Move on from Foveon

[DavidMillier]

20 Megapixels for a FFF is definitely not enough.

For whom?

It's much more than enough for me!

Entry level today is 24 Mpix, and customers expect a low price for that. It's no good saying "But these are better pixels" as this can be countered by "I don't want a weird camera."

If that were true, Nikon wouldn't be making this brand new camera:

https://www.bhphotovideo.com/c/product/1515330-REG/nikon_d6_dslr_camera_body.html

That's a special use model, meant for sports photographers who want very fast burst rates. The price is extremely high.

Is this also a "special use" model?

https://www.bhphotovideo.com/c/product/1547010-REG/canon_eos_r6_mirrorless_digital.html

The fact is there are many high-end digital cameras that come with a 20 MP sensor. There are hundreds of digital cameras listed at B and H, when I searched for cameras with 20 MP sensors, and they include many of the m4/3 cameras, the top-of-the-line Canon and Nikon cameras, including that one I first linked to, which is Nikon's newest professional camera, if I'm not mistaken, and all the cameras with 1" sensors, but there are still cameras on the market with sensors that capture fewer than 20 MP. I don't think it's accurate to say that entry level is more than 20 MP.

My idea of an entry level camera is the Sony A6000, now discontinued.

Here's my idea of an entry level camera:

https://www.bhphotovideo.com/c/prod...ar_vx054_blk_10_1mp_digital_camera_black.html

Here's a nice step up to a "quality" Kodak entry level camera:

https://www.bhphotovideo.com/c/product/1318517-REG/kodak_fz53_bk_pixpro_fz53_compact_digital.html

Here's my idea of an entry level pro-sumer camera (also a Kodak, surprisingly):

https://www.bhphotovideo.com/c/product/1391175-REG/kodak_az652bk_pixpro_az652_astro_zoom.html

Here's my idea of an entry level interchangeable lens pro-sumer

https://www.bhphotovideo.com/c/product/1351010-REG/olympus_v207072bu010_om_d_e_m10_mark_iii.html

Here's my idea of an entry level interchangeable lens pro-sumer DSLR camera:

https://www.bhphotovideo.com/c/product/1238183-REG/canon_1159c003_eos_rebel_t6_dslr.html

. . . and here's my idea of an entry level interchangeable lens pro-sumer full-frame camera:

https://www.bhphotovideo.com/c/prod...lce7sm2_b_alpha_a7sii_mirrorless_digital.html

The first version would have been a better choice, but they've been discontinued.

This might be a more appropriate choice for an entry level full-frame camera, because of the price:

https://www.bhphotovideo.com/c/product/1459282-REG/canon_eos_rp_mirrorless_digital.html

Of course, if you wanted to stick with Sony products, because you have a Sony TV and a Sony stereo system, you might decide on this, less-expensive Sony camera, with which you get twice the megapixels vs that more expensive Sony:

https://www.bhphotovideo.com/c/prod...ilce7m2_b_a7ii_mirrorless_digital_camera.html

You see . . . entry level can mean a lot of things to a lot of people.

If you are value conscious (which I am), having a historical view is very useful for setting expectations.

For example, this print comparison from a while back:

http://wyofoto.com/EOS_IQ_shootout_2008/EOS_shootout_2008.html by the late Miles Hecker, says that the Canon 5DMkii provides the best 30"x20" prints from the Canon stable of that time (assessment done from a viewing distance of 1 foot). The original 5D marginally lagged behind. Neither of these cameras are entry level.

Looking at DXOmark, my Lumix GX7 that I paid around £140 for in nearly new condition from MPB.com has a sensor that performs as well as the original 5D (in fact it has an extras stop of DR, the only area where the 5D wins is a stop advantage in noise at high iso). The GX7 has many convenience advantages and is half the size and weight.

A camera like the Nikon D3300 (and similar 24MP sensored cameras) outperforms the 5DMkii in everything accept high ISO and is only half a stop behind there.

If you are a serious (but inpecunious) landscape photographer printing at 24" to 30" wide, you can get yourself a nearly new camera with a 6 month warranty for £200 or less that outperforms a former top performing full framer and offers image quality superior to 645 medium format film and rivalling 6x7 film.

If you can avoid casting envious eyes at the latest and greatest, the image quality available today for very low price is quite extraordinary. Back in 2000 when I purchased a Coolpix 950 (2MP), today's largesse was unimaginable. My sense is that state of the art equipment today is already well into the low volume/high margin approach that is starting to make photography a very high priced hobby, but the availability of a good used market makes a big difference at the entry level end.

It would be nice if you could pick up SD1 or SD Q for those kinds of prices but rarity keeps the prices high even though they slow sellers.

--
DPReview gallery: https://www.dpreview.com/galleries/0286305481
Website: http://www.whisperingcat.co.uk/ (2018 - website revived!)
Flickr: http://www.flickr.com/photos/davidmillier/ (very old!)

 

[DavidMillier]

Scott, I don't believe Sigma have ever claimed any of their sensors were equivalent to Bayers if rge same the photosite count. When the SD9 was first marketed they were at first very careful to explain that the equivalence was about x2.

It was only later when they realised that 3.4MP sensors were hard to sell that they started calling them 10MP cameras. But even then they still suggested the resolution was 2x. They left retailers to market it as equivalent to 10MP bayers without correct their error in doing so.

In terms of resolution loss from CFA arrays the most independent way to draw a conclusion is to look at Leica Monochrome res charts. What advantage to they have over the identical colour model? The difference between them will the same ratio as Foveon to an AA filterless Bayer.

 

[DavidMillier]

Arrgh, that looks bad! Will it soldier on with superglue or gaffer tape? An honourable complement to the Leica brassed look....

 

[xpatUSA]

<>

The thing is, that less expensive Sony is not good in low-light situations, where a typical beginner might expect their camera to work. It's only able to meter in -1 EV, while the one I chose initially can meter to -4 EV.

"only"?!!

How about your SD1M ... "only" good for +1EV and that only with a 50mm f/1.4 at 100 ISO ... ;-)

--
WYSINWIG: what you see is not what I got.

 

[D Cox]

[snip]

"Digital SLRs with full-frame 35mm (44 mm diagonal) sensors are the cameras of choice for professionals and serious amateurs who can afford them. The best of them-- the Canon EOS-1Ds Mark II-- has image quality comparable to medium format film (645). The recently-announced EOS 5D comes close, and it is smaller, lighter, and less expensive. The number of pixels in full frame cameras will slowly increase, topping out around 24 megapixels. There would be little benefit from more pixels: resolution would be limited by the lens, and noise, dynamic range, and yield would deteriorate."

He also wrote:

"Fantasy 4: X3 sensor, same spacing as the D60, filling a 24x36mm frame.

4864x3242 (15.8MP)

resolution relative to 35mm film: 1.58

Performance may surpass medium format [film] if such a sensor could be built."

And you call the fp "low resolution"....

In support of which, long ago Clark wrote:

"Too few pixels are bad for image quality and too many pixels are bad for image quality. So there must be an optimum."

See graph here:

https://clarkvision.com/articles/digital.sensor.performance.summary/index.html#FSAIQ

Apparently 5 to 6 um pixel pitch is optimum. Not too far away from the current FFF proposal.

Old article though, before anyone rushes to point that out ...

I have seen no sign at all that image quality is lower on the high resolution cameras.

Hard to comment without a definition of "image quality".

Clark posts his definition in the reference provided, and shows the aforesaid optimum quite clearly in the referenced graph.

I on the hand have measured that sharpness per pixel decreases consistently with pixel pitch.

That would depend on the MTF of the lens. If the pixels are small enough to record the highest spatial frequencies in the image, the sharpness will be near zero.

As you increase the pixel pitch, you slide down the MTF curve.

 

[Scottelly]

20 Megapixels for a FFF is definitely not enough.

For whom?

It's much more than enough for me!

Entry level today is 24 Mpix, and customers expect a low price for that. It's no good saying "But these are better pixels" as this can be countered by "I don't want a weird camera."

If that were true, Nikon wouldn't be making this brand new camera:

https://www.bhphotovideo.com/c/product/1515330-REG/nikon_d6_dslr_camera_body.html

That's a special use model, meant for sports photographers who want very fast burst rates. The price is extremely high.

Is this also a "special use" model?

https://www.bhphotovideo.com/c/product/1547010-REG/canon_eos_r6_mirrorless_digital.html

The fact is there are many high-end digital cameras that come with a 20 MP sensor. There are hundreds of digital cameras listed at B and H, when I searched for cameras with 20 MP sensors, and they include many of the m4/3 cameras, the top-of-the-line Canon and Nikon cameras, including that one I first linked to, which is Nikon's newest professional camera, if I'm not mistaken, and all the cameras with 1" sensors, but there are still cameras on the market with sensors that capture fewer than 20 MP. I don't think it's accurate to say that entry level is more than 20 MP.

My idea of an entry level camera is the Sony A6000, now discontinued.

Here's my idea of an entry level camera:

https://www.bhphotovideo.com/c/prod...ar_vx054_blk_10_1mp_digital_camera_black.html

Here's a nice step up to a "quality" Kodak entry level camera:

https://www.bhphotovideo.com/c/product/1318517-REG/kodak_fz53_bk_pixpro_fz53_compact_digital.html

Here's my idea of an entry level pro-sumer camera (also a Kodak, surprisingly):

https://www.bhphotovideo.com/c/product/1391175-REG/kodak_az652bk_pixpro_az652_astro_zoom.html

Here's my idea of an entry level interchangeable lens pro-sumer

https://www.bhphotovideo.com/c/product/1351010-REG/olympus_v207072bu010_om_d_e_m10_mark_iii.html

Here's my idea of an entry level interchangeable lens pro-sumer DSLR camera:

https://www.bhphotovideo.com/c/product/1238183-REG/canon_1159c003_eos_rebel_t6_dslr.html

. . . and here's my idea of an entry level interchangeable lens pro-sumer full-frame camera:

https://www.bhphotovideo.com/c/prod...lce7sm2_b_alpha_a7sii_mirrorless_digital.html

The first version would have been a better choice, but they've been discontinued.

This might be a more appropriate choice for an entry level full-frame camera, because of the price:

https://www.bhphotovideo.com/c/product/1459282-REG/canon_eos_rp_mirrorless_digital.html

Of course, if you wanted to stick with Sony products, because you have a Sony TV and a Sony stereo system, you might decide on this, less-expensive Sony camera, with which you get twice the megapixels vs that more expensive Sony:

https://www.bhphotovideo.com/c/prod...ilce7m2_b_a7ii_mirrorless_digital_camera.html

You see . . . entry level can mean a lot of things to a lot of people.

If you are value conscious (which I am), having a historical view is very useful for setting expectations.

For example, this print comparison from a while back:

http://wyofoto.com/EOS_IQ_shootout_2008/EOS_shootout_2008.html by the late Miles Hecker, says that the Canon 5DMkii provides the best 30"x20" prints from the Canon stable of that time (assessment done from a viewing distance of 1 foot). The original 5D marginally lagged behind. Neither of these cameras are entry level.

Looking at DXOmark, my Lumix GX7 that I paid around £140 for in nearly new condition from MPB.com has a sensor that performs as well as the original 5D (in fact it has an extras stop of DR, the only area where the 5D wins is a stop advantage in noise at high iso). The GX7 has many convenience advantages and is half the size and weight.

A camera like the Nikon D3300 (and similar 24MP sensored cameras) outperforms the 5DMkii in everything accept high ISO and is only half a stop behind there.

If you are a serious (but inpecunious) landscape photographer printing at 24" to 30" wide, you can get yourself a nearly new camera with a 6 month warranty for £200 or less that outperforms a former top performing full framer and offers image quality superior to 645 medium format film and rivalling 6x7 film.

If you can avoid casting envious eyes at the latest and greatest, the image quality available today for very low price is quite extraordinary. Back in 2000 when I purchased a Coolpix 950 (2MP), today's largesse was unimaginable. My sense is that state of the art equipment today is already well into the low volume/high margin approach that is starting to make photography a very high priced hobby, but the availability of a good used market makes a big difference at the entry level end.

It would be nice if you could pick up SD1 or SD Q for those kinds of prices but rarity keeps the prices high even though they slow sellers.

Yes, and since Sigma does not replace their old bodies with almost identical bodies with a few upgrades, you won't often find used Sigma bodies for rock-bottom prices, like you will with Sony and others. I did find my SD14 very cheap though, which is why I bought it, after my Nikon D5000 was stolen out of my friend's van (in my backpack, with my computer and a bunch of other stuff).

 

[Scottelly]

Arrgh, that looks bad! Will it soldier on with superglue or gaffer tape? An honourable complement to the Leica brassed look....

I don't even know. I haven't tried turning it on. In fact, I haven't even cleaned it up and taken a good look at it yet. I guess I've been too disgusted with myself for letting that happen to my beautiful SD Quattro H. I will decide what to do eventually.

 

[Scottelly]

Scott, I don't believe Sigma have ever claimed any of their sensors were equivalent to Bayers if rge same the photosite count.

Go here and you will see that it says, "Resolution equal to 51MP Bayer sensor on resolution testing"

When the SD9 was first marketed they were at first very careful to explain that the equivalence was about x2.

I see you are right.

"We enlarged the sensor to APS-C size (1.5x focal length equivalent), while narrowing pixel pitch, thereby dramatically raising the pixel count to 46MP (4,800 x 3,200 x 3).

The luminance resolution of this sensor is, in fact, equivalent to that of a 30MP CFA sensor as measured on the standard B&W resolution chart used in conventional digital camera resolution testing."

That's on page 6 of this: http://www.sigma-sd.com/SD1Merrill/pdf/SD1-Merrill_Catalog2012_EN_Monitor.pdf

It was only later when they realised that 3.4MP sensors were hard to sell that they started calling them 10MP cameras. But even then they still suggested the resolution was 2x. They left retailers to market it as equivalent to 10MP bayers without correct their error in doing so.

In terms of resolution loss from CFA arrays the most independent way to draw a conclusion is to look at Leica Monochrome res charts. What advantage to they have over the identical colour model? The difference between them will the same ratio as Foveon to an AA filterless Bayer.

I actually find the resolution of the Sigma sensors to be slightly more than twice the megapixels with some subject matter, while slightly less with other subject matter. In my opinion, the 2x equivalency is good and realistic. It doesn't tell the whole story, but for an easy-to-understand, meaningful number, I think it says a lot, and is not just some inaccurate garbage. It has a good basis in fact too, considering the CFA sensors almost all have half as many green filters as their total, and green is supposed to be the most important color determining luminance resolution, right?

 

[Scottelly]

[snip]

"Digital SLRs with full-frame 35mm (44 mm diagonal) sensors are the cameras of choice for professionals and serious amateurs who can afford them. The best of them-- the Canon EOS-1Ds Mark II-- has image quality comparable to medium format film (645). The recently-announced EOS 5D comes close, and it is smaller, lighter, and less expensive. The number of pixels in full frame cameras will slowly increase, topping out around 24 megapixels. There would be little benefit from more pixels: resolution would be limited by the lens, and noise, dynamic range, and yield would deteriorate."

He also wrote:

"Fantasy 4: X3 sensor, same spacing as the D60, filling a 24x36mm frame.

4864x3242 (15.8MP)

resolution relative to 35mm film: 1.58

Performance may surpass medium format [film] if such a sensor could be built."

And you call the fp "low resolution"....

In support of which, long ago Clark wrote:

"Too few pixels are bad for image quality and too many pixels are bad for image quality. So there must be an optimum."

See graph here:

https://clarkvision.com/articles/digital.sensor.performance.summary/index.html#FSAIQ

Apparently 5 to 6 um pixel pitch is optimum. Not too far away from the current FFF proposal.

Old article though, before anyone rushes to point that out ...

I have seen no sign at all that image quality is lower on the high resolution cameras.

But my argument for having more than 20 Mpix on the FFF is about marketing. I don't think people will buy a low resolution camera in 2022, however magical the pixels.

At 24 Mpix, direct comparisons with the usual Bayer cameras can be made without having to scale one of the images up or down. Moire can be compared.

Remember that potential customers will be looking for reasons not to buy a Sigma camera.

I agree with the marketing thoughts. That's why years ago I said Sigma should make a full-frame "Merrill" sensor with 28 MP per layer, which would make the photo-sites larger than what is on the Merrill sensor, but the native images would be a step up from the images coming out of their flagship SD Quattro H camera, which makes 25.5 MP images. The S-Hi images from such a camera would be more than 50 MP, making it king, if the S-Hi mode would be similar to the one in the SD Quattro H. If it were similar to the one in the SD1 Merrill, then the S-Hi images would be crazy high MP, and it probably wouldn't matter either way (20 MP per layer or 28 MP per layer). At the time though, I think I said that it was the native images that were most important.

Today I would make a full-frame sensor that captures 30 MP per layer, to compete with the 60 MP images from the latest high-resolution Sony camera. I think Sigma should still do this, though that would no doubt delay the new sensor a few months.

 

[Johan Borg]

In terms of resolution loss from CFA arrays the most independent way to draw a conclusion is to look at Leica Monochrome res charts. What advantage to they have over the identical colour model? The difference between them will the same ratio as Foveon to an AA filterless Bayer.

That's a simplification where a weak or sparse CFA may appear to improve resolution while we know that the trade off is in color resolution. A bit like how a Quattro will measure better than a Merrill due to higher pixel count, but real life difference is less obvious.

 

[Scottelly]

<3

 

[xpatUSA]

"We enlarged the sensor to APS-C size (1.5x focal length equivalent), while narrowing pixel pitch, thereby dramatically raising the pixel count to 46MP (4,800 x 3,200 x 3).

The luminance resolution of this sensor is, in fact, equivalent to that of a 30MP CFA sensor as measured on the standard B&W resolution chart used in conventional digital camera resolution testing."

That's on page 6 of this: http://www.sigma-sd.com/SD1Merrill/pdf/SD1-Merrill_Catalog2012_EN_Monitor.pdf

A great pity that they don't tell us what chart and how it was measured. :-(

Oh well, that's Marketing for ya ...

I actually find the resolution of the Sigma sensors to be slightly more than twice the megapixels with some subject matter, while slightly less with other subject matter. In my opinion, the 2x equivalency is good and realistic. It doesn't tell the whole story, but for an easy-to-understand, meaningful number, I think it says a lot, and is not just some inaccurate garbage. It has a good basis in fact too, considering the CFA sensors almost all have half as many green filters as their total, and green is supposed to be the most important color determining luminance resolution, right?

Right. Which still make me wonder about your beloved Quattro top layer ... ;-)

 

[xpatUSA]

I on the [other] hand have measured that sharpness per pixel decreases consistently with pixel pitch.

That would depend on the MTF of the lens.

Many of my comparison measurements were made with the same lens. Some even with the same lens and camera (high versus low resolution mode).

If the pixels are small enough to record the highest spatial frequencies in the image, the sharpness will be near zero.

As you increase the pixel pitch, you slide down the MTF curve.

Are you sure? any credible reference for that? Sorry, Don, but I do have difficulty with one-line assertions ...

... in my tests, you slide up!

--
WYSINWIG: what you see is not what I got.

 

[DavidMillier]

Our car was broken into recently and they stole my daughter's homework. They didn't think much of it, though, left a trail of her stuff down the street. Next morning we got it all back.

 

[DavidMillier]

<>

The thing is, that less expensive Sony is not good in low-light situations, where a typical beginner might expect their camera to work. It's only able to meter in -1 EV, while the one I chose initially can meter to -4 EV.

"only"?!!

How about your SD1M ... "only" good for +1EV and that only with a 50mm f/1.4 at 100 ISO ... ;-)

Histogram works fine in any amount of dark

 

[DavidMillier]

Arrgh, that looks bad! Will it soldier on with superglue or gaffer tape? An honourable complement to the Leica brassed look....

I don't even know. I haven't tried turning it on. In fact, I haven't even cleaned it up and taken a good look at it yet. I guess I've been too disgusted with myself for letting that happen to my beautiful SD Quattro H. I will decide what to do eventually.

A few weeks ago, I was shooting long exposures at Southerness in Scotland. The wind blew the tripod over (the main drawback with carbon fibre tripods). Fortunately, it was a sandy beach so once I cleaned it off, no harm done. It interrupted the dark frame though and I lost the shot I'd been standing in the cold waiting for for 20 minutes. It served as a reminder that disaster can strike any minute. Sometimes luck favours you, sometimes not.

Sorry about your poor camera, I'd be gutted if it was mine.

 

[DavidMillier]

Scott, I don't believe Sigma have ever claimed any of their sensors were equivalent to Bayers if rge same the photosite count.

Go here and you will see that it says, "Resolution equal to 51MP Bayer sensor on resolution testing"

The H is 25MP, I think, so 2x 25 is c51MP. Sounds comparable with their usual marketing claim. They are not simply adding the photosites from the other layers together otherwise they would be claiming 38MP.

When the SD9 was first marketed they were at first very careful to explain that the equivalence was about x2.

I see you are right.

"We enlarged the sensor to APS-C size (1.5x focal length equivalent), while narrowing pixel pitch, thereby dramatically raising the pixel count to 46MP (4,800 x 3,200 x 3).

The luminance resolution of this sensor is, in fact, equivalent to that of a 30MP CFA sensor as measured on the standard B&W resolution chart used in conventional digital camera resolution testing."

That's on page 6 of this: http://www.sigma-sd.com/SD1Merrill/pdf/SD1-Merrill_Catalog2012_EN_Monitor.pdf

It was only later when they realised that 3.4MP sensors were hard to sell that they started calling them 10MP cameras. But even then they still suggested the resolution was 2x. They left retailers to market it as equivalent to 10MP bayers without correct their error in doing so.

In terms of resolution loss from CFA arrays the most independent way to draw a conclusion is to look at Leica Monochrome res charts. What advantage to they have over the identical colour model? The difference between them will the same ratio as Foveon to an AA filterless Bayer.

I actually find the resolution of the Sigma sensors to be slightly more than twice the megapixels with some subject matter, while slightly less with other subject matter. In my opinion, the 2x equivalency is good and realistic. It doesn't tell the whole story, but for an easy-to-understand, meaningful number, I think it says a lot, and is not just some inaccurate garbage. It has a good basis in fact too, considering the CFA sensors almost all have half as many green filters as their total, and green is supposed to be the most important color determining luminance resolution, right?

I think the resolution of Foveon sensors is subject independent. It's the resolution of CFA sensors that is variable so it would probably be better to say that Bayer sensors have about 1/2 their official count compared to Foveon. I suppose that seems too difficult to get across.

The actual relative performance is complicated. It varies shot to shot dependent on several variables: presence or absence of an optical AA filter, sensor size, spectral content of the subject (red/blue being weakest), and crucially aliasing. There is no doubt that because Foveon is sharp and uses no AA filter, aliasing artefacts are often/usually present. Sometimes these create horrible unwanted luminance moire patterns and jaggies that visually corrupt the image, sometimes they create useful fake detail that fools the eye into thinking there is extra detail. It's a lottery. It definitely effects resolution chart determination. Which is where imaging Resource are particularly careful. They cut off their rating as soon as they see any aliasing. Other testers are less diligent. Because aliasing causes repeating areas of clearly aliased patches, followed by false detail that look plausible, the more forgiving can wrongly ignore early aliasing and call the resolution higher than it is.

There is no difference between mono and CFA and Foveon and CFA as far as resolution goes, when there is no AA filter, so the resolution difference between the Monochrom and colour versions of the same Leica models looks like a great independent test of the effect of the CFA.

--
DPReview gallery: https://www.dpreview.com/galleries/0286305481
Website: http://www.whisperingcat.co.uk/ (2018 - website revived!)
Flickr: http://www.flickr.com/photos/davidmillier/ (very old!)

 

[DavidMillier]

Quattro's problem also the noise. I got green color noise even at iso200. 50 f1.4 art lens was really good, but even at 1.4 most times iso100 was just too low to get enough shutter speed handheld. Of course if you have tripod it's not problem. The sensor needs at least clean iso 400 to become competitive. Live only at iso100 after 2020 is a dead end.

I don't think so. As long as the camera is not too expensive, then it will have its uses, even if limited to ISO 100.

Agreed, Scott.

I've NEVER understood the high-ISO fetish exhibited by many digital camera users.

The plain fact is that, unlike film cameras, the digital camera comes loaded with a sensor that has only a fixed, unchangeable ISO (ignoring Aptina dual). Any setting above so-called "base" ISO underexposes the sensor giving a lower Signal-to-Noise Ratio. That is to say that if a camera provides a "clean iso 400" then an iso 100 shot will be 4X cleaner.

As I have said before, my cameras are super-glued in "base" ISO ...

... "not enough light" = no shot in my world.

I used to think that as well. Then I discovered auto-iso. And what i learned from that was that a noisier shot is easier to fix than a shot with camera shake. I would want my cameras shooting at base ISO wherever possible (I wouldn't want to be shooting at higher ISOs as a matter of course) but sometimes getting a shot is better than not getting a shot.

 

[DavidMillier]

The DP2M was launched in 2012. The only current Foveon Sigma camera, the dp and sp Quattro's, were launched in 2014 & 2016. respectively. For electronic devices, they are old. Earlier SD and DP cameras are now very old. Only Ricoh with its GR has had such a long periods without new models.

The cameras' ages alone, makes it no surprise that people "move on". These old cameras will fail entirely or end up with intermittent faults and their values make them irrational to repair, even if it is possible to do so. I've got a dead DP1s with a failed focus mechanism. It looks like a new camera but it's useless.

All the above are statements of the blindly obvious but because of them, Sigma have to acknowledge their loyal Foveon buyers will drift away. Getting them back will be hard, even if a full frame Foveon (FFF) sensor camera is launched.

There is no doubt that Sigma can produce a camera with a very high build quality cameras and the fp is an excellent example. I gather it's been popular and looking at some of the photos it produces with the Sigma Art lenses, especially in conditions where the Foveon cameras are hopeless, its popularity comes as no surprise.

If a FFF camera is delivered in 2021 - and I now sense that is a big if - and it is sensible money what is the expected market? I'd guess it would be tiny. If it ends up as a camera that only shines at ISO 100, then it's an even smaller market. To add to Sigma's FFF challenge, Bayer sensors have improved since the Quattro launch in 2014.

For me, I guess I'll carry on as an occasional amateur photographer using my current Sigma Foveon cameras when I can and until they pack up. I certainly wouldn't have them repaired. I'm potentially a prime customer for the FFF at the moment but that situation is fading.

Best, Steve

Steve,

I think you should email exactly what you have written here directly to Kazuto Yamaki at Sigma Japan, because I think you are right and people will continue to drop off unless some information is released soon regarding the FFF or some other new camera offering. I was tempted to copy your post myself but would not do so without your permission.

S

Hmm...there is a valley in the camera market. People are unable to travel safely with the virus so that is happening less.

I just traveled to 8 states over the last five weeks, and I ate at almost 100 restaurants, and did not get the virus! I don't agree that people can not travel safely.

Irresponsible, don't you think?

And how do you know you didn't get the virus? Were you tested every day for 5 weeks? In at least 20% of infections, it is asymptomatic - in which case you could have been a super spreader without knowing (as could anyone). If you were (and lots of people have been), there is a reasonable possibility that someone died as a result of your trip. And even if no did and no one caught it and passed it on to someone who did die, that would be blind luck.

The only responsible thing for all of us to do until vaccination kills this virus off, is to reduce all unavoidable social contact and make sure unavoidable contacts are properly socially distanced. Anything less could be, literally murder.

 

[DavidMillier]

There is very little real world difference between 40, 50, 60MP. The occasions when you could actually put those pixel count differences onto paper in a way that was unambiguously visible to most viewers must be very rare indeed.