25P trouble, is it just me?

OK - 100% there's strobing/motion judder here.

Now, the clip looks noisy (high gain?) and also it seems (I say this as I'm not viewing on a decent machine) that the strobe/judder becomes more pronounced as soon as the high contrast (outside) windows enter frame. This would make sense as the recording path (and whatever compression etc) would start to apply as the pan continues in motion - meaning that the progressive frame is gaining more high contrast elements in a 'cumalative' effect.

It starts off OK and then gets 'progressivley' (LOL) worse... maybe it's related to compression/sensor readings & in camera settings?

Just thinking out loud & basing on a couple of views.

It's a shame the pan doesn't continue further (or reverse) - as it would be interesting to see if the issues settle once the high contrast elements have exited frame (and sensor).
 
Any chance it could be to do with the type of sd card used?

One of the reasons I bought this camera was to avoid the edit room shock which my Canon DSLR would give me. Having been a bit fan of the HVX and HPX171 I'm a bit concerned.

It would be fantastic if the route cause can be found.
 
I doubt it has anything to do with the cards. We used Sandisc 32GB/16GB class 10 and Transcend 32GB class 10.
We are on location (shooting interlaced) so no time to do any more tests at the moment
 
Did you download the clip and play it locally? It looks much smoother. Flash playback through vimeo isn't a great method of accessing the cadence

Parsing each frame, there are no drops or repeats, and transitions are relatively smooth

There are parts where it slightly slows down & speeds up, but I think that is from the operator .
 
I posted it there. I also ask not to watch it from the small embeded Vimeo window. All you have to do is to download the 60MB / 25 second full HD clip.

http://vimeo.com/20429696

All the settings/parameters are described on the Vimeo page. No gain, 400 ISO.
The judder is less visible when contrast is lower (windows leave the frame) but it's still there. The whole point of this excercise was to test it with verticals at high contrast. After all windows entering the frame while panning are nothing unusual.
Please comment (after watching download) if you believe the judder level is normal for the pan speed and listed parameters.



MadMonkFish said:
OK - 100% there's strobing/motion judder here.

Now, the clip looks noisy (high gain?) and also it seems (I say this as I'm not viewing on a decent machine) that the strobe/judder becomes more pronounced as soon as the high contrast (outside) windows enter frame. This would make sense as the recording path (and whatever compression etc) would start to apply as the pan continues in motion - meaning that the progressive frame is gaining more high contrast elements in a 'cumalative' effect.

It starts off OK and then gets 'progressivley' (LOL) worse... maybe it's related to compression/sensor readings & in camera settings?

Just thinking out loud & basing on a couple of views.

It's a shame the pan doesn't continue further (or reverse) - as it would be interesting to see if the issues settle once the high contrast elements have exited frame (and sensor).
Click to expand...
 
Had not seen your full vimeo page (with details etc) - just watched the embedded vid posted earlier... looking over that now.

Don't misunderstand, it's a good test! One that shows a common shooting condition, so I'm not saying the varying low/hi contrast should be avoided... just noting that, in your clip, it was evident that this accumaulation was going on :) - and that was what you needed to show.

Going to grab the HD file later. If you make the original (card/camera) file available (so OP's can use it and convert to their various workflows) it might yield interesting results?

Thanks for testing.
 
Mmm - risked playback on the larger file on this machine... I see what nugat means now... the choppy breakup on the window pan section looks like tape dropout on old BetacamSP tapes (showing my age - LOL).

Still seems very noisy, but the issue with strobe/judder is very visible, sort of looks like interlacing mixed with progressive... very odd/bad... :(

Why not repeat the test in a variety of modes (24P/25P/30P) maybe even interlaced ones too, as it would be good to have a single shot recorded in the same conditions but in differing frame rates to see how each performs.

Anyway, not what I'd expect to see from 25P.

Wonder if the BBC regs on this are going to 'pass for use with tripod use and no camera motion'? LOL.

Lets hope we can get to the bottom of this.
 
That test it´s similar to the first shot we noticed the problems had the problem. It was a very slow pan at an airport with big glass windows divided by pillars.
I will download it now
 
I can see the judder on my laptop, altough it´s not too disturbing. I do think it will look much worse on a TV.
If you only deliver to web, this is probably not a problem, but if you broadcast, it certainly is.
 
Yoak said:
I can see the judder on my laptop, altough it´s not too disturbing. I do think it will look much worse on a TV.
If you only deliver to web, this is probably not a problem, but if you broadcast, it certainly is.
Click to expand...

When you watch it make sure it's 1920x1080 mapped 1:1 onto our screen. I noticed that even small rescaling can affect the strength of the pehomenon. I uploaded the +60MB file from iMovie to Vimeo as 1080/25p and it remained there the same size in megabytes, so hopefully neither Apple nor Vimeo do anything with the original SD card file nor codec itself. (container might change though .mov or .mp4)
 
anyone any further forward with this? i had noticeable judders when playing back on my computer (i assumed it was something to do with computer...but it now sounds like it wasnt).
note that i had already upgraded firmware to 1.15.
modes were....
1080 25p.
system freq at 50 hz.
camera was in 'film cam' mode but vfr was off.
gamma was cine-like d and matrix was cine-like.
i shot onto sdhc class 10 cards.
shutter was off (1/50).

be good to hear from Jan in regard to this. this is hugely disappointing if 25p is this bad under these conditions. fingers crossed there is a simple menu fix, or something wrong in my set up.

perhaps recording using nanoflash will alleviate it but i want to be able to shoot using cards.
nb i have only reviewed this footage using avccam viewer.


[side note......perhaps causing a conflict, tho i cant see why.
i am realising that the scene files are perhaps 'locked'. i had selected scene file 6 but withing the menu had opted for cine like v rather than d. it was d that showed on the screen display tho....surely they should tidy up the menu so that those options are blocked if they wont work?].
 
The discussion moved to "first impressions.... serious drawbacks..." thread here and "AF101 judder" at dvdoctor.net
 
The only update I have is that Holdan got together with Alan Roberts (the tech wizard who does all the camera evaluations for the BBC) to conduct tests specifically for this, and concluded that there is no issue and that the camera is performing exactly as it should and exactly as they expected. Holdan is now looking into hosting the video that they shot. Alan has already gone on record to say that he saw no problems, in the dvdoctor.net thread.
http://forums.dvdoctor.net/showpost.php?p=382023&postcount=82
 
as this has been still annoying, i went browsing my mac to see if there was anything i could change with the monitor display......didnt find what i didnt know what i was looking for but came across this log in the sync section of my computer.
does this mean anything to anyone? its gobbledygook to me but could it infer that the frames were there but the computer dropped them for some reason?

iMac [0x0-0x12e12e].com.panasonic.AVCCAMViewer[1840]: me 969 sleep time -411 < 0
[0x0-0x12e12e].com.panasonic.AVCCAMViewer[1840]: Skip frame 970 to reduce drop frame
iMac [0x0-0x12e12e].com.panasonic.AVCCAMViewer[1840]: Dangerous Frame 983 sleep time -2077 < 0
iMac [0x0-0x12e12e].com.panasonic.AVCCAMViewer[1840]: Skip frame 984 to reduce drop frame
iMac [0x0-0x12e12e].com.panasonic.AVCCAMViewer[1840]: Dangerous Frame 986 sleep time -1379 < 0
iMac [0x0-0x12e12e].com.panasonic.AVCCAMViewer[1840]: Skip frame 987 to reduce drop frame
iMac [0x0-0x12e12e].com.panasonic.AVCCAMViewer[1840]: Dangerous Frame 990 sleep time -5596 < 0
iMac [0x0-0x12e12e].com.panasonic.AVCCAMViewer[1840]: Skip frame 991 to reduce drop frame
{
 
Rick Burnett said:
Well that sounds like frames are getting dropped to me :)
Click to expand...

it does indeed rick.....tho i have never been to that part of my computer before, so really its all new ground.

fingers crossed it was simply my computer dropping the frames rather than the avccam player. (?!?)
i am waiting to hear back from an editor.....but he uses avid and needs to get some extra software to view things.

"being at the forefront of technology is just another way of saying we dont have a feckin clue....yet" ghandi.
 
If you go into Applications/Utilities there is an app called Console. When this is running, if applications are talkative in their design, you can learn a lot about what is going on with the app when there is trouble. That said, one has to remember that the information there is not qualitative. There could be errors or warnings shown that mean nothing because they are just really information for the developer to see. When I code, I definitely use the Console as the location to report things.

Also, it is probably a combination of both. Your computer might be having trouble keeping up with the playback and then AVCCAM player is dropping frames to keep from loosing sync with maybe the audio (or glitching).

In a couple months Panasonic will be releasing the Codec everyone has been waiting for on OSX so you can just use QuickTime to view the files without having to rewrap or transcode. That will be SOOOOOOO nice.
 
You must log in or register to reply here.
Back
Top