hyalinejim

The two shots are not taken from the same viewpoint.

Here I've moved the M43 shot so that the position of the background actually lines up with background of the full frame shot. You'll see that there is no background compression to speak of. OK, there is a slight difference in scale but we don't know if the photographer moved or not. A tripod would have helped this comparison a lot.

13 minutes ago, tweak said:

A 50mm is a 50mm, so soon as you use a 25mm to get that same FOV and a similar bokeh on the smaller sensor it changes the compression of things in the background.

The only thing that can can affect the relative geometry of objects is moving the position of camera. Once the camera is fixed, geometry is fixed. Light goes in straight lens. Focal length and sensor size only affect how much of the scene in front you're selecting for your composition, as if (as mentioned) you're cutting out a rectangle from a magazine cover.

I'm a bit rusty myself, but if the Isco can only go to about 80mm wide before vignetting is introduced it doesn't bode well for other anamorphics. On full frame, I remember that the Isco could do 50mm and the Kowas and Sankors of this world would do about 85 - but you have to remember they have differing crop factors, ostensibly 1.5x and 2x respectively although this open to interpretation (my isco is probably more like 1.4x).

I'm not familiar with the two models you mention.

Today I did a quick test regarding exposure methods for WideDR. I fed a +1, 0 and -1 exposure (using the camera's metering and an eye on the zebras for reference) into FilmConvert. When overexposed shots are brought back into range they become more contrasty. The opposite is true for underexposure. Nevertheless, I was able to match all three:

Looks similar right? Now look at the lower midtones / shadows:

It's clear to me how I'm going to go about exposing WideDR.

It is a bit heavy on the lens for sure and causes a bit of droop when the XC10 is at the telephoto end. I'm not sure that I'd ever use it for anything in particular because there's not a whole lot of bokeh to make it worthwhile. You still get flare, vignetting and slight barrel distortion but for me anamorphic bokeh is no.1!

Speaking of putting bits of glass in front of the lens, here's a frame grab with a pre-Isco 36 on the XC10. I shot at f5.6 and got an image more or less free from vignetting at around the 80mm mark. So not a lot of bokeh to give away the anamorphic look. But it's maybe just discernible on the figure entering the frame on the bottom left:

White vignetting occurs, however, when the light source is going fairly directly into the lens, faintly discernible here on the right (and easier to see against a black background):

10 minutes ago, Lintelfilm said:

ghosting

Don't mention the war!

That's interesting @kidzrevil because at the moment I'm using WideDr because I've found there's less banding in midtones. However,  the highlight roll off is more abrupt compared to CLog and less detail there for sure. 

But does diffusion actually increase purple fringing in some picture styles?

What diffusion filter did you get @kidzrevil? I ordered a 1/4 Black Pro Mist the other day.

If anyone is interested, there's a great comparison video that demistifies the differences between diffusions.

From Canon:

Quote

Thank you for providing the additional video files. I have passed this over to our product specialists for investigation.

No probs if you're too busy, mercer. I just think that Canon will like to see the original file so they know it's happening in camera and the XML file as well so they know what settings were used.

5 hours ago, kidzrevil said:

Here's the reply I received from Canon 

"a lot of the chatter consists of posts from people who... don't even own the camera"

Normally I agree. That's why I'm now compiling a folder of MXF files with XML metadata from owners here on EOSHD to send to Canon.

Thank you @Lintelfilm @kidzrevil for footage so far.

XC10 owners - PM me your ghosting clips

Wow, that is for next to nothing!

I'm expecting a different seasonally themed artifact to kick in on November 1st, maybe something like...

48 minutes ago, tomsemiterrific said:

There is a close XC10/15 discussion group on Facebook.

Do you think those guys are interested in discussing ghosting?

I don't think it's RS because this image shows RS and ghosting together:

Also, if it was RS why would it change with ISO? It's surely noise reduction.

@kidzrevil Your Canon rep is very responsive. Mine is much more terse: thanks for the video, send it in.

STOP THE PRESS, this just in:
 

Quote

 

Thank you for contacting us in regards to a ghosting issue you are experiencing with your XC10. We have passed your request onto our European Product Specialists for further investigation and they will analyse the video files you provided to see if there is an issue.

We will contact you again shortly after the video has been analysed. In the meantime, if you have any further updates to your request, please don’t hesitate to contact us by replying to this email.

 

I think we should really take your suggestion @kidzrevil and compile a folder of brief shots from everybody's XC10 (and include Tom's XC15 to show it doesn't have the problem). If we include the XML files they will see that they're all from different cameras and the variety of settings being used

@kidzrevil @tomsemiterrific @mercer @Lintelfilm @HugoS316

Who can host this shared folder? File sizes might get big so keep 4K clips very brief if possible.

EDIT: I cleared some space in my dropbox. I'm gonna PM you guys with the link.
 

Not looking for ghosting, just using the camera as it's intended in TV mode and walking around a bit.

ND was still on as I had just been outdoors so I guess the ISO ramped up when I hit the stairwell.

I see FOUR traces of the skirting board.

22 minutes ago, Lintelfilm said:

Any idea what 8.5 equates to in ISO on the XC10?

From flicking back and forth between ISO and (fine) gain in a non-Log picture style, ISO 400 = 7.5db  and ISO 500 = 10.5db.

Going down from 10.5db to 8.5 using fine gain controls is -2db, so that's a 1/3 of a stop. It seems to only make a difference in WideDR and CLog. Could be useful for another little bit of exposure latitude when shooting outdoors.

12 minutes ago, Lintelfilm said:

 

Personally I think the XC10 is still an awesome camera and although this ghosting issue should be acknowledged by Canon it's not keeping me awake at night.

 

I think this is a very sensible point of view - you can great footage from the cam as is. Nevertheless, if it is a NR issue I'd prefer to have the option of using Canon's NR processing in camera, or using my own in post if I could avoid ghosting. Or having it fixed somehow.

Sometimes people talk about motion cadence and its importance and even if you can't point out ghosting with normally exposed footage at base ISO without looking at individual frame grabs, it nevertheless is there and is performing a streaking of motion that affects our perception of movement. And, for me, the premise of this camera is that it is a camera that moves - small, light, handheld, image stabilisation, on the run, on a gimbal. So its rendering of movement should be good. And that is compromised at the moment.

Incidentally, while testing Mercer's suggestion of switching to gain rather than ISO (it made no difference) I discovered that gain lets you get a lower exposure. If you're on 500 in CLog or WideDr, switch to gain and select fine adjustment. When you get back you'll see you're at 10.5db. You can bring this down to 8.5db.

No, I'm outside the return period and I already have a big scratch on the barrel of the lens where it brushed up against my laptop in my bag.

I did sign up for CPN. Once you have the XC10 you become a video member and turnaround for repairs is 5 days not including:

1. The day the item is received.
2. Time in shipment.
3. Time taken to accept / decline an estimate or repair cost. (Customers whose items are not covered by a Fixed Price Repair charge are advised to provide us with a repair limit to prevent delays).
4. Time spent awaiting any customer information necessary to complete the repair (e.g. missing documentation).
5. All turnaround times are subject to spare part availability,

http://www.canon.co.uk/support/consumer_products/repair.aspx

So I could send it in and have it back in around a week or so, probably. Should I be the guinea pig?

I just spoke to CVP, where I bought the camera from. Their engineering dept were not aware of the issue. They have had no reports of ghosting from the XC10 until now and they've sold lots, they said. They encouraged me to send my unit in, stating that Canon would not ask you to send it back unless they thought there was something they could do about it.

At the same time, however, I've seen footage from 7 different XC10s that show the problem, and none that don't. BUT tomsemiterrific's XC15 doesn't have the problem so we know that there's a fix of some kind. Apparently Canon will sometimes secretly change firmware or swap out the insides in order to fix something and then send it back saying "We didn't find anything wrong with it."

5 hours ago, jpfilmz said:

I then went into the camera and turned off everything...all image stabilization features, all auto focus features, including on the body of the camera, flicker reduction off, shockless WB off, rolling shutter reduction off and with those setting I was able to greatly eliminate ghosting affects in my footage.  EDIT I also tried with with rolling shutter reduction turned back on and had similar results.  Could someone else try with these setting off and see if that helps.  I'll try to post some of my footage.

Thanks @jpfilmz for your test footage.

3 hours ago, Lintelfilm said:

I just turned everything off and the ghosting is still there.

I can confirm that ghosting exists at ISO500 with everything switched off. Here's a CLog frame grab with contrast expanded:

Look at the curved handle on the RHS - there's a ghost image. This is affecting motion cadence at all ISOs.

@jpfilmz @Lintelfilm @mercer @HugoS316 and all other XC10 owners:

It's time to contact Canon! Send them your ghosting footage and/or  links to the threads here and ask them to resolve the issue.

Thanks Mercer for doing this test! From looking at this and the examples in the other thread I'm convinced that it's in all XC10s and not in the XC15 and that there's no point in anyone returning an XC10 as Canon should be able to replicate this themselves. The fix lies in firmware. I'm going to email Canon now.

I sure hope so Tom. That's very exciting about a firmware update. Even though I don't know anything about this CPS colour space. Must google it.

Y'know I'm pretty sure you would have spotted ghosting on your XC10 if it had it. All the same, I'd still like to see a similar clip from an XC10 without ghosting.

@kidzrevil It's looking more likely that we should send ours in. We know now that not all XC15s have ghosting. But it could still be the case that all XC10s do, and the difference is due to software processing. We still need to see a high ISO XC10 clip without ghosting.

I tried one in a shop yesterday and it was like Ghostie McGhosterton. Who else owns an XC10 round here?

Ghost free!

Big thanks for that @tomsemiterrific. Looks clean as a whistle on my end, no motion ghosting. If I did what you did with my cam, it'd look like we were all on mushrooms. ..

@mercer or any other XC10 owner, can you check if your results are clean @20000 ISO like tom, or sick like kidzrevil and me?