Ad
Ad
Ad
Pages: « 1 [2] 3 4 5 »   Bottom of Page
Print
Author Topic: Sigma SD10  (Read 26150 times)
janus
Guest
« Reply #20 on: May 08, 2004, 10:42:07 AM »
ReplyReply

Actually, careful use of the fill-light feature in the Sigma SPP software, which can be used in a positive and negative direction, may have cured this as well.

There is also a way in PS to cure this: blend a slightly over and a slughtly underexposed image and blend the two to get rid of blow outs and block ups.

After all, only that part was overexposed, while the rest of the shot looks fine. Sometimes, in real life, though, a white part being struck by such strong sunlight will even be "blown out" to the human eye, let alone a camera.

While I do not necessarily disagree with you, I would love to see the raw file and play around with it.
Logged
61Dynamic
Sr. Member
****
Offline Offline

Posts: 1442


WWW
« Reply #21 on: May 25, 2004, 11:17:07 AM »
ReplyReply

[font color=\'#000000\']
Quote
Tell me, is this a good camera, or what?

I would say "what."

Actually I should be more specific. It is a good cam as you are capable of taking pictures with it. But it is plagued with issues wich keep it from competing properly with the rest of the cams out there. The SD10 is only the second generation of the sensor and corrisponding tech. Or perhaps even v1.5 of the tech as I think there have only been software improvements since the SD9.

I've gotten a chance to play with some of it's RAW files and the PP2 software. It' a good 3mp camera that can perhaps compete with a 4.5mp camera in amount of resolved detail and it has good color. PP2 is a great program.

Considering the cost of the camera compared to the capabilities of it's competition it's not the greatest camera out there. And it's certainly not as good as you've (Janus) been touting.[/font]
Logged
janus
Newbie
*
Offline Offline

Posts: 39


« Reply #22 on: June 04, 2004, 09:19:35 AM »
ReplyReply

Quote
[font color=\'#000000\']The down side as far as image enlargeability is concerned is that when the subject material has geometric lines at other than 90, 45 or 180 degree angles, the 3.5 megapixel file array produces visible stair-stepping aliasing which is readily apparent in printed enlargements.[/font]
[font color=\'#000000\']How much can aftermarket software overcome these "jaggies"? What do you think is the best program?

I have used spline, but have been disappointed. I have used bicubic at 5% intervals, which is better than Spline. I am not familiar with the results of Genuine Fractals or other programs out there. I havew also used Fred Miranda's StairInterpolation, and it's not bad, but not discernibly better than the 5% solution.[/font]
Logged
JJP
Full Member
***
Offline Offline

Posts: 208


« Reply #23 on: February 03, 2004, 05:05:27 PM »
ReplyReply

Michael did not review the Contax N Dig.  He wanted to, but could not get one from the distributors.
Logged

JJ
how786
Jr. Member
**
Offline Offline

Posts: 98


« Reply #24 on: February 06, 2004, 07:38:42 AM »
ReplyReply

Wow, Janus----gorgeous!!!
Howard
Logged
Ian
Guest
« Reply #25 on: February 10, 2004, 07:16:54 PM »
ReplyReply

I don't like the algorithms that have to guess during their interpolation. Yellowy blotchiness in skin & hair, purplely blotchiness on grey walls, structures. Pretty  poor software Mr Bayer! Huh? It's not a 6MP camera it's 2 MP except for green sensor, when it's 4MP and the photosites are not even adjacent. Every one  always misses this fact.


 
Logged
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #26 on: February 11, 2004, 09:58:54 AM »
ReplyReply

Here's my dream camera:

Son of 1Ds, 16MP, 8 FPS. 4 layer Foveon sensor, B+G+R+IR. (no kludging with filters for IR shots) Color accuracy/sensor noise better than current 1Ds even at ISO 6400. Oh, and less than $3000.
Logged

janus
Guest
« Reply #27 on: February 13, 2004, 11:33:58 PM »
ReplyReply

This page has had more than 1000 hits in about two weeks! Not bad!

Actually, the Japanese guy at d-shell.net is working on different adaptors, but needs a suficient number of orders to make it worth his while. The M42 adaptor has been the only one made so far, and a new version is coming in March.

He has also a prototype SD9 camera with a new front plate to accept Canon lenses, that work with autofocus and auto f-stop control. Basically, it is a one-time conversion to a Canon mount/cameras with a Foveon sensor. He showed me a movie of it. But nothing concrete yet, as he is looking for investors. A wild idea, that I am not sure will take that easily. You would loose any guarantee on the Sigma part of the camera, and I am not sure the Japanese guy will warrant the entire camera. But as SD9's drop in pice, it may be worth a gamble, unless the conversion is too darn expensive.

But, and this is interesting, the Sigma mount is apparently so close to the Pentax K mount, that if you make minor modifications to the K mount, by removing the plastic protruding piece on the bottom of the lens and also cut the f-stop pin, you can use it as a fully manually controlled lens on either Sigma SD9 or SD10 (and their film cameras for that matter). You loose 1.5mm of spacing, and I am not sure how it affects the infinity focus of the lens.

The Sigma SD10, for outdoor work, continues to amaze me; see my updated gallery (update to be completed before Feb. 18).


http://www.pbase.com/janus

Also check out Sigma Japan's new cool website in English language:

http://www.sigma-photo.co.jp/sd10/english/index.htm


SD10\" target=\"_blank\"][a href=\"http://www.pbase.com/image....RL=http\" target=\"_blank\"]http://www.pbase.com/image....RL=http[/a] photo[/a]
Logged
Alf B.
Newbie
*
Offline Offline

Posts: 3


« Reply #28 on: February 14, 2004, 03:35:51 PM »
ReplyReply

Ray you are way off the mark mate and Janus you should really know better....You are both mistaking the image size for the amount of resoluton data the chip outputs.

The Foveon X3 is a TRUE 10.2 megapixel sensor because it has 3.4 million green pixels, 3.4 million red pixels and 3.4 million blue pixels.

A 6.2mp EOS 10D, in comparison, has only 3.1 million green pixels, 1.55 million red pixels and 1.55 million blue pixels.

This means that both the SD9 and SD10 have slightly more green pixels and double the amount of red and blue pixels than an average 6mp Bayer sensored DSLR.

In fact the nearest directly comparable Bayer sensored DSLR is the Canon 11mp EOS 1Ds which has 5.5 million green pixels, but only 2.75 million red pixels and 2.75 million blue pixels.
You also have to pay 5-6 times more for the 1Ds so its not exactly value for money!!!

As you can see, compared with the 1Ds the Siggy SD's both have more red and blue pixels and only have a little less green pixels.

You might think this would help the 1Ds to resolve green detail such as foliage better than the SD's but in practice the opposite occurs due to the sharper image produced by the Foveon X3 sensor.

I have never heard of any Bayer sensored DSLR, whether it be the Canon EOS 1Ds or the Kodak 14N being compared to the resolution of a medium format film camera but I have with the Sigma SD9:

http://www.sd9.org/sd9vrsmed.htm

What does this mean in practice?....It means the Foveon X3 endowed cameras can produce huge high quality prints more than double the size that any 6mp Bayer sensored DSLR is capable of:

http://forums.dpreview.com/forums....7182672

Oh and if any of you here are into IR photography you should seriously think about getting the SD10 as if you remove the "dust protector" you have probably the worlds most sensitive digital IR camera. (The dust protector is also the the IR filter on the SD10)
Note: this will not work with the SD9 as its IR filter is not removable.

Alf B.
Logged
janus
Guest
« Reply #29 on: February 15, 2004, 02:22:34 PM »
ReplyReply

Michael you wrote:

"In any event, I'm no expert in this area, and have no reason to believe that you are either. But I have talked extensively with experts in this regard and am comfortable with my understanding and explanation."

So you are more right that the other non-expert? Becasue you talked to experts? So, they really know? And how can you judge the person you are responding to? What experts did he talk to?

One of the flaws in modern sioence is that too many scientists ignore that with which they are uncomfortable. So they frantically hold on to what they feel is comfortable for their own understaning. A kind of "fit" between what they would like to believe and what they have chosen to fit into that category.

But, obviously, this is bogus science. That is belief, not science. How many people "believe" in evolution, even though it is still only a theory? Same with the big bang theory. It is a theory, with plenty of dissenters. Many people take theory for fact. nothing could be more wrong.

Hence, what you say is utterly illogical: how can you on the one hand affirm that are not an expert, but yet hold on to what you "have heard from other so-called experts", even though you also affirmed that the person you were responding to is probablky not an expert either. Are his expert friends "inferior?"

And you urge everyone else not to disucss the issue any further. That is not a very democratic attitude, by the way, and it is simply illogical.

I will not go into the technical discussion about pixels. The SD10 is capabale of taking some "####" good pictures. That is enough. The 17x24 prints I have made on an aging Epson 3000 speak for themselves. Once I get the new Epson these prints should look even better!

Nuf said. Nothing personal. Love this website!
Logged
Guest
« Reply #30 on: February 16, 2004, 07:21:08 AM »
ReplyReply

This is not a simple topic. Also, there isn't much good information available on it for the layman.

Here has a very good explanation of how the Bayer Matrix does its work and how luminance and chrominance information are interrelated.

A couple of points where I disagree with the author. Firsly, the ability to get chromanance information from the R/B sites is now a lot more than a "hack". Some of the math has become very sophisticated. That's why Capture One is such a great RAW converter. These folks do matrix conversion as good as anyone and better than most. But, that's a digression.

The point is that folks that I've spoken with who eat sleep and breath this stuff on an engineering level seem to agree that the reduction in resolution from a Bayer matrix is at worst about 25% these days. This means that a Foveon chip would need to be about 75% the size of a Bayer-based chip to have equal resolution.

To therefore compare a 3.4MP Sigma to an 11MP Canon is not supported by the facts.

Michael
Logged
janus
Guest
« Reply #31 on: February 16, 2004, 07:07:38 PM »
ReplyReply

Well, Sigma now has a OS lens (Optical Stabilizer) an 80-400, or on the Sigma a 135-680mm lens.


See http://www.pbase.com/rickdecker/sigma_80400_os for some test shots by a fellow SD10 owner.
Logged
janus
Guest
« Reply #32 on: February 18, 2004, 04:06:45 PM »
ReplyReply

I haven't done any testing on dynamic range.

But this I can tell you about latitude: I accidentally totally overexposed an image, and even in SPP it showed up nearly all washed out. Yet, with the software controls I was able to almost perfectly restore the image back to normal.

Although the Sigma SD9/10 has auto bracketing, it only makes sense to set it to its maximum setting, as the latitude is so good, that slight differences can easily be overome or mimicked in the software.

As with film, dynamic range depends a lot on the light: the softer and more even the light, the more dynamic range, and the harsher, the less of a range.

I took some shots of white marble buildings today and when I slide the fill light of SPP all the way to the right to make the image really light, I still retain detail in my whites.

In another shot, when I sldie the exposure setting in SPP to the right, I still have a gopod image at 1.5+, and somewhat washed out at 2.0.

Go to Sigma's Japanese website and play with the controls yourself!

http://www.sigma-photo.co.jp/sd10/english/picture/index.htm

(click on the demo bar to activate it and then click on any of the control sliders)
Logged
Guest
« Reply #33 on: March 15, 2004, 08:59:06 PM »
ReplyReply

Too many cameras, too little time.

That's all there is to it.

Maybe one day.

Michael
Logged
Erik M
Guest
« Reply #34 on: March 19, 2004, 12:09:17 PM »
ReplyReply

I would just add that to get the same pixel level sharpness from a 10D or 300D as from a SD9/10 you have to use L glass, which makes the system a bit more expensive. However, these debates are silly, as all the current DSLRs basically are bad cameras--bad either in resolution or dynamic range or good in resolution (1Ds) but bad in price. The real camera we all want is probably several years away.
Logged
janus
Guest
« Reply #35 on: March 27, 2004, 06:43:16 PM »
ReplyReply

RE: M42 lenses:

That's absolutely correct ! ! !
And there are some gems out there; I got hold of a few rare, but excellent MC lenses.

I bought the M42 adaptor from www.d-shelll.net and am using a set of screw mount lenses. I use a lightmeter to set exposure.

Actually, a lot of SD9 and SD10 owners are using all kinds of M42 lenses, with great results, and simply because....well, because it's a lot of fun...

As Bugs Bunny would aptly say to M42 lens users: "Screwy ain't it?"
Logged
drhiii
Newbie
*
Offline Offline

Posts: 1


« Reply #36 on: April 12, 2004, 05:44:41 PM »
ReplyReply

For grins, here is a gallery with sd9 and sd10 pics.  I love the camera(s), a personal choice of course.

drh


http://david.oldcolo.com/gallery/sd9
Logged
janus
Guest
« Reply #37 on: April 22, 2004, 12:38:42 PM »
ReplyReply

Needless to say, the change to your SD10 will be permanent with this kind of operation. Does anyone read Japanese and can decipher this? Is this being done by the same guy from www.d-shell.net who has been working on this, or by someone else as well?
Logged
Alf B.
Guest
« Reply #38 on: April 22, 2004, 03:54:55 PM »
ReplyReply

I have already decipherd the text with: www.babelfish.altavista.com
And its actually in simplified Chinese not Japanese (The japanese use Chinese characters)
The grammar does not translate 100% into English so it can be a bit hard to understand  but some imformation reads very clearly, such as this:

"The Canon and Sigma lens electronics (electrical connections) are identical in configuration"

So it should be fairly easy to make an adapter that will allow AF and camera controlled aperture to be used.
The resulting adapter, the C/SA Teleconverter adapter (shown in the pictures fitted between the lens and the camera)
It is made from a Canon 1.4x teleconverter and a Sigma Teleconverter or an old disposable Sigma lens.
The Canon male bayonet plate is removed from the Canon teleconverter and replaced with a male SA plate from Sigma teleconverter.
I think they have used packing and longer screws to bring the Canons elctrical contacts at the camera end up to meet the contacts inside the SA mount on the SDx.  
A commercialy viable adapter would need the lens elements replaced or re-configured to allow a 1:1 magnification ratio.
Logged
Ray
Guest
« Reply #39 on: April 26, 2004, 08:15:32 PM »
ReplyReply

Quote
A difference is not a deficiency.
Erik,
I can't quite follow your reasoning here. If you check out the pros & cons conclusion on the dpreview site you'll find that the main thing the SD10 has got going for it is its image quality is on a par with the 10D, and by implication the cheaper 300D.

Significant noise at long shutter speeds, color shifts and bleeding above ISO 400 and restriction to Sigma mount lenses are both differences and[/b] deficiencies.

As regards features, the SD10's lack of them is noted as a 'Con', although to be fair there are a couple which the 10D doesn't have (or at least my D60 doesn't), namely 'undelete' and a split histogram.
Logged
Pages: « 1 [2] 3 4 5 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad