Ad
Ad
Ad
Pages: « 1 [2] 3 4 ... 11 »   Bottom of Page
Print
Author Topic: Sigma SD1 review  (Read 38992 times)
ErikKaffehr
Sr. Member
****
Online Online

Posts: 6890


WWW
« Reply #20 on: July 19, 2011, 06:54:41 AM »
ReplyReply

Hi,

25 MP APS-C is said to be around the corner from Sony (Alpha 77) and Nikon D400, would be an improvement over 16 MP on APS-C in some regards, like better interpolation and less need for AA-filtering. Both cameras may show up in August (this year) according to rumors.

It will be interesting to see some test charts from the SD1, but it may be the case that the SD1 works better on "real subjects" than test charts.

Best regards
Erik

30mp on a cropped format is significantly different from 24mp on full frame. I've made many comparisons with my Canon cameras comparing full frame with cropped frame at higher pixel count. The Sigma SD1 is definitely unique in terms of resolution for a cropped format camera, assuming it is true that the pixel count on the Foveon is equal to double the pixel count of a Bayer array.
Logged

dreed
Sr. Member
****
Offline Offline

Posts: 1168


« Reply #21 on: July 19, 2011, 04:16:02 PM »
ReplyReply

Sigma's desire to go it alone with RAW software processing is absurd. I can remember life before Lightroom and I'm not going back there. If a camera's (RAW) photos cannot be integrated with a work flow that involes using only Lightroom then I'm simply not interested in even thinking about comparing it to other caemras and it may as well not exist.

I can't help but wonder what the decision makers at Sigma were thinking when they decided on (a) the price and (b) not to work with companies like Adobe for the benefit of Lightroom.

Is it really just a demonstration camera that is not actually meant to sell or be used?
Logged
Sareesh Sudhakaran
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #22 on: July 19, 2011, 11:04:18 PM »
ReplyReply

30mp on a cropped format is significantly different from 24mp on full frame. I've made many comparisons with my Canon cameras comparing full frame with cropped frame at higher pixel count. The Sigma SD1 is definitely unique in terms of resolution for a cropped format camera, assuming it is true that the pixel count on the Foveon is equal to double the pixel count of a Bayer array.

So is 30MP on a cropped sensor better than 24MP on full frame? In your comparisons with Canon cameras, did higher resolution APS-C sensors beat lower resolution FF sensors?
Logged

Get the Free Comprehensive Guide to Rigging ANY Camera - one guide to rig them all - DSLRs to the Arri Alexa.
ErikKaffehr
Sr. Member
****
Online Online

Posts: 6890


WWW
« Reply #23 on: July 19, 2011, 11:40:44 PM »
ReplyReply

Hi,

I'm a bit familiar with Ray's reasoning. A full frame camera needs to be stopped down more to achieve equivalent depth of field. That is absolutely correct.

My view is that larger sensor size is an advantage. If the sensor pitch is the same a full frame camera will have higher resolution because of having more pixels. This would be the case if we compare the Nikon D300 and the Nikon D3X. The sensor used in the two cameras is pretty closely related. My guesses:

  • The Nikon D3X has better resolution than the D300 when neither is cropped.
  • If Nikon D3X is cropped to DX size the resolution would be the same

On the other hand, if we compare the D3S or D700 it will probably not resolve significantly better than the D300! The reason is that the pixels are bigger. Theoretically the lens would transmit a higher MTF trough the lens. BUT, there is an optical low pass (OLP)  filter that blurs the image slightly so to reduce aliasing close to Nyquist. Larger pixels mean a stronger filter! The aim of the OLP/AA filter is to keep MTF (or microcontrast) down at what is know as the Nyquist limit to around 20-30%.

The result is by and large that regarding sharpness a 12 MP sensor will behave the same, independent of size (within limits). The discussion presumes pretty decent lenses.

A larger sensor will collect more light and will therefore work better at high ISO and have better dynamic range. On the other hand, all full size sensors are around three years old. There is a new generation of sensor from Sony, used in the Nikon D7000, Pentax K5 and the Sony Alpha 580 that has extremely low readout noise and therefore excels in dynamic range.

A knowledgable poster on this forum really indicated that the K5 has better DR than his Phase One P45+! He even posted raw images!

No doubt, new sensor will be coming. Sony and Nikon will probably announce new APS-C cameras based on Sony's latest sensor technology this autumn, maybe in August according to rumors. Rumors say 25 Mpixels, and a brand new sensor.

Will the new 25 MP APS-C sensor beat the old 24.6 MP full frame sensor? Maybe, maybe not! Especially Nikon did a very good job on the D3X, the AA-filter on that camera is weak. No doubt, the new 25 MP APS-C sensor need less AA-filtering.

Finally lenses are most important. Shrinking the pixels take us close to physical limits.

Best regards
Erik


So is 30MP on a cropped sensor better than 24MP on full frame? In your comparisons with Canon cameras, did higher resolution APS-C sensors beat lower resolution FF sensors?
Logged

deejjjaaaa
Sr. Member
****
Offline Offline

Posts: 743


« Reply #24 on: July 20, 2011, 12:02:44 AM »
ReplyReply

I can't help but wonder what the decision makers at Sigma were thinking when they decided on (a) the price and (b) not to work with companies like Adobe for the benefit of Lightroom.

why do you think it is Sigma to blame and not Adobe Labs developers having no time to fix their code ? which, by the way, was never updated to reflect the changes that Adobe Labs did w/ ACR v6.x/LR3.x for NR, etc even for the Sigma cameras that were supported ! It is just not on their list of priorities...
Logged
ErikKaffehr
Sr. Member
****
Online Online

Posts: 6890


WWW
« Reply #25 on: July 20, 2011, 01:10:10 AM »
ReplyReply

Hi,

Sigma is different as it doesn't have filters and Bayer Matrix. So processing is very much different from what ACR/LR does. Adobe could build a different processing engine for the Foevon, if given enough information by Sigma but the question is if there is enough return on investment?

Best regards
Erik


why do you think it is Sigma to blame and not Adobe Labs developers having no time to fix their code ? which, by the way, was never updated to reflect the changes that Adobe Labs did w/ ACR v6.x/LR3.x for NR, etc even for the Sigma cameras that were supported ! It is just not on their list of priorities...

Logged

deejjjaaaa
Sr. Member
****
Offline Offline

Posts: 743


« Reply #26 on: July 20, 2011, 09:14:22 AM »
ReplyReply

Hi,

Sigma is different as it doesn't have filters and Bayer Matrix. So processing is very much different from what ACR/LR does. Adobe could build a different processing engine for the Foevon, if given enough information by Sigma but the question is if there is enough return on investment?

Best regards
Erik



the mere fact that Adobe does support Sigma cameras prior to SD1 says that Adobe has all the information...

also in terms of processing... ACR/R deal w/o issues w/ linear DNGs and Tiffs - and those are not much different from Foveon (once you will do the color transformations to bring the Foveon data into RGB realm) Grin

the simple truth is that nobody there (in Adobe Labs) has time or desire to finally update the code.
« Last Edit: July 20, 2011, 09:17:38 AM by deejjjaaaa » Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8811


« Reply #27 on: July 20, 2011, 09:28:33 AM »
ReplyReply

So is 30MP on a cropped sensor better than 24MP on full frame? In your comparisons with Canon cameras, did higher resolution APS-C sensors beat lower resolution FF sensors?

Absolutely! This new Foveon sensor is slightly larger than previous Foveon sensors. It has a crop factor of about 1.5X, ie. a full frame sensor is 2.5x the area.

A full frame 35mm Foveon sensor with the pixel pitch of the SD1 would be the equivalent of a 67.5mp Bayer sensor, so your question is a bit like, "Is a 67mp full frame sensor higher resolution than a 24mp full frame?"

Of course, the result will depend on the use. There will not be much difference between 30mp and 24mp when lenses of different focal lengths are used to maintain equal FoV using the the full area of each sensor. However, when using the Bigma 300-800 with the SD1 at 800mm, which becomes effectively 1200mm, there should be a very noticeable increase in resolution compared with a 24mp full frame cropped to the same FoV. One is then effectively comparing 10mp with 30mp.
Logged
Graeme Nattress
Sr. Member
****
Offline Offline

Posts: 582



WWW
« Reply #28 on: July 20, 2011, 09:55:44 AM »
ReplyReply

From a linear raw RGB file, all you need is a colour correction matrix so that's relatively simple for Adobe to support linear DNGs and Tiffs. There's a lot of work needed to get a Foveon raw to linear RGB though - you need special processing to extract RGB from the three layer data (which diagrams depict as RGB, but it's not really) and some strong chroma NR to cope with the noise induced by the above processing. If the Foveon did produce good RGB data as is often assumed, indeed there'd be little work for Adobe or anyone else to support it's raws, but producing an RGB image from a Foveon raw needs more processing than it takes to get RGB from a Bayer mosaic colour filter array camera. It's worth reading the Foveon white papers: http://www.foveon.com/article.php?a=74 to understand more about their chip.

http://www.foveon.com/files/CIC13_Hubel_Final.pdf is perhaps the most interesting. Gems include "Using a direct image sensor such as Foveon’s X3 technology, reduces the requirement of a blur filter because aliasing occurs in all color channels equally giving no brightly colored aliasing fringing." - reduces, not eliminates, and the obvious conclusion from this is that without an OLPF aliasing will occur on a Foveon. Obviously Sigma omit this aspect of the camera an allow for luma aliasing which can look like a sharper image, but is really just artifacts. Given their removable dust shield IR filter, I'm surprised that now that they have a good native resolution they don't supply an OLPF/IR combo filter for this position to remove aliasing for photographers where such artifacts would be an issue.

On page two of that paper, under "New Image Processing Methods" we see the comparison of a typical image processing pipeline "A" for a Bayer CFA, with "B" and "C" from a Foveon. "the color correction transformation matrix required to convert the native sensor signals to a standard color space is aggressive compared with conventional camera sensors and this amplifies noise." - the diagrams show the "noise suppression" blocks necessary in the Foveon pipeline, and although they show such a block in "A", from experience I know that such a block is not particularly necessary. They also employ decimation tricks to get more performance out of the slow NR blocks, because as we know, good NR is not fast.

I don't agree with Michael's statement in the review "Some people may not be aware of it, but Color / Chroma Noise and its associated removal tool in various processing programs is there to remove noise introduced by Bayer Filter Array decoding. One would therefore assume that a Foveon type X3 sensor, which doesn't use a Bayer Matrix, wouldn't have Color Noise, and indeed it doesn't." Chroma noise comes from the colour correction matrix amplifying sensor noise. Without such a matrix a Bayer CFA image looks a bit under-saturated, and has some hue errors. As noted, the Foveon has a very aggressive colour correction matrix, and you can see how necessary that is by looking at the paper above with the spectral responses of the three layers and by how much they overlap each other. It's the overlap that needs to be removed by the CC matrix, and the more overlap, the stronger the matrix, and the more noise amplification, hence the necessary NR steps in the Foveon processing mentioned above. In a Bayer CFA, chroma noise does not come from the demosaic process, but instead from the CC matrix, which due to the use of good dye filters on the CFA is a much less noise inducing matrix than that used on Foveon.

I think the colour anomalies noticed by Michael, the red hue in the shadows, the colours not being "right" are fully due to the issues with extracting colour information from the Foveon sensor. As all things are a balance, these colour anomalies and associated NR processing are the price paid for lack of chroma moire. If you don't get chroma moire in your shots (and even if you do, it's usually quite easy to remove), you'll achieve much better colour accuracy from a Bayer CFA.

Graeme
Logged

www.nattress.com - Plugins for Final Cut Pro and Color
www.red.com - Digital Cinema Cameras
bjanes
Sr. Member
****
Offline Offline

Posts: 2714



« Reply #29 on: July 20, 2011, 10:56:29 AM »
ReplyReply

Perhaps it didn't come across clearly in the review, but we did that. The results were exactly as Michael described. The 645D pulled clearly ahead, to my eye the Leica was in second place, and the A900 and SD1 were very close with the A55 bringing up the rear by a small margin.

- N.

I enjoyed reading the review and found it to be excellent overall. However, one criticism is that Light Room sharpening was used exclusively. As is well known, cameras with low pass filters require more sharpening than those lacking such a filter. Michael did take this into account, but some authorities have reported that deconvolution sharpening is the best method to restore micro-contrast and loss of resolution secondary to the use of a low pass filter.

For example, in his comparison of the Nikon D3X to the Leica S2 Digilloyd (a pay site well worth its modest cost) found that deconvolution sharpening with the Richardson-Lucy (RL) algorithm brought the results obtained with these two cameras nearly to parity and that he was unable to obtain equivalent results for the Nikon with ACR/light room. According to Eric Chan, the latest iterations of ACR do enable deconvolution sharpening, but the number of iterations must be relatively low, since the conversion is much faster than programs using RL.

Bart vanderWolf has also advocated deconvolution sharpening as giving the best results.in my own testing with the Nikon D3and the latest version of ACR, I have been able to obtain micro-contrast and resolution very similar to that obtained with Focus Magic, which is another deconvolution algorithm.however, focus magic requires considerably less fooling around with the settings.

Regards,

Bill

Logged
deejjjaaaa
Sr. Member
****
Offline Offline

Posts: 743


« Reply #30 on: July 20, 2011, 12:09:15 PM »
ReplyReply

\but producing an RGB image from a Foveon raw needs more processing than it takes to get RGB from a Bayer mosaic colour filter array camera. It's worth reading the Foveon white papers: http://www.foveon.com/article.php?a=74 to understand more about their chip.

you can just see the dcraw code instead.
Logged
ErikKaffehr
Sr. Member
****
Online Online

Posts: 6890


WWW
« Reply #31 on: July 20, 2011, 12:14:33 PM »
ReplyReply

Hi,

The issue is more to interprete color information and handling perception of color than programming.

Best regards
Erik

you can just see the dcraw code instead.
Logged

douglasf13
Sr. Member
****
Offline Offline

Posts: 546


« Reply #32 on: July 20, 2011, 04:35:38 PM »
ReplyReply

I'd be curious to see how the Bayer cameras would do against the SD1 with a floating point raw converter, like RPP.  The amount of detail one gets out of RPP compared to ACR is pretty stunning.
Logged
BernardLanguillier
Sr. Member
****
Offline Offline

Posts: 7522



WWW
« Reply #33 on: July 20, 2011, 08:34:51 PM »
ReplyReply

For example, in his comparison of the Nikon D3X to the Leica S2 Digilloyd (a pay site well worth its modest cost) found that deconvolution sharpening with the Richardson-Lucy (RL) algorithm brought the results obtained with these two cameras nearly to parity and that he was unable to obtain equivalent results for the Nikon with ACR/light room. According to Eric Chan, the latest iterations of ACR do enable deconvolution sharpening, but the number of iterations must be relatively low, since the conversion is much faster than programs using RL.

Indeed, the D3x files sharpen real well through deconvolution alogs thanks to the weak AA filter of the body, yet have very few digital artifacts in them.

I am still unclear what it would take for Nikon, or any other DSLR manufacturer, to release a body that would deliver enough real world additional value to convince me to spend the cash. The truth of the matter is that the only thing preventing someone using a D3x + stitching to deliver the best landscape images in the world at any print size is... lack of skills.

If I need to spend money on equipment, then the replacement of my old B&W speakers appears to be of higher priority. Smiley

Cheers,
Bernard
« Last Edit: July 20, 2011, 08:38:15 PM by BernardLanguillier » Logged

A few images online here!
John Camp
Sr. Member
****
Offline Offline

Posts: 1251


« Reply #34 on: July 20, 2011, 10:22:44 PM »
ReplyReply

to deliver the best landscape images in the world

But what about the best skin tones? And what about tornado chasers -- would their shots be best done with stitching? What about motocross racing...this site may have "landscape" in its name, but most people here shoot more than landscapes. So "best" is, at best, a fairly slippery concept.

That said, I still don't see the reason for all the excitement about Foveon, or even why people see so much "potential" in it. What, exactly, would it do better that what we have? What *could* it do better? I think some people like it for the theoretical efficiency of its engineering, rather than for imaging reasons. Does anybody here except Ray (excuse me, Ray) really believe in that 3x multiplier effect, or whatever it is?
Logged
Sareesh Sudhakaran
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #35 on: July 21, 2011, 01:28:55 AM »
ReplyReply

Absolutely! This new Foveon sensor is slightly larger than previous Foveon sensors. It has a crop factor of about 1.5X, ie. a full frame sensor is 2.5x the area.

A full frame 35mm Foveon sensor with the pixel pitch of the SD1 would be the equivalent of a 67.5mp Bayer sensor, so your question is a bit like, "Is a 67mp full frame sensor higher resolution than a 24mp full frame?"

Of course, the result will depend on the use. There will not be much difference between 30mp and 24mp when lenses of different focal lengths are used to maintain equal FoV using the the full area of each sensor. However, when using the Bigma 300-800 with the SD1 at 800mm, which becomes effectively 1200mm, there should be a very noticeable increase in resolution compared with a 24mp full frame cropped to the same FoV. One is then effectively comparing 10mp with 30mp.

Thank you, Ray, for that explanation. Its utility makes a little more sense now, especially in relation to telephoto work.
Logged

Get the Free Comprehensive Guide to Rigging ANY Camera - one guide to rig them all - DSLRs to the Arri Alexa.
Aku Ankka
Newbie
*
Offline Offline

Posts: 24


« Reply #36 on: July 21, 2011, 03:45:43 AM »
ReplyReply

I don't agree with Michael's statement in the review "Some people may not be aware of it, but Color / Chroma Noise and its associated removal tool in various processing programs is there to remove noise introduced by Bayer Filter Array decoding. One would therefore assume that a Foveon type X3 sensor, which doesn't use a Bayer Matrix, wouldn't have Color Noise, and indeed it doesn't." Chroma noise comes from the colour correction matrix amplifying sensor noise. Without such a matrix a Bayer CFA image looks a bit under-saturated, and has some hue errors. As noted, the Foveon has a very aggressive colour correction matrix, and you can see how necessary that is by looking at the paper above with the spectral responses of the three layers and by how much they overlap each other. It's the overlap that needs to be removed by the CC matrix, and the more overlap, the stronger the matrix, and the more noise amplification, hence the necessary NR steps in the Foveon processing mentioned above. In a Bayer CFA, chroma noise does not come from the demosaic process, but instead from the CC matrix, which due to the use of good dye filters on the CFA is a much less noise inducing matrix than that used on Foveon.

I think the colour anomalies noticed by Michael, the red hue in the shadows, the colours not being "right" are fully due to the issues with extracting colour information from the Foveon sensor. As all things are a balance, these colour anomalies and associated NR processing are the price paid for lack of chroma moire. If you don't get chroma moire in your shots (and even if you do, it's usually quite easy to remove), you'll achieve much better colour accuracy from a Bayer CFA.

I think that the biggest problem with this "review" was that Mr Reichmann is not competent to write reviews invoving the digital imaging technology. That is not his strength. There were numerous errors and inaccuracies in the article - the one you quoted here is a great one (I had to read it twice to make sure he really said what he said and then I had a good long old fashioned  laugh rolling down the floor  Cheesy) - yet it was posted.

Here's a quick short list of some problems in the article:
  • Foveon is on a sensor with stacked RGB, but instead the top layer gathers pretty much all violet, almost all blue, plenty of green and many red photons as well, the middle one most of the remaining green, plenty of orange and red photons as well, while the bottom tries to capture the remaining photons. Calling this color soup "RGB" is misleading.
  • No decent demosaicing algorighm uses mostly green pixels for luminance.
  • He mixes pixel count and resolution. Resolution is one dimensional, thus making a statement like "pixel resolution on a Bayer sensor should be reduced (for purposes of comparison) by about a third" is problematic as some read it as the pixel count should be about 2/3 in comparisons, while the word resolution itself indicates that this should be read as 4/9 of the pixel count. Both are of course wrong. The demosaicing itself reduces resolution potential, depending on the algorithm used, by about 10%. And then he repeats the baseless claims that X3 resolution should be double (when comparing to Bayer CFA imagers). Adding these two claims together makes a Foveon pixel worth more than four Bayer CFA pixels for resolution. The reality is of course that the difference is very small and it is the AA-filter which makes most Bayer CFA equipped imagers look less detailed (and to a certain extent be less detailed as well).
  • He talks about sharpening, images being optimally sharpened, but the sample crops show that the images from the AA-filtered cameras are clearly undersharpened. I added some sharpening (with no visible sharpening artifacts) and suddenly the A55 image looked easily as sharp and detailed as the SD1 image. Proper capture sharpening is critical for cameras with anti-alias filter.
  • His understanding on why the Foveon colors are inaccurate is nonexistent. There are two problems with this sensor and neither he knows of, but instead wonders if the color "palette" is Sigma's choise or software error - the Foveon (technology) has extremely poor color separation and mememerism issues (ie. when two spectra look the same color to human eye, they may look very different to Foveon and vice versa) which are impossible to fix.
  • He implies that the sensor is great when he says it deserves great lenses. I imagine the worst colors in the industry, low dynamic range and high noise are not relevant when reviewing a quality of a sensor.
  • He praises the smoothness of color transitions and talks about 12 and 14 bits - the sample files are noise reduced, but since they're JPGs, it's hard to say how much is because of the raw-conversion. The noise reduction makes the transitions appear smoother - bits have nothing to do with it as most imagers could do well without the 13th and 14th bit as they're just noise even at the base ISO (yes, including medium format, Pentax and Phase One included).
  • His eyes tell him that upto ISO 800 the IQ is "excellent" and then suddently at ISO 1600 marginal - if such a drop happens with one level adjustment, there is something wrong somewhere in the analysis. In this case it is almost certainly the noise reduction used. I imagine the (edge detecting) algorihm Sigma uses doesn't work well on noisy images, thus ISO 1600 is bad. Also to my eyes the fine details started deteriating almost immediately after the base ISO. Too bad the crops he provided are of the worst kind. The only thing the flat surfaces tell is that there is lots of noise reduction being done somewhere as pushing and twisting the images doesn't show even photon shot noise. He doesn't see this, but thinks the sensor is great.

The Leica lens used in the sharpness test seems awesome, easily the best of the bunch. I want one Wink
Logged
michael
Administrator
Sr. Member
*****
Offline Offline

Posts: 4729



« Reply #37 on: July 21, 2011, 07:04:40 AM »
ReplyReply

Well Aku, you certainly seem to have fun dissing me and my write-up.

The problem is that while your comments certainly sound definitive, each one can be rebutted and even discounted – or at least is the subject for some debate.

Since I have no time or desire to refute your comments on a point-by-point basis, it'll leave it by saying that the article was reviewed prior to publication by several knowledgeable industry commentators, including a couple with extensive Foveon background and experience. There were no substantive errors noted.

Let's just leave it at that.

Michael
 
« Last Edit: July 21, 2011, 12:03:31 PM by michael » Logged
fredjeang
Guest
« Reply #38 on: July 21, 2011, 07:34:43 AM »
ReplyReply

I have to say that I'm giving full authority to MR experience.

This comes to a simple fact: each time (without just one exception) I had the oportunity to work (not only try but work) with equipment he reviewed,
it completly matched with what I was experiencing.

If I had to read reviews in order to help me choose a gear, I'd just have to open this website and could blindly trust the content with confidence.
That's what makes Lu-La's reputation and I think it's well deserved.

The last time I had a doubt on Michael's evaluation was on this Panasonic GH2. I honestly didn't beleived his enthousiasm but finaly purchased one for motion
and after a while with the camera, I have nothing else to admit that his comments where again spot-on.

The trustable Michael's experience comes IMO from 2 direction: a deep knowledge of the image industry and a truth knowledge as a photographer
who's been working with almost all equipment produced on earth. Yes, he may not be an engineer but I couldn't care less because each time
testings are done following a rigourous "scientific" pattern, they simply don't match (or rarely match) with the field results while the field evaluations
of mister Reichmann yes do match. In this sense I find his writings a trustable source of information, much more reliable than most of the testing websites
with the exception of a few, that come to the same conclusions anyway.

We shouldn't forget that MR is in touch regularly with techs and engineers within the profession, used to run an art gallery, is a master of prints, used to
worked in the industry as a profesional (most of the people who present themselves under the scientis label and think they know have never worked as a professional,
neither photographers nor engineers in this industry. The exception here is Graeme and maybe a few more we could count on the finger's hand), and has been using more gear of all kind that I would probably never see in my entire life.
He is in the confidence of some manufacturers and knows things we don't.

So yes, personaly, I give him all the credits he deserves when it comes to gear and image (quality) knowledge.



Logged
Graeme Nattress
Sr. Member
****
Offline Offline

Posts: 582



WWW
« Reply #39 on: July 21, 2011, 08:36:39 AM »
ReplyReply

Interesting thing is, when I first started interacting with sensor designers, one of the first questions I asked is about the Foveon design. I wanted to know why it looked as it did, and why such a technique was not more wildly used. The simple answer is that silicon depth is a very poor colour filter. You can see that from the published Foveon response curves that I referenced above. Almost everything about the image stems from this fundamental element of the design of how it senses colour, and the necessary math needed to untangle a proper colour response from the heavily over-lapping curves.

Graeme
Logged

www.nattress.com - Plugins for Final Cut Pro and Color
www.red.com - Digital Cinema Cameras
Pages: « 1 [2] 3 4 ... 11 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad