What would the FFF have to do in order to be competative?

Stillton wrote:

Doppler9000 wrote:

Stillton wrote:

Doppler9000 wrote:

Stillton wrote:

Doppler9000 wrote:

Stillton wrote:

"This looks like aliasing to me…"
It is mostly color blotching because it is showing on is400, but not on is100.

It was ISO 100.

I meant, this one. It was stated that this green-magenta blotching was the result of aliasing which is not the case. It is pretty obvious that unlike Bayer, foveon, especially 1:1:1 version of the tech, does not have that rainbow aliasing. This is what my previous post was comparing.

Yes, victorgv made this point 19 hours ago, and I agreed.
Do you think the subject you took the quote from is aliasing or color blotching (or processing)?

Disagree....

Disagree with what? I asked you if you thought it was aliasing, color blotching or processing. Do you disagree with all three? If so, what’s you thesis?

Because if that was the case, in this example I am attaching below, it would be all going rainbow colors in the center of that circle. Bayer sensor does show that, but even Foveon Q ( merrill even more so) behaves really well there, with no color aliasing. Brightness aliasing is still present because it is innate feature of any discrete/pixelated measuring system..

This looks like aliasing, but then it should show itself in every aliasing-prone scenario but it just doesnt. Instead it looks like it is either some chromatic aberration or some other effect.

Also, even individual lines below show red fringe colors... this is not how aliasing works. It needs repetitive patterns smaller than a pixel size to create those rainbow colors. This is not possible with single lines and especially not possible on a 1:1:1 sensor. This clearly is something else other than just a typical rainbow-colored aliasing....

I literally wrote a whole paragraph on what I disagree with there.
In one of the images (the one with is400 from Foveon sensor) it was clearly color blotching. On the converging lines though, it looked like fringe aberrations or some other such effect, since it was visible even on single lines.

Foveon sensor does not have a process by which color aliasing is generated, which is easily demonstrated by the image of "circles" with the center generating color aliasing in Bayer sensors configurations, but not in Foveon ones.

If THAT test does not generate rainbow pattern on foveon, nothing will.

Hence, it cant be a case of simple color aliasing and it has to be something else. This is what my whole logic/point was.

If you guys think it is color aliasing, that is fine, but you have to explain why "circle aliasing" test does not show that effect, while simple lines test - supposedly does.

This is why I said that color aliasing does not work like that.

I totally agree with you. I've shot lots of my own test images with my own cameras that have Foveon sensors, and I've never seen color moire, but I've seen enough real life examples from my own non-Foveon cameras (i.e. my Nikon D810), that I know color moire is a problem with "Bayer" sensor cameras. I'm hoping to get a 40 MP Fuji in the next few months, and I'll try to do a comparison using 20 MP shots from a Foveon (Quattro) sensor. When I do, I'll see if I can make some photos with color moire using the Fuji.

Scottelly wrote:

Stillton wrote:

Doppler9000 wrote:

Stillton wrote:

Doppler9000 wrote:

Stillton wrote:

Doppler9000 wrote:

Stillton wrote:

"This looks like aliasing to me…"
It is mostly color blotching because it is showing on is400, but not on is100.

It was ISO 100.

I meant, this one. It was stated that this green-magenta blotching was the result of aliasing which is not the case. It is pretty obvious that unlike Bayer, foveon, especially 1:1:1 version of the tech, does not have that rainbow aliasing. This is what my previous post was comparing.

Yes, victorgv made this point 19 hours ago, and I agreed.
Do you think the subject you took the quote from is aliasing or color blotching (or processing)?

Disagree....

Disagree with what? I asked you if you thought it was aliasing, color blotching or processing. Do you disagree with all three? If so, what’s you thesis?

Because if that was the case, in this example I am attaching below, it would be all going rainbow colors in the center of that circle. Bayer sensor does show that, but even Foveon Q ( merrill even more so) behaves really well there, with no color aliasing. Brightness aliasing is still present because it is innate feature of any discrete/pixelated measuring system..

This looks like aliasing, but then it should show itself in every aliasing-prone scenario but it just doesnt. Instead it looks like it is either some chromatic aberration or some other effect.

Also, even individual lines below show red fringe colors... this is not how aliasing works. It needs repetitive patterns smaller than a pixel size to create those rainbow colors. This is not possible with single lines and especially not possible on a 1:1:1 sensor. This clearly is something else other than just a typical rainbow-colored aliasing....

I literally wrote a whole paragraph on what I disagree with there.
In one of the images (the one with is400 from Foveon sensor) it was clearly color blotching. On the converging lines though, it looked like fringe aberrations or some other such effect, since it was visible even on single lines.

Foveon sensor does not have a process by which color aliasing is generated, which is easily demonstrated by the image of "circles" with the center generating color aliasing in Bayer sensors configurations, but not in Foveon ones.

If THAT test does not generate rainbow pattern on foveon, nothing will.

Hence, it cant be a case of simple color aliasing and it has to be something else. This is what my whole logic/point was.

If you guys think it is color aliasing, that is fine, but you have to explain why "circle aliasing" test does not show that effect, while simple lines test - supposedly does.

This is why I said that color aliasing does not work like that.

I totally agree with you. I've shot lots of my own test images with my own cameras that have Foveon sensors, and I've never seen color moire, but I've seen enough real life examples from my own non-Foveon cameras (i.e. my Nikon D810), that I know color moire is a problem with "Bayer" sensor cameras. I'm hoping to get a 40 MP Fuji in the next few months, and I'll try to do a comparison using 20 MP shots from a Foveon (Quattro) sensor. When I do, I'll see if I can make some photos with color moire using the Fuji.

Interestingly enough, X-tran handles color aliasing a bit better than Bayer. I am guessing, irregular sized pixels help with breaking up aliasing pattern generation.

DMillier wrote:

Stillton wrote:

dellaaa wrote:

Reading a recent post that included a 60M pixel Bayer sensor shot, the the discussion turned into a debate as to whether the FFF would ever arrive.

Looking at the Bayer photo what do you believe the FFF sensor would have to do better than the current generation of Bayer sensors? What are the design criteria the FFF design team have?

I don't know but given the physics of the design and the past Foveons, low light performance would most likely not be on par with the Bayer, agreed?

Historically, Sigmas have required extensive image processing both while taking the shot and afterwords (SPP). Will the FFF shoot at the frame rates of the current Bayers? Again, given the history, this seems doubtful.

So with low light and speed out of the equation, whats left, detail and color rendition.

As for detail, how would the specs of the proposed FFF stack up against the 60M plus Bayer sensors? Would the advantage in detail make up for the shortcoming?

The remaining design criteria, is color. Would Sigma's color rendition be enough for people to buy it?

I don't know and would like rational, opinions as just what will the FFF bring to the table, and why any non L mount people in 2023 would buy it when 100M pixel Bayers are on the horizon that take great low light picture at high frame rates.

At this point, fanboys aside, the fact that we are still discussing this seems a bit absurd to me, so why did I start this thread lol?.

FFF needs to exist at least in its original (and not in Quattro) form retaining (or rather - rediscovering) its 1:1:1 Merrill spirit.

Combining that with better infrastructure (akin to FP or quattro platform) it will win any comparison within its limits (aka low iso performance and other stuff like this)

In fact, if Sigma could create 6x4 digital back with their sensor, they will destroy any bayer competition in that market too.

Recently looked at some 2010 hassel's images (100Mpx images) they were horribly blurred and meh colors. If foveon could only step into that segment, they would be by far the most "magical" sensor.

I've been in a weeding process of my back catalogue (going back to 1999) for the last few weeks. The standout (as in sticks out like a sore thumb) thing I've noticed is the really appalling colour from all my Foveon cameras. Colour casts of one hue or another across all the cameras.

I've seen enough images from other photographers to know that Foveon is prone to this (especially the early models that I have), it's not just me. I've also seen enough photos from other Sigma photogs to know that Foveon is capable of beautiful colour in the right hands and right light - skillful post processing plays an important part.

But even so... just looking at my images, it is clearly a major struggle with my workflow and (for me) the results are not worth the effort. I don't think I'll be using my Sigmas ever again or investing in the FFF if it ever appears.

Nonetheless, it is important that there are alternatives in the market and there is clearly a desire from fans for the FFF; I wish Sigma every success, I'm sure it will make plenty of people happy.

I wonder which cameras you have ?

The colours from the DP3M (processed in SPP) have always seemed to me to be very pleasing, and those from the sdQH DNG files depend on the input profile that I use with Adobe Camera Raw.

I think this is a JPG processed by the camera. Warm morning light well rendered, IMO.

The TIF version converted from the raw file looks the same.

Don

DMillier wrote:

Stillton wrote:

dellaaa wrote:

Reading a recent post that included a 60M pixel Bayer sensor shot, the the discussion turned into a debate as to whether the FFF would ever arrive.

Looking at the Bayer photo what do you believe the FFF sensor would have to do better than the current generation of Bayer sensors? What are the design criteria the FFF design team have?

I don't know but given the physics of the design and the past Foveons, low light performance would most likely not be on par with the Bayer, agreed?

Historically, Sigmas have required extensive image processing both while taking the shot and afterwords (SPP). Will the FFF shoot at the frame rates of the current Bayers? Again, given the history, this seems doubtful.

So with low light and speed out of the equation, whats left, detail and color rendition.

As for detail, how would the specs of the proposed FFF stack up against the 60M plus Bayer sensors? Would the advantage in detail make up for the shortcoming?

The remaining design criteria, is color. Would Sigma's color rendition be enough for people to buy it?

I don't know and would like rational, opinions as just what will the FFF bring to the table, and why any non L mount people in 2023 would buy it when 100M pixel Bayers are on the horizon that take great low light picture at high frame rates.

At this point, fanboys aside, the fact that we are still discussing this seems a bit absurd to me, so why did I start this thread lol?.

FFF needs to exist at least in its original (and not in Quattro) form retaining (or rather - rediscovering) its 1:1:1 Merrill spirit.

Combining that with better infrastructure (akin to FP or quattro platform) it will win any comparison within its limits (aka low iso performance and other stuff like this)

In fact, if Sigma could create 6x4 digital back with their sensor, they will destroy any bayer competition in that market too.

Recently looked at some 2010 hassel's images (100Mpx images) they were horribly blurred and meh colors. If foveon could only step into that segment, they would be by far the most "magical" sensor.

I've been in a weeding process of my back catalogue (going back to 1999) for the last few weeks. The standout (as in sticks out like a sore thumb) thing I've noticed is the really appalling colour from all my Foveon cameras. Colour casts of one hue or another across all the cameras.

I've seen enough images from other photographers to know that Foveon is prone to this (especially the early models that I have), it's not just me. I've also seen enough photos from other Sigma photogs to know that Foveon is capable of beautiful colour in the right hands and right light - skillful post processing plays an important part.

But even so... just looking at my images, it is clearly a major struggle with my workflow and (for me) the results are not worth the effort. I don't think I'll be using my Sigmas ever again or investing in the FFF if it ever appears.

Nonetheless, it is important that there are alternatives in the market and there is clearly a desire from fans for the FFF; I wish Sigma every success, I'm sure it will make plenty of people happy.

Hmmm . . .

This is an interesting post, with some interesting observations David. Good luck with whatever camera(s) you find in your hands frim now on.

DMillier wrote:

Currently SD9 and SD14, formerly DP1 and DP2 Merrill.

DP2M was fine.

David,

The SD9 was too yellow and the SD14  had a green cast.

I would not regard either one as a standard to judge all Sigma cameras. I think each one from the SD15 have very acceptable colour with the sd Qs being the best of the Foveon sensor cameras.

The fps are in a different league.

S

DMillier wrote:

Quite noticeable luminance aliasing (eg diagonal hatching on the rectangular frame). This part of the chart is cruel to anything but very high resolution cameras.

Shame about the general trend of dropping AA filters. My GFX50s doesn't actually show up well at suppressing artefacts, sadly. Pixel shift cameras do better.

The GFX 100S does better -- and much better with pixel shift. But in the end a medium format camera that can capture all three colours at each pixel location and doesn't suffer from colour aliasing would be even better -- and wouldn't need an AA filter. Or even a full frame camera with those same specs...

Stick to your wonderful monochrome long exposure shots with the GFX50s, Dave, and everything will be fine and dandy.

I too am using the horses for courses approach, and plan to get a Fuji X camera soon. Crazy maybe, but hey, nothing else can shoot 8K video, and I want to do that. It'll make for a good camera for shooting distant action shots too, and really any action that's too fast for my Nikon. I'll still use my Sigma for most things though, no doubt. My Nikon may get relegated to high-ISO shooting, which is something I rarely do.

Possibly. But a quick review of other shots by Quattro sensors on this forum also demonstrate the purple / green colors in gray sections of the image such as roads.

https://www.dpreview.com/forums/post/66859304?image=1

Don't know how the .jpg was created, but the result is not clean ...

DMillier wrote:

Do you use your S1r in pixel shift mode? If so, what do you think?

Hello --

Not normally, as the resolution is usually adequate for my print sizes.  It is an interesting question though -- I'll try a tripod test with a section of road in late afternoon to see the pixel shift vs. normal mode makes any difference.

Here's a recent example of a shot I'm happy with from the Fryeburg fair last fall - both the pavement and the boy's grey parka show no signs of color fringing, and even on an overcast day  I find the color balance very pleasing, even down to the iridescence of the bubbles.

DBE wrote:

Possibly. But a quick review of other shots by Quattro sensors on this forum also demonstrate the purple / green colors in gray sections of the image such as roads.

https://www.dpreview.com/forums/post/66859304?image=1

Don't know how the .jpg was created, but the result is not clean ...

At 7680x3296, that image has obviously been resized. I suspect the issue has as much to do with processing as anything else, and any image from any camera can look bad, and seem to indicate a problem with the camera, if the processing isn't done very carefully.

Scottelly wrote:

DBE wrote:

Possibly. But a quick review of other shots by Quattro sensors on this forum also demonstrate the purple / green colors in gray sections of the image such as roads.

https://www.dpreview.com/forums/post/66859304?image=1

Don't know how the .jpg was created, but the result is not clean ...

At 7680x3296, that image has obviously been resized. I suspect the issue has as much to do with processing as anything else, and any image from any camera can look bad, and seem to indicate a problem with the camera, if the processing isn't done very carefully.

Well said, Scott!

I am no fan of the use of post-processed images to illustrate camera performance.

By "post-processed images" I mean any other than the converter's initial default review image that appears when a raw or an OOC-jpg is opened.

xpatUSA wrote:

Scottelly wrote:

DBE wrote:

Possibly. But a quick review of other shots by Quattro sensors on this forum also demonstrate the purple / green colors in gray sections of the image such as roads.

https://www.dpreview.com/forums/post/66859304?image=1

Don't know how the .jpg was created, but the result is not clean ...

At 7680x3296, that image has obviously been resized. I suspect the issue has as much to do with processing as anything else, and any image from any camera can look bad, and seem to indicate a problem with the camera, if the processing isn't done very carefully.

Well said, Scott!

I am no fan of the use of post-processed images to illustrate camera performance.

By "post-processed images" I mean any other than the converter's initial default review image that appears when a raw or an OOC-jpg is opened.

Well, the problem with default settings is those settings are not the optimal settings for all scenes/photos Ted. Right?

(Of course, optimal is a word that opens up a can of worms, isn't it? 🥴)

Let's take post processing out of the conversation.  The Quattro was able to save a .jpg in S-HI mode, 7680 x 5120.  Here is a straight .jpg from the camera taken at ISO 200 in a shed used for trolley car restoration - mostly shade from afternoon daylight - you can check the metadata if you wish.

Full size .jpg in S-HI

Lots of corrugated metal and gravel with false colors and a high degree of noise that really shouldn't be in an ISO 200 image.

If the rumored FFF Foveon can produce a clean shot under these conditions, I will be very interested.

Scottelly wrote:

xpatUSA wrote:

Scottelly wrote:

DBE wrote:

Possibly. But a quick review of other shots by Quattro sensors on this forum also demonstrate the purple / green colors in gray sections of the image such as roads.

https://www.dpreview.com/forums/post/66859304?image=1

Don't know how the .jpg was created, but the result is not clean ...

At 7680x3296, that image has obviously been resized. I suspect the issue has as much to do with processing as anything else, and any image from any camera can look bad, and seem to indicate a problem with the camera, if the processing isn't done very carefully.

Well said, Scott!

I am no fan of the use of post-processed images to illustrate camera performance.

By "post-processed images" I mean any other than the converter's initial default review image that appears when a raw or an OOC-jpg is opened.

Well, the problem with default settings is those settings are not the optimal settings for all scenes/photos Ted. Right?

obviously Scott

on the other hand, a perfectly-processed rendition of a blurred, poorly WB'd and exposed shot tells me nothing about the camera at all ...

(Of course, optimal is a word that opens up a can of worms, isn't it? 🥴)

as does the phrase "camera performance", I suppose.