Foveon sensor (SIGMA cameras)

Started by scrax, April 21, 2019, 05:20:29 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

scrax

Anybody has some experiences with it?
From what I understood it has only 100 ISO real because of no amplification before A/D
But no bayer and a lot more details than CMOS

It's something new for me, I was aware only of CMOS (ok the foveon is a particular CMOS by the way) and CCD sensors...

With only 100ISO I'm curious about long shutter time noise
I'm using ML2.3 for photography with:
EOS 600DML | EOS 400Dplus | EOS 5D MLbeta5- EF 100mm f/2.8 USM Macro  - EF-S 17-85mm f4-5.6 IS USM - EF 70-200mm f/4 L USM - 580EXII - OsX, PS, LR, RawTherapee, LightZone -no video experience-

ArcziPL

Look at samples. The dynamic range (=> shadow noise) sucks. People like it due to its colors, not ISO performance.

There were another variations in the past from different manufacturers. Honeycomb instead of standard Bayer, additional fourth color and differently sized sensels to increase DR, to name just a few.
M50.110 [main cam] | G7X III [pocket cam] | 70D.112 [gathers dust] | M.202 [gathers dust] | waiting for M5II

scrax

Quote from: ArcziPL on April 21, 2019, 08:56:20 PM
Look at samples. The dynamic range (=> shadow noise) sucks. People like it due to its colors, not ISO performance.

There were another variations in the past from different manufacturers. Honeycomb instead of standard Bayer, additional fourth color and differently sized sensels to increase DR, to name just a few.

Is what I was thinking, I've dowloaded the sigma app (that is not fast at all in render edits previews) and some raw files to see how they are.
Noise and ISO are ok only at 100 - 200 and 400 and from reading how it works on dedicated forums it's because is like a camera with only 100 ISO film.
Changing it is just a "memo" for the raw processor to push up exposure in post, that's why it gives so much noise at high ISOs
But sharpness and details are great.
Seems also to have a lot of potentiality in recovering highligts (at least).

Another thing I've found on dedicated forum is that it is good with natural lights but has big problem with artificials lights like tungsten because they have more red than blue frequencies...

I'll like to try one (mainly fo B/W), but if real ISO are only 100, it is really a camera good only for landscapes (with good weather)...
I'm using ML2.3 for photography with:
EOS 600DML | EOS 400Dplus | EOS 5D MLbeta5- EF 100mm f/2.8 USM Macro  - EF-S 17-85mm f4-5.6 IS USM - EF 70-200mm f/4 L USM - 580EXII - OsX, PS, LR, RawTherapee, LightZone -no video experience-

Luther

Quote from: scrax on April 22, 2019, 01:58:28 AM
I'll like to try one (mainly fo B/W), but if real ISO are only 100, it is really a camera good only for landscapes (with good weather)...

Anything you can use tripod and is static (not moving) this can be a good camera. Would be nice to test together with HDRMerge...
I think Foveon is a great technology and much better than Bayer. The noise issue could be solved if there was more people producing/researching it, but AFAIK, SIGMA patented the technology and now no one can use it, except them. Too bad.

You can simulate the Foveon, doing a technique called Drizzle. This software (supposedly) does that:
https://github.com/LucCoiffier/DSS/

scrax

Quote from: Luther on April 22, 2019, 05:37:39 AM
Anything you can use tripod and is static (not moving) this can be a good camera. Would be nice to test together with HDRMerge...
I think Foveon is a great technology and much better than Bayer. The noise issue could be solved if there was more people producing/researching it, but AFAIK, SIGMA patented the technology and now no one can use it, except them. Too bad.

You can simulate the Foveon, doing a technique called Drizzle. This software (supposedly) does that:
https://github.com/LucCoiffier/DSS/
Yes Sigma bought Foveon and are the only one who can make it but both Sony and Canon are trying to find how to make it, and recently there is this that could be the evolution of the Foveon and maybe can be bought and developed by Canon

thank's for the DSS link going to check it now  :D
:'( win only
I'm using ML2.3 for photography with:
EOS 600DML | EOS 400Dplus | EOS 5D MLbeta5- EF 100mm f/2.8 USM Macro  - EF-S 17-85mm f4-5.6 IS USM - EF 70-200mm f/4 L USM - 580EXII - OsX, PS, LR, RawTherapee, LightZone -no video experience-

ArcziPL

Quote from: scrax on April 22, 2019, 01:58:28 AM
Noise and ISO are ok only at 100 - 200 and 400 and from reading how it works on dedicated forums it's because is like a camera with only 100 ISO film.
Changing it is just a "memo" for the raw processor to push up exposure in post, that's why it gives so much noise at high ISOs
But sharpness and details are great.
If noise is "ok only at 100 - 200 and 400" then the dynamic range is bad. Nikon, Sony and Pentax are also having so called ISO-less sensors, i.e. it doesn't matter if you dial in ISO 3200 or ISO 100 and push the file by +5EV in postprocess: the results will be the same.

I believe this comparison should tell all:
https://www.dpreview.com/reviews/image-comparison/fullscreen?attr144_0=sigma_sdquattroh&attr144_1=nikon_d850&attr144_2=canon_eosm50&attr144_3=canon_eos5dmkiv&attr146_0=100_6&attr146_1=100_6&attr146_2=100_6&attr146_3=100_6&normalization=compare&widget=645&x=0.1367195625070196&y=0.5167449200029397

sd Quattro H vs. D850 vs. M50 vs. 5D Mark IV. All with same subject, same exposure, ISO100 + 6EV.


Quote from: scrax on April 22, 2019, 01:58:28 AM
Seems also to have a lot of potentiality in recovering highligts (at least).
If there is potential to recover highlights, then the photos are probably not ETTR-er. No surprise, as all DSLRs I know tend to underexposure.
M50.110 [main cam] | G7X III [pocket cam] | 70D.112 [gathers dust] | M.202 [gathers dust] | waiting for M5II

scrax

Quote from: ArcziPL on April 22, 2019, 11:35:23 PM
If noise is "ok only at 100 - 200 and 400" then the dynamic range is bad. Nikon, Sony and Pentax are also having so called ISO-less sensors, i.e. it doesn't matter if you dial in ISO 3200 or ISO 100 and push the file by +5EV in postprocess: the results will be the same.
If there is potential to recover highlights, then the photos are probably not ETTR-er. No surprise, as all DSLRs I know tend to underexposure.

From my readings (and experiences on bayer) what you say is right for bayer kind sensor but foveon is not working same way because how it works and how the RAW files are generated (they are not really the RAW data from the sensor, for basically two reason one tecnical and one purely commercial (Sigma hides the algoritms to reconstruct colors by not giving out pure raw data from sensor).

For what I do now (cave/mines photography) my tipical subject is in totally dark enviroment with tripod and a lot of flashes or old magnesium bulb. So I have direct control on scene light like in studio and can shot at low ISO's.
That's basically why I'm looking at those sensors, but also seems that they are working on a new FF sensor for 2020 with L-mount, so I think I'll wait and try to learn more about this before making a decision.

I have basically only EF lens so going to the new mount R, M or L (the new mount Sigma, Leica and Panasonic are working on together) will be almost the same for me (all them will work with a 600$ adapter for EF  ::) )
I'm using ML2.3 for photography with:
EOS 600DML | EOS 400Dplus | EOS 5D MLbeta5- EF 100mm f/2.8 USM Macro  - EF-S 17-85mm f4-5.6 IS USM - EF 70-200mm f/4 L USM - 580EXII - OsX, PS, LR, RawTherapee, LightZone -no video experience-

Luther

Quote from: scrax on April 22, 2019, 02:08:40 PM
[...] and recently there is this that could be the evolution of the Foveon and maybe can be bought and developed by Canon

Nice research. I hope no one closes it with patents. This could represent a new evolution in color precision, if the sensor comes with the spectral sensitivity table (to use as IDT on ACES - as each camera has variation on color absorption).

Quote from: ArcziPL on April 22, 2019, 11:35:23 PM
I believe this comparison should tell all:
https://www.dpreview.com/reviews/image-comparison/fullscreen?attr144_0=sigma_sdquattroh&attr144_1=nikon_d850&attr144_2=canon_eosm50&attr144_3=canon_eos5dmkiv&attr146_0=100_6&attr146_1=100_6&attr146_2=100_6&attr146_3=100_6&normalization=compare&widget=645&x=0.1367195625070196&y=0.5167449200029397

sd Quattro H vs. D850 vs. M50 vs. 5D Mark IV. All with same subject, same exposure, ISO100 + 6EV.

You can clearly see Foveon has more fine details when more light is available. On 1/5 speed, look on the grass areas.

Quote from: scrax on April 23, 2019, 12:28:26 AM
For what I do now (cave/mines photography) my tipical subject is in totally dark enviroment with tripod and a lot of flashes or old magnesium bulb. So I have direct control on scene light like in studio and can shot at low ISO's.
That's basically why I'm looking at those sensors, but also seems that they are working on a new FF sensor for 2020 with L-mount, so I think I'll wait and try to learn more about this before making a decision.

Your best bet for now would be Hasselblad... if you can afford it.

scrax

Quote from: Luther on April 23, 2019, 03:12:35 AM
Your best bet for now would be Hasselblad... if you can afford it.

I'm feared just to check the prices  :P

EDIT:
Here the link to the paper of the research mentioned before...
First thinkg noticed: "The IR fraction of light remains unabsorbed" so no IR photo seems

attached pic of the one pixel sensor they made  :D
I'm using ML2.3 for photography with:
EOS 600DML | EOS 400Dplus | EOS 5D MLbeta5- EF 100mm f/2.8 USM Macro  - EF-S 17-85mm f4-5.6 IS USM - EF 70-200mm f/4 L USM - 580EXII - OsX, PS, LR, RawTherapee, LightZone -no video experience-

ArcziPL

After downloading RAWs from the above mentioned linke, setting "all to 0", no sharpening, no denoising, white balance set on a gray card in the middle for each shot separately:
- sd Quattro H has an amazing sharpness and nearly no moire/aliasing artifacts. The text below the color checker is readable to the last line. Way better than D850 and 5DIV, despite much lower resolution of the final image (45 and 30 vs 25MPx).
- color checker looks virtually idenctical on D850 and 5DIV but sd Quattro H is very different. Blues, oranges, browns, violets are heavily affected. Despite I somehow "like it" (does it matter? or should the photo just match reality perfectly?), I don't know what is closer to real, as I can't compare it with the color checker itself.
- on sd Quattro H the blue channel is noisy already on well exposed parts (look at blues of the color checker!). I wonder how is it usable for landscape shots with a lot of blue sky...
- response curve is different; has to push shadows by +0,5EV to match exposure of D850 and 5DIV
- +1EV and the image from sd Quattro H already gets awful, large colored spots over the whole dark gray background. Mostly prominent close to image edges. Just with +1EV! For me, such a cam would be useless.

I can live with Bayer, for me it's a sweet spot. No wonder it's commonly adopted. No breath-holding what near future brings. :)


EDIT:
If someone argues about the choice of competitors: just checked the prices and ok, fine, comparing a 1,3k€ cam with a 3k€ might be unfair. Just take then 80D instead of 5DIV and D7500 instead of D850. The outcome stays same.
M50.110 [main cam] | G7X III [pocket cam] | 70D.112 [gathers dust] | M.202 [gathers dust] | waiting for M5II

scrax

Quote from: ArcziPL on April 23, 2019, 11:51:39 AM
After downloading RAWs from the above mentioned linke, setting "all to 0", no sharpening, no denoising, white balance set on a gray card in the middle for each shot separately:
- sd Quattro H has an amazing sharpness and nearly no moire/aliasing artifacts. The text below the color checker is readable to the last line. Way better than D850 and 5DIV, despite much lower resolution of the final image (45 and 30 vs 25MPx).
- color checker looks virtually idenctical on D850 and 5DIV but sd Quattro H is very different. Blues, oranges, browns, violets are heavily affected. Despite I somehow "like it" (does it matter? or should the photo just match reality perfectly?), I don't know what is closer to real, as I can't compare it with the color checker itself.
- on sd Quattro H the blue channel is noisy already on well exposed parts (look at blues of the color checker!). I wonder how is it usable for landscape shots with a lot of blue sky...
- response curve is different; has to push shadows by +0,5EV to match exposure of D850 and 5DIV
- +1EV and the image from sd Quattro H already gets awful, large colored spots over the whole dark gray background. Mostly prominent close to image edges. Just with +1EV! For me, such a cam would be useless.

I can live with Bayer, for me it's a sweet spot. No wonder it's commonly adopted. No breath-holding what near future brings. :)


EDIT:
If someone argues about the choice of competitors: just checked the prices and ok, fine, comparing a 1,3k€ cam with a 3k€ might be unfair. Just take then 80D instead of 5DIV and D7500 instead of D850. The outcome stays same.

I'm totally agree, even foveon users are divided because the foveon quattro is not like the predecessor and a lot of them like the older (merrill) more (quattro has 4 time more pixel in the blue layer than the other two, to let the camera be more faster, maybe that's also why blue is not responding the same as other channel in your tests).
Another thing to consider is that the dng raw gives worst result than the x3f edited in ProPhoto app from sigma (if understood right it has some specific settings like x3f fill light and something similar to picture styles that can be edited only from the proprietary file and not on the dng).

I think that is interesting the dp1 merril, a sort of compact camera with aps-c, fixed lens and great detail around 300€ (used and really hard to find) but limited a lot in ISO, AF and other things like it takes like 3sec to save a pic and can't be reviewed during this time...
It's not a "good" camera or an all around camera but for me is interesting at leat the concept.
But the poor performances at high ISO's are really a pita for me too.

I'll just keep an eye on it for the future, maybe in some years there will be some evolution that will make the foveon or another layered cmos better than a bayer not only on details and colors (the thing that I like more than bayer, in the comparisons I saw, too)

AH one thing that is nice and could be interesting also on canon cameras is the IR filter of the sigma mirrorless is easily removable with a pair of tweezers. So you can remove it, put a filter for visible light on the lens and have a session of IR photography and when finished put back in the IR filter in the body and you are back to standard photography without permanently mod the camera
I'm using ML2.3 for photography with:
EOS 600DML | EOS 400Dplus | EOS 5D MLbeta5- EF 100mm f/2.8 USM Macro  - EF-S 17-85mm f4-5.6 IS USM - EF 70-200mm f/4 L USM - 580EXII - OsX, PS, LR, RawTherapee, LightZone -no video experience-

Luther

Quote from: ArcziPL on April 23, 2019, 11:51:39 AM
Despite I somehow "like it" (does it matter? or should the photo just match reality perfectly?)

Should always match reality, but hold as much information as possible. Post-production is here if you want to change the look.

Quote
- +1EV and the image from sd Quattro H already gets awful, large colored spots over the whole dark gray background. Mostly prominent close to image edges. Just with +1EV! For me, such a cam would be useless.

That's why you should always expose right  ¯\_(ツ)_/¯
Seriously now, sometimes even with artificial light and fast lenses you can't get enough... in those cases, CMOS Bayer seems to work better.

Quote
I can live with Bayer, for me it's a sweet spot. No wonder it's commonly adopted. No breath-holding what near future brings. :)

Indeed, we can live with Bayer, but we should go always beyond.
I think more important than asking "how" to create better sensors is also "why". What cameras need more than a Alexa 65 can give now? Signal to noise ratio? Dynamic range? Well, Alexa already have a great ISO performance and Kodak films already have enormous DR. So what can we push more? Foveon offers a solution for aliasing, color precision and fine detail preservation... that's nice. Alexa/film is too expensive? Then the point is to make the technology not "better" but "cheaper".
I think people should ask those questions more frequently.

Quote from: scrax on April 23, 2019, 12:32:14 PM
like it takes like 3sec to save a pic and can't be reviewed during this time...

This is just a matter of adapting to other advancements in industry. For example, the issue you mentioned could be solved with better ASIC, using parallel programming, better/faster compression algorithms and better flash drives:
http://parallel.princeton.edu/openpiton/
http://fadu.io/
https://github.com/facebook/zstd


BTW, I'm not saying Foveon is ready. It's clearly not and, in my opinion, never will be because of patents.

ps: Anyone knows if AXIOM project is still going?

scrax

Quote from: Luther on April 23, 2019, 01:08:03 PM
This is just a matter of adapting to other advancements in industry. For example, the issue you mentioned could be solved with better ASIC, using parallel programming, better/faster compression algorithms and better flash drives:
http://parallel.princeton.edu/openpiton/
http://fadu.io/
https://github.com/facebook/zstd

Also with the old way bigger/faster cpu instead of proper software like the computer world has used us  ;D :D
(just joking)

I agree with you a lot to be serious.

I discovered the Foveon by looking for something like the Leica M monochrom, that is someway better than bayer and foveon together just because it takes only B/W pic, no bayer no layer no colors to deal with only luminosity...
And with some old tecnique you can have color pics like this  :P
I'm using ML2.3 for photography with:
EOS 600DML | EOS 400Dplus | EOS 5D MLbeta5- EF 100mm f/2.8 USM Macro  - EF-S 17-85mm f4-5.6 IS USM - EF 70-200mm f/4 L USM - 580EXII - OsX, PS, LR, RawTherapee, LightZone -no video experience-

ArcziPL

Quote from: Luther on April 23, 2019, 01:08:03 PM
Indeed, we can live with Bayer, but we should go always beyond.
I think more important than asking "how" to create better sensors is also "why".
(...)
I think people should ask those questions more frequently.
Generally I agree with you. But speaking for myself: I started my adventure with photography in 2004 with Olympus C4000 Zoom. A compact with pretty good IQ at that time, f/2.8 and RAW format with a hack. At that time I was very unhappy with the DR and noise performace. One had to use ISO100, ETTR perfectly and the DR was still too low to cover very typical daylight scenarios. Burnt-out sky and heavy noise in shadows were typical for outdoor photos. Cropping in post was not really cool as there won't be much left from 3,92Mpx afterwards. :) And the cam was also soooo slooow (especially shooting the 6MB RAW monsters ;P). Yes, this setup was limiting me. I was dreaming about super-high DR and also low light performace. Fujifilm Finepix S3 Pro with the special type of sensor consisting of S- and R-pixels for increased DR was a dream. Same as lenses f/1.8 and faster. I remember checking out some samples from the net and being amazed how much it was possible to recover from highlights.

With my actual equipment I am the limit, not my hardware. Especially that I hardly find time for photographing outside holiday travels. The HW is nearly perfect. Better cam won't make my photos better (with the exception of damn BF/FF using the phase AF in DSLRs, which in my case improved with 70D due to AFMA and is finally gone with M50). That's why I don't look what future brings. I have the HW which already is good enough. Just need to use it more and better. :)

BTW: after all the years I've just started reviewing and tidying up my photo collection few days ago, which was only growing in last 15 years. I didn't delete bad shots or shots I don't need anymore until now. There are also many sessions which were never converted from RAW. :) And seeing the photos from C4000Z: yeah, really decent cam as for 2004. The shots are still perfectly usable. But the DR is scary low. :) There is so small room for postprocessing... One can really see the progress of 15 years.
M50.110 [main cam] | G7X III [pocket cam] | 70D.112 [gathers dust] | M.202 [gathers dust] | waiting for M5II