Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
04-28-2014, 01:28 AM   #46
Veteran Member
carrrlangas's Avatar

Join Date: Mar 2012
Location: Joensuu (Finland)
Photos: Gallery
Posts: 1,761
QuoteOriginally posted by Rob01 Quote
Carrrlangas, I am pretty sure the resolution is based NOT on full colour resolution. I do not see how the Quattros can resolve more colour when they have only 5 MP of red and green compared to 15 MP for the Merrills. The way I see it the resolution is based on the blue colour or gray scale image overlayed with the other colours... there is no other way. It will have a higher SNR for the 5 MP colours but that is all.
Sigma should hire you, you already know so much about a camera that hasnīt been released yet!
Now seriously, despite you sounding very rude, this is what Iīve read: Foveon supports the resolution statement with two bullet points:
- Human eye doesnīt respond to chroma and luminance resolution in the same way and so does the new Quattro sensor, so it is good trade off that should provide imperceptible difference.
- The chroma information is not obteined from each layer independently. i.e.: There is some red information on blue and green channels as well as on red channel of course, and they take advantage of this through complex math/processing

QuoteOriginally posted by Rob01 Quote
If Sigma does not supply the kind of full colour resolution I want... I will simply go for cameras from another company. Period. In a year or two we will have a Nikon successor with around 50MP thatwill match the colour resolution of the Merrills even through a Bayer sensor, or a Pentax full frame, or maybe a Foevon type of sensor from another company.
Better hope sigma delivers what you want? Period? I hope all you say will come true, would be really fun and nice to have such options

04-28-2014, 01:43 AM   #47
Junior Member




Join Date: Apr 2014
Posts: 32
QuoteOriginally posted by carrrlangas Quote
Better hope sigma delivers what you want? Period? I hope all you say will come true, would be really fun and nice to have such options
"Period" - means I am not attached to the Sigma brand. If they do not deliver, I will switch.

Plus you should read the info supplied elsewhere. Sigma people openly admit that they sacrificed some colour resolution for better high ISO performance and data management and read out. They say the extra colour information does not matter - I disagree with them. This fact should be obvious to everyone by noticing that 5 Mega pixels of a colour CANNOT have more resolution than 15 Megapixels... I would think this is obvious to everyone. LOL ... but apparently not

Last edited by Rob01; 04-28-2014 at 02:27 AM.
04-28-2014, 02:26 PM   #48
Veteran Member




Join Date: Sep 2013
Location: Pasadena, CA
Photos: Albums
Posts: 1,125
I agree, there was an interview with sigma, which said they sacrifice color resolution for luminance. They said that Bayer sebspr do it sp its no big deal. We will see how I is in practice.
04-28-2014, 02:50 PM   #49
Veteran Member
carrrlangas's Avatar

Join Date: Mar 2012
Location: Joensuu (Finland)
Photos: Gallery
Posts: 1,761
I dont care to be corrected but I havent read other articles that contradict what I am saying. Do 'ou have any sources?
Also, read my post again. You cant compare 5mp of red channel with very low noise to 15mp of red channel with very high noise. Also take the interleaving of channels into account. The difference in resolution would be insignificant if this they deliver

04-28-2014, 09:23 PM   #50
Veteran Member
Aegon's Avatar

Join Date: Jan 2008
Location: Portland, OR
Posts: 1,414
QuoteOriginally posted by rrstuff Quote
I agree, there was an interview with sigma, which said they sacrifice color resolution for luminance. They said that Bayer sebspr do it sp its no big deal. We will see how I is in practice.
Here's an interview that says something similar...
Sigma Q&A Part II: Does Foveon’s Quattro sensor really out-resolve conventional 36-megapixel chips?

But you see that the interviewer, not Sigma, says that Quattro has "lower resolution in the red and green channels".
In response, the Sigma rep says "None of them are a pure color, so in reality, what it allows you do to -- to separate out the detail information, but also be able to get the color out, and the detail, or the color out of the rest".

Since the channels have wide band spectral response, and since color information is calculated based on a subtractive system that heavily emphasizes the so-called blue layer, I don't think that Sigma themselves would say that they are sacrificing color resolution for luminance. If anything, they sacrifice for sensitivity. That may sound like splitting hairs, but at least I've linked to a source.
04-28-2014, 09:38 PM   #51
Veteran Member




Join Date: Sep 2013
Location: Pasadena, CA
Photos: Albums
Posts: 1,125
I think it's not even resolution, more like color selectivity. I am not sure if it will matter much in practice though, because lower layers of the foveon chip were pretty noisy and color resolution was poor. So while technically, they will do an interpolation for red/green channels, perhaps the amount of information in these channels was not that great in the first place.
From talking to users of foveon's they seemed to welcome the changes, because they were unhappy with color noise in the r/g channels. I would wait and see how it turns out, I am mostly just glad that they are innovating and developing something else than a bayer sensor. As a bonus, it makes the old DP-series cameras very inexpensive .
04-29-2014, 01:23 AM   #52
Junior Member




Join Date: Apr 2014
Posts: 32
The quote in question is:

The main claim of Bryce Bayer, and the fact that most TV formats and image and video compression algorithms rely on, is that the visual system doesn't care nearly as much about chroma resolution as about luma resolution. Unfortunately, trying to exploit that factor with a one-layer mosaic sensor has these awkward aliasing problems. Doing it with the Foveon 1:1:4 arrangement works better, requiring no AA filter, no filtering compromises. So, yes, the chroma resolution is less than the luma resolution, but you'd be hard pressed to see that in images.

Read more at: https://www.pentaxforums.com/forums/76-non-pentax-cameras-canon-nikon-etc/220...#ixzz30GB0yFXg

(look for a post by Mr Bassie that quotes extensive statments from Sigma TPTB)

... by comparison, and in the terminology used above, for the Merrils the chroma resolution is the same as the luma resolution (at 15 Megapixels for all).

You really CANNOT reconstruct the colour you do not measure, i.e. the Merrill with 15 megapixel will measure more colour resolution than Quattro with 5 Megapixels (for green and red anyway)... even if all the colours are partially absorbed in all the layers. I could write the equations if people can follow basic algebra and a set of simultaneous equations. Of course none of this is simple because the blue resolution of the Quattro will be higher and thus from the 20/5/5 Megapixels arrangement will have more colour than a 5/5/5 Megapixels ones... I think of it as partial colour resolution through the extra blue component (for shades that have blue in them anyway).

Now, is the sacrifice of the unique colour resolution the Merrills had worth the extra high ISO sensitivity? Not to me. I shoot any camera at ISO 100 if I at all can... and at this ISO the Merrills had NO problems with noise in green or red. So, it is NOT entirely correct that the 20/5/5 Megapixels arrangement reduces noise in the green and red... the staement is ONLY correct at high ISOs. In my case I gain nothing but lose the colour resolution. To me the statement that losing the colour resolution does not matter much is an opinion or a judgment and not a fact. The truth is that it may not matter to some or in some applications, but it may in others.

My only reason to invest in the Sigma system is the colour performance of the Merrills as for all other purposes I am perfectly happy with my Pentax K5IIs. Accordingly with the Quattros my reasons for moving to Sigma disappear... and I have to wait for some other manufacturer to provide a solution that does colour as well as the Merrills The Quattros may be fine cameras but they lose the unique advantage the Merrills had... so why would anyone buy them now? Other brands do better what the Quattros are attempting to do: high ISO performance, resolution and speed or performance. The one remaining IQ advantage is that the 4/1/1 Foevon sensor is claimed not to suffer from moire... so while it is equivalent in resolution to a Bayer sensor of between 20 and 40 Megapixels (depending on whether you mean B&W or colour etc) without an anti-alias filter, the Bayer sensor will be subject to moire artefacts and the Foevon will not be (assuming the 4/1/1 configuration does not cause moire or other artefacts).

In addition I have to think whether it now makes sense to invest in Sigma mount lenses if all the new SLRs have Foevon sensors with this 4/1/1 ratio of colours. Of course I could get the SD1M and keep using it for many years but with no upgrade path it makes very little sense to buy a set of good lenses for it. I will think more on this, but I may just stick to a full set of DP1M, DP2M and DP3M and see how the future develops... someone advised me I could just learn to shoot 4x5

I am now looking into how well Pentax mount lenses can be adapted to the Sigma mount as then I may just get the SD1M.

P.S. To demonstrate if colour resolution makes a difference, look below at the very well done comparison image from dpreview, especially check out the tree or the white wall next to the tower... and this is without even knowing if the jpeg is encoded with half or full colour resolution:

http://www.dpreview.com/forums/post/53152669?image=0




Last edited by Rob01; 04-29-2014 at 04:22 AM.
04-29-2014, 06:59 AM   #53
Veteran Member
carrrlangas's Avatar

Join Date: Mar 2012
Location: Joensuu (Finland)
Photos: Gallery
Posts: 1,761
QuoteOriginally posted by Aegon Quote
Here's an interview that says something similar...
Sigma Q&A Part II: Does Foveon’s Quattro sensor really out-resolve conventional 36-megapixel chips?

But you see that the interviewer, not Sigma, says that Quattro has "lower resolution in the red and green channels".
In response, the Sigma rep says "None of them are a pure color, so in reality, what it allows you do to -- to separate out the detail information, but also be able to get the color out, and the detail, or the color out of the rest".

Since the channels have wide band spectral response, and since color information is calculated based on a subtractive system that heavily emphasizes the so-called blue layer, I don't think that Sigma themselves would say that they are sacrificing color resolution for luminance. If anything, they sacrifice for sensitivity. That may sound like splitting hairs, but at least I've linked to a source.
Thatīs the interview I posted. I am saying this exactly, but you put it better.

QuoteOriginally posted by rrstuff Quote
I am not sure if it will matter much in practice though, because lower layers of the foveon chip were pretty noisy and color resolution was poor. So while technically, they will do an interpolation for red/green channels, perhaps the amount of information in these channels was not that great in the first place.
Thatīs the catch, me thinks

QuoteOriginally posted by rrstuff Quote
As a bonus, it makes the old DP-series cameras very inexpensive .
Thatīs how I bought a DP2 Merrill

---------- Post added 04-29-14 at 11:13 AM ----------

QuoteOriginally posted by Rob01 Quote
The quote in question is:

The main claim of Bryce Bayer, and the fact that most TV formats and image and video compression algorithms rely on, is that the visual system doesn't care nearly as much about chroma resolution as about luma resolution. Unfortunately, trying to exploit that factor with a one-layer mosaic sensor has these awkward aliasing problems. Doing it with the Foveon 1:1:4 arrangement works better, requiring no AA filter, no filtering compromises. So, yes, the chroma resolution is less than the luma resolution, but you'd be hard pressed to see that in images.

Read more at: The Foveon Club: Life Unfiltered

(look for a post by Mr Bassie that quotes extensive statments from Sigma TPTB)

... by comparison, and in the terminology used above, for the Merrils the chroma resolution is the same as the luma resolution (at 15 Megapixels for all).

You really CANNOT reconstruct the colour you do not measure, i.e. the Merrill with 15 megapixel will measure more colour resolution than Quattro with 5 Megapixels (for green and red anyway)... even if all the colours are partially absorbed in all the layers. I could write the equations if people can follow basic algebra and a set of simultaneous equations. Of course none of this is simple because the blue resolution of the Quattro will be higher and thus from the 20/5/5 Megapixels arrangement will have more colour than a 5/5/5 Megapixels ones... I think of it as partial colour resolution through the extra blue component (for shades that have blue in them anyway).

Now, is the sacrifice of the unique colour resolution the Merrills had worth the extra high ISO sensitivity? Not to me. I shoot any camera at ISO 100 if I at all can... and at this ISO the Merrills had NO problems with noise in green or red. So, it is NOT entirely correct that the 20/5/5 Megapixels arrangement reduces noise in the green and red... the staement is ONLY correct at high ISOs. In my case I gain nothing but lose the colour resolution. To me the statement that losing the colour resolution does not matter much is an opinion or a judgment and not a fact. The truth is that it may not matter to some or in some applications, but it may in others.

My only reason to invest in the Sigma system is the colour performance of the Merrills as for all other purposes I am perfectly happy with my Pentax K5IIs. Accordingly with the Quattros my reasons for moving to Sigma disappear... and I have to wait for some other manufacturer to provide a solution that does colour as well as the Merrills The Quattros may be fine cameras but they lose the unique advantage the Merrills had... so why would anyone buy them now? Other brands do better what the Quattros are attempting to do: high ISO performance, resolution and speed or performance. The one remaining IQ advantage is that the 4/1/1 Foevon sensor is claimed not to suffer from moire... so while it is equivalent in resolution to a Bayer sensor of between 20 and 40 Megapixels (depending on whether you mean B&W or colour etc) without an anti-alias filter, the Bayer sensor will be subject to moire artefacts and the Foevon will not be (assuming the 4/1/1 configuration does not cause moire or other artefacts).

In addition I have to think whether it now makes sense to invest in Sigma mount lenses if all the new SLRs have Foevon sensors with this 4/1/1 ratio of colours. Of course I could get the SD1M and keep using it for many years but with no upgrade path it makes very little sense to buy a set of good lenses for it. I will think more on this, but I may just stick to a full set of DP1M, DP2M and DP3M and see how the future develops... someone advised me I could just learn to shoot 4x5

I am now looking into how well Pentax mount lenses can be adapted to the Sigma mount as then I may just get the SD1M.

P.S. To demonstrate if colour resolution makes a difference, look below at the very well done comparison image from dpreview, especially check out the tree or the white wall next to the tower... and this is without even knowing if the jpeg is encoded with half or full colour resolution:

Re: Myth of FF Resolution Superiority...: Sigma Camera Talk Forum: Digital Photography Review
Now I get what you are trying to say, the new Quattro design kind of takes the same path as Bayer array. I am going to try the DP2 merrill for atwork reproduction (continuous or flash lighting) and that is an example where the difference you talk about could be seen. But Merrills do have problems in shadows areas not just for recovery but they tend to be really special.Very good post by Mr. Bassie, in his words:
QuoteQuote:
My words and yours are not going to convince anybody of anything until they see it...
04-29-2014, 04:49 PM   #54
Junior Member




Join Date: Apr 2014
Posts: 32
Now I get what you are trying to say, the new Quattro design kind of takes the same path as Bayer array. I am going to try the DP2 merrill for atwork reproduction (continuous or flash lighting) and that is an example where the difference you talk about could be seen. But Merrills do have problems in shadows areas not just for recovery but they tend to be really special.Very good post by Mr. Bassie, in his words: Quote:
My words and yours are not going to convince anybody of anything until they see it...





Yes, I agree on this point... the tech is hard to understand, it took me a week to compose it in my brain and images speak louder than words. I most likely will be getting a DP2M as well (I am starting to play with DP1M for landscapes)... I think DP2M is an amazing camera. I think the highlights/vs shadows question can be worked around. I have heard people shooting at ISO 200 with the Merrills and working on it in post processing... you would have to search out the details. For static tripod work you may also consider HDR and/or multiple exposure stacking or even super-resolution processing.

For me I am hoping that that the SD1M gets upgraded with a larger buffer, faster electronics, etc, but that the sensor is not touched. Or that it gets a new 1//1/1 colour ratio sensor and not the Quattro. We shall see...
05-24-2014, 10:30 AM   #55
Senior Member




Join Date: May 2013
Location: Lima-Peru
Photos: Albums
Posts: 117
QuoteOriginally posted by Rob01 Quote
The quote in question is:

The main claim of Bryce Bayer, and the fact that most TV formats and image and video compression algorithms rely on, is that the visual system doesn't care nearly as much about chroma resolution as about luma resolution. Unfortunately, trying to exploit that factor with a one-layer mosaic sensor has these awkward aliasing problems. Doing it with the Foveon 1:1:4 arrangement works better, requiring no AA filter, no filtering compromises. So, yes, the chroma resolution is less than the luma resolution, but you'd be hard pressed to see that in images.

Read more at: https://www.pentaxforums.com/forums/76-non-pentax-cameras-canon-nikon-etc/220...#ixzz30GB0yFXg

(look for a post by Mr Bassie that quotes extensive statments from Sigma TPTB)

... by comparison, and in the terminology used above, for the Merrils the chroma resolution is the same as the luma resolution (at 15 Megapixels for all).

You really CANNOT reconstruct the colour you do not measure, i.e. the Merrill with 15 megapixel will measure more colour resolution than Quattro with 5 Megapixels (for green and red anyway)... even if all the colours are partially absorbed in all the layers. I could write the equations if people can follow basic algebra and a set of simultaneous equations. Of course none of this is simple because the blue resolution of the Quattro will be higher and thus from the 20/5/5 Megapixels arrangement will have more colour than a 5/5/5 Megapixels ones... I think of it as partial colour resolution through the extra blue component (for shades that have blue in them anyway).

Now, is the sacrifice of the unique colour resolution the Merrills had worth the extra high ISO sensitivity? Not to me. I shoot any camera at ISO 100 if I at all can... and at this ISO the Merrills had NO problems with noise in green or red. So, it is NOT entirely correct that the 20/5/5 Megapixels arrangement reduces noise in the green and red... the staement is ONLY correct at high ISOs. In my case I gain nothing but lose the colour resolution. To me the statement that losing the colour resolution does not matter much is an opinion or a judgment and not a fact. The truth is that it may not matter to some or in some applications, but it may in others.

My only reason to invest in the Sigma system is the colour performance of the Merrills as for all other purposes I am perfectly happy with my Pentax K5IIs. Accordingly with the Quattros my reasons for moving to Sigma disappear... and I have to wait for some other manufacturer to provide a solution that does colour as well as the Merrills The Quattros may be fine cameras but they lose the unique advantage the Merrills had... so why would anyone buy them now? Other brands do better what the Quattros are attempting to do: high ISO performance, resolution and speed or performance. The one remaining IQ advantage is that the 4/1/1 Foevon sensor is claimed not to suffer from moire... so while it is equivalent in resolution to a Bayer sensor of between 20 and 40 Megapixels (depending on whether you mean B&W or colour etc) without an anti-alias filter, the Bayer sensor will be subject to moire artefacts and the Foevon will not be (assuming the 4/1/1 configuration does not cause moire or other artefacts).

In addition I have to think whether it now makes sense to invest in Sigma mount lenses if all the new SLRs have Foevon sensors with this 4/1/1 ratio of colours. Of course I could get the SD1M and keep using it for many years but with no upgrade path it makes very little sense to buy a set of good lenses for it. I will think more on this, but I may just stick to a full set of DP1M, DP2M and DP3M and see how the future develops... someone advised me I could just learn to shoot 4x5

I am now looking into how well Pentax mount lenses can be adapted to the Sigma mount as then I may just get the SD1M.

P.S. To demonstrate if colour resolution makes a difference, look below at the very well done comparison image from dpreview, especially check out the tree or the white wall next to the tower... and this is without even knowing if the jpeg is encoded with half or full colour resolution:

Re: Myth of FF Resolution Superiority...: Sigma Camera Talk Forum: Digital Photography Review

If the merrills deliver why would you have yo look for another manufacter?
05-24-2014, 11:12 AM   #56
Pentaxian
normhead's Avatar

Join Date: Jun 2007
Location: Near Algonquin Park
Photos: Gallery | Albums
Posts: 37,334
Everyone wants to equate what the Foveon is doing with what a Bayer sensor does. Big mistake.

This seems really easy to me.

In a Bayer sensor you read 3 different pixels for their colour values and then combine them to make one pixel, for 3 pixels with the same values.
With a Foveon Sensor you read the total luminance, read the blue and green and subtract to get the red. You are still taking all three readings from one pixel site, even though it's indirect for the red. With a Bayer system, you are still reading blue , green , and red from different pixel sites.

No matter who you state it, Bayer needs to read at least 3 pixels to obtain the red , blue and green values. The Foveon needs to read one.

Here's the thing, total light 133, Blue = 40, Green=42 Red must equal 41
Total light value in a bayer 133, Blue=40.. there are an incredible number of combinations of what Red and Green values may be, and essentially , the software guesses on values based on adjacent pixels. It may seem that the processes are related, except if you understand how it works, one is precise and one is imprecise.

AN extrapolation of one value is precise. Reading red, blue and green but only sampling red, blue or green, means you are extrapolating two values, that's called guess work.
Why is this necessary? Because at more than 16 MP diffraction is already affecting your red channel, and is probably somewhat affecting your green channel, your blue channel should be good up to 100 MP so it's the only one where you are actually getting good clean resolution from an APS-c sensor over 16 Mp. After 16 Mp, you can increase pixel density if you want, it's not going to get you equivalent resolution in the red channel. I assume that's why Foveon chose this route.

I expect that Foveon hasn't decreased their quality at all, and that with my DP2M being very close to D800 quality the 19.5 Mp Quattro will equal or surpass it.

Really this is all guess work on my part, but it's an explanation that from my reading is closer to the truth than the speculation I've seen so far.
05-24-2014, 10:59 PM   #57
Banned




Join Date: May 2010
Location: Back to my Walkabout Creek
Photos: Gallery | Albums
Posts: 2,535
Original Poster
People are splitting hairs. I believe Quattro is indeed a pocket powerhouse, and the closest to the ideal sensor — if there ever was one.
It will be interesting to see how Sigma handles video, for example, as video on Foveons would be a dream come true.
05-28-2014, 01:46 PM   #58
Pentaxian
ivoire's Avatar

Join Date: Mar 2007
Location: chicago burbs
Photos: Gallery
Posts: 3,381
DP Quattro news:

Sharpness: the Sigma dp2 Quattro can reach the level of the Pentax 645Z
05-29-2014, 02:28 PM   #59
Banned




Join Date: May 2010
Location: Back to my Walkabout Creek
Photos: Gallery | Albums
Posts: 2,535
Original Poster
QuoteOriginally posted by ivoire Quote
Perhaps in some circumstances, and using basic ISO. But sharpness isn't everything, as MF cameras are purchased for many other advantages, one of most important being MF lenses. MF is purchased mainly because of the image qualities (various unique aspects) delivered thanks to its optics.
However, all that aside, little Quattro is indeed a very interesting camera and a little powerhouse on the budget.
06-12-2014, 07:00 AM   #60
Pentaxian
ivoire's Avatar

Join Date: Mar 2007
Location: chicago burbs
Photos: Gallery
Posts: 3,381
Japan site posted these hires jpegs
http://www.sigma-global.com/jp/cameras/dp-series/gallery/

They don't look much diff than images produced by the Merrills. Have to wait and see what other features the Quattros offer before I decide on a Quattro vs another Merrill. Faster AF, longer battery life, faster read /write time and higher iso might sway my decision. I suspect the camera will sell for $1200-1500 which will make a DP1,2,3M a bargain

Last edited by ivoire; 06-12-2014 at 08:16 AM.
Reply

Bookmarks
  • Submit Thread to Facebook Facebook
  • Submit Thread to Twitter Twitter
  • Submit Thread to Digg Digg
Tags - Make this thread easier to find by adding keywords to it!
bayer, button, color, colour, dp2m, grip, information, iso, jpeg, layer, light, megapixels, mp, mpix, quattro, resolution, sensor, sigma, sigma quattro dp2, vs
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
Humorous comparison: Q vs. Sigma DP2 Merrill panoguy Pentax Q 23 03-27-2013 10:54 AM
Awesome Camera Concept mervis Pentax K-01 22 03-31-2012 10:22 PM
New Sigma DP1 and DP2 Merrill with the same 46MP Foveon X3 sensor from the Sigma SD1 UdonUdon Non-Pentax Cameras: Canon, Nikon, etc. 29 02-20-2012 12:30 PM
For Sale - Sold: Sigma DP2 FOVEON Digital Camera with accessories (CONUS) EsBee Sold Items 2 05-25-2011 09:16 AM



All times are GMT -7. The time now is 11:40 PM. | See also: NikonForums.com, CanonForums.com part of our network of photo forums!
  • Red (Default)
  • Green
  • Gray
  • Dark
  • Dark Yellow
  • Dark Blue
  • Old Red
  • Old Green
  • Old Gray
  • Dial-Up Style
Hello! It's great to see you back on the forum! Have you considered joining the community?
register
Creating a FREE ACCOUNT takes under a minute, removes ads, and lets you post! [Dismiss]
Top