 2 million views and not a single mention of the rolling shutter in 4K

[Andrew Reid]

If you like this forum and want to stay, you better apologise then.

6 minutes ago, Jimmy said:

It was clearly a tongue in cheek joke ffs.

I'll leave you to it... wasn't attacking you in the slightest. 

I like this forum... you've done a good job... but you have seriously thin skin... it is obvious why manufacturers don't reach out to you and it's nothing to do with "telling the truth"... you're just way too erratic and emotional.

Nah mate, it's not a thin skin - it is that I am fed up with dealing with so many fuckwits.

It is not that manufacturers "don't reach out to me", Panasonic, Leica, Sony all did previously. Even Canon did once upon a time. I am not here to defend myself. I don't need to. I do the right thing for the direction I want to take the blog in, and if you can't see that because you're too stupid, that ain't my problem!

[Jimmy]

I'm not apologising for make a daft joke or pointing out the z7 had very bad rs. I apologise if I've said something else that is offensive though.

You are in a minority of people that own both cameras... so you could make a good post about RS, with the same focal length and amount of movement.. and no matter the result, the Nikon is shooting FF, so will have an excuse, unlike canon.

I like your passion and it comes from the right place.... wanting what's best while the big guns hold tech back... just fight the fight against every camera maker

[Dan Sherman]

1 hour ago, Andrew Reid said:

I don't really want to be drawn by your trolling. I don't need to defend myself. It's my house and I can pick and choose who I invite into the community and who I consider bad for it, who should go. You aren't offering much either apart from snide anti-Andrew Reid sentiment, in my own living room and on my phone.

 

You can call it whatever you want. If I read what I think is hypocritical bs, then I'm going to call out who ever said it. I don't care who it is, no one gets a free pass.

 

1 hour ago, Andrew Reid said:

You're spinning things and you know you are. This site is about learning information about the gear and filmmaking, so when I mentioned the rolling shutter within a day of getting my EOS R. you should thank me for it. 

 

Now you are being somewhat disingenuous.  You sell luts, and you have adds for them on every thread for them. The Forum and your articles (including the rants), drive page loads, and thus add views, it's SEO 101. So please don't try and act like your mother Teresa, or some benevolent parental figure who's just looking out for all of us

Yes the forum has educational benefits as just about every other forum on the net does, but you are not the sole person people learn from here. Personally I've probably learned more from other members than I have from you.

 

1 hour ago, Andrew Reid said:

 The other article about a completely different camera, that isn't even a review, with rolling shutter that isn't problematic, does not contain a rolling shutter mention - Tell me what's wrong with that. I can't see it.

I know it's a different camera. However i think you missed the point ( @jonpais, and mine ), you didn't mention RS, and now you are saying you didn't mention it because in your opinion it "isn't problematic".

DPR has a different view.

https://***URL removed***/reviews/nikon-z7-first-impressions-review/7

Quote

On our pre-production Z7, rolling shutter performance in the Super 35/DX crop region looks identical to what we saw on the D850 in both its full-frame and crop modes, which is impressive considering the Z7 is oversampling. We generally find this level of rolling shutter isn't too bothersome, mainly posing problems in fast pans and the like. Once you switch the Z7 into full-frame 4K mode, the rolling shutter effect is unfortunately magnified, and is more noticeable in casual shooting.

 

Should DPR publish and article on their take on your "initial thoughts"?

 

 

 

[Nikkor]

Oh wow the Canon defense force NPCs are attacking.

[Jimmy]

14 minutes ago, Nikkor said:

Oh wow the Canon defense force NPCs are attacking.

No one is defending canon .... having terrible RS and a crop is the worst of both worlds.

[Mokara]

On ‎10‎/‎16‎/‎2018 at 11:04 AM, Andrew Reid said:

Well at least until the Panasonic S1 and S1R come out, but their Achilles heel is still likely going to be the AF - especially via an adapter to Canon EF.

I really think the EOS R is going to please a lot of people... and upset an equal amount as well.... both perspectives are equally right depending on who you are and what you need from the camera.

Correct, doesn't change the fact. However as time passes you expect constant improvement. Canon are infamous for refusing to give us that. They used the same sensor in some of the APS-C cameras for 7 years! Speaking purely in a technological sense, to give you an idea of how much slower Canon's sensor is, that 30ms is a readout of 2160 lines and 3840 columns. Whereas on the NX1 4 years ago, the readout is 3648 lines, 6480 columns!! Quite a bit more to read in one sweep. If that had a 1:1 crop mode of 3840 x 2160 instead of doing the full-width full-pixel readout at 6.4K, it would have far less jello than the EOS R.

The A6300 again is doing a 6K readout so it's processing much more data than the EOS R in the same time it takes the EOS R to do 4K, it is doing 6K.

So Canon have technological problems but actually they don't because the 1D X Mark II does 4K 60p and for that you need sub-17ms rolling shutter, so they have a sensor which is fast enough to avoid the worst rolling shutter issues on the market. It is also a problem in silent stills mode on the EOS R as well so not just a video limitation.

So they caught up but won't give it us. How cynical is that!?

RS on the NX1 in HD mode is quite a lot better (although you get other artifacts in 1080p), The main reason you see it in 4K is the sheer amount of data that has to be read.

The issue really is that the processor is being stretched to its limits in terms of band width. I think the sensor itself is capable of delivering an image with lower RS, it is just that the processor is struggling. The 1D XII has two processors (three, if you include the older Digic used for the focussing/exposure system) and does relatively little processing, which is likely why it has better RS. The EOS R probably has a single processor that does everything, so even though it is newer it has less computational power than the older 1D as a system.

[Andrew Reid]

NX1 was down at 2ms like an Alexa in 1080p for rolling shutter, I once saw, can't remember where.

I am surprised no camera has a distortion correction mode in firmware. Why don't they correct rolling shutter in-camera, just crop-in slightly and re-align the image.

[KnightsFan]

10 minutes ago, Andrew Reid said:

NX1 was down at 2ms like an Alexa in 1080p for rolling shutter, I once saw, can't remember where.

NX1 is around 8ms in 1080p, according to that DVXuser thread, which is right near the maximum you can have for 120p.

13 minutes ago, Andrew Reid said:

I am surprised no camera has a distortion correction mode in firmware. Why don't they correct rolling shutter in-camera, just crop-in slightly and re-align the image.

I bet that's a really difficult problem to solve. It might be straightforward on whip pans across static scenery, but I don't see how a camera would be able to correct a fast moving object across the frame, in real time. Or correct for a fast zoom, where you can see zooming on half the image before the other.

[Yurolov]

I asked for the focal length because at the 90mm equivalency the RS will be exaggerated. It would be helpful if you did another test at 35mm or 50mm equivalency - as I would imagine most people using crop glass will be using it at these focal lengths.     

[Snowbro]

[Andrew Reid]

When you're moving the frame, a longer lens will of course make the contents of the frame move through quicker.

Of course even if you use a 35mm equiv. wide frame, completely still on a tripod, stuff can still move in the frame - a sparring boxer, a train, a swift movement from an actor, and trigger the distortion.

The important thing is to do comparison tests. All cameras at the same focal length (equivalent, so not all same lens but same FOV), on same tripod, moved at same speed at exact same moment.

I just did a test for myself of the Z7 vs EOS R and the Z7 is nowhere near as bad.

That is both in 4K (full frame on Z7).

I could share it, but you don't deserve it

[Snowbro]

I usually like some of this guys videos, but his was a hack job imo. He tweaked the HD footage and 4k footage to try and make them look as similar as possible. Never seen the 4k that soft from anything i've shot with Canon 4k. People do anything to justify their purchases, or get views. 

 

[Django]

DPreview (Jordan Drake aside), Kai (without Lok), Tony & Chelsea, Fro etc.. are all general photography based channels. They don't really know/care about video stuff , beyond the basics.. 

The few video channels that exist are either biased to a system, or sponsored by camera stores / manufacturers (ie shills).

Not many independant, cross platform reviewers out there..aside from Andrew. 

I thought his review of EOS R was rather balanced, was expecting him to just rip it to shreds honestly. But he gave it some credit where it was due.

..and indeed pointed out it's giant flaws for video, rightfully so. i wanna hear the good & the bad in a review. not just one side & not just the obvious stuff.

RS is a serious issue. Bigger issue then the crop factor imo. I'm only using 4K on EOS R for locked tripod shots, but if it weren't the case.. it would be basically unusable.

Anyways let's all take a deep breath, we're all in here with a common passion & these gear wars are getting out of hand..

[Yurolov]

54 minutes ago, Snowbro said:

I usually like some of this guys videos, but his was a hack job imo. He tweaked the HD footage and 4k footage to try and make them look as similar as possible. Never seen the 4k that soft from anything i've shot with Canon 4k. People do anything to justify their purchases, or get views. 

 

He used the movie crop mode in 1080p. Which is disingenuous cause later he said he got the camera so he could use it in full frame 1080p. So rightly speaking he should be comparing the full frame 1080p, being the reason he bought the camera, against the 4k. I think there is a reason why he didnt do that. 

 

1 hour ago, Andrew Reid said:

When you're moving the frame, a longer lens will of course make the contents of the frame move through quicker.

Of course even if you use a 35mm equiv. wide frame, completely still on a tripod, stuff can still move in the frame - a sparring boxer, a train, a swift movement from an actor, and trigger the distortion.

The important thing is to do comparison tests. All cameras at the same focal length (equivalent, so not all same lens but same FOV), on same tripod, moved at same speed at exact same moment.

I just did a test for myself of the Z7 vs EOS R and the Z7 is nowhere near as bad.

That is both in 4K (full frame on Z7).

I could share it, but you don't deserve it

I just noticed it because on dave's footage of the eos r in the background you can see cars whizzing by and there is no issue with skewing. That was at a pretty wide angle though, so I guess I wanted to see if it was acceptable at a 35mm equivalent. 

[androidlad]

4 hours ago, Andrew Reid said:

NX1 was down at 2ms like an Alexa in 1080p for rolling shutter, I once saw, can't remember where.

I am surprised no camera has a distortion correction mode in firmware. Why don't they correct rolling shutter in-camera, just crop-in slightly and re-align the image.

Alexa's rolling shutter is about 4.5ms, the minimum required to reach 200fps full readout.

How would a camera correct rolling shutter when a train flies through the frame while the camera is static?

 

[Andrew Reid]

14 minutes ago, Yurolov said:

I just noticed it because on dave's footage of the eos r in the background you can see cars whizzing by and there is no issue with skewing. That was at a pretty wide angle though, so I guess I wanted to see if it was acceptable at a 35mm equivalent. 

Do you mean a small car in the background? That is not going to show up the rolling shutter.

Needs to be a larger object closer to the camera and moving quite fast sideways, or the whole frame needs to be moving.

7 minutes ago, androidlad said:

Alexa's rolling shutter is about 4.5ms, the minimum required to reach 200fps full readout.

How would a camera correct rolling shutter when a train flies through the frame while the camera is static?

Rolling shutter correction when the whole frame moves is fine, that is how Premiere can do it in post. But you're right it wouldn't work for moving objects that are distorted. The correction would have to be applied only to moving subjects in a static frame and not the background. Impossible. Background would have to be filled-in around the corrected object and you'd have all sorts of artefacts.

1 hour ago, Snowbro said:

I usually like some of this guys videos, but his was a hack job imo. He tweaked the HD footage and 4k footage to try and make them look as similar as possible. Never seen the 4k that soft from anything i've shot with Canon 4k. People do anything to justify their purchases, or get views. 

 

Something very odd about that test!!

Almost as if he's converted both to a 1080p file, before digitally zooming both for the crop-ins.

Also, lens seems very soft.

Maybe he should share the original files.

[jonpais]

On 10/17/2018 at 10:53 PM, Andrew Reid said:

The earliest initial impression I had of the EOS R was the rolling shutter. The earliest impressions I had of the Z7 was that it was nicer than my Sony A7R III in the hand, and in terms of the image, and the rolling shutter didn't stand out as anything abnormal, unlike on the EOS R. It just wasn't an issue for me. Ditto on the D850. It wasn't immediately obvious from the get go, the minute I opened the box.

So we're 10 minutes into Kai's video, out of 13 minutes and yes it seems there's a 2 second mention. So I should probably give him the benefit of the doubt. But when I watched it the first time just after release as a subscriber to his channel, I am pretty sure it wasn't in. There are such a thing as ninja edits and re-uploads. But let's be clear, I've nothing personal with Kai, he could be a nice guy, but he doesn't help our cause as customers - he doesn't give the problem any gravitas. He's not even a proper reviewer. He's an entertainer. He turns it into a positive in those few seconds of a mention, which are just soft pawing out claims to the effect of the 1080p "has an advantage" because it has less rolling shutter than 4K, 'just like that wonderful best selling 5D IV you all bought'. That's how it comes across to me. It may technically be a 'mention of rolling shutter' but it's not a mention in the sense I meant it in my article - a mention of just how BAD the rolling shutter is on this camera. He just sweeps it under the rug in less than 5 seconds in a 13 minute video and doesn't really criticise it. That's the *whole point* I was making in the article - none of them mentioned the rolling shutter in terms of it being amongst the worst on the market in 2018 and immediately OBVIOUS straight out of the box in 4K.

So whilst @jonpais is calling EOSHD "fake news" and telling me how to run the site, perhaps he should realise one thing. I am a bit disappointed by his remarks. All the squabbling on the forums over the past few months from him and the Sony fanboyism in every subject, treating some threads like his own private inbox for chats and arguments, all leave a bitter taste for me because he's a long-time member. He seems to have gone from productive posting in the early days of the forum about Panasonic, up through being rewarded by being promoted to a mod, back down to being a non-mod due to bad behaviour, and now he's cruising for a bruising or even a ban. I just find that incredibly self destructive.

I see what you did just there. Kai ‘ninja-edited’ his video the moment after you finished watching it - a video uploaded over a month ago, with some half-dozen contemporaneous comments about the awful rolling shutter!

Are we also to believe that Kai added those half-dozen remarks in the comments section of his video just seconds after you hastily skipped through it?

Let's get real, Andrew - do you really expect us to believe that Kai edited his review just because of your stinking blog?

What gives you the divine authority to decide for each and every living human being which shortcomings are the least forgivable anyhow? Your shitty YouTube channel?

#bringiton #i'llbeingoodcompany

[IronFilm]

On 10/17/2018 at 5:34 AM, Andrew Reid said:

To be fair, the rolling shutter is also pretty shit on the NX1 - but that is doing a 6K readout, no crop, and it's 4 years old!!

If the EOS R were a laptop, it'd run DOS.

On the upside, the Samsung NX1 in 1080 mode has some of the most minimal rolling shutter ever seen. 

So at least NX1 shooters have a "get out of jail card", shoot in 4K until you're doing a high moving action scene (or whatever) then switch over to 1080 just for that segment. 

 

[Mattias Burling]

Youtube doesn't allow re uploads as far as I know. Its not like Vimeo. Unless they have changed it recently, when you take down a video it is down. As are the stats, comments, likes and dislikes.

I for one thought the test above was ok. I did my first 4K vs HD a couple of years ago and just like now, 4K vs HD on youtube is hardly visible, depending on the camera used.

28 minutes ago, IronFilm said:

On the upside, the Samsung NX1 in 1080 mode has some of the most minimal rolling shutter ever seen. 

So at least NX1 shooters have a "get out of jail card", shoot in 4K until you're doing a high moving action scene (or whatever) then switch over to 1080 just for that segment. 

 

In HD they don't have as much aliasing either.

[jonpais]

@Mattias Burling I'll be banned shortly, but want you to know I appreciate your comments.