Forgot Password
Pentax Camera Forums Home
 

Reply
Show Printable Version Search this Thread
05-26-2011, 07:28 AM   #121
Forum Member




Join Date: Dec 2010
Posts: 53
QuoteOriginally posted by ogl Quote
It seems to me you are troll. Sensor could has different dimension, but it's the same TECHNOLOGY. Have a look at official samples - the same problem with colours.

LL is MUCH clever than you and more professional
Weird, to me it seems you're a troll. Almost one line answers with shouting as if the people in this forum could not read otherwise. The last statement you said is half-insulting and clearly a troll-behavior.

Also you don't seem to read my posts, but just react. Otherwise you might have noticed that one of the issues of the whole silicon based color separation (ie. three layers of photodiodes at different depths, relying in the wavelength dependant probabilistic nature of photon absorbion) is color and color noise. I may or may have brought up the observer metameric failure issue as well - the organic human eyes and inorganic filtering of the Foveon do not see colors the same way - there was a good post on the issue by joseph Wisniewski on the DPR a couple of days ago. (And also Foveon engineers themselves have ignologed this long time ago.)

05-26-2011, 07:35 AM   #122
Forum Member




Join Date: Dec 2010
Posts: 53
QuoteOriginally posted by Pentaxor Quote
yeah, the blue end seems to be the major issue in the Foveon. it interests me to know how Sigma would be able to overcome that on the SD1.
It is not just one color, but it is two separate issues. One is the noise from the low color separation. The bigger issue is the metameric failure - one may be able to correct the colors well for some parts of the image, but doing so may wreck a havoc to the another part and vice verse. Again, Joseph Wisniewski understand this better than I do, so quickly browsing through his last few days posts on the DPR allows one to find the relevant post on the issue I am sure.

Anyhow, since it interestes you to know how SD1 behaves regading this, I am sorry to report that this is not fixable, as this is how a Foveon-style sensor behaves by nature. (I am not sure how much adding more layers would help though, but that would not be too sensible solution for a consumer mass product.)
05-26-2011, 07:39 AM   #123
Forum Member




Join Date: Dec 2010
Posts: 53
QuoteOriginally posted by thibs Quote
Yep, it's not that Digma really competes with Nikon. Or do they?
Well, they both make lenses for Nikon-cameras and Sigma's often competertive pricing certainly hampers Nikon (and Canon et al) profits.
05-26-2011, 07:47 AM   #124
Forum Member




Join Date: Dec 2010
Posts: 53
QuoteOriginally posted by Class A Quote
Hi Falk,

stunning model! Hope we'll see more of her.


Something is odd. When I view the original Sigma Sigma-SDIM7925AA.jpg file side to side with your crop, the eye appears significantly larger than in your 100% crop. That shouldn't be the case?!?
Seems like even Falk is a mortal and makes mistakes

QuoteOriginally posted by Class A Quote
BTW, the part where the lashes are the sharpest in the SD 1 image (towards the outer corner of the eye) is not visible in your crop. The focus is pretty forward, making the iris look dull. Any shot with a detailed iris will look more impressive in comparison.
I disagree a bit on the docus - looking at the top SD1 eye lashes, it is clear that the ends (parts closest to he camera) are already OOF, meaning that the focus is closer to the face than that. It does however look a bit like the left side of the eye is also OOF, so the area in focus should be somewhere in-between.

05-26-2011, 07:48 AM   #125
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,862
QuoteOriginally posted by Class A Quote
Something is odd. When I view the original Sigma Sigma-SDIM7925AA.jpg file side to side with your crop, the eye appears significantly larger than in your 100% crop. That shouldn't be the case?!?

BTW, the part where the lashes are the sharpest in the SD 1 image (towards the outer corner of the eye) is not visible in your crop. The focus is pretty forward, making the iris look dull. Any shot with a detailed iris will look more impressive in comparison.
Class A, thanxs for spotting this.

The link I provided did still include some rescaling done by the image hoster (rescaled to 75% of what I uploaded). I fixed it now.

You're right about the sharpest part. But the eyelashes have quite some depth meaning that my crop contains the focal plane as well. I actually looked at the respective eyelashes when comparing shots.

Overall, I think I won't revise my humble opnion expressed above

P.S.
My model is Patty from Munich ( Pet des Monats Mai: Patricia - Penthouse.de ). Don't know about the model in the Sigma shot. I have more shots but pentaxforums always comes back with their NSFW acronym
05-26-2011, 07:52 AM   #126
Pentaxian
JohnBee's Avatar

Join Date: Sep 2008
Location: front of computer
Posts: 4,496
QuoteOriginally posted by falconeye Quote
In conclusion, I don't think the SD1 does any better in good light compared to a K-5. Resolution-wise. Despite the K-5's AA filter

OTOH, I'm sure the 645D smashes the SD1 resolution wise and the K-5 is far superior in almost all other aspects (speed, high ISO, weather etc.).

So, I am sorry to say but at the moment, I can't see the point in Sigma's market proposition. Which is sad for an otherwise interesting technology.
Actually, I've been debating the strangeness of Sigma's promotion images for several days now. However, in all fairness, there are a few points that have yet to be addressed.

1. None of the sample(safe one) show the actual pixels.
2. The lens in the mentioned samples are at higher focal lengths(100mm). However, I did find one 70mm sample that looks much more like the K-5 at 1:1 respectively. However again, we don't get to see the pixel level(re-sampled?).

The only image(I found) that showed pixel level was of the dam, which was shot at 16mm. And I must admit. It does look pretty impressive. However without RAW, there is no way to tell just how well it scales to 45mp. And so I for one would love to see that.

However, in defense of the K-5, I must say, I am very impressed with what can be squeezed out of the K-5 with a typical lens and casual setting:

Throwaway image w/comparable framing:



First the SD1:


Then the lowly K-5: - (cropped to match)


A little larger: - just to be sure


Obviously at this level the K-5 pixel radius is much smaller than any of the SD1 samples. Which raises a rather controversial question... if Sigma chose these samples to launch the product, why didn't they provide high res. samples?

And I guess only time will tell.
But something tells me that this won't be a true 45mp camera.
Or at least not in comparison to existing 45mp camera's anyways.

My two cents.

PS. Having gone through several images, I found several shots that exhibit sharpness like this. And I never really examined them before, but it after having done that, I was like... hmmm, those are pretty damned sharp for a lowly 16mp camera

Last edited by JohnBee; 05-26-2011 at 07:59 AM.
05-26-2011, 08:08 AM   #127
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,862
QuoteOriginally posted by JohnBee Quote
1. None of the sample(safe one) show the actual pixels.
2. The lens in the mentioned samples are at higher focal lengths(100mm).
I don't get your point.

The SD1 is a 15MP camera. Period. So what we see are actual pixels.

Sigma's argument is that the SD1 pixel quality is superior because they live w/o Bayer and AA filter.

And I challenged this exact statement by making my comparison post.

There would be no additional information in an SD1 RAW file, except the 16Bit depth. Both, SD1 RAW and JPG are 15MP x 3 colors each.


BTW, the #1 way to sharp images is good lighting. Which is why I selected a studio image from my archive. Any serious comparison should be with similiar lighting. The differences in focal length (100mm vs 70mm and aperture f/7.1 vs. f/8) are neglegible though.


I said above that IMHO the SD1 doesn't beat the K-5 resolution wise in good light.

In all fairness though, there are situations where an already blurred image (motion blur, defocus etc.) would turn out sharper from an SD1 than a K-5. Because then part of the resolution would be lost irreversibly.

Last edited by falconeye; 05-26-2011 at 08:14 AM.
05-26-2011, 09:01 AM   #128
Pentaxian
JohnBee's Avatar

Join Date: Sep 2008
Location: front of computer
Posts: 4,496
QuoteOriginally posted by falconeye Quote
I don't get your point.
What I'm saying is that I don't think were seeing the full potential of the sensor in these images. The minimum detail radius between those two images is dramatically different. And though I'm sure the lens plays a part in this, the dam is showing much higher level of detail(minimum radius) or overall line width, than with the model.

Additionally, the second model taken with a 70mm lens is likely a better candidate, though again... since the image has already been processed, it is unlikely that we could ever draw any sound conclusion from that one, beyond what is being shown. Though I didn't find any issues upsampling the image to 25mp myself, and so something tells me that RAW files would likely yield better results.





And so I'm waiting for the RAW
And the first thing I would do is to try and push this to 45mp and see how things look.

PS. I edited pixel with detail which I think better explains it(minimum detail radius).

QuoteOriginally posted by falconeye Quote
The SD1 is a 15MP camera. Period. So what we see are actual pixels.
Sure, but that's not really that important is it?
Apparently there was a 3mp camera(D30) that could rival film way back in the day.
I never saw any RAW samples(couldn't find any) but according to those who lived it, apparently it was every bit as detailed.
And so I'm thinking Sigma believes their 15mp sensor can rival a 45mp. Which remains to be seen, but.. it will definitely be interesting to watch as things unfold

However some of the samples open in ACR, and so they can be upsampled, but since they are preprocessed(sharpened etc), I think that also presents limitations.


Last edited by JohnBee; 05-26-2011 at 09:26 AM.
05-26-2011, 11:42 AM   #129
Veteran Member
Pentaxor's Avatar

Join Date: May 2009
Location: Vancouver, B.C.
Photos: Gallery
Posts: 6,513
QuoteOriginally posted by Aku Ankka Quote
I do have access to Sigma images, I do understand how the sensor works, I do understand how the relevant physics work, yet all that is irrelevant because I don't own a Sigma camera? Will the same be true for every other product and technology in the world - only those owning a product can know anything about them?
I thought you said you have better resources? I'm asking for more basis aside from what you possibly borrowed from theories of others that coincidentally don't have the SD1 for testing such theory is indeed holds true.

QuoteQuote:
You don't seem to understand that the Foveon technology has certain hard limitations - no actual hand holding of a Sigma camera is needed to understand them. One of theses limitations is the noise factor (for color imagery). This is because of the very weak color separation which forces the noise up. This is something that no mumbo jumbo will overcome.
same argument as above.



QuoteQuote:
On the other hand, if the new sensor in the Sigma is based on the recent patent presented in many forums, then the sensor will have correlated double sampling which takes away so-called reset-noise. This is somehting that is indeed fixable, and what we do not know if it's happened or not. If that patent describes the SD1 sensor, then it'll also have only a quarter of the red and green photodiodes compared to the blue ones.
again, this is assuming. there is still no definitive answer due to lack of resource or practical application of the not yet released SD1.




QuoteQuote:
You should learn to be more critical of photography sites nstead of laughing at other people. Laughing at others is not polite.

LL has a long history of very subjective articles which carefully avoid reality when it comes to hard performance figures. However, they also do have plenty of articles of interest, but the ones which should have some scientific accuracy do not usually have much.
I believe I'm entitled to laugh at what you posted. LL is in the dark as much as you do concerning the SD1 since neither of you owns the SD1. atleast LL gives it's reservations rather than blatantly proclaim that they knew more about the SD1.


QuoteQuote:
I am not sure if you just want to insult me with you lack of reason and logic, or what? There is nothing that will defeat my argument in me saying that I might buy SD1 if the price is under 1k. As I said, it is an interesting product (well, the sensor is, the camera itself is not), and should be very good for black and white photography where the noise from the color separation does not come to play.
can't other cameras do B&W as effective or good as other cameras?
do you really know what the Foveon sensor emphasizes on or atleast what it's main strength is ?

QuoteQuote:
It seems like you live in a world of absolutes. I don't. While I am critical on the Sigma cameras and Foveon sensors, that does not mean I fail to see their merits. I live in a world where one may be interested in a flawed product and even buy one - your words strongly indicate that such behavior is not possible in your world.

I'm sorry to hear that you think of such. it is not about absolute but rather finding solutions or ways to overcome things that seem finite already or something that hold true as far as physics and logic goes as of the moment. much like discovering that the world is really round and not flat. also remember the time where MP count is already maxed out it's optimal performance at 12-14MP resolution for the APS-C sensor, and yet disprove that it could still perform extremely well over those MP range.
05-26-2011, 11:42 AM   #130
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,862
John, I get you now. Thanks for explaining your point.

However, I really think you are complicating matters here.

The point is that both, K-5 and SD1, are ~15MP cameras. You may want to read my paper about understanding image sharpness (found via my blog). As soon as a camera yields ~1.2 pixels 10-90% edge blur w/o sharpening artifacts at the raw pixel level, there can nothing be gained anymore except by a higher Nyquist frequency. And SD1 has NO higher Nyquist frequency.

Now, what Sigma marketing is implying, is that an APS-C camera with Bayer and AA filter is not reaching this ideal limit (of 1.2 px edge blur). And then they even fail to demonstrate this capability for their own samples except maybe the dam sample as you properly point out.

My point (in my post above) was that the K-5 actually DOES reach the 1.2 px edge blur limit and provided a sample image (rather than a chart) for a change

This means that an SD1 gains nothing over a K-5, at least not in the studio or with good light.

Of course, you can upsample an SD1 image to 45MP. And so you can for a K-5 image. But this cannot create detail which is beyond the Nyquist limit. So, it is a desparate trial by marketing to justify the ridiculous SD1 pricing.


As for what we see demonstrated here...

It is very hard to fully exploit a 15MP APS-C camera. Autofocus erros, VF blur, lens abberations and body vibration are all magnified compared to full frame or medium format. Therefore, the ability of photographers to deliver tac-sharp samples with a given camera is a meaningful test. And the photographers selected by Sigma seem to have had trouble to do so. More than me with the K-5, it appears.

It doesn't matter so much how sharp a camera's images can be. It matters more how sharp they typically turn out to be. And there, the SD1 doesn't excel and can't compete with FF or MF.

To make it even more clear: A 5DmkII would win hands down against a SD1 in any category I can imagine, including resolution.

Last edited by falconeye; 05-26-2011 at 11:49 AM.
05-26-2011, 11:54 AM   #131
Veteran Member
Pentaxor's Avatar

Join Date: May 2009
Location: Vancouver, B.C.
Photos: Gallery
Posts: 6,513
QuoteOriginally posted by falconeye Quote
John, I get you now. Thanks for explaining your point.

However, I really think you are complicating matters here.

The point is that both, K-5 and SD1, are ~15MP cameras. You may want to read my paper about understanding image sharpness (found via my blog). As soon as a camera yields ~1.2 pixels 10-90% edge blur w/o sharpening artifacts at the raw pixel level, there can nothing be gained anymore except by a higher Nyquist frequency. And SD1 has NO higher Nyquist frequency.

Now, what Sigma marketing is implying, is that an APS-C camera with Bayer and AA filter is not reaching this ideal limit (of 1.2 px edge blur). And then they even fail to demonstrate this capability for their own samples except maybe the dam sample as you properly point out.

My point (in my post above) was that the K-5 actually DOES reach the 1.2 px edge blur limit and provided a sample image (rather than a chart) for a change

This means that an SD1 gains nothing over a K-5, at least not in the studio or with good light.

Of course, you can upsample an SD1 image to 45MP. And so you can for a K-5 image. But this cannot create detail which is beyond the Nyquist limit. So, it is a desparate trial by marketing to justify the ridiculous SD1 pricing.


As for what we see demonstrated here...

It is very hard to fully exploit a 15MP APS-C camera. Autofocus erros, VF blur, lens abberations and body vibration are all magnified compared to full frame or medium format. Therefore, the ability of photographers to deliver tac-sharp samples with a given camera is a meaningful test. And the photographers selected by Sigma seem to have had trouble to do so. More than me with the K-5, it appears.

It doesn't matter so much how sharp a camera's images can be. It matters more how sharp they typically turn out to be. And there, the SD1 doesn't excel and can't compete with FF or MF.

To make it even more clear: A 5DmkII would win hands down against a SD1 in any category I can imagine, including resolution.

this could be the scenario Falk. although it would had been better I think if we had RAW samples to play with and certain scenarios that would show it's advantages over the K-5.

Sigma already made a booboo on the pricing, guess they kept the boobooness consistent by providing inadequate and unsatisfactory samples.
05-26-2011, 04:33 PM   #132
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 8,934
QuoteOriginally posted by falconeye Quote
But the eyelashes have quite some depth meaning that my crop contains the focal plane as well.
Yes, you are right, but personally I don't think that any of the visible lashes impart a sensation of high acuity. The latter happens more (but probably not to the extent possible) with lashes not visible in the crop.

QuoteOriginally posted by falconeye Quote
Overall, I think I won't revise my humble opnion expressed above
Well, I guess there is no need to. Your K-5 image is very impressive.

However, I think the jury is still out a bit. I feel that your image shows signs of oversharpening. There is some noise-like structure in the white of her eyeballs that doesn't belong there. In contrast, the SD 1 shot seems strangely coarse. As if the pixels had been much larger and then cleverly blurred to hide the pixel boundaries. As John Bee said, I doubt we are seeing a fair pixel-to-pixel comparison, but then I cannot really put my finger on what the problem could be.

QuoteOriginally posted by falconeye Quote
My model is Patty from Munich ( Pet des Monats Mai: Patricia - Penthouse.de ).
Nice!

QuoteOriginally posted by falconeye Quote
I have more shots but pentaxforums always comes back with their NSFW acronym
Maybe you could post them to one of your special sections?
05-26-2011, 06:51 PM   #133
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,862
QuoteOriginally posted by Class A Quote
I feel that your image shows signs of oversharpening. There is some noise-like structure in the white of her eyeballs that doesn't belong there.
...
Maybe you could post them to one of your special sections?
My image is certainly sharpened significantly (actually, it has a microcontrast effect applied with 25% strength). I had a certain visual impact in mind when doing so (like stunning iris) and not the 100% crop appearance. The image is a couple months old.

I'll build up a selection of photos on the page you suggested but it is taking time
05-26-2011, 09:24 PM   #134
Pentaxian
Class A's Avatar

Join Date: Aug 2008
Location: Wellington, New Zealand
Posts: 8,934
QuoteOriginally posted by falconeye Quote
I'll build up a selection of photos on the page you suggested but it is taking time
Looks like it will be worth the wait.

All good regarding your sharpening. I'm also very sceptical that the SD 1 will be able to justify its current price. The Foveon principle appeals to me, though, and I wish Sigma well in further refining the implementation and succeeding with it.
05-27-2011, 05:00 AM   #135
Pentaxian
falconeye's Avatar

Join Date: Jan 2008
Location: Munich, Alps, Germany
Photos: Gallery
Posts: 6,862
QuoteOriginally posted by Class A Quote
All good regarding your sharpening. I'm also very sceptical that the SD 1 will be able to justify its current price. The Foveon principle appeals to me, though, and I wish Sigma well in further refining the implementation and succeeding with it.
I think, there is a good side to a sad story here.

If SD1 becomes the failure we all think it will (given the price policy), then Sigma may want to sell off Foveon Inc. I imagine there are better partners than Sigma to exploit the technology.
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!
foveon x3, pentax news, pentax rumors, sensor, technology
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
K-x Sony Sensor vs K-7 Samsung Sensor karl79 Video and Pentax HDSLRs 9 09-23-2010 09:35 AM
the sensor in k-x techmulla Pentax DSLR Discussion 1 05-23-2010 08:05 PM
K-7 sensor vs K20d sensor Mystic Pentax News and Rumors 33 06-21-2009 03:01 AM
Sensor cleaning: Pec-Pads or Sensor Swabs gadgetnu Pentax Camera and Field Accessories 29 09-24-2007 10:52 AM
Sensor cleaning > Sensor Swab > void warranty? Twinky Pentax DSLR Discussion 2 07-28-2007 01:10 AM



All times are GMT -7. The time now is 12:08 PM. | See also: NikonForums.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