Excessive GH1 FPN?

[mimirsan]

Lol - why would a burn in work? Think about it logically.... At any rate your post is very telling - the burn in worked for you... BUT.... "Also fpn vastly improved when I stopped using the slow kit f3.5/f4. lense for indoor/lowlight shooting. Shoot with a fast lense with correct exposure at lower iso's will give you clean result"

Ummm yeah.... the noise reduced... when you started using a faster lens. Then with your retrospectoscope you have put 2+2 together to reach 5.

I am firmly of the belief that all GH1's have this problem (which should be called banding noise - not fixed pattern noise.) I have personally seen less banding on my shots as time has gone on - but whaddya know? i have also been much more concious of what conditions create it and what lens to use in what situation! I can re create terrible banding with no bother at all.

If people really want to save a banded shot then I have recently been experimenting with device noise profiles in neat video - will put another post up about this soon.

Oh dear someones a grumpus....right...when I originally had my first GH1 a year and a half ago I had the kit lense...I had bad fpn shooting in the day indoors with this lense....I wanted some faster lenses.

I eventually sold the kit lense...bought some primes including the 20mm pancake...sure enough it helped with fpn but with higher isos were clearly visable.
about four months ago I bought the 14-45mm cheap...not a fast lense 3.5...indoors in the day? no fpn...iso 800 a little. I can see that over time there has been some kind of improvement. This from shooting general home videos in my badly lit house not from setting up shots to avoid fpn/banding. Dont get opportunity to do that with a little one.

To be clear Im not saying burning in works...I have my doubts...but I have a fpn riddled cam here cooking right now to find out in a few days.

Im sure if it did work you will call it fake anyway.
Me im treating it as a fun experiment...and trying to help out someone. Nothing wrong with that.

If it doesnt work then I call it a myth and say well it must be my eyes getting old....that or my gh1s are magical.
All I know is right now is that my old chinese made GH has no fpn /banding below 800 iso...when it was new it did....even with fast lenses.

 

[mimirsan]

@framerate
park cameras yes? saw them for £299...Good bargain...good company. I grabbed one from a currys digital auction site £199...Im too like "do I need it? but it was so friggon cheap!" hahah

Yeah bad QC by panasonic and from the looks of things theyre still at with GH2.

 

[Framerate]

@framerate
park cameras yes? saw them for £299...Good bargain...good company. I grabbed one from a currys digital auction site £199...Im too like "do I need it? but it was so friggon cheap!" hahah.

No, from Harrison Cameras. The first one @ £249 came as a tip-off in here. They had a few bodies from Panasonic UK and I was lucky enough to get one. At the time I'd sold my previous GH1 and had a GH2 on order, but £249 was irresistible for new hackable model.

I was very surprised when it proved to be a good 'un - so when they had a few more back in over Christmas I bought the second one. Speaking to then on the phone last week, they mentioned that they'd sold GH2's at launch for £625 body-only - which was even cheaper than Mathers.

Yeah bad QC by panasonic and from the looks of things theyre still at with GH2.

Well, my own theory is that there is a sensor problem that Panasonic know about, and are trying to fix - but that the lure of Christmas sales made them ship some units before they had been modded.

I know that they told 'kleinkleinklein' that no no fault was acknowledged and no fix was planned - but I still wouldn't be surprised if future GH2 batches turned out to be mysteriously better

I base my reasoning on the fact that Vitaly spoke of rumoured sensor production issues in his '43 Rumors' interview last year - which was then followed by widely publicised, but undefined, rumours of a delay in shipping. When they turned up on time it was tempting to think that it had all been smoke & mirrors - except that banding reared its head (once again) in the first retail units, which seemed to reinforce the idea that *something* wasn't quite right about it all.

But, again, I emphasise that's only my theory. I do want a GH2 (unless Sony come up with the rumoured Nex7) but I decided to wait a bit and see if later production models solved some of the reported issues.

 

[the doc]

Well - for whatever reason the banding exists (and I personally go along with this excellent explanation...)

http://www.eoshd.com/content.php?175-Panasonic-recognising-GH1-banding-issue

i.e. electrical interferance with the sensor from other components and something inherent to CMOS unfortunately.

The only fix that I have found so far is the neat video plugin. It is not perfect - but it helps. I have posted my findings here.

http://www.dvxuser.com/V6/showthread.php?234142-Neat-video-for-GH1&p=2209376#post2209376

 

[nathankw]

I have a friend who recently bought a gh1 (ironically WE0A) and hes just popped round with it...sure enough bad FPN...Sooo Ive lent him one of mine and with his agreement I am going to try the burn in.:

Any chance of shooting some comparison shots - and finally we could get some real evidence of variation and/or burn in.

Be great to have the same scene with all the settings identical on the good and bad cameras. Then the same scene after the burn in.

Despite all the discussion on banding/FPN, I've still never seen a clear side by side comparison showing the same shot on two GH1s with significantly different banding.

 

[mimirsan]

Any chance of shooting some comparison shots - and finally we could get some real evidence of variation and/or burn in.

Be great to have the same scene with all the settings identical on the good and bad cameras. Then the same scene after the burn in.

Despite all the discussion on banding/FPN, I've still never seen a clear side by side comparison showing the same shot on two GH1s with significantly different banding.

Thats what im currently doing ;-) Ive shot pre burn in process...the cams going through the cook right now lol will post comparison in a few days. ;-)

 

[Matthew P]

Framerate, could you perhaps post some samples of your GH1? I'd really like to see how a good one compares to a bad one. Just a short underexposed clip of a plain wall or something?

 

[Framerate]

Framerate, could you perhaps post some samples of your GH1? I'd really like to see how a good one compares to a bad one. Just a short underexposed clip of a plain wall or something?

This link is to samples i posted a couple of months ago http://vimeo.com/16894125

To be honest, I didn't think that the sample you provided was that bad - I've seen much worse. The banding on the still image is nasty - but the video looked quite reasonable.

 

[mimirsan]

I may have encouraging news...today I shot same conditions same settings too see if anything has changed. The banding is still there but I can say is less pronounced!

Cameras back on burn... Im going to shoot again tomorrow again same conditions settings etc

HOPEFULLY the final demo showing a gradual difference each day...tbh I dont think 2 days will be enough so going for the weekend....if theres some improvement even minimally
then we could well say that over a long period of time you will see an major improvement (like my year and half old gh)

BTW
This cam im testing here has bloody awful fpn and banding...one band literally looks like a dark thick transparent bar over the right side.

No vids yet you will see when I up the final experiment.

 

[Matthew P]

Thanks for the info mimirsan. Glad you're having success!

I tried the burn in myself, but it did absolutely nothing to reduce the vertical streaks (FPN). I've also been experimenting with photos, and high ISO performance is TERRIBLE compared to some guy's GH1 on DP Review (which HE himself was complaining about!). It's like my ISO1600 or even ISO800 is what ISO3200 should be.

I also did some filming today for an internet series, and the lines were visible ALL OVER the skin tones! Terrible, really. I did get this for £260 which was great, but this FPN is just driving me nuts. I've contacted the seller on ebay, and he's quite happy for me to return it. I'm on the hunt for a GOOD GH1 now. Very hard to do on ebay, I reckon.

 

[Matthew P]

This link is to samples i posted a couple of months ago http://vimeo.com/16894125

To be honest, I didn't think that the sample you provided was that bad - I've seen much worse. The banding on the still image is nasty - but the video looked quite reasonable.

It's quite hard to see FPN on Vimeo.

If you say that mine isn't too bad (the permanent vertical streaks; FPN), does that mean that all GH1s suffer from FPN on people's faces and such? I'd be very surprised if that's so, because I've not heard many people complain about it. If you say, recorded someone's face, going down from properly exposed to under exposed, would you see any FPN at all on ISO 100? Not banding, mind, just the unmoving vertical streaks.

 

[the doc]

Not heard many people complain about it? There are multiple threads - some of them many many pages long.

 

[Framerate]

It's quite hard to see FPN on Vimeo.

If you say that mine isn't too bad (the permanent vertical streaks; FPN), does that mean that all GH1s suffer from FPN on people's faces and such? I'd be very surprised if that's so, because I've not heard many people complain about it. If you say, recorded someone's face, going down from properly exposed to under exposed, would you see any FPN at all on ISO 100? Not banding, mind, just the unmoving vertical streaks.

Hi. I downloaded your street shot, and although there was some digital noise, I wasn't aware of any vertical banding. I did see VB on the still image, and it was pretty bad.


To answer your second question; No, absolutely not. If you are getting that (my own inexpert eyes notwithstanding) then you do seem to have a problem.

 

[mimirsan]

I have finished my burn in test...really took a risk as left the cap off for 10 hours leaving the sensor exposed (72 hours total)...In all honesty I dont recommend doing it but it worked for me.
Will link video tonight.
My friends gonna love me. Or hate me if his GH1s sensor dies over a week or so lol

 

[mimirsan]

Here you this is what this is... make your own conclusions. You can still see a little bit of banding (right side area) left. But you can see it has cleared most of that big band.
Watch full screen to see the banding better (cant miss that big band on the right anyway).

Camera was a red WEOA serial.

 

[mojo43]

Thanks for doing the test. The after shot looks a lot brighter. Was the lighting and shutter and ISO exactly the same before and after?

 

[mimirsan]

exactly the same...same time of the day 8pm same lighting conditions,same iso,same shutter (started each record at 30 then two clicks to 50) all at f1.7 and 4000k wb

 

[mojo43]

exactly the same...same time of the day 8pm same lighting conditions,same iso,same shutter (started each record at 30 then two clicks to 50) all at f1.7 and 4000k wb

Wow, then there is quite a difference! That's really great and strange....

 

[PJF]

I have finished my burn in test...really took a risk as left the cap off for 10 hours leaving the sensor exposed...

Are you saying you are concerned because you left the sensor physically exposed (body cap off) for 10 hours, or because you left the sensor exposed to light (lens cap off) for 10 hours? I presume it's the former.

This test is very encouraging. It did cross my mind that burning a sensor in with light might be more productive than leaving it in the dark. It fitted in with user reports of GH1s improving with time.

I thank you for this test, and so do the family of silverfish who are enjoying their new home and currently tunneling a new access through that large flat plate with the pretty, multi-coloured surface.

 

[Ian-T]

Well...I'll be a monkey's uncle..!! That's the first time I've seen a before and after and...YES....something definitely changed. There was a particular vertical band that I kept my eyes on till the end...and by the time 72hrs was notated it was barely there. I think leaving the cap off in the light actually did the trick.

It's funny how some folks ridicule the idea of the "burn-in" trick...but in this case something you did made the resulting image change within 72 hrs. I didn't completely get rid of it...but it wasn't nearly as bad.

Now...the question for those GH-2 owners out there is...are you willing to try this to see if it works on that "one" annoying band at high ISO's?