OM-D EM-5 - High Iso Banding?

Started May 8, 2012 | Discussions
Gianluca Grossi
Forum MemberPosts: 84Gear list
Like?
Re: That's not what JPEG compression looks like (nt)
In reply to Vlad S, May 8, 2012

let's see the full res pics and then we could talk about it...

Vlad S wrote:

Gianluca Grossi wrote:

in low res pic you could have banding or other other artifact due to the lost information of jpeg compression.

l

 Gianluca Grossi's gear list:Gianluca Grossi's gear list
Olympus OM-D E-M5 Panasonic Lumix G 14mm F2.5 ASPH Panasonic Leica Summilux DG 25mm F1.4 Olympus M.Zuiko Digital 45mm 1:1.8 Samyang 7.5mm F3.5 UMC Fisheye MFT +1 more
Reply   Reply with quote   Complain
cbran
New MemberPosts: 23
Like?
Re: OM-D EM-5 - High Iso Banding?
In reply to cbran, May 8, 2012

cbran wrote:

I noticed something similar yesterday when I used the 20mm on my OM-D and your thread prompted me to investigate more deeply.

MY CONCLUSIONS:

At high ISOs (primarily around 6400 for me) with RAW files I see banding ONLY WITH THE 20mm f1.7 LENS . Sometimes there is no banding at all, but most of the time it's there.

It is quite variable, however, both in distribution across the frame and in spacing .

For example, sometimes the banding affects the entire image; sometimes only the top 50%; sometimes only the top 85%. When looking at the thickness/spacing of the banding with 100% crops, sometimes I see 3 bands per inch on my monitor, sometimes 5 bands per inch, sometimes 8 bands per inch, and sometimes 15-20 bands per inch.

In addition to testing the 20mm lens, I tested a manual focus legacy lens (Zeiss 35mm f1.4) plus 3 other Panasonic lenses: the 14mm f2.5, the 45mm f2.8, and the 25mm f1.4. I also tested the Olympus 45mm f1.8. NONE OF THESE OTHER LENSES EXHIBITED BANDING in my tests.

Could it be that the 20mm lens itself is causing some kind of interference pattern? I haven't tested the lens at high ISOs in any other room than my office, but the fact that I don't see banding with other lenses suggests to me that any interference is coming from the lens itself, not other equipment in the room, though perhaps there is some weird interaction going on.

I will be happy to show crops of sample pictures as soon as I prepare some but wanted to share my results in the meantime.

I'd love to hear what others think about this and see results from other people's tests.

In reference to my post above, here are some examples of the banding using the 20mm f1.7 lens on my OM-D at iso 6400 (from RAW), 100% crops. The first picture shows the more widely spaced bands I referred to; the second picture shows more tightly spaced bands (spacing on other samples can be even tighter). The third picture shows how the bands are sometimes distributed over only the top part of the image.

Reply   Reply with quote   Complain
toscha_seidel
Regular MemberPosts: 197
Like?
Re: Maybe the banding is caused by the sensor
In reply to JL Auch, May 8, 2012

Isn't there a strong electromagnet that suspends the sensor?

Reply   Reply with quote   Complain
Jonas B
Forum ProPosts: 14,596
Like?
A combination of 20/1.7 and E-M5 problem?
In reply to cbran, May 8, 2012

cbran wrote:

I noticed something similar yesterday when I used the 20mm on my OM-D and your thread prompted me to investigate more deeply.

MY CONCLUSIONS:

At high ISOs (primarily around 6400 for me) with RAW files I see banding ONLY WITH THE 20mm f1.7 LENS . Sometimes there is no banding at all, but most of the time it's there.

It is quite variable, however, both in distribution across the frame and in spacing .

Interesting indeed. Btw, I recall some Canon lens causing problems just this way.

Was the camera set to continues AF?

Jonas

Reply   Reply with quote   Complain
Lights
Veteran MemberPosts: 3,270Gear list
Like?
Re: OM-D EM-5 - High Iso Banding?
In reply to JL Auch, May 8, 2012

I have no idea on this, since I do not have this camera, but did just read this thread:
http://forums.dpreview.com/forums/read.asp?forum=1041&message=41456595
--

http://www.pbase.com/madlights
http://barriolson.aminus3.com/

Reply   Reply with quote   Complain
Bryan Biggers
Veteran MemberPosts: 3,511Gear list
Like?
Re: A combination of 20/1.7 and E-M5 problem?
In reply to Jonas B, May 8, 2012

OK, I had to try this, and cbran is right, the banding is much worse with the 20mm lens. I see almost nothing with the kit lens, but the 20mm causes banding even at ISO 6400. OK, now I have to try some more lenses...

 Bryan Biggers's gear list:Bryan Biggers's gear list
Olympus Tough TG-1 iHS Panasonic Lumix DMC-LX7 Olympus PEN E-P3 Olympus OM-D E-M5 Olympus PEN E-PM2
Reply   Reply with quote   Complain
JL Auch
Forum MemberPosts: 84
Like?
Re: OM-D EM-5 - High Iso Banding?
In reply to cbran, May 8, 2012

Okay. I have done extensive testing today to try and get to the bottom of this. I tested many variables. The best that I can come up with is this:

It seems to be related to the halogen light and/or electric box above the couch, that creates some sort of interference when the light is on. When I turn off the light, no banding with the 20mm. Also the banding seems most pronounced at 1/60sec and 1/30sec but it is still there at other shutter speeds.

Here is the kicker; I swapped lenses to the 14-42 kit lens, and bang! NO banding with the lights on at either 1/60sec or 1/30sec. The exact same lighting, tripod position, iso, and settings.

So I think cbran might be onto something. But if this is true, it is a real shame. The 20mm is a real gem of a lens and this is simply ridiculous.

I'll leave you to your own conclusions.

But what to do now? Return the 20mm lens?

-j

cbran wrote:

cbran wrote:

I noticed something similar yesterday when I used the 20mm on my OM-D and your thread prompted me to investigate more deeply.

MY CONCLUSIONS:

At high ISOs (primarily around 6400 for me) with RAW files I see banding ONLY WITH THE 20mm f1.7 LENS . Sometimes there is no banding at all, but most of the time it's there.

It is quite variable, however, both in distribution across the frame and in spacing .

For example, sometimes the banding affects the entire image; sometimes only the top 50%; sometimes only the top 85%. When looking at the thickness/spacing of the banding with 100% crops, sometimes I see 3 bands per inch on my monitor, sometimes 5 bands per inch, sometimes 8 bands per inch, and sometimes 15-20 bands per inch.

In addition to testing the 20mm lens, I tested a manual focus legacy lens (Zeiss 35mm f1.4) plus 3 other Panasonic lenses: the 14mm f2.5, the 45mm f2.8, and the 25mm f1.4. I also tested the Olympus 45mm f1.8. NONE OF THESE OTHER LENSES EXHIBITED BANDING in my tests.

Could it be that the 20mm lens itself is causing some kind of interference pattern? I haven't tested the lens at high ISOs in any other room than my office, but the fact that I don't see banding with other lenses suggests to me that any interference is coming from the lens itself, not other equipment in the room, though perhaps there is some weird interaction going on.

I will be happy to show crops of sample pictures as soon as I prepare some but wanted to share my results in the meantime.

I'd love to hear what others think about this and see results from other people's tests.

In reference to my post above, here are some examples of the banding using the 20mm f1.7 lens on my OM-D at iso 6400 (from RAW), 100% crops. The first picture shows the more widely spaced bands I referred to; the second picture shows more tightly spaced bands (spacing on other samples can be even tighter). The third picture shows how the bands are sometimes distributed over only the top part of the image.

Reply   Reply with quote   Complain
Bryan Biggers
Veteran MemberPosts: 3,511Gear list
Like?
Re: OM-D EM-5 - High Iso Banding?
In reply to JL Auch, May 8, 2012

Tried 2 more lenses, they were fine, only the Panasonic 20mm has banding problems. Must be poorly shielded or something.

 Bryan Biggers's gear list:Bryan Biggers's gear list
Olympus Tough TG-1 iHS Panasonic Lumix DMC-LX7 Olympus PEN E-P3 Olympus OM-D E-M5 Olympus PEN E-PM2
Reply   Reply with quote   Complain
Vlad S
Senior MemberPosts: 3,075Gear list
Like?
Most likely not sensor.
In reply to toscha_seidel, May 8, 2012

toscha_seidel wrote:

Isn't there a strong electromagnet that suspends the sensor?

The equipment is usually shielded from the elecromagnetic fields. In addition, it is not equally sensitive to all frequencies - like an antenna inside your radio or cell phone it resonates with certain frequencies, and doesn't with others.

It seems that the 20mm might have poor shielding, or resonates with certain household frequencies, and that current is passed through the electric contacts to the body, which is otherwise shielded.

There is a remote possibility that the IBIS is the source of the radiation, but it runs contrary to the OP's observation that when the fluorescent lights are out the banding disappears.

Vlad

 Vlad S's gear list:Vlad S's gear list
Olympus OM-D E-M5 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Olympus M.Zuiko Digital ED 9-18mm 1:4.0-5.6 Panasonic Lumix G 20mm F1.7 ASPH Panasonic Lumix G Vario 100-300mm F4-5.6 OIS +2 more
Reply   Reply with quote   Complain
Vlad S
Senior MemberPosts: 3,075Gear list
Like?
The lens needs a tin foil hat! (nt)
In reply to Bryan Biggers, May 8, 2012

Bryan Biggers wrote:

Tried 2 more lenses, they were fine, only the Panasonic 20mm has banding problems. Must be poorly shielded or something.

 Vlad S's gear list:Vlad S's gear list
Olympus OM-D E-M5 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Olympus M.Zuiko Digital ED 9-18mm 1:4.0-5.6 Panasonic Lumix G 20mm F1.7 ASPH Panasonic Lumix G Vario 100-300mm F4-5.6 OIS +2 more
Reply   Reply with quote   Complain
JL Auch
Forum MemberPosts: 84
Like?
Re: Most likely not sensor.
In reply to Vlad S, May 8, 2012

So Vlad,

Given your knowledge. Is there a possible software/firmware fix for this type of thing, or is this a hardware issue?

If the second, I can't have a lens randomly malfunctioning on me.

Total bummer.

-J

Reply   Reply with quote   Complain
azazel1024
Senior MemberPosts: 1,093Gear list
Like?
Re: The lens needs a tin foil hat! (nt)
In reply to Vlad S, May 8, 2012

Only losely related to this, but I've found on my wife's E-PL1 when I use my cactus V4 radio trigger on it I get horrible banding (about 10x stronger than the worst examples I have seen here) along the right fifth of the frame, but ONLY at ISO's higher than 200. At ISO200 there is absolutely no banding no matter how much I play with contrast, shadow light levels, etc.

The only thing I can come up with there is some bad EM interferance as well. Again, only losely related to this, but digital cameras can deffinitely be succeptible to EM interference, and in some wierd ways too.

It would be a shame if the 20mm lens for some reason was strangely vulnerable to a fair amount of different kinds of EM.

 azazel1024's gear list:azazel1024's gear list
Sigma 30mm F2.8 EX DN +10 more
Reply   Reply with quote   Complain
Bryan Biggers
Veteran MemberPosts: 3,511Gear list
Like?
Re: Most likely not sensor.
In reply to JL Auch, May 8, 2012

It COULD be that they "clock" the Panny lens differently and that is causing the problem. For example, maybe they clock it slower because it is not an official Oly lens and they want to be conservative. In that case they could change the clock frequency a little and maybe help it if it is a harmonic or due to beating against another frequency used in the camera power supply or CCD readout or something, but that is probably a long shot.

 Bryan Biggers's gear list:Bryan Biggers's gear list
Olympus Tough TG-1 iHS Panasonic Lumix DMC-LX7 Olympus PEN E-P3 Olympus OM-D E-M5 Olympus PEN E-PM2
Reply   Reply with quote   Complain
Detail Man
Forum ProPosts: 14,950
Like?
Re: OM-D EM-5 - High Iso Banding ? - Lighting Pulsations / Scan Rates
In reply to JL Auch, May 8, 2012

I'm the fellow who posted the double high-pass sharpened B&W crop early on in this thread. Regarding the theories about "JPG artifacts". JPEGsnoop reveals that your original (larger sized) image that the crop was taken from did have a lot of JPEG Quatization data-compression. Quality Factor around 90%. A mere 1.0 Bits/Pixel or so. The most prevalent appearance of the phenomena seemed to occur ar a radius of 8 pixels (thus, a diameter 16 pixels) setting of the high-pass sharpening process. Note that JPEG encoding splits the image-data into 8x8 pixel blocks for processing. So, possibly some cause to suspect some relation to JPEG encoding processes ...

JL Auch wrote:

Okay. I have done extensive testing today to try and get to the bottom of this. I tested many variables. The best that I can come up with is this:

It seems to be related to the halogen light and/or electric box above the couch, that creates some sort of interference when the light is on. When I turn off the light, no banding with the 20mm. Also the banding seems most pronounced at 1/60sec and 1/30sec but it is still there at other shutter speeds.

Rathe than the idea of RFI at over 2 GHz (from cell phones, Wi-Fi, etc.) affecting largely digital circuitry directly (which overall has high noise-immunity), save for power-supply lines, and the reference voltages derived from such more "analog" potential entry-points ...

... it sounds (from what you observe and report) that the relation to 60 Hz AC line-frequency (at which your flourescent/incandescent indoor light sources are to some extent modulated in luminance by) is interacting with the image-sensor's scanning rates/processes, etc. under certain conditions.

Here is the kicker; I swapped lenses to the 14-42 kit lens, and bang! NO banding with the lights on at either 1/60sec or 1/30sec. The exact same lighting, tripod position, iso, and settings.

That sounds like it surely must be significant . Exactly why seems unclear ...

So I think cbran might be onto something. But if this is true, it is a real shame. The 20mm is a real gem of a lens and this is simply ridiculous.

I'll leave you to your own conclusions.

But what to do now? Return the 20mm lens?

Maybe not so fast - unless you are within some retailer return time constraint where you can receive a full (or nearly full) refund ? I say that because there may be a less lens-specific phenomena going on that you have yet to discover ? Dunno.

I would say that the 60 Hz line-frequency affecting your lighting, coupled with the camera's image-sensor scanning process and rates, sounds like the most likely hypothesis here. Note that incandescent lighting pulsates directly with 120 Hz half-wave rectified sinusoidal (and some harmonics at of 60 Hz at 120, 180, 240 Hz, etc.) current. Flourescent lighting involves a sustained illumination of the suspended mercury-vapor in inert gas - but I think (though I'm no expert on this) that it is likely to (also, still) pulsate somewhat at 120 Hz (with some energy at harmonics frequencies of 60 Hz, as mentioned above for the incandescent case) ...

cbran wrote:

I noticed something similar yesterday when I used the 20mm on my OM-D and your thread prompted me to investigate more deeply.

MY CONCLUSIONS:

At high ISOs (primarily around 6400 for me) with RAW files I see banding ONLY WITH THE 20mm f1.7 LENS . Sometimes there is no banding at all, but most of the time it's there.

It is quite variable, however, both in distribution across the frame and in spacing .

For example, sometimes the banding affects the entire image; sometimes only the top 50%; sometimes only the top 85%. When looking at the thickness/spacing of the banding with 100% crops, sometimes I see 3 bands per inch on my monitor, sometimes 5 bands per inch, sometimes 8 bands per inch, and sometimes 15-20 bands per inch.

In addition to testing the 20mm lens, I tested a manual focus legacy lens (Zeiss 35mm f1.4) plus 3 other Panasonic lenses: the 14mm f2.5, the 45mm f2.8, and the 25mm f1.4. I also tested the Olympus 45mm f1.8. NONE OF THESE OTHER LENSES EXHIBITED BANDING in my tests.

Could it be that the 20mm lens itself is causing some kind of interference pattern? I haven't tested the lens at high ISOs in any other room than my office, but the fact that I don't see banding with other lenses suggests to me that any interference is coming from the lens itself, not other equipment in the room, though perhaps there is some weird interaction going on.

I will be happy to show crops of sample pictures as soon as I prepare some but wanted to share my results in the meantime.

I'd love to hear what others think about this and see results from other people's tests.

In reference to my post above, here are some examples of the banding using the 20mm f1.7 lens on my OM-D at iso 6400 (from RAW), 100% crops. The first picture shows the more widely spaced bands I referred to; the second picture shows more tightly spaced bands (spacing on other samples can be even tighter). The third picture shows how the bands are sometimes distributed over only the top part of the image.

Reply   Reply with quote   Complain
Anders W
Forum ProPosts: 16,945Gear list
Like?
Re: OM-D EM-5 - High Iso Banding?
In reply to cbran, May 8, 2012

cbran wrote:

I noticed something similar yesterday when I used the 20mm on my OM-D and your thread prompted me to investigate more deeply.

MY CONCLUSIONS:

At high ISOs (primarily around 6400 for me) with RAW files I see banding ONLY WITH THE 20mm f1.7 LENS . Sometimes there is no banding at all, but most of the time it's there.

It is quite variable, however, both in distribution across the frame and in spacing .

For example, sometimes the banding affects the entire image; sometimes only the top 50%; sometimes only the top 85%. When looking at the thickness/spacing of the banding with 100% crops, sometimes I see 3 bands per inch on my monitor, sometimes 5 bands per inch, sometimes 8 bands per inch, and sometimes 15-20 bands per inch.

In addition to testing the 20mm lens, I tested a manual focus legacy lens (Zeiss 35mm f1.4) plus 3 other Panasonic lenses: the 14mm f2.5, the 45mm f2.8, and the 25mm f1.4. I also tested the Olympus 45mm f1.8. NONE OF THESE OTHER LENSES EXHIBITED BANDING in my tests.

Could it be that the 20mm lens itself is causing some kind of interference pattern? I haven't tested the lens at high ISOs in any other room than my office, but the fact that I don't see banding with other lenses suggests to me that any interference is coming from the lens itself, not other equipment in the room, though perhaps there is some weird interaction going on.

I will be happy to show crops of sample pictures as soon as I prepare some but wanted to share my results in the meantime.

I'd love to hear what others think about this and see results from other people's tests.

There was a thread a couple of weeks ago that was not about banding but where several people nevertheless spotted it in the sample shown. I searched for it when I read about your suspicions and, again, the image in question turns out to be shot with the 20.

http://forums.dpreview.com/forums/read.asp?forum=1041&message=41341115

I'd appreciate any further experimentation you and others can do in order to nail the problem: Does it happen only/mainly with the 20? And does it happen with the 20 under all circumstances or only in certain cases, e.g., with certain sources of electromagentic interference in the vicinity.

I love my 20 so I am really concerned about this.

 Anders W's gear list:Anders W's gear list
Panasonic Lumix DMC-G1 Olympus OM-D E-M5 Olympus E-M1 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Panasonic Lumix G Vario 7-14mm F4 ASPH +21 more
Reply   Reply with quote   Complain
Vlad S
Senior MemberPosts: 3,075Gear list
Like?
Re: OM-D EM-5 - High Iso Banding ? - Lighting Pulsations / Scan Rates
In reply to Detail Man, May 8, 2012

Detail Man wrote:

Flourescent lighting involves a sustained illumination of the suspended mercury-vapor in inert gas - but I think (though I'm no expert on this) that it is likely to (also, still) pulsate somewhat at 120 Hz

I doubt that light pulsation itself is the cause. But the electric discharges within the lamp are a source of radio waves, and the oscillators in the ballast also might be a source.

 Vlad S's gear list:Vlad S's gear list
Olympus OM-D E-M5 Panasonic Lumix G Vario 14-45mm F3.5-5.6 ASPH OIS Olympus M.Zuiko Digital ED 9-18mm 1:4.0-5.6 Panasonic Lumix G 20mm F1.7 ASPH Panasonic Lumix G Vario 100-300mm F4-5.6 OIS +2 more
Reply   Reply with quote   Complain
markintosh13
Senior MemberPosts: 1,414
Like?
Re: OM-D EM-5 - High Iso Banding?
In reply to Anders W, May 8, 2012

Is this a fiasco?

Reply   Reply with quote   Complain
cbran
New MemberPosts: 23
Like?
Re: A combination of 20/1.7 and E-M5 problem?
In reply to Jonas B, May 8, 2012

Jonas B wrote:

Was the camera set to continues AF?

Jonas

No, camera was on S-AF.

Reply   Reply with quote   Complain
Bryan Biggers
Veteran MemberPosts: 3,511Gear list
Like?
Re: OM-D EM-5 - High Iso Banding ? - Lighting Pulsations / Scan Rates
In reply to Vlad S, May 8, 2012

Vlad S wrote:

Detail Man wrote:

Flourescent lighting involves a sustained illumination of the suspended mercury-vapor in inert gas - but I think (though I'm no expert on this) that it is likely to (also, still) pulsate somewhat at 120 Hz

I doubt that light pulsation itself is the cause. But the electric discharges within the lamp are a source of radio waves, and the oscillators in the ballast also might be a source.

I did a test with an LED lamp that shows the effect a lot using a dark piece of cardboard. Using light from just the lamp, the effect is there. Opening a shade and using sunlight from a nearby window, but with the lamp on and being the same distance from the lamp but with the light from the lamp blocked, the effect is not there. So it is the 60 Hz light variation and probably not RFI.

 Bryan Biggers's gear list:Bryan Biggers's gear list
Olympus Tough TG-1 iHS Panasonic Lumix DMC-LX7 Olympus PEN E-P3 Olympus OM-D E-M5 Olympus PEN E-PM2
Reply   Reply with quote   Complain
Bryan Biggers
Veteran MemberPosts: 3,511Gear list
Like?
Re: A combination of 20/1.7 and E-M5 problem?
In reply to cbran, May 8, 2012

cbran wrote:

Jonas B wrote:

Was the camera set to continues AF?

Jonas

No, camera was on S-AF.

The effect is there even in manual focus.

 Bryan Biggers's gear list:Bryan Biggers's gear list
Olympus Tough TG-1 iHS Panasonic Lumix DMC-LX7 Olympus PEN E-P3 Olympus OM-D E-M5 Olympus PEN E-PM2
Reply   Reply with quote   Complain
Keyboard shortcuts:
FForum MMy threads