GH1 Banding Comparison... help me evaluate

G

GrgurMG

Active member
Okay, I'm aware all GH1s and most DSLRS for that matter, experience some level of banding. Some say all GH1s are about the same, others claim a bit of variance. It seems that there's at least some rare cases where a GH1 has been found to be operating out of spec in terms of banding and so forth.

The problem I am having is with being absolutely sure mine falls in line among the norm when it comes to the intensity of banding. My best thought for judging banding degree on other GH1s would be to ask the following...

With the lens cap ON with the dial set to movie mode..
Exposure Compensation set to Max (+3)..
Shutter Priority Set ( 1/30, though any should suffice)...
and ISO set to 1600...

... I get visible banding on a the black screen (slighty more visible on the right hand side I think). Even when hitting record and the picture filtering/NR process kicks in I still see some banded noise and its recorded to the clip. My question for anyone who cares to help is... do you also get banding on black in those settings and furthermore is it the same intensity? Here is a link to a sample clip (zipped) of my results... recorded at MJPEG...

http://www.jerseyforged.com/sampleclips/banding_1600iso_maxexpcomp.zip

Please take careful note that I'm not saying I think I should be shooting at those settings.. I'm simply using it as a theoretical basis for comparison. All I'm really looking for is some people to chime in with "Yup dude.. mine looks exactly the same" or "Be happy dude.. mines worse" or "Eh.. it's a little worse then mine" etc etc. I am fairly happy with my GH1, but since I typically have to shoot in scenerios with very limited lighting... I'd like to know if their may be fruition in exchanging my GH1 for another. If anyone gets different results... no matter how slightly.. I'd be interested to know.

Also, at still mode "S".. when the picture is obviously being gained up for preview visual.. I get really exteme crazy red banding/noise... is that the norm as well?
 
Last edited:
your link doesnt work for me.

just tested and I get a decent amount of magenta banding. with exposure compensation 0 it is still there but not as bad more dark blue than red. I think if you are hunting for it you will find it. a better test would be to set the iso at a more usable level say 800 to see if one gh1 is more prone to banding than the other? i dunno. maybe everyone shoot on standard profile too? or smooth/ nostalgic?
 
Last edited:
Wow.. thanks for letting me know about the broken link. Made a mistake while copying the post over... fixed it so the link should work now.

I went with a high ISO setting because I felt it might provide the best baseline. My thinking was something more visible would be easier to compare... this is assuming the banding phenomenom works linearly like that.

Your magenta banding... that was on the "S" dial setting without recording or during recording? My footage is with the dial to movie mode and shutter priority set in the menu. Even with the dial set to "S".. once I start recording it pretty much goes back to what I get in the movie mode. And yeah.. I was shooting normal color as well. Take a look at the clip if you have the time.

Thank again for the info.
 
Last edited:
You do realize that you're shooting in almost complete darkness? Are you getting banding in normal shooting conditions? A proper exposed image would be a good baseline to measure against...
 
GrgurMG said:
Wow.. thanks for letting me know about the broken link. Made a mistake while copying the post over... fixed it so the link should work now.

I went with a high ISO setting because I felt it might provide the best baseline. My thinking was something more visible would be easier to compare... this is assuming the banding phenomenom works linearly like that.

Your magenta banding... that was on the "S" dial setting without recording or during recording? My footage is with the dial to movie mode and shutter priority set in the menu. Even with the dial set to "S".. once I start recording it pretty much goes back to what I get in the movie mode. And yeah.. I was shooting normal color as well. Take a look at the clip if you have the time.

Thank again for the info.
Click to expand...


I shot in creative movie shutter priorty as per your settings. Your clip is what my banding looks like with exposure at 0. When I up the exposure comp to 3 i get far more visible red banding. does that mean mine suffers more banding...maybe.

I agree that a better test would be on a subject rather than just pitch black but we would all have to be in the same spot shooting the same thing..ha.
 
Isaac_Brody said:
You do realize that you're shooting in almost complete darkness? Are you getting banding in normal shooting conditions? A proper exposed image would be a good baseline to measure against...
Click to expand...

Not almost complete darkness... completely complete darkness ;) Lens cap on.

Again, the point is to test something consistent, not try and get a good picture. It's not a matter of not being able to get a good picture on the my camera at all.. it's a matter of discerning the level of banding on my camera compared to others in order to identify if I should try exchanging it or not in attempt to get the most for my dollar.

As MR Fanny pointed out.. it's kinda of hard to have a bunch of people shoot the exact same thing with the exact same lighting conditions... that's more then I'd ask people to go out of their way anyway. Im going on the assumption that if in pitch black at worste settings, one person gets more banding then another.. then the results will scale so in better lighting at better settings, if banding is still present it will be worse on the same one as when on black. I could be wrong about this testing methodology... but it was the easiest possibility for precise comparison.

And yes, I'm getting some banding in certain normal (though perhaps not ideal) shooting conditions at lower ISOs.


MR Fanny said:
When I up the exposure comp to 3 i get far more visible red banding. does that mean mine suffers more banding...maybe.
Click to expand...

That is interesting. My GH1 is recently purchased so perhaps I might have one of the supposed better models (newer factory or whatever the myth is). It's interesting that there's still some level of uncertainty from many users as to whether differences in GH1s exist... because as someone mentioned on another thread... no one has ever shot two side by side on the same settings... or at least no ones ever posted said results.

Not to toot my own horn, but this little lens cap test thing is actually the most concise and methodical attempt at comparison that I could find on a forum.... everything else I could find are just clips of footage that are nothing like one another (even if the same settings.. surely not the same thing being shot with the same EXACT lighting).

Also, for those who claim the banding lessens over time... wouldn't shooting with the cap on be the best way to retest frequently to compare results over time? Again, perhaps the type of sensor system imperfections that cause banding are a bit more complex and variable then I give them credit for and results on black might not scale perfectly with results in other shooting situations.. but I haven't heard anyone infer otherwise either.
 
Last edited:
I see, but then perhaps it might make sense to do this test at all the available iso settings to see if banding is better or worse at different settings. This was done on the 7D/5D which lead to the discovery of the odd iso settings being less noisy. Have you tested with lens cap on for all iso's? If I had camera I would try this in all the available shooting modes, and with NR turned on and off.
 
Nope, there is definately lighter toned noise/banding in there (particularly to the right)... though concievably people with different monitor/player settings might have their blacks cutoff sooner. I'm viewing in Quicktime on a Windows PC... fairly regular monitor display settings, I believe.
 
I had to turn brightness and contrast up very high to see anything but black in your video. Once I did, the banding patterns looked typical to me. If you're seeing banding in this video during normal playback, you might want to check the calibration of your monitor.
 
I think my monitor settings are fine. The results I see are what I get with the newest version of Quicktime on my PC. Windows Media Player and a few other players do indeed get more of a pitch black though.

I see that even this attempt at comparison is a bit troublesome. I might try heading out to the camera shop tomorrow and seeing about comparing mine with their floor model.
 
Using QuickTime Player on a PC to play MOV files is notorious for showing elevated Gamma/Black levels which is not correct. Windows Media Player will play the MOV back at the correct Gamma/Black levels. Having said that, the banding you are seeing looks completely normal to me.

You can actually "fix" the specific MOV file to play correct Gamma levels with QuickTime Player by going to the top menu bar and selecting "Window" and choosing "Show Movie Properties" to open up a new window.

Highlight "Video Track" then click on the "Visual Settings" tab. Now look down and find the "Transparency" area with a drop-down selection menu. From those choices, select "Straight Alpha" and close the window. Go back to the Top Menu Bar under "File" and select "Save As" to save it as a new file. This will not recompress the video file at all so it is not altering it in any way.

Now play the video with QuickTime Player and you should have the proper Gamma/Black level during playback. If you now use Windows Media Player it will look the same as the original file since nothing was altered in the video and it wasn't recompressed.

Randy
 
EDIT:

It's 1280x720 jpeg. Two layer of compression on a low level signal, and there are possible gamma isses with the conversion and the players. On VLC is shows pitch dark. On Media Player and Quicktime, noise is visible on the right.

A single lightly compressed frame from the original file is enough for seeing sensor noise.

By boosting the image, I see compression artifacts, and horizontal lines moving vertically which are electronic interference.

If there was any banding, it would be so low that compression wouldn't be able to represent it. The PSNR is not high enough to judge banding with the lens cap on.

Try shooting a grey wall at ISO400, 800, 1600 at 1/60 shutter and without exposure compensation.
 
Last edited:
GrgurMG said:
Nope, there is definately lighter toned noise/banding in there (particularly to the right)... though concievably people with different monitor/player settings might have their blacks cutoff sooner. I'm viewing in Quicktime on a Windows PC... fairly regular monitor display settings, I believe.
Click to expand...

ok- yup see it now, left side looks smooth- right 2/3's of screen definite noise banding there. Looks completely 'normal' for gh1.
 
I would like to thank everyone for their input on this matter.

After closely considering the opinions of everyone here as well as doing some extensive testing and comparison today, I've decided my GH1 certainly does seem to fall well enough into spec and I won't be bothering exchanging it for another.

Like I've mentioned, I just had to be sure I was getting the most for my dollar, but I'm very impressed with the GH1 in general. The higher degree of banding in some shooting conditions is an unfortunate querk, but it's a trade off for it having excellent noise characterstics... not to mention a great video feature set. I was running some test comparisons between my GH1 at 1600 ISO and an old FX1 and the Sony needed pretty much no gain to stay under the GH1s noise levels.. although in poorer lighting that again was offset by some managable banding. I know some people downplay the banding, but it hurts usability in some lighting/exposure situations and furthermore seems inconsistent on some models, so I feel safe enough to be amonst those who view it as a small defect (not a normal characterstic) and really it seems to be just a matter of some refinement Panasonic did not put into it... which will probably go into the GH2.

Many of the threads I've read have discussed only using low ISOs when shooting video to avoid banding, however after running various tests, I'm not so sure it's as simple/linear as that. Banding occurs at certain exposure situations in low detail areas of a scene and it seems to me that there are some (perhaps rarer) situations where bumping up the ISO could even minimize the appearance of banding... or at least certainly brighten up the picture without really increasing it's appearance further. Though of course there's other situations where things are quite the opposite. Playing with the exposure compensation will often effect banding far more then ISO setting and again, the noise that comes thru due to high ISO puts many camcorders to shame. It's a really querky sort of thing.

In any case, thanks again. I'll stay very happy with the GH1.. at least until I trade it in for a GH2. ;)
 
Last edited:
You must log in or register to reply here.
Back
Top