E-330 horizontal banding and lackadaisical Oly support
- Thread starter cjtune
- Start date
- Status
Wait.. there is something that's been improved, though the banding severity hasn't been fixed.
I'm noticing that after awhile, the coarse bands sort of dissipate into finer, thin, bands that are less noticeable, regardless of whether liveview was on or not. But still an eyesore, just less of an eyesore.
All pix below at ISO1600, EV-1 smart metering:
At time=0 sec.
At time=20 sec.
At time=60 sec.
After turning off the camera and letting it rest like 2-3 minutes, the picture quality at high ISO reset back to that of the picture 'at time=0 sec'.
I'm noticing that after awhile, the coarse bands sort of dissipate into finer, thin, bands that are less noticeable, regardless of whether liveview was on or not. But still an eyesore, just less of an eyesore.
All pix below at ISO1600, EV-1 smart metering:
At time=0 sec.
At time=20 sec.
At time=60 sec.
After turning off the camera and letting it rest like 2-3 minutes, the picture quality at high ISO reset back to that of the picture 'at time=0 sec'.
Maybe I need litigation this time... not luck.
That is if you leave it hanging around you neck often. What if you're one of the careful kind like me who likes to tuck away the camera in the bag?
What if you had an opportune moment and have no choice but to use high ISO to capture it? Can you wait 60 secs? "Hold on, hold on everybody! My camera's not warmed-up yet!"
Hi there, just read this thread.
To confirm, Raw has the banding or not? JPEG may have sharpening+contrast applied that's why it's more obvious. The banding can be any angle depending on the frequency of that noise. Try tuning your TV to encoded cable channels and you will get my meaning.
If it's always present...
Banding is always because of interference, which means the noise comes regularly from somewhere between the sensor to the ADC, or caused by a part near the sensor giving out too much radiation (esp high speed components).
If it's always there it could be electrical noise, but you said the camera needs to warm up. If it needs to warm up, it could be that some capacitor-type of part inside is either out-of-spec/faulty or not connected/soldered properly.
To add a bit, the Sony Alpha also has banding issues when shooting ISO1600. It leads me to think that having high ISO picks up all sorts of noise not just visible. Must remember that sensors not only see visible light, some can see IR some can see UV.
LOL if it's out of wty u might try to slip in a piece of antistatic wrapper from a hard disk around the back of the NMOS if possible. lol.
If it's only JPEG... then it's really a bug in the "engine". Something simple like putting an extra 00 can cause banding. There was an old bug in Netscape that added regularly carriage returns to mp3 files causing regular blips. Not related but just to show you the idea.
Sorry I can't help except to give more info as far as I know. I really empathise with you, when the device is a lemon by design or poor QC. Writing a personal email higher up can help. Or maybe try Phil Askey. But my experience with Olympus is that they just replace the whole PCB, no real diagnostics done.
To confirm, Raw has the banding or not? JPEG may have sharpening+contrast applied that's why it's more obvious. The banding can be any angle depending on the frequency of that noise. Try tuning your TV to encoded cable channels and you will get my meaning.
If it's always present...
Banding is always because of interference, which means the noise comes regularly from somewhere between the sensor to the ADC, or caused by a part near the sensor giving out too much radiation (esp high speed components).
If it's always there it could be electrical noise, but you said the camera needs to warm up. If it needs to warm up, it could be that some capacitor-type of part inside is either out-of-spec/faulty or not connected/soldered properly.
To add a bit, the Sony Alpha also has banding issues when shooting ISO1600. It leads me to think that having high ISO picks up all sorts of noise not just visible. Must remember that sensors not only see visible light, some can see IR some can see UV.
LOL if it's out of wty u might try to slip in a piece of antistatic wrapper from a hard disk around the back of the NMOS if possible. lol.
If it's only JPEG... then it's really a bug in the "engine". Something simple like putting an extra 00 can cause banding. There was an old bug in Netscape that added regularly carriage returns to mp3 files causing regular blips. Not related but just to show you the idea.
Sorry I can't help except to give more info as far as I know. I really empathise with you, when the device is a lemon by design or poor QC. Writing a personal email higher up can help. Or maybe try Phil Askey. But my experience with Olympus is that they just replace the whole PCB, no real diagnostics done.
Yup, it's also present in the ORF (RAW) format files as previously mentioned. Another member NMSS_2 and I have inspected the ORFs and verified that they are still there.
Possibly. Something within my E-330's body that hasn't been replaced. Otherwise, why is the loaned E-330 unit much better? Interference from lenses also can be ruled out, I guess, as it appears when using my 14-45, 14-54, 35macro, 40-150.
Highly suspected... that's why I find the old story about the old Nikon DSLR having banding quite intriguing.
I hope that the case of the Sony Alpha does not lower the industry standard and one day everyone insists (well, maybe to the exception of Canon) that banded noise is 'normal', and all high ISO pics look like bad TV reception.
I'm getting crazy enough to try it... but note also that interference also travels by conduction, and it can be more significant compared to radiation EM intereference.
That would be a really simple fix to the FW... but alas I doubt it is. Besides, JPEG compression algorithms are so mature and commonplace I doubt that what's used in the FW was developed in-house or from scratch by Olympus.
My feeling is that it's already a small lemon by design. Mild banding is widely reported. I can live with what I've seen of those mild banding.
My unit (and that of Blu-by-U's) is a big lemon. It's banding is far from mild, and I'm surprised it didn't got caught at the factory QC level. Maybe they should not be taking high ISO test shots at bright areas....
From my own experience with product development, at the field level, they'll just diagnose to the level of a whole subsystem that's suspected of being faulty and then they'll swap out the whole subsystem and retest for the defect to reoccur. Detailed diagnosis is left to engineers who may be 2nd or 3rd generation in the that product's development... and if it's escalated enough to seem serious. Usually, there's a detailed step-by-step diagnosis and repair guides prepared for both field repair and factory rework (ie. repair units that failed QC to salvage yield).
Blu-By-U
Senior Member
It cannot be the JPEG compression algorithms. It appears also in the RAW files. Think we should ask Tomcat to try it on his L1. I am suspecting the NMOS. Both E-330 and the L1 share the same sensor. If it is the sensor, then tough luck...it's gonna be in all E-330. Where is that Uncle Tomcat??
Just PM both Uncle Tomcat (L1 and E-300) and Olyflyer (E-500)
Just PM both Uncle Tomcat (L1 and E-300) and Olyflyer (E-500)
That's great! Good to know they've solved the pre-production L1's banding problems.
Not sure if Olympus managed to silent-roll in some fixes on the later production E-330s too.
I want one of those. I should have gotten one of those.
Silly early adopter I am.
Hi,
On a special request from Blu-By-U I did a test of my E-500. After taking and analyzing 11 SHQ (3264x2448) images I can safely say there is no banding issue with my camera. I think that is a E-330 problem. I have no time to check all that is written on the subject, maybe others has come to the same conclusion. I think it is a problem with the sensor. Maybe that is why Oly went back to Kodak for the E-400. I think the E-330 sensor is a Panasonic NMOS type. Maybe Pana folks could confirm if that is a generic problem or just E-330 specific.
Anyway, during these experiments I discovered a lens issue. I will start a new thread on that one because it is OT.
On a special request from Blu-By-U I did a test of my E-500. After taking and analyzing 11 SHQ (3264x2448) images I can safely say there is no banding issue with my camera. I think that is a E-330 problem. I have no time to check all that is written on the subject, maybe others has come to the same conclusion. I think it is a problem with the sensor. Maybe that is why Oly went back to Kodak for the E-400. I think the E-330 sensor is a Panasonic NMOS type. Maybe Pana folks could confirm if that is a generic problem or just E-330 specific.
Anyway, during these experiments I discovered a lens issue. I will start a new thread on that one because it is OT.
You're welcome.
Sorry, no I had 0 so I guess I'll do it again but it won't be before saturday.
What is the issue you have with the 14-45? Please comment my thread on my 14-45 issue.
Blu-By-U
Senior Member
Sometime back when I first got the lens, I mentioned that the lens's zoom ring was just nice and the 40~150 a bit tight. Now the ring is way too loose. In fact it can actually move even when it's focusing.
tomcat
Senior Member
Personally, I don't see the point in shooting at lower than optimal EV. Why would anyone shoot at -1 or -2 EV at ISO 1600 and still expect the outcome to be good ? But since Blu asked, here is the same scene at -1 EV.... very under-exposed and noisy :what:
Hi tomcat. I purposely underexpose to gain shutter speed and the trade-off is the loss of brightness but not to the point of getting shadow pixels. As long as I don't get shadow pixels, I can still photoshop the shot back to health. Blurness from slower shutter speeds, is harder to fix though.
You must ask, EV -1 with what sort of metering? Smart metering? Centre-weighted? Spot? So, proper exposure will depend on your area of interest within a shot, and the exposure of everything else will be relative to it, over or under-exposed. The problem with banding is, even with proper exposure for your subject, there are bound to be some shadowy areas where the bands will turn up and spoil the whole picture.
I have read on some forums, and for their model of cameras affected by banding, that slight overexposure may help, and at the risk of highlighting pixels beyond any recoverable detail. Plus this kind of defeats purposeful underexposure to gain shutter speed. Even at ISO1600.
Blu-By-U
Senior Member
Uncle tomcat, thank you. the banding in my E-330 comes only when it's -EV. when it's set to 0, not noticeable. anyway, me just helping cjtune on his problem.
I hardly shoot at such high ISO anyway. I personnely won't risk anyone fiddling inside the cam. I would rather change the cam. But the hot pixel, so far only 2 on my IR shots. If more and too troublesome to cleanup, then I most likely sell the cam and change to another one.
Now back to your take with the -ev. there is slight red groups on the background towards the right. around the cat's ears. (Is that only on my screen only??) Don't see any banding so Pano may have fixed the problem.
So cjtune, a Pano L1 for you or you rather wait for the Leica unit?? :think:
I hardly shoot at such high ISO anyway. I personnely won't risk anyone fiddling inside the cam. I would rather change the cam. But the hot pixel, so far only 2 on my IR shots. If more and too troublesome to cleanup, then I most likely sell the cam and change to another one.
Now back to your take with the -ev. there is slight red groups on the background towards the right. around the cat's ears. (Is that only on my screen only??) Don't see any banding so Pano may have fixed the problem.
So cjtune, a Pano L1 for you or you rather wait for the Leica unit?? :think:
- Status
Similar threads
