Full frame

Started Jul 5, 2012 | Discussions
Mike Stratil Contributing Member • Posts: 873
Full frame

I don't own an SD1 but I like the foveon look. I'm wondering if there are any indications about whether Sigma is going to come out with a full frame version any time soon.

If this has been discussed in previous posts, perhaps someone could point me toward them. I did a search and didn't find anything among the recent posts.
--
Michael S

Sigma SD1
If you believe there are incorrect tags, please send us this post using our feedback form.
maple Veteran Member • Posts: 3,215
A rather remote prospect

I'm afraid, if at all. Most would agree?

If the SD1 sensor does not do well against Bayer competition of same size, there's little reason to think it would fare better (or worse) scaling up to FF, or scaling down, for that matter. It's too much money for a too risky venture.
--
Maple

Paul Petersen
Paul Petersen Senior Member • Posts: 1,295
We can only dream for now.

Sorry but there is no know activity that we know of. Sigma is running seriously thin in the margin department building the Foveon only for its own cameras that sell in small volumes. Also the sensor is difficult to manufacture using multiple epitaxial layers to create the vertical color array.

The CEO Yamaki-sam has publicly stated it as their dream to build the sensor as a full frame. It would be a world beater considering the resolution the SD1 can generate. With 2.25X the surface area no current FF sensor could begin to touch it.
Pete

-- hide signature --

A bad day of train chasing is better than a good day at work.
http://peterzpicts.smugmug.com/

 Paul Petersen's gear list:Paul Petersen's gear list
Sigma SD14 Nikon D7500 Sigma 18-200mm F3.5-6.3 DC OS HSM Nikon AF-S DX Nikkor 18-140mm F3.5-5.6G ED VR Sigma 18-35mm F1.8 DC HSM Art +2 more
OP Mike Stratil Contributing Member • Posts: 873
Re: Full frame

Thanks for these comments.

Is there any way to use Canon EF lenses on the SD1 M? I've looked for adapters but have not found any.

-- hide signature --

Michael S

amdme127 Regular Member • Posts: 483
Re: We can only dream for now.

Paul Petersen wrote:

Sorry but there is no know activity that we know of. Sigma is running seriously thin in the margin department building the Foveon only for its own cameras that sell in small volumes. Also the sensor is difficult to manufacture using multiple epitaxial layers to create the vertical color array.

The CEO Yamaki-sam has publicly stated it as their dream to build the sensor as a full frame. It would be a world beater considering the resolution the SD1 can generate. With 2.25X the surface area no current FF sensor could begin to touch it.

I agree with Paul, it is a wish they (Sigma) have and are working towards, but nobody knows the time table. Why you ask? Sigma recently changed their process in creating a sensor about 2 years ago after they purchased Foveon which should shorten their plan to product transition phase is shorten, but it is still too early to tell by how much and start to make predictions on how long it will take.

I personally hope that their full frame sensor (if that is their next sensor and not another crop like a 1.3) is not 2.25x the resolution of their current sensor. I think the pixel density is too high for me, most of my EX (Pro Glass, like L series) lenses wouldn't be great for such a high resolution sensor. I rather a little sacrifice of pixels for a more usable sensor overall that would allow a little better ISO sensitivity with higher quality (each ISO step will have better image quality) and faster operation (more photos fit in the buffer). 25mp to 28mp full frame would be awesome, I would jump at the camera very quickly
--
Matt

"Knowledge, absolutely sure of its infallibility is faith" -Yevgeny Zamyatin

http://ureye-studio.com

 amdme127's gear list:amdme127's gear list
Sigma dp2 Quattro Sigma SD15 Sigma SD9 Sigma SD14 Sigma sd Quattro +15 more
rf-design
rf-design Contributing Member • Posts: 522
Re: We can only dream for now.

I think FF add a second complicated processing step to the sensor manufacturing. The image size of the 10:1 DUV-optical steppers is too small for FF and also 24x16 need special, more relexed, layout design rules for the chip edges. In addition to the three deep charge collection implants they have to process the optical stepping outside of the current foundry.

It is true that the current lenses are better matched to FF if they are designed for. But they are mostly not designed for the actual pixel density. So I expect that new lens designs are prefered for APS-C. To me it seems that FF provide only in low light situations a one stop advantage. A 1.4/50mm gives in principle not more resolution than a 1.4/30mm wide open. That is because if both are similar they are limited by geometrical abberations. So you get one stop advantage. If you need at the same time more deep of focus there is no significant advantage for FF, both in terms of noise and resolution.

 rf-design's gear list:rf-design's gear list
Panasonic LX100 Sigma SD14 Sigma SD15 Sigma SD1 Merrill Sony SLT-A57 +57 more
Joerg V Contributing Member • Posts: 636
Won't happen. Alas!

A full frame foveon would be a dream but keep in mind that Sigma is a large lens making company but a small camera making company. The investment for such a sensor would not pay off.

And image what extraordinary demands such a sensor (let's assume current Foveon * 2 = 30MP FF) would put on the lens!

 Joerg V's gear list:Joerg V's gear list
Sigma DP1 Nikon D800 Nikon AF-S Nikkor 50mm f/1.8G Sigma 35mm F1.4 DG HSM Art Nikon AF-S Micro-Nikkor 60mm F2.8G ED +3 more
HBowman
HBowman Senior Member • Posts: 1,237
Re: Full frame

Building a FF foveon will be not more complicated than building the actual 1.5 Foveon.

Actually, our 1.5 foveon do have very small sensels. Carving very small sensels is costly and a pain in the ass (especially when you do it kinda alone like SIGMA).

Building actual FF like D700 or 5DII is, by far, as difficult as building actual Merrill Foveon because of the 3 layers.

Actually, on hearth at least, the 1.3 APS-H format made by Canon (and some other brands) is the biggest sensor who can be made in 1 ONE "step". All sensors, from tiny pinhole to APS-H are made in only one step, which is why they "should be inexpensive".

For FF, there is more than one step to produce 1 sensor. There is 3 steps, like the 3 layers in the foveon. This is why they are more expensive than APS-C and APS-H (maybe not anymore with the venture of "cheap-eco" sensors sold by Sony to every body who is ok back the pants).

A 15 MP FF foveon can be less difficult to achieve than the actual Merrill generation.

Just need probably new machines and new Body.

MESSAGE TO SIGMA:

If you go in this venture, you must do a FF DP. This is obligatory !!

-- hide signature --
Michael Kaminski Regular Member • Posts: 280
Re: Won't happen. Alas!

The current SD1 is stuck in relation to it's Bayercompetitors where the SD14 was as it came out and where the SD10 was at it's best times. You get 1,5 of it's nominal resolution plus the Foveon clarity, minus high ISO capacity, minus writespeed and fps. That's been the equation pretty much all the time a new SD came out. THe SD1 had the chance to break this cycle and have an advantage over Bayer for at least a year or so before Bayer cameras with better specs came out like the D800. Pretty soon the D600 will arrive for let's say 1500-1800 bucks, delivering equal Bayer resulution with way better High ISO capabilities, fast fps, large buffer, dual card slot and software compatibility.

The SD1 launch was goofed up by the high price, which not only made it a non seller but also took away customers conidence.

If SIGMA want's to get attractive for customers again, they shoud come off their track trying to become a high priced company and build good products for a widely more attractive price in comparison to their competitors. Also lenses should get cheaper get more traction in the market.

If the next SD is full frame, it will likely function according to the above equation. If high ISO can get better to deliver clean shadows with good detail, low noise and good color from 400-1600 it would be a big improvement. They have to try to launch ot at least a year before Bayer competition and at a very attractive pricepoint. Otherwise they will never get out of the hole.

If the marketing strategy of the SD1 is what goes on behind the scenes, SIGMA will loose more customers. Since going to astronomic prices with a product of which you think it will be a winner will place them in a position where all customer trust will be lost, since no one will buy from a company that leaves you in the rain.

For my part I'm into NIKON and I'm really torn between the D800 and SD1. If the SD1 price fiasco would not have taken place the SD1 had a clear advantage. But now I'm suspicious something like this will happen again and I'm left with SIGMA Glass and a new body option at an unaffordable price range. It's not directly blackmailing, but not very customer friendly either. So for the time beeing it's clearly advantage NIKON.
Greets
Michael

rf-design
rf-design Contributing Member • Posts: 522
Re: Full frame

HBowman wrote:

Building a FF foveon will be not more complicated than building the actual 1.5 Foveon.

Wrong: Wafer processing of 36mmx24mm is significant more complicated because there is no stepper outside Sony and Canon

Actually, our 1.5 foveon do have very small sensels. Carving very small sensels is costly and a pain in the ass (especially when you do it kinda alone like SIGMA).

Building actual FF like D700 or 5DII is, by far, as difficult as building actual Merrill Foveon because of the 3 layers.

Actually, on hearth at least, the 1.3 APS-H format made by Canon (and some other brands) is the biggest sensor who can be made in 1 ONE "step". All sensors, from tiny pinhole to APS-H are made in only one step, which is why they "should be inexpensive".

For FF, there is more than one step to produce 1 sensor. There is 3 steps, like the 3 layers in the foveon. This is why they are more expensive than APS-C and APS-H (maybe not anymore with the venture of "cheap-eco" sensors sold by Sony to every body who is ok back the pants).

The wafer processing involves about 20-30 mask steps. If the image size of the stepper is smaller than FF you need 2,4,6 and so on steps per mask and stitching them. Not 3 or any odd number. FF is stitched 2 times but there are drawbacks. If you stitch you loose accuracy and resolution. That is because w/o stitching there could be some amount of image distortion, significant more than resolution. But if you stitch you need to align a left hand edge of the first to a right hand edge of the second mask. Because the distortion is not perfect radial but have also have decentering you loose stitching accuracy. That is a secret to optical stepper makers Nikon and Canon. And I think it has less to do with cost than control.

A 15 MP FF foveon can be less difficult to achieve than the actual Merrill generation.

Just need probably new machines and new Body.

MESSAGE TO SIGMA:

If you go in this venture, you must do a FF DP. This is obligatory !!

-- hide signature --
 rf-design's gear list:rf-design's gear list
Panasonic LX100 Sigma SD14 Sigma SD15 Sigma SD1 Merrill Sony SLT-A57 +57 more
HBowman
HBowman Senior Member • Posts: 1,237
Re: Full frame

rf-design wrote:

HBowman wrote:

Building a FF foveon will be not more complicated than building the actual 1.5 Foveon.

Wrong: Wafer processing of 36mmx24mm is significant more complicated because there is no stepper outside Sony and Canon

Actually, our 1.5 foveon do have very small sensels. Carving very small sensels is costly and a pain in the ass (especially when you do it kinda alone like SIGMA).

Building actual FF like D700 or 5DII is, by far, as difficult as building actual Merrill Foveon because of the 3 layers.

Actually, on hearth at least, the 1.3 APS-H format made by Canon (and some other brands) is the biggest sensor who can be made in 1 ONE "step". All sensors, from tiny pinhole to APS-H are made in only one step, which is why they "should be inexpensive".

For FF, there is more than one step to produce 1 sensor. There is 3 steps, like the 3 layers in the foveon. This is why they are more expensive than APS-C and APS-H (maybe not anymore with the venture of "cheap-eco" sensors sold by Sony to every body who is ok back the pants).

The wafer processing involves about 20-30 mask steps. If the image size of the stepper is smaller than FF you need 2,4,6 and so on steps per mask and stitching them. Not 3 or any odd number. FF is stitched 2 times but there are drawbacks. If you stitch you loose accuracy and resolution. That is because w/o stitching there could be some amount of image distortion, significant more than resolution. But if you stitch you need to align a left hand edge of the first to a right hand edge of the second mask. Because the distortion is not perfect radial but have also have decentering you loose stitching accuracy. That is a secret to optical stepper makers Nikon and Canon. And I think it has less to do with cost than control.

Nikon, Canon and Sony

... And I continue to think that SIGMA have hard time crafting Foveons, which let me think that they will not have "very big" problems to build FF ones.

Foveon is one (if not the most) complicated sensor to craft.

-- hide signature --
Rodrigo Cunha Contributing Member • Posts: 701
Re: We can only dream for now.

rf-design wrote:

It is true that the current lenses are better matched to FF if they are designed for. But they are mostly not designed for the actual pixel density. So I expect that new lens designs are prefered for APS-C. To me it seems that FF provide only in low light situations a one stop advantage. A 1.4/50mm gives in principle not more resolution than a 1.4/30mm wide open. That is because if both are similar they are limited by geometrical abberations. So you get one stop advantage. If you need at the same time more deep of focus there is no significant advantage for FF, both in terms of noise and resolution.

A 50mm f/2 full frame and a 30mm f/1.4 aps/c might be similar as far as resolution, light gathering power and depth of field goes, but the 50 f/2 is simpler to design, cheaper, and ends up having better overall image quality, is more telecentric with less micro-lens vignetting, etc

A large exit aperture is a problem! This has implications in lens design and availability. For example, where is the high quality 24-105 f/4 equivalent for aps/c, which would have to be 16-70 f/2.6 ?

As you go up into medium format you can work with modest relative apertures and still get shallow depth of field. This is a real advantage in practice and translates into better overall image quality.

So, size matters and might even reduce total system cost while upgrading quality

Sigma Nick Forum Member • Posts: 95
Re: We can only dream for now.

Sigma USA replied to me in an unofficial email that they have no intention of putting out a full frame foveon sensor anytime "soon". Basically because of cost vs. demand and also because their line of "DC" lenses would not work on a FF sensor. This was in February 2010 when I was waiting for the SD15 to come out.

Halldor Eiriksson Senior Member • Posts: 1,989
Re: Full frame

Looks like there will not be a FF Foveon. Now I'm tempted to wait for the D600 and some Samyang lenses to go with it, I want Full Frame.
--
Halldor
http://www.flickr.com/photos/eirod/

 Halldor Eiriksson's gear list:Halldor Eiriksson's gear list
Nikon D7200 Sigma 18-35mm F1.8 DC HSM Art Nikon AF-S Nikkor 70-200mm F4G ED VR Nikon AF-S DX Nikkor 18-140mm F3.5-5.6G ED VR
yvind Strm Veteran Member • Posts: 4,130
Re: Full frame

Hi Mike

Your dreams will come through.
I am sure Foveon/Sigma is working on it now.

I predicted that they would announce a FF SD based on the resolution in the SD14/15, but instead they announced the SD1.

Remember that they (Sigma/Foveon) anticipated where the competition would be, and designed the new camera after that. Expect that to happen again.
Sigma HAS said that they will launch a FF.
As the D800/D800E has pushed the SD1 from the throne, Sigma will retalliate

I will not be surprised if Sigma announce a FF this autumn, at Photokina, in september. But the big question is lenses, and getting the buffer large enough.

Mike Stratil wrote:

I don't own an SD1 but I like the foveon look. I'm wondering if there are any indications about whether Sigma is going to come out with a full frame version any time soon.

If this has been discussed in previous posts, perhaps someone could point me toward them. I did a search and didn't find anything among the recent posts.
--
Michael S

D Cox Forum Pro • Posts: 21,340
Re: We can only dream for now.

amdme127 wrote:

I agree with Paul, it is a wish they (Sigma) have and are working towards, but nobody knows the time table. Why you ask? Sigma recently changed their process in creating a sensor about 2 years ago after they purchased Foveon which should shorten their plan to product transition phase is shorten, but it is still too early to tell by how much and start to make predictions on how long it will take.

It would be surprising to see anything before 2015.

I personally hope that their full frame sensor (if that is their next sensor and not another crop like a 1.3) is not 2.25x the resolution of their current sensor. I think the pixel density is too high for me, most of my EX (Pro Glass, like L series) lenses wouldn't be great for such a high resolution sensor.

But the sensor should be better than your best lens. Then you know that you are getting all the detail the lenses can give.

I rather a little sacrifice of pixels for a more usable sensor overall that would allow a little better ISO sensitivity with higher quality (each ISO step will have better image quality) and faster operation (more photos fit in the buffer). 25mp to 28mp full frame would be awesome, I would jump at the camera very quickly
--

Certainly the Foveon sensors need less noise (= higher ISO).

Matt

"Knowledge, absolutely sure of its infallibility is faith" -Yevgeny Zamyatin

Ignorance, absolutely sure of its infallibility, is faith

http://ureye-studio.com

amdme127 Regular Member • Posts: 483
Re: We can only dream for now.

D Cox wrote:

amdme127 wrote:

I agree with Paul, it is a wish they (Sigma) have and are working towards, but nobody knows the time table. Why you ask? Sigma recently changed their process in creating a sensor about 2 years ago after they purchased Foveon which should shorten their plan to product transition phase is shorten, but it is still too early to tell by how much and start to make predictions on how long it will take.

It would be surprising to see anything before 2015.

I think it is overall hard to tell, since SD1 seems to be the only sensor Sigma has made with the new process and we don't know exactly when they started and also what all they will need to account for in developing a full frame. You may very well be right, I have no basis to make an estimate.

I personally hope that their full frame sensor (if that is their next sensor and not another crop like a 1.3) is not 2.25x the resolution of their current sensor. I think the pixel density is too high for me, most of my EX (Pro Glass, like L series) lenses wouldn't be great for such a high resolution sensor.

But the sensor should be better than your best lens. Then you know that you are getting all the detail the lenses can give.

If like the SD1/SD1M in being that much better, the lens will actual blur details if it is not up to it. I would like the sensor to outperform the lenses, but hopefully not at the amount that I feel I need to replace my collection of good full frame EX glass lenses.

I rather a little sacrifice of pixels for a more usable sensor overall that would allow a little better ISO sensitivity with higher quality (each ISO step will have better image quality) and faster operation (more photos fit in the buffer). 25mp to 28mp full frame would be awesome, I would jump at the camera very quickly
--

Certainly the Foveon sensors need less noise (= higher ISO).

Exactly, I want a combinition of higher ISO capability, but I really also want to see much better performance at each ISO level (i.e. 100, 200, 400, 800...etc). It doesn't need to compete with Bayer, but if you could get a decent image at ISO 6400 and be able to fully use ISO 9600 or ISO 12800 for black and white, that would be awesome.

Matt

"Knowledge, absolutely sure of its infallibility is faith" -Yevgeny Zamyatin

Ignorance, absolutely sure of its infallibility, is faith

Too true, too true indeed sir

http://ureye-studio.com

-- hide signature --

Matt

"Knowledge, absolutely sure of its infallibility is faith" -Yevgeny Zamyatin

http://ureye-studio.com

 amdme127's gear list:amdme127's gear list
Sigma dp2 Quattro Sigma SD15 Sigma SD9 Sigma SD14 Sigma sd Quattro +15 more
Keyboard shortcuts:
FForum MMy threads