GH2 - solutions for the band problem...

It looks like a scan line as a retult of the sync with the light source. If it's 60Hz, try multiples of 60Hz shutter, and other numbers to see if it goes away.

I have a feeling that bulbs with high refresh rates (20KHz florescents) you won't see this. Nor with sun.
 
Which brings up a good point. A good test might be to underexpose at night not near ANY human made light sources. At least, this way you can see if in fact it is a light modulation issue or not.

Of course, if you cannot duplicate it, that makes it hard to test! :)
 
polfilmblog said:
It looks like a scan line as a retult of the sync with the light source. If it's 60Hz, try multiples of 60Hz shutter, and other numbers to see if it goes away.

I have a feeling that bulbs with high refresh rates (20KHz florescents) you won't see this. Nor with sun.
Click to expand...

As I've stated in the comments on that video (the original vimeo video is mine) I've done this test in natural lighting, so unless you can suggest a shutter speed that doesn't cause flicker issues with THE SUN, then I state again, that this is reproduceable at any of the shutter speeds I shot at in the video, in any dimmer lit scenes. You can go and watch my other video that I shot, which isn't of a wall, and you will definitely see it in various scenes (http://vimeo.com/17369205)

And yes, in well lit scenes, the majority of what you'll all shoot, its next to unnoticabe. But I shoot a lot of wedding videography, and need to be able to push to higher ISOs and lower lighting, so its why I tested it to these extremes.
 
Do you have an idea - why it is not seen in some videos from the GH2. For example - I don't see it in the videos of PB or that one here:

http://vimeo.com/17464042

Is it only a problem of some cams and not a general problem?
 
memphishills said:
As I've stated in the comments on that video (the original vimeo video is mine) I've done this test in natural lighting, so unless you can suggest a shutter speed that doesn't cause flicker issues with THE SUN, then I state again, that this is reproduceable at any of the shutter speeds I shot at in the video, in any dimmer lit scenes. You can go and watch my other video that I shot, which isn't of a wall, and you will definitely see it in various scenes (http://vimeo.com/17369205)

And yes, in well lit scenes, the majority of what you'll all shoot, its next to unnoticabe. But I shoot a lot of wedding videography, and need to be able to push to higher ISOs and lower lighting, so its why I tested it to these extremes.
Click to expand...

Ahhh! I didn't see that you had tested it in the sun, well, I guess there goes that. :) That is very bizarre. Does the line remain exactly the same as you change the light intensity slightly as well?

I guess the next question is, do you know anyone else with a GH2, you could test both of them and at least see if the results are slightly different, non-existant, or the same.
 
Joho said:
Do you have an idea - why it is not seen in some videos from the GH2. For example - I don't see it in the videos of PB or that one here:

http://vimeo.com/17464042

Is it only a problem of some cams and not a general problem?
Click to expand...
That's a good question. And yes...that video is very clean with no artifacts. Could this be the issue that delayed Panasonic from releasing the cam on time maybe??!!
 
Dont forget PB's video's were shot with a fantastic f/0.95 lense....I get no banding with the GH1 with my cheap senko f0.95 cctv lense...like the gh1... I theory it could be a sensor/exposure conflict.
That damn kit lense was a real bugger with making banding worse.

I doubt Panasonic delayed the GH2 to "fix this" they just had delays because they just didnt make enough to meet preorder demand.
Its already been said that a UK model had the banding also...and theyve only just been released here in small quantities in blighty....were usually the last to get these things.
Looks to me like they just made the same boo boo as they did with GH1...bless em!
 
Last edited:
I've personally tried two separate US GH2 bodies and the lines appeared in the exact same locations, at the exact same shutter speeds on both bodies. I'm currently waiting for a 3rd body to test, which appears to be coming from a new shipment from Panasonic, so we'll see.

The line stays the same, and its location is related to shutter speed, its just not really noticeable unless the video is 1-3 stops underexposed, its a non-issue in a correct or slightly overexposed shot.
 
I was just reminded that you got your first body from Henry's. Was your second body from the same place?
 
Last edited:
Camera Expert said:
I was just reminded that you got your first body from Henry's. Was your second body from the same place?
Click to expand...

Yes my second one was from Henrys as well.

Ian-T said:
memphishills....I'm finding that hard to believe.
Click to expand...

Full disclosure here, since at this point it seems necessary, I am an employee of Henrys Camera, and as such pulled a second unit from another store in my area immediately once I noticed the issue. Had both cameras side by side at home, and filmed the exact same scenes. Banding, and its visibility were identical with each camera. The lines happened in the same places at the same shutter speeds.

I really don't know what you want from me guys, I'm a professional photographer, and I've been working in a camera store for 5 years. I'm quite capable of fulling testing and diagnosing issues, and I promise nothing I managed to produce was on account of user error. There is definitely a bigger issue in this camera, whether it's hardware of software related, well I'm not an engineer, so I couldn't really say.

I'm open to whatever tests or suggestions you have, but just trust me when I say that I know what I'm doing.
 
Please correct me if I'm off base here but maybe Ian-T was commenting on your handle, mate. Not the results you achieved from the cam.

Edit: or maybe I'm an idiot.... or maybe both
 
Last edited:
....just finding it hard to believe he had access to that many GH-2's. But it now makes sense that he works at a camera shop.

I'm just trying to find where to buy one...
 
memphishills said:
Yes my second one was from Henrys as well.



Full disclosure here, since at this point it seems necessary, I am an employee of Henrys Camera, and as such pulled a second unit from another store in my area immediately once I noticed the issue. Had both cameras side by side at home, and filmed the exact same scenes. Banding, and its visibility were identical with each camera. The lines happened in the same places at the same shutter speeds.

I really don't know what you want from me guys, I'm a professional photographer, and I've been working in a camera store for 5 years. I'm quite capable of fulling testing and diagnosing issues, and I promise nothing I managed to produce was on account of user error. There is definitely a bigger issue in this camera, whether it's hardware of software related, well I'm not an engineer, so I couldn't really say.

I'm open to whatever tests or suggestions you have, but just trust me when I say that I know what I'm doing.
Click to expand...

Do you have access to an HD television? Might be interesting to see if you see it there as well before recording, or, during playback. Also, any settings affect it as well like the noise and such?
 
grimepoch said:
Do you have access to an HD television? Might be interesting to see if you see it there as well before recording, or, during playback. Also, any settings affect it as well like the noise and such?
Click to expand...

I tried it through hdmi output (computer and television) and yes it is visible. I've tried multiple color profiles on camera, as well as adjusting NR and such on a per film mode basis, no change.

It should be noted, that you never see the band before recording, never, it only shows up once you hit that record button, and once you watch the clip.
 
memphishills said:
I tried it through hdmi output (computer and television) and yes it is visible. I've tried multiple color profiles on camera, as well as adjusting NR and such on a per film mode basis, no change.

It should be noted, that you never see the band before recording, never, it only shows up once you hit that record button, and once you watch the clip.
Click to expand...

Interesting. Well if you don't see it before record but only after you hit record, this lends me to believe it is not a sensor problem (which I think is at least one good thing). When you press stop does it go away also after you have recorded? (meaning in the monitored image shown on the HDMI, not whatever was recorded)

What baffles me is that shutter speed just controls how much delay there is between when the line is reset and when it is read, which since the line moves downward as you make the shutter speed longer, then it appears that the line timing is consistent roughly from the start time of recording the rest and when you record the pixels of that row. And going back, I swear I can still see faint lines that are VERY similar above and below the line. Given when you move to 1/30, this pushes that line right out of the frame based on the other results.

Do you have some sort of color chart you can shoot and see if it's just the line going darker or if there is some sort of chromatic change. Looks to me like there is some sort of firmware bug. Maybe the reset code is a little off and hitting a few more lines relative to the shutter speed and since the delta of when they are hit is SO small, you only see it on highly underexposed over gained footage. That would be my guess anyways.
 
grimepoch said:
Interesting. Well if you don't see it before record but only after you hit record, this lends me to believe it is not a sensor problem (which I think is at least one good thing). When you press stop does it go away also after you have recorded? (meaning in the monitored image shown on the HDMI, not whatever was recorded)

What baffles me is that shutter speed just controls how much delay there is between when the line is reset and when it is read, which since the line moves downward as you make the shutter speed longer, then it appears that the line timing is consistent roughly from the start time of recording the rest and when you record the pixels of that row. And going back, I swear I can still see faint lines that are VERY similar above and below the line. Given when you move to 1/30, this pushes that line right out of the frame based on the other results.

Do you have some sort of color chart you can shoot and see if it's just the line going darker or if there is some sort of chromatic change. Looks to me like there is some sort of firmware bug. Maybe the reset code is a little off and hitting a few more lines relative to the shutter speed and since the delta of when they are hit is SO small, you only see it on highly underexposed over gained footage. That would be my guess anyways.
Click to expand...

To be honest, I was thinking the same thing, that it could be fixed in firmware, because it wasn't present in still images, and wouldn't show up before or after recording while monitoring. It could have something to do with the interpolation perhaps when its bringing the 16MP down to HD res, but again, I'm not an engineer, so its all just a guess for me.

The line goes away on monitoring after you stop, and then shows up again on record. It's definitely an on record-off record thing. From what I've been able to tell, I think it's just making things darker, but I'll see if I can get a color chart to test. Problem with that is, since it needs to be underexposed anyways to really see it, that's going to have an effect on color representation as well, so it might be hard to tell from a test like that. But I'll see if I can get something and record it for you.
 
Haha, well I am an electrical engineer and I work on CMOS microelectronics daily, and am a programmer, but that makes me no more knowledgeable about what's happening than you :) Just doing some basic troubleshooting methods to try and narrow down what we think it is. You'd think whatever method is reducing the image to display even when not recording would be the same. Given the HDMI out is before the compression, hence the reason people want to use it to avoid the AVCHD inside, something else is strange going on. From a processing standpoint I'd find it hard to believe they'd compress to AVCHD, then decompress and show it while recording, which of course would mean the quality of the output HDMI would change during record (although in this case it does anyways!). Unless of course the AVCHD encode and decode are handled by different parts of hardware that could run in parallel, but really, that just sounds like a lot more work for no gain.
 
You must log in or register to reply here.
Back
Top