Banding in IQ4 images (and low-light, hand-held stage photographer with IQ4+TC!)

[onasj]

I recently used my IQ4 with a Rodenstock HR50 lens and a small tech camera body (Alpa TC12) to capture hand-held, low-light images of moving subjects during a stage performance, with the idea that while such an application is way outside the usual application scope of such a system, its small size, excellent optics, and amazing high-ISO performance might result in decent images.

The good news is I was quite happy with how some of the images turned out, with one major caveat.

I noticed quite pronounced vertical banding in three of the five images I chose to post-process, even though I shot them at a relatively modest (for me) ISO of 1600 to 2000. You can see the images, and the raw files, here:

Processed JPEGs: https://www.dropbox.com/sh/fhb1l1lei06yyd3/AACGWje_HGESNKxvdHIQUgwWa?dl=0
Raw files: https://www.dropbox.com/sh/lqwzq481p28zgre/AAAEvZ39c-UR4yPVcxoNcfwwa?dl=0





One without apparent banding:


I noticed that in the landscape-orientation images that showed banding, the banding is now horizontal rather than vertical. Best example (a bad shot due to back focus, but shows the banding very well):
https://www.dropbox.com/s/2odnmja0e0qvthu/P0001387-horizontal banding.jpg?dl=0

Since I was using the electronic shutter the entire time, I wonder if banding might be caused by LED stage lighting having a certain strobe frequency that can't be detected by human eyes, but that can be picked up by the sequential readout of the sensor during an ES shot. This theory is supported by the fact that the banding appears to be localized to certain lit regions of the image, rather than present across the entire image or across all parts of the image of a certain brightness level.

Unfortunately I didn't use the Copal shutter during this shoot for comparison, largely because the several extra steps involved in framing the image in live view with the diaphragm open, then closing the diaphragm, waking up the back, opening the diaphragm, and cocking and releasing the Copal shutter wasn't really amenable to this application.

Have any other users experienced banding caused by ES + LED (or other non-100% duty cycle) lighting? I'm hoping that's the cause of the problem, and not an issue with the IQ4.

 

[earburner]

Firstly I would like to say "brave" I do a lot of low light theatre photography and would never dream of using my iq4 and tc or even my xf
I think the lines are likely to come from how the electronic shutter work and movement in the lights and your hand. The ES setting returns the image line by line and depending on what file setting you have the slower it is... so things could have changed at the top of the image while the bottom is still being captured. Also a lot of led lighting is dimmed with PWM technology so depending on the relationship between the pulse width modulation and the censor scanning the image could be effected

 

[TheDude]

I recently used my IQ4 with a Rodenstock HR50 lens and a small tech camera body (Alpa TC12) to capture hand-held, low-light images of moving subjects during a stage performance... was using the electronic shutter the entire time

Banding aside, which hopefully may be resolved with a firmware update or with a different file setting, these photographs demonstrate well what is possible with electronic shutter. Would have been hard (impossible) to take these photographs using Copal shutter.

 

[Boinger]

I recently used my IQ4 with a Rodenstock HR50 lens and a small tech camera body (Alpa TC12) to capture hand-held, low-light images of moving subjects during a stage performance, with the idea that while such an application is way outside the usual application scope of such a system, its small size, excellent optics, and amazing high-ISO performance might result in decent images.

The good news is I was quite happy with how some of the images turned out, with one major caveat.

I noticed quite pronounced vertical banding in three of the five images I chose to post-process, even though I shot them at a relatively modest (for me) ISO of 1600 to 2000. You can see the images, and the raw files, here:

One without apparent banding:
View attachment 139446

I noticed that in the landscape-orientation images that showed banding, the banding is now horizontal rather than vertical. Best example (a bad shot due to back focus, but shows the banding very well):
https://www.dropbox.com/s/2odnmja0e0qvthu/P0001387-horizontal banding.jpg?dl=0

Since I was using the electronic shutter the entire time, I wonder if banding might be caused by LED stage lighting having a certain strobe frequency that can't be detected by human eyes, but that can be picked up by the sequential readout of the sensor during an ES shot. This theory is supported by the fact that the banding appears to be localized to certain lit regions of the image, rather than present across the entire image or across all parts of the image of a certain brightness level.

Unfortunately I didn't use the Copal shutter during this shoot for comparison, largely because the several extra steps involved in framing the image in live view with the diaphragm open, then closing the diaphragm, waking up the back, opening the diaphragm, and cocking and releasing the Copal shutter wasn't really amenable to this application.

Have any other users experienced banding caused by ES + LED (or other non-100% duty cycle) lighting? I'm hoping that's the cause of the problem, and not an issue with the IQ4.

You can't use electronic shutter with artificial modern lighting. I suspect tungsten based lighting would work okay but not leds / florescent.

You will always have banding with these lights and electronic. Nothing you can do about it, not an error with the XF. Will happen on every electronic shutter in every camera.

It is a function of the frequency of the power in your region either 50 Hz or 60 Hz. You can shoot at theoreticly 1/60th or 1/50th and not get banding or in slower ratios of those 2 shutter speeds. But in practice I have found that you can still see faint banding under close inspection. Just a problem with modern lights. That is why we need global shutters.

Also the reason it is horizontal is that is the way the sensor reads each line.

 

[MrSmith]

Banding aside, which hopefully may be resolved with a firmware update or with a different file setting, these photographs demonstrate well what is possible with electronic shutter. Would have been hard (impossible) to take these photographs using Copal shutter.

As stated it’s only going to be solved by a global shutter or different light source. If I handed over files to a client with banding like that it would be a reshoot and/or a lost client.

Is that sensor even capable of global readout?

 

[dougpeterson]

You can't use electronic shutter with artificial modern lighting. I suspect tungsten based lighting would work okay but not leds / florescent. You will always have banding with these lights and electronic. Nothing you can do about it, not an error with the XF. Will happen on every electronic shutter in every camera.

Phase One's implementation of Electronic Shutter (ES) provides for anti-flickering of either 50hz or 60hz lighting sources.

However anti-flickering is not available on the IQ4 yet; the menu option is there, but it does not have an effect. I would expect a future firmware update to make the IQ4 behave as the IQ3 does, allowing for deflickering.

(see update below: I was wrong; this is already enabled/working on the IQ4)

Is that sensor even capable of global readout?

As far as I can tell global shutters are still many years away from being part of the photographic camera market in larger-than-cell-phone sized sensors. They currently cost a lot in terms of image quality.

 

[Boinger]

Phase One's implementation of Electronic Shutter (ES) provides for anti-flickering of either 50hz or 60hz lighting sources.

However anti-flickering is not available on the IQ4 yet; the menu option is there, but it does not have an effect. I would expect a future firmware update to make the IQ4 behave as the IQ3 does, allowing for deflickering.



As far as I can tell global shutters are still many years away from being part of the photographic camera market in larger-than-cell-phone sized sensors. They currently cost a lot in terms of image quality.

I don't know how it worked in the IQ3. But with other camera's I have tried anti flickering, and it still doesnt really correct the full banding. Have had rather unsatisfactory results with anti-flickering. But will be happy to give it a go when it is released on the IQ4

 

[dougpeterson]

Update: with current firmware the anti-flicker feature of the IQ4 works as expected. My statement that it wasn't yet working was based on earlier firmware. My mistake.

See attached. Note that 50 points contrast is added to make the banding more apparent; it's easiest to see in the out of focus area at the bottom.

Left image = With [Menu > IQ settings > Antiflickering] set to 60hz
Right image = With [Menu > IQ settings > Antiflickering] set to Off

1/125th sec using electronic shutter (ES). No other change between the frames.