HPX371e small Jello effect?

G

Garion

Well-known member
Have had my new 371 for a few days and im liking it a lot, not really seeing any noise/skew issues in 1080 25p.

I have noticed a small thing,i was practising using macro on some leaves the other day and on some shots i could see a "shimmering" effect on some leaves.

This was using 1080 25p/180 deg shutter and handheld, it seemed that very small movements caused the effect. You can also see the effect in both the viewfinder and lcd screen, i tried all shooting modes and its still there.

Wierd thing is thats its not like the Jello ive seen on other cameras, it ignores the main body of the object and affects the backgrounds/shaded colours of the insides.

i uploaded a very quick and dirty video to vimeo which shows it quite well, look at the poster of sarah michelle Geller and look at the moon behind her. Whilst i had the camera on full wide and slowly moved the camera back and forth, the moon moves at a completely differant rate to the borders of he poster.

If it was normal jello i would expect the whole video frame to do this not just one small part?

it might be nothing and im probably pixel peeping too much ( it is late here and im tired lol) just wondered if any other 370 owners had noticed this too.

heres the link

http://vimeo.com/11697695

password is testvid

cheers

Andy
 
What codec were you using?

Is this just in the viewfinder? In the edit system or when fully rendered to a good quality master? and what was the workflow to the vimeo clip?

You will still get some jello with CMOS and maybe like the 301 they are using you as a beta tester to find the faults and will then release a half fix in firmware and then the 372 in 11 months time?
 
Hi Gary, i was using 1080 50i/25p avc100intra and the cineg scene file (also does it on bbc film too). I can see this effect in both the viewfinder,lcd and on rendered footage. I use edius5.5 and rendered to a 1920x1080p wmf file at 6mp cbr.

It doesnt happen much only when theres a "small" amount of movement and when im on full wide.

Although i didnt have time to render it i tried changing to 1080i,720p dvcprohd etc and in viewfinder and lcd the effect was still there.

To me it looks almost like the camera is seeing the moon as noise and processing it differantly to the rest of the image. I was getting the same when doing closeups on leaves with insects, most leaves would be fine and 1 or 2 would shift if they had a subtle vein pattern.

Your right about the beta testing bit :cheesy:

Apart from this i am very happy with the look of the video so far and very rarely come off 1080 50i, i just need to get a decent tripod now and ill be sorted.

cheers

Andy
 
Wow, that is weird!
I was ready to say that it'd be haze, or that you just can't make anything out on the web - but it's exactly as you describe, in fact if it April 1st I'd have thought you might have superimposed the moving moon as a trick! That is so strange, I have no answers, sorry. I get the feeling you may be along the right lines with it being the computer trying to process something and making a mess of it. Look forward to finding the answer to this one. Any other comments of the camera would be much appreciated.
Steve
 
Noticed this exact same effect in the HPX300 too.

We were shooting AVC-I 1080 24pN in a boardroom. The camera was sitting on the table, propped-up on a sandbag for a reaction shot of one of the attendees. The door in the shoot-off behind the subject was filling about 70% of the frame behind her. There was no skew in the doors as we were mostly static and pretty wide. However, the entire contents of the door (stained wood-grain) seemed to be shifting a bit, but were still contained within the geographic area of the door itself.
 
steve phillipps said:
Wow, that is weird!
I was ready to say that it'd be haze, or that you just can't make anything out on the web - but it's exactly as you describe, in fact if it April 1st I'd have thought you might have superimposed the moving moon as a trick! That is so strange, I have no answers, sorry. I get the feeling you may be along the right lines with it being the computer trying to process something and making a mess of it. Look forward to finding the answer to this one. Any other comments of the camera would be much appreciated.
Steve
Click to expand...

Having come from an HMC151 the HPX371 is a great camera, i can see why so many people love the 300. It balances perfectly on the shoulder, i had the Anton bauer plate swapped for the new IDX one and using an IDX 10s its just right on the shoulder.

Like i said earlier, i havent noticed any skew when im shooting using full zoom on 1080. I altered the standard scene files taking the coring to zero and its still clean for me.

I usually use either the BBC scene files recommended by Gary above or CINEG,i have detail at -2 which looks very good. I tried detail at 0 on CINEG but i could see anti-aliasing on some edges so ive backed it down 2 notches.

What has surprised me is how vibrant and lush foliage and landscapes look with the HPX.

One thing to be aware of if you ever demo the 370/371 is that the viewfinder will show skew when there is non at all in the footage. Freaked me out when i demo'd it in the store, but we saw that there was no skew in the lcd or on the card. Guessing the refresh rate of the viewfinder is a little slow lol but its damn sharp and very easy to focus with :)
 
It's very nice to see some feedback about 370/371 while i'm on the edge of buying or not...:violin:
 
We're using the BBC recommended settings as well - at least as a baseline. With the exception of zebras, that is.
 
I wonder how many of the bad reports of skew and noise on the 300/301 have been caused by people doing quick demo's and judging it on what they see in the viewfinder not what renders out of the edit suite?

I certainly don't have problems with my two 301's and they are used as B camera's to the 2700 a lot and not a single client has noticed. OK I am not doing sport or anything other than drama or features.

The noise levels are acceptable and with the BBC settings and the latest firmware I see a lot worse on some broadcast content. Once again I am not shooting in bad light most of the time so it is not a problem.
 
Erik Olson said:
We're using the BBC recommended settings as well - at least as a baseline. With the exception of zebras, that is.
Click to expand...

Agree Erik, the BBC film settings really do look like film and yes the zebra settings they have seem a little low.

I have also backed the master pedestal down to -6 as per Barry's recommendation. Think the standard files are a bit high but then i guess it does preserve the shadows and helps grading more.

One thing i did notice last night as i was saving my camera files to card was that even on a PAL HPX the scene files showed a frame rate of 24 instead of 25. I've changed them now just to be sure.
 
It would be interesting to see if you get this effect with the DVCPRO HD codec.
I have a hunch that it's an artifact of the AVC motion prediction algorithm.
jeff
 
GaryNattrass said:
What codec were you using?
You will still get some jello with CMOS and maybe like the 301 they are using you as a beta tester to find the faults and will then release a half fix in firmware and then the 372 in 11 months time?
Click to expand...

This isn't rolling shutter (IMHO).. this really looks like a motion
prediction codec artifact.

Curious one at that .. I suspect the soft image of the moon is such
that the edge detection just sees it as a blob of pixels.

jeff
 
Wow, it gives a 3D feel, Geller and the Moon seem on two different planes. So it looks like you got an added 3D feature for free.
 
It absolutely manifests in the MXF files and in the rendered footage. I watched it again today. We lensed that project entirely in AVC-I 100. I have not noticed it in the DVCProHD, but that is entirely anecdotal. We were all in front of a 42" monitor in the edit suite when the door footage caught someone's attention.
 
Are you kidding me? This is the weirdest thing I've ever seen. It looks matted.
Buhhhhh zarrrrrrr.
 
Bear in mind that the BBC settings are certainly not Gospel. Alan Roberts would I'm sure admit that he doesn't get to spend that much time and in real situations with most of the cameras he tests.
The big exception is probably the tape Varicam which had massive testing for the Planet Earth series and it's very likely that the Alan and the team's settings get the very best out of that camera.
Steve
 
You must log in or register to reply here.
Back
Top