SD9&Medium Format

[ChromeLight]

Zone,

I agree with everything you wrote. I'm merely conducting these 'tests' out of my own curiosity. I'm not at all attempting to prove anything. Based upon subject content and enlargement size there are, I imagine, a number of scenarios where ther SD9 will produce Super A3 prints of such quality that I can use the SD9 instead of medium format. Through these 'tests' and the production and examination of prints I'm hoping to establish a few 'rules of thumb' that will guide me as to when I can use the SD9 and when I must use MF.

Erik

 

[whiteOrangedesign]

I shot some landscape shots hand held that were sharper than the ones posted but I wasn't anywhere near 1/15sec. I was probably closer to 1/500. But I didn't use a tripod or MLU. They were in broad daylight which is perfectly acceptable. Agreed that a tripod would of course make it even sharper.

Fact- if shooting with the Sigma using 1/15 to 1 sec exposures you
have to use mirror lock up otherwise you will see noticeable
movement, Sigma made a 3rd rate mirror mechanism that damn vibrates.
I'm talking about sharpness and resolution , you can't use the
Sigma hand held and claim the images are sharp if you do you are
plain blind.
Just try it and make a analogue comparison.

--



http://www.whiteorangedesign.com/

 

[centuar]

As I said it is not my intention to offend ,I'm a pro photographer and my boss asked me to test a variety of cameras as we wanted to get our feet wet and make a start in digital photography.

The criteria was camera and lens quality ,price and work flow time it was no point spending 10,000 dollars and next year a 22 mega pixel camera for half the price.

I spent a couple of months testing cameras in the 5 mega pixel to 6 mega pixel range and for the studio I could only recommend the Fuji S 2 pro

The 6 mega pixel range is suitable for portraits and wedding photography but is limited in so far as you can't shoot large groups of people , we still have to use medium format for that as we don't like lying to our customers like camera manufacture's like lying to us.

The Fuji was reasonably priced , a excellent range in high quality lens unlike the Sigma lenses and working with the Adode raw plugin with photoshop saves a lot of time ,a easy workflow and what I see is what I get as it is is all done in photoshop.
That can't be said with SPP what you see in SPP is not what you get in

photoshop as SPP does not integrate your monitor profile so that means damn more work and in a studio time is money.

I bought the Sigma for my private work as I had very specific needs that the Fuji could not give me and I have the time to play around with colour correcting.

I have read many articles on digital photography for the past 2 years

and it would be great to read and see a real valid digital vs analogue test for a change.

Zone,

I agree with everything you wrote. I'm merely conducting these
'tests' out of my own curiosity. I'm not at all attempting to prove
anything. Based upon subject content and enlargement size there
are, I imagine, a number of scenarios where ther SD9 will produce
Super A3 prints of such quality that I can use the SD9 instead of
medium format. Through these 'tests' and the production and
examination of prints I'm hoping to establish a few 'rules of
thumb' that will guide me as to when I can use the SD9 and when I
must use MF.

Erik

 

[ChromeLight]

It is not my intention to offend but ultimately it will.
I really have have to laugh about tests made to compare digital to
analogue
In two years of reading no one has made any real tests of substances.
The reason being if you look at any article written you will notice
that the photos are invalid ,you have digital data from the camera
and a scan from a negative or positive and that is called a valid
comparison.
A valid comparison is to print the digital data and the analogue
negative and then scan both prints together.

Centuar,

I actually made prints and provided verbal descriptions of those prints, though I did not scan them. I further informed the readers that I thought it was not a good idea to draw any conclusions based upon viewing crops. I provided the crops because I knew people would want to look at them. I have seen a test on the web where a Mamiya 7 II 20x24 inch print was scanned and compared to a 20x24 inch 1Ds print. There was of course little difference in the prints, though I"m sure the 100% crops would have looked quite different. In my very own comparison I state that many of the things you see in a 100% percent crop are not apparent (or there at all) in the final print, and the final print is what you should evaluate. So I think it's flying in the face of the facts to portray my comparison as relying upon inspection of the crops. In fact, I counsel against that very path.

Fact is Sigma doesn't manufacture a digital lens, there lens are
made for full frame 35 mm ,just think about it the sensor is
exactly 1/3 the size of a 35 mm film. There zoom lens are hopeless
with the Sigma SD9 and only a few F stops are usable. The only lens
that can be used with the Sigma are the fixed lens but with a wider
F stop range but still limited compared to full frame 35 mm if they
were real digital lens like they claim this would not be the case.
Fact- if shooting with the Sigma using 1/15 to 1 sec exposures you
have to use mirror lock up otherwise you will see noticeable
movement, Sigma made a 3rd rate mirror mechanism that damn vibrates.
I'm talking about sharpness and resolution , you can't use the
Sigma hand held and claim the images are sharp if you do you are
plain blind.
Just try it and make a analogue comparison.

I'm not sure why I would try it. As I indicated in the text of my comparison I used a tripod and mirror lockup.

If you are going to compare the Sigma with 645 format then at least
make sure you are photographing static objects and the same crop in
either the width or the length and as close to the same lens factor

I think my crops were close enough. Of course the aspect ratio between the Sigma sensor and 645 was a little different.

You can't compare a 1st generation image with a 2nd generation
image and call that a valid test as I mentioned no one has made a
valid comparison as yet , I ask what has happened to logical
thinking.

Transparancies don't exist in a usable form, unless you are autistic and like to look at shiny things on a light table. So who cares if they are a 1st generation? They're a useless first generation. You can either put your transparancy in a box or destroy it with repeated exposure in a slide projector. To get a print from a transparancy requires the modern workflow of having your transparancy either drum scanned or scanned on your own scanner. In my opinion transparancies are merely a temporary holding mechanism. They come alive once scanned and properly worked on in Photoshop. If you prefer traditional Ciba prints and analogue techniques, fine. I don't. I see nothing wrong with a scan of a transparancy. There's no point in us arguing this. We'll never agree. But I do thank you for your input.

 

[Roland Karlsson]

In two years of reading no one has made any real tests of substances.
The reason being if you look at any article written you will notice
that the photos are invalid ,you have digital data from the camera
and a scan from a negative or positive and that is called a valid
comparison.
A valid comparison is to print the digital data and the analogue
negative and then scan both prints together.

This issue has been discussed zillions of times on rec.photo.digital.
The opinions differs. Som claim what you do and some don't.
The fact is though - what are your plans with the picture? If
you shall print it, it will always be via a digital step, so then the
scanning is 100% valid. Some claims that the absolute best
quality prints from film you get if you drum scan it and then print
it on photographic paper.

Several tests with drum scanning vs direct dicital can be found
in posts in rec.photo.digital. So ... I advice you to do a google
search for those tests and read them. They are perfectly
valid IMHO.

And the verdict?

Depends on what you want. A 11 Mpixel Bayer picture (1Ds) and a
drum scanned 35 mm high quality film is about equal when
it comes to resolution. The direct digital picture is much smoother,
has better color rendition but suffer from aliasing. In the case of
Bayer this aliasing is of the ugly colored kind

Roland

 

[Zone8]

A valid comparison is to print the digital data and the analogue
negative and then scan both prints together.

No it's not. Scanning any print made via an inkjet printer is not a valid way to make any end print due dpi of print and d[pi of scanner. Problems occur if a film scanner is used to scan siler B+W negs - end result is a grainy appearance due interference between dpi of scanner and the silver grains. There is noit equivalent problem using a dye image as the source (e.g. colour neg, XP2-type B+W chromogenic film) The valid test is to print the digital via inkjet and the neg via conventional means. Or even have the digital file output via a laser on to conventional material.

There zoom lens are hopeless with the Sigma SD9 and only a few F stops are usable. The only lens

that can be used with the Sigma are the fixed lens but with a wider
F stop range but still limited compared to full frame 35 mm if they
were real digital lens like they claim this would not be the case.
Fact- if shooting with the Sigma using 1/15 to 1 sec exposures you
have to use mirror lock up otherwise you will see noticeable
movement, Sigma made a 3rd rate mirror mechanism that damn vibrates.
I'm talking about sharpness and resolution , you can't use the
Sigma hand held and claim the images are sharp if you do you are
plain blind.

I use the SD9 with the 24-70 EX lens and results are excellent. My criteria for judging sharpness and quality is, even if I say it myself, somewhat renowned after many years of teaching both advanced amateurs and professionals. I find no problems with mirror movement but I do use a decent Manfrotto tripod. I have also found the design of the SD9 excellent for handholding too. Very simple, in fact and on a par with the Leica M3 I used for years.

The sensor is equivalent to 10.5 mega pixels that is another
fantasy it is equivalent to current 6 mega pixels cameras on the
market.
Do the tests yourself if you get a chance example Fuji S pro2,
Canon EOS-1D(11 mega pixels) and Sigma and blow them up to 50 X 70
cm prints. You will see how the sigma is better in some things than
the Fuji and the canon is better than the both of them.

No way. I chose the SD9 because it was better than anything else on the market as regards end prints (Fuji, Nikon, Canon included) and I am VERY fussy about quality and sharpness matters.

Of course you are stating what YOU have found but like the tests you say are not valid, I would say the same about your findings and statements compared to my own objective tests before I chose the SD9. Touché, mon ami!

Zone8

 

[sg10]

Wow, I wouldn't say that. I think he did a great job and I really
appreciate him doing this test. I don't know of any other test
like this currently. My only fear is that something was wrong or
happened with the SD9 when shooting and it made the results skewed.

It does look like the SD-9 pic is rather blurry. Could be all the SD-9 file manipulation/resampling discussed in the dialog.

Seems to me the right test would be a normally autofocused, standard-output, same-size TIF, printed using the same colorspace at whatever dimension you'd like to compare, with absolutely no Photoshop (or any other program) pixel resampling against a film print which is the same size.

Then I'd suggest doing the same with double-size output just for grins.

 

[Roland Karlsson]

If
you shall print it, it will always be via a digital step, so then the
scanning is 100% valid.

Oops - in Swedish printing (e.g. in a book) or
printing (on photographic paper) are different words.
I mean the printing in a book.

Roland

 

[sg10]

As I said it is not my intention to offend ,I'm a pro photographer
and my boss asked me to test a variety of cameras as we wanted to
get our feet wet and make a start in digital photography.
The criteria was camera and lens quality ,price and work flow time
it was no point spending 10,000 dollars and next year a 22 mega
pixel camera for half the price.
I spent a couple of months testing cameras in the 5 mega pixel to 6
mega pixel range and for the studio I could only recommend the Fuji
S 2 pro
The 6 mega pixel range is suitable for portraits and wedding
photography but is limited in so far as you can't shoot large
groups of people , we still have to use medium format for that as
we don't like lying to our customers like camera manufacture's like
lying to us.
The Fuji was reasonably priced ,

Maybe when the SD-9 was selling for the same the S2 could be considered a competitive buy, but with the SD-9 selling for less than $1000 and the S2 at $1800, I can imagine anyone who doesn't mind not being mainstream considering the S2.

I used SuperCCD3 for my last 6000 shots, I have to say I wasn't impressed by it--very artifact prone. SCCD3 is actually the main reason I bought my SD-9.

a excellent range in high quality
lens unlike the Sigma lenses

The selection isn't great, but the value is better with Sigma. Unless, like I said, you need to share.

and working with the Adode raw plugin
with photoshop saves a lot of time ,a easy workflow and what I see
is what I get as it is is all done in photoshop.
That can't be said with SPP what you see in SPP is not what you get in
photoshop as SPP does not integrate your monitor profile so that
means damn more work and in a studio time is money.

Honestly, I don't find batch processing a big deal.

And if you are interested in powerful, fast RAW workflow, the SD-9 is really king. Especially considering the bandwidth hunger of the S2, combined with the relatively poor image quality of SCCD3.

I bought the Sigma for my private work as I had very specific needs
that the Fuji could not give me and I have the time to play around
with colour correcting.

I have read many articles on digital photography for the past 2 years
and it would be great to read and see a real valid digital vs
analogue test for a change.

Zone,

I agree with everything you wrote. I'm merely conducting these
'tests' out of my own curiosity. I'm not at all attempting to prove
anything. Based upon subject content and enlargement size there
are, I imagine, a number of scenarios where ther SD9 will produce
Super A3 prints of such quality that I can use the SD9 instead of
medium format. Through these 'tests' and the production and
examination of prints I'm hoping to establish a few 'rules of
thumb' that will guide me as to when I can use the SD9 and when I
must use MF.

Erik

 

[DavidMillier]

Please note that Michael has not abandoned the 1Ds because the MF back provides significantly more resolution! He rates them similarly but chose the MF back with its square sensor for the extra flexibility in cropping...

Thanks so much for the comparison. I do have to ask though. Your
SD9 image looks quite bad and not really the same quality I am used
to seeing. The images I'm getting out of the camera are way
sharper. I am also upsampling my images. I'm wondering if you had
something wrong with the lens or camera or if it was just the
compressed image online. It could have even been the upsampling.
There is just almost no detail in that shot which is what the SD9
is known for.

Hi. There is a big difference between fashion, product, portrait
and landscape shots, the former requiring less resolution than the
latter. I chose a landscape shot with an almost pure foliage
content because it is the most challenging for an imaging system.
(Think of the resolution you need to resolve all those pine needles
one hundred yards away!) I think what you're seeing is that even
3.54 true color megapixels is not enough for landscape use for this
particular type of shot. (I expect my tight field of view and
street shots to be much more favorable to the SD9. Also, keep in
mind that alough the crops may look quite different, the difference
is not as apparent in the print. That's why I discourage people
from making evaluations based entirely on lookinig at 100% crops.)
Michael Richeman, of the Luminous Landscape web site, no longer
uses his Canon 1Ds for landscape use. He has now moved up to a
Contax 645 system with a Kodak 16MP DSC ProBack. So even the 1Ds
was found a bit wanting in resolution for landscape use. However,
your concern is noted. This weekend, when I take more comparison
shots, I will attempt to broaden my comparison images.

I would encourage you to download the two large crops. Open both
of them in Photoshop and view at print size. When you do that
you'll see that the two crops don't look that much different.

 

[DavidMillier]

Centaur

Interesting comments about the testing methodologies.

However....

...surely what really matters is not the absolute theoretical performance under optimised conditions but the real world performance under the conditions a photographer is likely to encounter?

I don't have an SD9 (use a D100) so I can't comment on this camera but I do have a 645 and a 5*4 camera and a scanner. The trouble is, big format scanners cost a lot of money and are just not a practical option for a lot of people. I make do with the high value but not perfect Epson 3200 for my scanning. My hope was (is?) that the 645 + 3200 will provide better fine detail than my D100 for 18*12inch landscapes. I'm still testing so I haven't yet decided whether there is an advantage to the film.

It seems to me that that the whole point of scanning a 645 image is to get it into the digital domain for post processing and printing. The idea of chemical printing a 645 image then scanning it then printing it again on an inkjet seems a little silly for anything other than academic testing!

I bought the Sigma for my private work as I had very specific needs
that the Fuji could not give me and I have the time to play around
with colour correcting.

I have read many articles on digital photography for the past 2 years
and it would be great to read and see a real valid digital vs
analogue test for a change.

Zone,

I agree with everything you wrote. I'm merely conducting these
'tests' out of my own curiosity. I'm not at all attempting to prove
anything. Based upon subject content and enlargement size there
are, I imagine, a number of scenarios where ther SD9 will produce
Super A3 prints of such quality that I can use the SD9 instead of
medium format. Through these 'tests' and the production and
examination of prints I'm hoping to establish a few 'rules of
thumb' that will guide me as to when I can use the SD9 and when I
must use MF.

Erik

 

[Roland Karlsson]

A valid comparison is to print the digital data and the analogue
negative and then scan both prints together.

As already have been stated - all do not agree that this is a valid test.

This test do not correspond to any sensible use of your photos.
Why print and then scan and then compare? Sounds rather
weird use of photos IMHO.

You introduce lots of possible errors in this test. Different
prints differs in how easy it is to scan them. In particular,
you may get interference patterns when scanning something
that already has been digitized.

Nope - I think that this test is lest valid than scanning the
negative with the best possible scanning method.

Roland

 

[ChromeLight]

Please note that Michael has not abandoned the 1Ds because the MF
back provides significantly more resolution! He rates them
similarly but chose the MF back with its square sensor for the
extra flexibility in cropping...

He's abandoned the 1Ds when it comes to pure landscape photography because it provides enough extra resolution for cropping, as you said. The point I was trying to make is that the photographic 'arms race' is far from over. So no one should really be at all bothered if a particular digital camera doesn't quite equal its counterpart in the film world. There are still many more years of development yet to come. The 1Ds and DCS Proback aren't the end of the line. It will be interesting to see if Michael compares some of the 20MP backs he will be reviewing to 4x5. That would be a fun test!

 

[centuar]

"And if you are interested in powerful, fast RAW workflow, the SD-9 is really king. Especially considering the bandwidth hunger of the S2, combined with the relatively poor image quality of SCCD3."

What a load of rubbish , SPP is made for amateurs it is not for a pro workflow.

As I mentioned it ignores your monitor profile so all the calibrating you do is for nothing,so what you see is not what you get.

To get closer to what you see in SPP you have to save your files in S-RGB and then use assign profile in Photoshop to change it to Adobe RGB or ColourMatch RGB and that comes from the horses mouth at Foveon.

Just do a test and use say the colour match rgb in SPP and open it in photoshop it is nothing to what you see in SPP

When you save the file in S-RGB and use assign profile to colour match rgb or adobe rgb is is close to what you see in SPP but you still have to tweak it further.
Batch processing is fine for amateur work but not for a pro workflow.

"S2, combined with the relatively poor image quality of SCCD3." Another stupid statement

Not all pixels are the same, not all camera software is the same and not all sensors are the same.

In fact prints made with the S2 pro are sharper than the Sigma , the Foveon sensor is better than the Fuji but Fuji did some very clever manipulating that makes the images seem sharper but that is a clever
optical trick on there part.

We did extensive tests and surveyed our customer's to tell us which was the better print ( 20 X 30 cm samples of identical crops )and 90%

picked the prints made with the Fuji. So in other words the eye can be easily fooled and if our customers say that the Fuji prints are better

then they are right after all they keep us in business and I'm not going to argue with them.

As I said it is not my intention to offend ,I'm a pro photographer
and my boss asked me to test a variety of cameras as we wanted to
get our feet wet and make a start in digital photography.
The criteria was camera and lens quality ,price and work flow time
it was no point spending 10,000 dollars and next year a 22 mega
pixel camera for half the price.
I spent a couple of months testing cameras in the 5 mega pixel to 6
mega pixel range and for the studio I could only recommend the Fuji
S 2 pro
The 6 mega pixel range is suitable for portraits and wedding
photography but is limited in so far as you can't shoot large
groups of people , we still have to use medium format for that as
we don't like lying to our customers like camera manufacture's like
lying to us.
The Fuji was reasonably priced ,

Maybe when the SD-9 was selling for the same the S2 could be
considered a competitive buy, but with the SD-9 selling for less
than $1000 and the S2 at $1800, I can imagine anyone who doesn't
mind not being mainstream considering the S2.

I used SuperCCD3 for my last 6000 shots, I have to say I wasn't
impressed by it--very artifact prone. SCCD3 is actually the main
reason I bought my SD-9.

a excellent range in high quality
lens unlike the Sigma lenses

The selection isn't great, but the value is better with Sigma.
Unless, like I said, you need to share.

and working with the Adode raw plugin
with photoshop saves a lot of time ,a easy workflow and what I see
is what I get as it is is all done in photoshop.
That can't be said with SPP what you see in SPP is not what you get in
photoshop as SPP does not integrate your monitor profile so that
means damn more work and in a studio time is money.

Honestly, I don't find batch processing a big deal.

And if you are interested in powerful, fast RAW workflow, the SD-9
is really king. Especially considering the bandwidth hunger of the
S2, combined with the relatively poor image quality of SCCD3.

I bought the Sigma for my private work as I had very specific needs
that the Fuji could not give me and I have the time to play around
with colour correcting.

I have read many articles on digital photography for the past 2 years
and it would be great to read and see a real valid digital vs
analogue test for a change.

Zone,

I agree with everything you wrote. I'm merely conducting these
'tests' out of my own curiosity. I'm not at all attempting to prove
anything. Based upon subject content and enlargement size there
are, I imagine, a number of scenarios where ther SD9 will produce
Super A3 prints of such quality that I can use the SD9 instead of
medium format. Through these 'tests' and the production and
examination of prints I'm hoping to establish a few 'rules of
thumb' that will guide me as to when I can use the SD9 and when I
must use MF.

Erik

 

[Zone8]

What a load of rubbish , SPP is made for amateurs it is not for a
pro workflow.
As I mentioned it ignores your monitor profile so all the
calibrating you do is for nothing,so what you see is not what you
get.

It is on my two monitors and computers and on my laptop, too.

Sounds as though you are missing something somewhere in your computer/display set-up. I'm a long-established Pro and find PhotoPro an excellent programme that is very simple to use and set-up to give exactly what I require - with no - or very little - adjustment for most images. Even if it did take longer, it would be worth it for the sheer extra quality.

Old Chinese proverb (related to cooking, actually but seemed appropriate!). "Anything worthwhile is worth making a little extra effort for".

As for the Fuji image quality - well, it's OK and clearly satisfies your own requirements but the Sigma SD9 gives ME the sort of professional quality I require - that's why I chose the SD9 above the other offferings.

Each to his own, I guess but your continuing assertion that the Fuji is better is a bit wearying, to say the least. Do try and give others credit for knowing what quality they want and how best to achieve it, please!

Zone8

 

[sg10]

"And if you are interested in powerful, fast RAW workflow, the SD-9
is really king. Especially considering the bandwidth hunger of the
S2, combined with the relatively poor image quality of SCCD3."

What a load of rubbish , SPP is made for amateurs it is not for a
pro workflow.

Ahh, but reducing images to but a small percentage of their original color information in-camera is for pros, right? I don't think so. A 100-JPEG (better than most JPEG-fine options in DSLRs) captures about half the color information of a 24-bit TIF, which also has only a fraction of the 36-bit color captured in the RAW image itself.

As I mentioned it ignores your monitor profile so all the
calibrating you do is for nothing,so what you see is not what you
get.
To get closer to what you see in SPP you have to save your files in
S-RGB and then use assign profile in Photoshop to change it to
Adobe RGB or ColourMatch RGB and that comes from the horses mouth
at Foveon.
Just do a test and use say the colour match rgb in SPP and open it
in photoshop it is nothing to what you see in SPP

I think pro's would understand that RAW has no associated color profile. The color profile is tagged only after the RAW visualization is saved for the first time with such a tag. Obviously a calibrated monitor is perfect for such a process, and if SPP appears on that monitor, that is the best one can ever do with any RAW visualization.

When you save the file in S-RGB and use assign profile to colour
match rgb or adobe rgb is is close to what you see in SPP but you
still have to tweak it further.
Batch processing is fine for amateur work but not for a pro workflow.

"S2, combined with the relatively poor image quality of SCCD3."
Another stupid statement

SCCD3 is simply not competitive IMO. Sorry, but that is the reason I switched to the SD-9, and I am so glad I did. Here is a reasonably well controlled sample, it does a good job of capturing the resolution and artifact deficiencies of a 12MP SCCD image compared to a 3.5MP Foveon image...

http://www.pbase.com/imageprocessing/test

Not all pixels are the same,

Fuji proves that every day, when they develop pixels based on 1/6th the amount of sensor information used in the Foveon standard.

Foveon - 3 color sensors per pixel
Fuji SCCD3 - 2 pixels per color sensor

Relatively poor images result. I have taken about 6000 SCCD3 that prove that. And that is at about 2x the rated MP.