Is Foevon better than Bayer sensors in IR and UV photography?

[Plastek]

Are Sigma Foevon sensors better than regular bayer sensors of infrared or ultraviolet photography?

After all here every pixel collects all incoming light, so by my logic - there should be some advantage from that and removing filters from in front of the sensor should allow capturing broader spectrum of light with much higher quality then in regular bayer sensors.

With Bayer while shooting IR only red pixels respond, so your actual resolution is only 1/4th of the one camera have, so from 16 MPx color camera you have 4 MPx IR camera. Which is rather pathetic.

Same with UV - when using filter only blue pixels register UV light, so your actual resolution also falls down to 1/4th.

Does Sigma Foevon behave any better in that respect? Is it capable of capturing 4 times the true resolution of bayer filter in IR and UV spectrum?

Just imagine the implications for astrophotography. This would mean that basically your noise levels fall down by 75% when it comes for IR photography. This could be a real holy grail (next to modified Leica Monochrome).

 

[Roland Karlsson]

Plastek wrote:

Are Sigma Foevon sensors better than regular bayer sensors of infrared or ultraviolet photography?

After all here every pixel collects all incoming light, so by my logic - there should be some advantage from that and removing filters from in front of the sensor should allow capturing broader spectrum of light with much higher quality then in regular bayer sensors.

With Bayer while shooting IR only red pixels respond, so your actual resolution is only 1/4th of the one camera have, so from 16 MPx color camera you have 4 MPx IR camera. Which is rather pathetic.

Same with UV - when using filter only blue pixels register UV light, so your actual resolution also falls down to 1/4th.

Does Sigma Foevon behave any better in that respect? Is it capable of capturing 4 times the true resolution of bayer filter in IR and UV spectrum?

Just imagine the implications for astrophotography. This would mean that basically your noise levels fall down by 75% when it comes for IR photography. This could be a real holy grail (next to modified Leica Monochrome).

All three filters (RGB) in Bayer cameras lets IR through.

In Foveon, the red layer is the main source for IR sensitivity.

 

[Laurence Matson]

Plastek wrote:

Are Sigma Foevon sensors better than regular bayer sensors of infrared or ultraviolet photography?

After all here every pixel collects all incoming light, so by my logic - there should be some advantage from that and removing filters from in front of the sensor should allow capturing broader spectrum of light with much higher quality then in regular bayer sensors.

With Bayer while shooting IR only red pixels respond, so your actual resolution is only 1/4th of the one camera have, so from 16 MPx color camera you have 4 MPx IR camera. Which is rather pathetic.

Same with UV - when using filter only blue pixels register UV light, so your actual resolution also falls down to 1/4th.

Does Sigma Foevon behave any better in that respect? Is it capable of capturing 4 times the true resolution of bayer filter in IR and UV spectrum?

Just imagine the implications for astrophotography. This would mean that basically your noise levels fall down by 75% when it comes for IR photography. This could be a real holy grail (next to modified Leica Monochrome).

Sigma dSLR cameras certainly are out of the box.

For years, I have wondered about all of the fools investing in converted dSLR for dedicated IR photography when at various times, they could have picked up an SD body with lens and some IR filters for less.

In any case, all you have to do is remove the hot mirror and you are good to go. There are ooddles of samples all over pbase. Here are my early efforts during the proof of concept phase:

Loading…

www.pbase.com

 

[mike earussi]

Laurence Matson wrote:

Plastek wrote:

Are Sigma Foevon sensors better than regular bayer sensors of infrared or ultraviolet photography?

After all here every pixel collects all incoming light, so by my logic - there should be some advantage from that and removing filters from in front of the sensor should allow capturing broader spectrum of light with much higher quality then in regular bayer sensors.

With Bayer while shooting IR only red pixels respond, so your actual resolution is only 1/4th of the one camera have, so from 16 MPx color camera you have 4 MPx IR camera. Which is rather pathetic.

Same with UV - when using filter only blue pixels register UV light, so your actual resolution also falls down to 1/4th.

Does Sigma Foevon behave any better in that respect? Is it capable of capturing 4 times the true resolution of bayer filter in IR and UV spectrum?

Just imagine the implications for astrophotography. This would mean that basically your noise levels fall down by 75% when it comes for IR photography. This could be a real holy grail (next to modified Leica Monochrome).

Sigma dSLR cameras certainly are out of the box.

For years, I have wondered about all of the fools investing in converted dSLR for dedicated IR photography when at various times, they could have picked up an SD body with lens and some IR filters for less.

In any case, all you have to do is remove the hot mirror and you are good to go. There are ooddles of samples all over pbase. Here are my early efforts during the proof of concept phase:

http://www.pbase.com/lmatson/just_infrared

 

[Kendall Helmstetter Gelner]

mike earussi wrote:

Actually Laurence, that's not the case. I did a test of the SD1 vs the NEX 7 and because of live view (and the 11.7x mag) which enables totally accurate focusing even in IR frequencies, the NEX 7 shot came out much better/sharper. Now when Sigma makes a camera with live view with magnification that situation might change in favor of Foveon.

That makes it easier but you can get the SD-1 just as sharp, you simply have to learn how to focus for IR. It's not like you can't take a shot and examine sharpness to verify if it's correct or not.

 

[Kendall Helmstetter Gelner]

Plastek wrote:

Are Sigma Foevon sensors better than regular bayer sensors of infrared or ultraviolet photography?

After all here every pixel collects all incoming light, so by my logic - there should be some advantage from that and removing filters from in front of the sensor should allow capturing broader spectrum of light with much higher quality then in regular bayer sensors.

As someone else pointed out, Bayer cameras when capturing IR are actually bypassing all of the filters so they get full resolution.

However because those filters are designed to block IR to some degree, it means that taking IR images with a non-converted camera means VERY long exposures. With the IR filter removed on the SD-1 I can do (and have done) handheld IR that has normal exposure times, making a lot more IR photography practical.

Same with UV - when using filter only blue pixels register UV light, so your actual resolution also falls down to 1/4th.

For some reason we have not seen many attempts at UV photography with Sigma cameras. It requires more specialize filters.

Just imagine the implications for astrophotography. This would mean that basically your noise levels fall down by 75% when it comes for IR photography. This could be a real holy grail (next to modified Leica Monochrome).

I've been experimenting a little bit with that but have nothing solid to report yet. It does help though.

 

[Erik Magnuson]

Laurence Matson wrote:

For years, I have wondered about all of the fools investing in converted dSLR for dedicated IR photography when at various times, they could have picked up an SD body with lens and some IR filters for less.

A "converted" SLR has the IR blocking filter above the sensor replaced with an IR pass filter. So unlike a Sigma SLR, the converted camera does not require a filter on the lens and so has a normal bright viewfinder. It means you can use any lens including those like the Sigma 12-24mm that are difficult to add external filters to. The autofocus is also typically adjusted for IR (but only for a single simple lens like a 50mm prime.)

 

[Erik Magnuson]

Kendall Helmstetter Gelner wrote:

As someone else pointed out, Bayer cameras when capturing IR are actually bypassing all of the filters so they get full resolution.

Yes and no. Most people do not use the few specialized converters that bypass demosaicking.

However because those filters are designed to block IR to some degree, it means that taking IR images with a non-converted camera means VERY long exposure.

"Those" (i.e. the filters used for the color mosaic) don't block IR, a separate IR filter does. However, the color filters do have very slight differences in IR pass - this is why you see so many digital IR images with blueish skies and/or reddish foliage. (Actually, these images have the red/blue channels swapped - straight from the camera the skies are reddish and foliage has a blue tint.) You only get this effect with a "light" IR filter like an 89B; if you use a dark IR filter like a 87C there is no channel differential.

 

[Laurence Matson]

Mike,

Essentially, your spin is in agreement with what I said.

The NEX7, once converted, can only do IR. Better, perhaps, if pixel count is the only game in town. But, I am pretty sure you will not be shooting too many of those IR images hand-held unless something radical has happened in the Bayer world since last I checked.

IR on any of the Sigma dSLRs beginning with the SD10, is shot at normal times. Typically, 1/125 at f5.6. This means, I can even shoot hand-held if I am willing to pray and shoot.

Also, Seng had a series of hot mirror replacement filters made, which allowed one to have a temporarily dedicated IR SD camera and use any lens you wanted. For $400, I am pretty sure that effort could be repeated with filters at different cut-offs.

If you want a dedicated, focussing DPxM camera with live view and focus magnifier, I am quite sure that could be done relatively easily and certainly for less than a NEX7. Merrill did a series of those with the original DP for Seng; of course he knew what to do. I am pretty sure someone else could do that too.

As far as I know, the Foveon imagers are the most IR responsive imagers on the planet in consumer cameras. Correct me if I am wrong. That makes it a no-brainer to start there.

mike earussi wrote:

Laurence Matson wrote:

Plastek wrote:

Are Sigma Foevon sensors better than regular bayer sensors of infrared or ultraviolet photography?

After all here every pixel collects all incoming light, so by my logic - there should be some advantage from that and removing filters from in front of the sensor should allow capturing broader spectrum of light with much higher quality then in regular bayer sensors.

With Bayer while shooting IR only red pixels respond, so your actual resolution is only 1/4th of the one camera have, so from 16 MPx color camera you have 4 MPx IR camera. Which is rather pathetic.

Same with UV - when using filter only blue pixels register UV light, so your actual resolution also falls down to 1/4th.

Does Sigma Foevon behave any better in that respect? Is it capable of capturing 4 times the true resolution of bayer filter in IR and UV spectrum?

Just imagine the implications for astrophotography. This would mean that basically your noise levels fall down by 75% when it comes for IR photography. This could be a real holy grail (next to modified Leica Monochrome).

Sigma dSLR cameras certainly are out of the box.

For years, I have wondered about all of the fools investing in converted dSLR for dedicated IR photography when at various times, they could have picked up an SD body with lens and some IR filters for less.

In any case, all you have to do is remove the hot mirror and you are good to go. There are ooddles of samples all over pbase. Here are my early efforts during the proof of concept phase:

http://www.pbase.com/lmatson/just_infrared

 

[DavidMillier]

I used a Nikon D50 converted to IR (sold now). The camera cost £300, the filter £50 from life pixel and i paid a camera centre service engineer £50 to do the job privately. So £400 all in. SD14 cost about £800 at the time and I already had a Nikon lens outfit. Economically, for me Sigma was not the cheap way to do it.

Also, once converted I had a camera I could use handheld with a fully functional viewfinder. I've tried using my SD14 for IR but you have to remove the filter (a nerve wracking experience because they are so fragile and expensive) and you have to shoot through a black filter which makes the viewfinder unusable. For all practical purposes it means shooting from a tripod and if you are going to do that you might as well just use an unconverted camera with a IR filter and long exposure times. My DP1 made a pretty good IR camera with nothing more than a 10 stop Hitech ND filter with 15 sec exposures.


I have to confess to not feeling a fool for having the D50 converted.

The Sigmas give very good results, mind you, if you don't mind the inconvenience.

 

[DavidMillier]

Why do you say this?

Exposure times on my converted D50 in IR were exactly the same as for visible liight. The sensitivities were about the same.

 

[DavidMillier]

Nonsense.

On occasions, exposure times on my converted D50 hit the maximum shutter speed limit in strong sunlight.

The main difference between the D50 before and after was that dynamic range in IR is lower. You would sometimes overexpose even though the metering and the histogram thought everything was fine.

 

[DavidMillier]

Some example shots from D50 - see exif for exposure details.

DMillier wrote:

Why do you say this?

Exposure times on my converted D50 in IR were exactly the same as for visible liight. The sensitivities were about the same.

--
Galleries and website: http://www.whisperingcat.co.uk/
Flickr: David Millier

--
Galleries and website: http://www.whisperingcat.co.uk/
Flickr: David Millier

 

[Lin Evans]

All except #4 have red overtones. It would be easy enough to remove it, but it does depict the issues with bayer and IR.

Best regards,

Lin

 

[Raist3d]

Laurence Matson wrote:

Mike,

Essentially, your spin is in agreement with what I said.

That's not what I read he wrote.

The NEX7, once converted, can only do IR. Better, perhaps, if pixel count is the only game in town. But, I am pretty sure you will not be shooting too many of those IR images hand-held unless something radical has happened in the Bayer world since last I checked.

He also mentioned they would have that per pixel detail *and more* resolution. Both. So at least in the pre-current X3 era, where, even people as yourself admitted would like more megapixels, many of those conversions are a clear win for some photographers who wanted to print bigger.

IR on any of the Sigma dSLRs beginning with the SD10, is shot at normal times. Typically, 1/125 at f5.6. This means, I can even shoot hand-held if I am willing to pray and shoot.

Also, Seng had a series of hot mirror replacement filters made, which allowed one to have a temporarily dedicated IR SD camera and use any lens you wanted. For $400, I am pretty sure that effort could be repeated with filters at different cut-offs.

If you want a dedicated, focussing DPxM camera with live view and focus magnifier, I am quite sure that could be done relatively easily and certainly for less than a NEX7. Merrill did a series of those with the original DP for Seng; of course he knew what to do. I am pretty sure someone else could do that too.

As far as I know, the Foveon imagers are the most IR responsive imagers on the planet in consumer cameras. Correct me if I am wrong. That makes it a no-brainer to start there.

You also fail to add any other usability workflow issues in using the cameras themselves. I am not saying there aren't considerations in using a Sigma camera, but it's hardly the slamdunk "geesh, those people are fools" that your marketing spin makes it out to be (your words, you said fools).

mike earussi wrote:

Laurence Matson wrote:

Plastek wrote:

Are Sigma Foevon sensors better than regular bayer sensors of infrared or ultraviolet photography?

After all here every pixel collects all incoming light, so by my logic - there should be some advantage from that and removing filters from in front of the sensor should allow capturing broader spectrum of light with much higher quality then in regular bayer sensors.

With Bayer while shooting IR only red pixels respond, so your actual resolution is only 1/4th of the one camera have, so from 16 MPx color camera you have 4 MPx IR camera. Which is rather pathetic.

Same with UV - when using filter only blue pixels register UV light, so your actual resolution also falls down to 1/4th.

Does Sigma Foevon behave any better in that respect? Is it capable of capturing 4 times the true resolution of bayer filter in IR and UV spectrum?

Just imagine the implications for astrophotography. This would mean that basically your noise levels fall down by 75% when it comes for IR photography. This could be a real holy grail (next to modified Leica Monochrome).

Sigma dSLR cameras certainly are out of the box.

For years, I have wondered about all of the fools investing in converted dSLR for dedicated IR photography when at various times, they could have picked up an SD body with lens and some IR filters for less.

In any case, all you have to do is remove the hot mirror and you are good to go. There are ooddles of samples all over pbase. Here are my early efforts during the proof of concept phase:

http://www.pbase.com/lmatson/just_infrared

 

[Raist3d]

Lin Evans wrote:

All except #4 have red overtones. It would be easy enough to remove it, but it does depict the issues with bayer and IR.

Yea, because Sigma's wouldn't have any issues at all compared to the bayer cameras for IR. They are so perfect.

Best regards,

Lin

 

[Kendall Helmstetter Gelner]

DMillier wrote:

Why do you say this?

Exposure times on my converted D50 in IR were exactly the same as for visible liight. The sensitivities were about the same.

He says that because he was talking about UNCONVERTED shooting times for IR photography on Bayer cameras - because if you have not removed the CFA you have to leave the shutter open a long time to let enough IR light come through the filter to capture.

Of course converting the bayer camera to IR will give you normal shooting times. But it also fixes that camera as shooting mostly IR henceforth.

 

[Laurence Matson]

Nice shots. Which cutoff?

 

[Laurence Matson]

Raist3d wrote:

Laurence Matson wrote:

Mike,

Essentially, your spin is in agreement with what I said.

That's not what I read he wrote.

Why does this not surprise me?

The NEX7, once converted, can only do IR. Better, perhaps, if pixel count is the only game in town. But, I am pretty sure you will not be shooting too many of those IR images hand-held unless something radical has happened in the Bayer world since last I checked.

He also mentioned they would have that per pixel detail *and more* resolution. Both. So at least in the pre-current X3 era, where, even people as yourself admitted would like more megapixels, many of those conversions are a clear win for some photographers who wanted to print bigger.

Another amazing surprise. My pushing for something long before anyone else except for Dominic is somehow construed as . . . What is your point except to preach in contrarianisms? I am quite sure I can print an IR shot from my SD1 to fill the width of my Epson 9900 using the short side. And the quality will be just fine, thank you very much. Ask Epson France if you need confirmation. Need a contact number?

IR on any of the Sigma dSLRs beginning with the SD10, is shot at normal times. Typically, 1/125 at f5.6. This means, I can even shoot hand-held if I am willing to pray and shoot.

Also, Seng had a series of hot mirror replacement filters made, which allowed one to have a temporarily dedicated IR SD camera and use any lens you wanted. For $400, I am pretty sure that effort could be repeated with filters at different cut-offs.

If you want a dedicated, focussing DPxM camera with live view and focus magnifier, I am quite sure that could be done relatively easily and certainly for less than a NEX7. Merrill did a series of those with the original DP for Seng; of course he knew what to do. I am pretty sure someone else could do that too.

As far as I know, the Foveon imagers are the most IR responsive imagers on the planet in consumer cameras. Correct me if I am wrong. That makes it a no-brainer to start there.

You also fail to add any other usability workflow issues in using the cameras themselves. I am not saying there aren't considerations in using a Sigma camera, but it's hardly the slamdunk "geesh, those people are fools" that your marketing spin makes it out to be (your words, you said fools).

That's your issue. I have worked with SPP for 10 years plus now (since 1.0), and have bellyached less about that tool over that period than some do in a single posting. It is the same as with all devices: Learn to work with it or move on.

mike earussi wrote:

Laurence Matson wrote:

Plastek wrote:

Are Sigma Foevon sensors better than regular bayer sensors of infrared or ultraviolet photography?

After all here every pixel collects all incoming light, so by my logic - there should be some advantage from that and removing filters from in front of the sensor should allow capturing broader spectrum of light with much higher quality then in regular bayer sensors.

With Bayer while shooting IR only red pixels respond, so your actual resolution is only 1/4th of the one camera have, so from 16 MPx color camera you have 4 MPx IR camera. Which is rather pathetic.

Same with UV - when using filter only blue pixels register UV light, so your actual resolution also falls down to 1/4th.

Does Sigma Foevon behave any better in that respect? Is it capable of capturing 4 times the true resolution of bayer filter in IR and UV spectrum?

Just imagine the implications for astrophotography. This would mean that basically your noise levels fall down by 75% when it comes for IR photography. This could be a real holy grail (next to modified Leica Monochrome).

Sigma dSLR cameras certainly are out of the box.

For years, I have wondered about all of the fools investing in converted dSLR for dedicated IR photography when at various times, they could have picked up an SD body with lens and some IR filters for less.

In any case, all you have to do is remove the hot mirror and you are good to go. There are ooddles of samples all over pbase. Here are my early efforts during the proof of concept phase:

http://www.pbase.com/lmatson/just_infrared

 

[mike earussi]

No argument, in fact I'm surprised Sigma hasn't made a series of different cutoff filters to replace the easily removable hot mirror. A series of IR filters of different frequency cutoffs, and a UV one as well, would certainly be easier than having to deal with all the different filter sizes now. Then a built-in AF offset for each filter would make the SD line the most popular alternative frequency camera around (and probably increase sales, as well). Maybe buy them in a set in a nice little wooden box.

Sigma, are you listening?

Laurence Matson wrote:

Mike,

Essentially, your spin is in agreement with what I said.

The NEX7, once converted, can only do IR. Better, perhaps, if pixel count is the only game in town. But, I am pretty sure you will not be shooting too many of those IR images hand-held unless something radical has happened in the Bayer world since last I checked.

IR on any of the Sigma dSLRs beginning with the SD10, is shot at normal times. Typically, 1/125 at f5.6. This means, I can even shoot hand-held if I am willing to pray and shoot.

Also, Seng had a series of hot mirror replacement filters made, which allowed one to have a temporarily dedicated IR SD camera and use any lens you wanted. For $400, I am pretty sure that effort could be repeated with filters at different cut-offs.

If you want a dedicated, focussing DPxM camera with live view and focus magnifier, I am quite sure that could be done relatively easily and certainly for less than a NEX7. Merrill did a series of those with the original DP for Seng; of course he knew what to do. I am pretty sure someone else could do that too.

As far as I know, the Foveon imagers are the most IR responsive imagers on the planet in consumer cameras. Correct me if I am wrong. That makes it a no-brainer to start there.