Bench Marking Your DSLR

[Loris Medici]

...From a personal perspective I am strongly considering purchase this year or next of a DSLR and welcome every bit of valid information I can get on the subject to guide me on how to best spend my hard earned money...

Sandy, I don't know it you have an arsenal of lenses for Canon or Nikon bodies, but if you don't I'd suggest that you definitely consider Sony Alpha 900 or the next full frame model from them:

http://www.dpreview.com/reviews/sonydslra900/
Features are very good, price is even better...

I have the little brother A700 (along with a 17-50 2.8 Tamron) and I'm very happy with it (especially after the latest firmware upgrade)...

Regards,
Loris.

 

[sanking]

Loris,

Thanks for your comments.

The only thing I have in the way of 35mm film cameras at this time is an older Nikon FG and a couple of zoom lenses so I could go for the Sony just as easily as a Nikon or Canon. I have read some good things about the Alpha 900 but have not had a chance to actually see and hold one.

Sandy King

Sandy, I don't know it you have an arsenal of lenses for Canon or Nikon bodies, but if you don't I'd suggest that you definitely consider Sony Alpha 900 or the next full frame model from them:

http://www.dpreview.com/reviews/sonydslra900/
Features are very good, price is even better...

I have the little brother A700 (along with a 17-50 2.8 Tamron) and I'm very happy with it (especially after the latest firmware upgrade)...

Regards,
Loris.

 

[Loris Medici]

You'll fall in love with the VF.

Loris,

Thanks for your comments.

The only thing I have in the way of 35mm film cameras at this time is an older Nikon FG and a couple of zoom lenses so I could go for the Sony just as easily as a Nikon or Canon. I have read some good things about the Alpha 900 but have not had a chance to actually see and hold one.

Sandy King

 

[jd callow]

I like the the Sony as well. A freind has the 12mp version, my father the last Minolta DSLR (8mp?). Both cameras are very nice and I think the lenses are excellent. I have the Nikon 300d (the 700 wasn't out otherwise...) and really like the Camera, but I bought it because I had lenses to fit. If I were in Sandy's shoes or when I'm ready to buy again I'll check this site and may very well switch.

 

[ann]

interesting as i have people ask all the time about what to buy and basically i really lean toward companies that are know as camera "makers"; Sony not being one of them. Just goes to show.

i will certainly keep this in mind for people who already have Minolta lenses that would fit.

 

[Loris Medici]

Sony's DSLR are based on Minolta design (Sony purchased Konica Minolta's camera business in 2006), which is a pretty "legitimate" maker in my view. And yes, I agree the Sony name may be a little "repellent" to people who don't know this fact. On the other hand the merging of Sony's digital (sensor etc) know-how and Minolta's camera/lens making know-how is a pretty nice concept. In the end, both A700 and A900 are fine cameras...

Regards,
Loris.

interesting as i have people ask all the time about what to buy and basically i really lean toward companies that are know as camera "makers"; Sony not being one of them. Just goes to show.

i will certainly keep this in mind for people who already have Minolta lenses that would fit.

 

[ann]

again, this is a great bit of information and without this thread i would have been clueless.

just an old dog learning new tricks

 

[sanking]

The 24 mp Sony A900 costs more than I was expecting to spend. Looks like it is about $3000 USD for body alone so you would need to about double that for a two-lens outfit.

Right now I could pick up a 12.8 mp Canon EOS 5d with, which is full sensor, with a couple of Canon lenses for less than $1500.

Question is, since I shoot MF primarily for B&W prints, could any full sensor DSLR compete with my Mamiya 7II outfit? I have a Canon G9 so am well aware of the creative possibilities of digital with color, but I think I would be trading a lot of detail with digital.

On the other hand, perhaps my question should be can the full sensor DSLR compete in quality with my Fuji GA645zi (6X4.5 cm) since this is what I would probably replace.

These questions are closer for me than for others since I am able to scan MF at 5080 spi with drum scanner quality.

Sandy

 

[ann]

"These questions are closer for me than for others since I am able to scan MF at 5080 spi with drum scanner quality."


smartie pants she says with envy.

 

[Marco B]

Ray and Marko,

If you bother to take the time to read through the dxomark.com Q&A section all of these questions are addressed.

The technology developed by DxO Optics is globally accepted as the industry leader in measuring image quality. That technology is licensed and used by companies and organizations such as Kodak, Olympus, and NASA. If their findings aren't adequate or acceptable for your own use and questions then I suppose no information source will be.

Don Bryant

Don, I am not questioning the sense of it all as regards to the parts of the test setup that I have seen and read about. I do think that much of that makes perfect sense and that DxO is doing a much better job than many others. However, that same reason makes it all the more baffling to me that they DO NOT specify clearly what lenses they are using, or, if I understand the Q&A section correctly, that in fact they do not use lenses at all (Am I right here, Don?) If you spend so much time on standardization and document so much, how could you forget to clearly include it, or state clearly you don't use lenses / optics at all... :confused: . I just don't get it...

But apart from that huge oversight, the website does make some sense.

Anyway, the FAQ is also terribly difficult to find, after going through a dozen or so pages in the "Technologies" pages section, where I would have expected to find it, or even better, straight in the main menu next to the other main topics like "Technologies" / "News", i finally discovered a Q&A in the "News" section... of all places!

Well, I think they are better in testing than in webdesign

 

[Marco B]

On the other hand, perhaps my question should be can the full sensor DSLR compete in quality with my Fuji GA645zi (6X4.5 cm) since this is what I would probably replace.
Sandy

Hi Sandy,

Well, I think it comes awfully close, but that is also highly depended on the film you're using...

After the release of the new Ektar 100 film by Kodak, and picking up a roll at the Photokina, I decided to try to test it along side Velvia 100, Tmax 100, Porta 160VC and... a Sony Alpha 900!

Although I am not a professional tester, so don't expect DxO level of testing facilities , I have given it my utter best using an Imacon Flextight scanner for scanning the film, and shooting a test target poster picked up at the Photokina as well.

You can read more about it here, and download the full sized files for all film types and the Sony Alpha 900... and yes, that last one looks impressive, but so does at least one of the films.

So here it is (Ektar 100 link for you to follow):

http://www.boeringa.demon.nl/menu_technic.htm

Marco

 

[tom_micklin]

Question is, since I shoot MF primarily for B&W prints, could any full sensor DSLR compete with my Mamiya 7II outfit? I have a Canon G9 so am well aware of the creative possibilities of digital with color, but I think I would be trading a lot of detail with digital.

On the other hand, perhaps my question should be can the full sensor DSLR compete in quality with my Fuji GA645zi (6X4.5 cm) since this is what I would probably replace.

These questions are closer for me than for others since I am able to scan MF at 5080 spi with drum scanner quality.

Sandy

Hi Sandy,
I went from the Mamiya 7 to the Canon 5D and am very satisfied with the results (I only shoot B&W).
However, My scans were done on my Microtek i800, a capable scanner, but nowhere near the quality of a drum scanner.
My opinion, for what it's worth, is that you would need to go to the Canon 5D MarkII or the Sony (both over 20mp and full frame) to really be satisfied.
I've done no testing, just an opinion.
Regards,
Tom

 

[Loris Medici]

The 24 mp Sony A900 costs more than I was expecting to spend. Looks like it is about $3000 USD for body alone so you would need to about double that for a two-lens outfit.

Right now I could pick up a 12.8 mp Canon EOS 5d with, which is full sensor, with a couple of Canon lenses for less than $1500.

Sandy, there's the new 5D II (out in December) also to consider. To match MF quality (in terms of detail) I think you need something higher than 12Mp...

Regards,
Loris.

 

[Marco B]

Again guys, for some impression of film versus the Sony Alpha 900, see here:

http://www.boeringa.demon.nl/menu_technic.htm

Follow the Ektar 100 link, it will show you Alpha 900 images as well. Should be interesting for all of you...

Marco

 

[Loris Medici]

...My opinion, for what it's worth, is that you would need to go to the Canon 5D MarkII or the Sony (both over 20mp and full frame) to really be satisfied.
I've done no testing, just an opinion.

Tom sorry, I haven't noticed that you already mentioned 5D MkII. I agree, what I've read about A900 says that it outresolves many lenses; a crude calculation will reveal that you need a lens with resolving power of ca. 80lp/mm to match the resolution of A900's sensor -> that's a pretty serious figure, especially in the MF lens arena (where resolution tolerances are more relaxed due to less magnification ratio), also when considering Sandy's style too (landscapes, architectural w/ great DoF = high diaphragm figures = diffraction limited...). Maybe I'm wrong here -> just telling what I feel...

Regards,
Loris.

 

[Marco B]

I agree, what I've read about A900 says that it outresolves many lenses; a crude calculation will reveal that you need a lens with resolving power of ca. 80lp/mm to match the resolution of A900's sensor -> that's a pretty serious figure

Loris, you're right about the 80 lp/mm, but I don't think this is a major challenge or issue with most top quality lenses. The Alpha 900 certainly did not "outresolve" my Minolta F1.7 50mm lens, it was just a very fine match. If it started outresolving anything, it was actually the 60x90 cm large test chart, as it was capable of resolving almost all detail in that chart. A major feat of the Alpha 900!

 

[Loris Medici]

Loris, you're right about the 80 lp/mm, but I don't think this is a major challenge or issue with most top quality lenses. The Alpha 900 certainly did not "outresolve" my Minolta F1.7 50mm lens, it was just a very fine match. If it started outresolving anything, it was actually the 60x90 cm large test chart, as it was capable of resolving almost all detail in that chart. A major feat of the Alpha 900!

But you tested w/ JPEG -> RAW performance and JPEG are apples and oranges and you can bet I won't use JPEG with such a camera unless I'm shooting family snapshots and such Plus, have you turned off NR? (Probably couldn't, because A900 does noise reduction on JPEGS whereas it's optional / can be turned off for RAWs - or was that in case of A700? I'm not sure anymore...) NR does very nasty things in terms of "detail".

BTW nice page / effort...

Regards,
Loris.

 

[Marco B]

But you tested w/ JPEG -> RAW performance and JPEG are apples and oranges and you can bet I won't use JPEG with such a camera unless I'm shooting family snapshots and such Plus, have you turned off NR? (Probably couldn't, because A900 does noise reduction on JPEGS whereas it's optional / can be turned off for RAWs - or was that in case of A700? I'm not sure anymore...) NR does very nasty things in terms of "detail".

BTW nice page / effort...

Regards,
Loris.

Eeeeh Loris, did you actually take the time to have a look at the images??? I think the XFine JPEG images just look fantastic. How much more do you want? You can even see the fine paper structure in the grey area's of the 60x90 cm test chart, all the details in the cobbles... ... that's just unbelievable!

And as much as I read about NR and the Alpha 900 (which isn't that much I admit), it shouldn't kick in at 100 ISO, that would be ridiculous. But truth to be told, I didn't go through the entire manual to find out how to switch it of. Anyway, I still don't see the point, if the images look just fine in XFine JPEG setting...

If even this can't satisfy you, than nothing will!

 

[Loris Medici]

Marco, the point was "if you want to reveal the true possibilities of a digital camera you have to shoot RAW", JPEGs are too much manipulated in-camera, to satisfy average Joe. Maybe you'd see more quality difference between Velvia and A900 if you had done it shooting RAW (that's why I wrote that; you say Velvia is very close, I say there could be some more difference when shooting RAW and doing fine post-processing)... To me even the A700 is better than 35mm ISO 400 B&W film using the same full frame lenses BTW.

Regards,
Loris.

 

[sanking]

Marco,

I looked at the comparisons. Very interesting. I have one comment and a question.

1. Comment -- As you already anticipated, I would quibble a bit with the choice of the Imacon to scan the film. A drum scanner, or even a high end flatbed that allows fluid mounting, would have subdued the grain in the film scans quite a bit. The Imacon is a fine scanner but tends to emphasize grain more than would be the case with a drum. I am assuming that you did not fluid mount with the Imacon? I know some people who do this, but it is not common.

2.Question -- Perhaps I missed this in the explanation, but exactly how did you set up the target with reference to the field view? Did you set up the two cameras so as to fill the frame with the target, and if so, what focal length did you use with the A900 to match the field with the lens on the 35mm camera?

Sandy

Hi Sandy,

Well, I think it comes awfully close, but that is also highly depended on the film you're using...

After the release of the new Ektar 100 film by Kodak, and picking up a roll at the Photokina, I decided to try to test it along side Velvia 100, Tmax 100, Porta 160VC and... a Sony Alpha 900!

Although I am not a professional tester, so don't expect DxO level of testing facilities , I have given it my utter best using an Imacon Flextight scanner for scanning the film, and shooting a test target poster picked up at the Photokina as well.

You can read more about it here, and download the full sized files for all film types and the Sony Alpha 900... and yes, that last one looks impressive, but so does at least one of the films.

So here it is (Ektar 100 link for you to follow):

http://www.boeringa.demon.nl/menu_technic.htm

Marco

 

[pschwart]

The 24 mp Sony A900 costs more than I was expecting to spend. Looks like it is about $3000 USD for body alone so you would need to about double that for a two-lens outfit.

Right now I could pick up a 12.8 mp Canon EOS 5d with, which is full sensor, with a couple of Canon lenses for less than $1500.

Question is, since I shoot MF primarily for B&W prints, could any full sensor DSLR compete with my Mamiya 7II outfit? I have a Canon G9 so am well aware of the creative possibilities of digital with color, but I think I would be trading a lot of detail with digital.

On the other hand, perhaps my question should be can the full sensor DSLR compete in quality with my Fuji GA645zi (6X4.5 cm) since this is what I would probably replace.

These questions are closer for me than for others since I am able to scan MF at 5080 spi with drum scanner quality.

Sandy

Hi Sandy. I think you should get the full sensor Canon and sell me your Mamiya 7 really cheap
Seriously, I think of the digital SLR as another tool in the arsenal, but I just can't give up film. I am still using Rollei TLRs and I just acquired a Fuji 6x9 rangefinder. Print size and printing process are also determining factors. I make smaller prints (< 8x10) so I can get sufficient quality for both palladium and carbon transfer shooting digital originals or shooting film and scanning.

 

[donbga]

Don, I am not questioning the sense of it all as regards to the parts of the test setup that I have seen and read about. I do think that much of that makes perfect sense and that DxO is doing a much better job than many others. However, that same reason makes it all the more baffling to me that they DO NOT specify clearly what lenses they are using, or, if I understand the Q&A section correctly, that in fact they do not use lenses at all (Am I right here, Don?) If you spend so much time on standardization and document so much, how could you forget to clearly include it, or state clearly you don't use lenses / optics at all... :confused: . I just don't get it...

But apart from that huge oversight, the website does make some sense.

Anyway, the FAQ is also terribly difficult to find, after going through a dozen or so pages in the "Technologies" pages section, where I would have expected to find it, or even better, straight in the main menu next to the other main topics like "Technologies" / "News", i finally discovered a Q&A in the "News" section... of all places!

Well, I think they are better in testing than in webdesign

Marko,

First of all the dxomark website is labeled as beta so I would assume that might mean that changes will be made in future versions of their website.

I do agree that finding information isn't totally convenient but I didn't find it that difficult either. Any complaints you have about their website should be directed to DXO Optics.

In my original post I simply posted the link to their site without visiting every page or reading every word there. However, I did read enough to know that that their tests were based on metrics and evaluation of the RAW output from a cameras sensor.

Anyway the site has so much information that I found it pointless for me to parrot what could be found there, preferring that people take the time to read about the testing protocols and the data obtained from their tests drawing whatever conclusions they thought relevant.

In short I have confidence in their testing protocol and I'm glad such a site was created as I find sites such as dpreview.com sorely lacking in one way or another with their evaluations of cameras.

Don

 

[donbga]

Hi Sandy,

Well, I think it comes awfully close, but that is also highly depended on the film you're using...

After the release of the new Ektar 100 film by Kodak, and picking up a roll at the Photokina, I decided to try to test it along side Velvia 100, Tmax 100, Porta 160VC and... a Sony Alpha 900!

Although I am not a professional tester, so don't expect DxO level of testing facilities , I have given it my utter best using an Imacon Flextight scanner for scanning the film, and shooting a test target poster picked up at the Photokina as well.

You can read more about it here, and download the full sized files for all film types and the Sony Alpha 900... and yes, that last one looks impressive, but so does at least one of the films.

So here it is (Ektar 100 link for you to follow):

http://www.boeringa.demon.nl/menu_technic.htm

Marco

Marko,

In regard to Kodak Ektar 100, Kodak never advertised that Ektar 100 is the world's finest grained color film but rather the world's finest grained color negative film.

Don

 

[donbga]

Eeeeh Loris, did you actually take the time to have a look at the images??? I think the XFine JPEG images just look fantastic. How much more do you want? You can even see the fine paper structure in the grey area's of the 60x90 cm test chart, all the details in the cobbles... ... that's just unbelievable!

And as much as I read about NR and the Alpha 900 (which isn't that much I admit), it shouldn't kick in at 100 ISO, that would be ridiculous. But truth to be told, I didn't go through the entire manual to find out how to switch it of. Anyway, I still don't see the point, if the images look just fine in XFine JPEG setting...

If even this can't satisfy you, than nothing will!

I agree with Loris. Comparing JPG output is not an objective comparison. WIth JPEGS there is too much in camera processing at play regardless of the selected ISO.

I did think your presentation nice but inconclusive.

Don

 

[sanking]

BTW, I am getting a loaner Canon EOS 5d and will be able to make some direct comparisons of 12.8 mp with the Fuji GA645zi 6X4.5 cm negatives.

As I indicated earlier, this is the comparison that is appropriate for me as I am fairly certain that a 12.8 mp camera, perhaps not even a 22mp or 24mp one, , is going to beat Mamiya 7II in print size over 16X20, especially not for raw detail.

Sandy King