You can't do That with medium format

[Eric Myrvaagnes]

At last this thread is getting useful.
What I want to know from the resident pundits is this: Will an MF chainsaw get me better results than a plain old 35mm (full frame, of course)? I need the info right away as I'm about to advertise my brain surgery specials on Craig's List.

Eric

[Rob C]

At last this thread is getting useful.
What I want to know from the resident pundits is this: Will an MF chainsaw get me better results than a plain old 35mm (full frame, of course)? I need the info right away as I'm about to advertise my brain surgery specials on Craig's List.

Eric

Never forget: advertise your services for frontal lobotomies and somebody here will be perfectly willing to give it a try! What are your charges...?

Reminds me of that funny Western custom: the ballot box.

Just imagine, asking someone who has never personally been in government to vote in a candidate! Is that that funny thing called democracy? It'll never last, juging by what one reads...

Rob C

[eronald]

Fred

 I find the DR tests relevant - sorry - in real life. Take your favorite camera to its limit, then the graph will tell you if another camera will help you more or less. If your camera (and the new one) are not near a limit, and your exposure will be correct, then the test is simply not practically relevant.

Edmund

Oh yes, this DxO...unforgetable truly usable informations for our works indeed. The little Pentax KX has almost the same DR as the 645D and the Leica M9 is somewhere average in their testings...
They have little buttons that when you move then, put some noise to the pic according to the Isos. How exciting!
And what about the demonstration of DR on a masterpeice picture of the Eiffel Tower, so you can see (actually you can see nothing of course) how important is one point of DR.
Can't wait their newsletter to see the last testing. That will help a lot choosing my next gear and make good pictures.
Others that have found a cool way to earn their lives with vaccum. Only a great publicist would have done so good.

[bjanes]

Yup.

The OP has probably penned more words on MFD systems than any other contributor to this forum and all of this without ever having had any meaningful experience of actually using them.

Thanks, indeed.

I have been following this thread with some interest and amusement. I too have medium format experience but only with film and have never seen a MFDB or even a large print from such a camera. In the 40 years that I have been doing photography at the enthusiast level, a larger format gives better image quality at increased cost and with less convenience and portability, and I am sure that this applies to MFDBs.

However the arrogance and non-credible claims of some of the MFDB users grate some of us mere mortals the wrong way and results in some heated discussions. Prices may come down, but the cost of current MFDBs limits their use to top flight professionals and very rich enthusiasts. Many enthusiasts have advanced degrees in the physical sciences or engineering and have a better understanding of technical issues relating to digital photography than many professional photographers who lack a rigorous scientific background and may have received their photographic education during the film era. Therefore, their analysis should not be dismissed out of hand.

Certainly no one with any rigorous scientific knowledge would claim that MFDBs have orders of magnitude more dynamic range than full frame dSLRs. The MFDBs do have larger sensors and can thus collect more photons than a dSLR, but their CCDs also have more read noise than the best CMOS designs used in dSLRs. A stop or possibly even two of DR over a full frame dSLR would be reasonable, but the inflated claims of some can be dismissed as physically impossible without having seen or used such a camera. 

MFDBs do have more pixels than dSLRs, but it remains to be seen if their pixels are any better than those of the top dSLRs, such as the Nikon D3x. The former do lack low pass filters (probably since such filters would be impractical for large sensors) and their images do appear sharper, but suffer from aliasing artifacts. The blurring of low pass filters on dSLRs can largely be overcome by deconvolution sharpening as Diglloyd has shown on his web site.

[Dennis Carbo]

"MFDBs do have more pixels than dSLRs, but it remains to be seen if their pixels are any better than those of the top dSLRs, such as the Nikon D3x"

No.....That is a fact 16 bit, larger 9 micron photosite...etc  They are better at base ISO question is ...do you really care ?  D3X is an awesome tool for the right job, just as a MFDB is DEPENDING ON THE JOB  .... sheesh

[eronald]

bjanes,

 I agree, what we are seeing here in this forum interesting, it's a dialogue between those who understand the structure of the tools, and those who actually use them.

 But in this dialogue, there is one thing that the "pros" forget: Some tools will increase the pleasure of the process of photography.  In other words, it's not only about the amateur owning the camera, or the pro being able to wring pictures from it, sometimes it is about having fun with the camera while you use it, or at the very least its not getting in your way. And about all this DR cr*p which Fredjean dislikes, well it's about getting back to the computer and having a file that works with you, not one which just tears up as soon as you apply a curve.

Edmund

I have been following this thread with some interest and amusement. I too have medium format experience but only with film and have never seen a MFDB or even a large print from such a camera. In the 40 years that I have been doing photography at the enthusiast level, a larger format gives better image quality at increased cost and with less convenience and portability, and I am sure that this applies to MFDBs.

However the arrogance and non-credible claims of some of the MFDB users grate some of us mere mortals the wrong way and results in some heated discussions. Prices may come down, but the cost of current MFDBs limits their use to top flight professionals and very rich enthusiasts. Many enthusiasts have advanced degrees in the physical sciences or engineering and have a better understanding of technical issues relating to digital photography than many professional photographers who lack a rigorous scientific background and may have received their photographic education during the film era. Therefore, their analysis should not be dismissed out of hand.

Certainly no one with any rigorous scientific knowledge would claim that MFDBs have orders of magnitude more dynamic range than full frame dSLRs. The MFDBs do have larger sensors and can thus collect more photons than a dSLR, but their CCDs also have more read noise than the best CMOS designs used in dSLRs. A stop or possibly even two of DR over a full frame dSLR would be reasonable, but the inflated claims of some can be dismissed as physically impossible without having seen or used such a camera. 

MFDBs do have more pixels than dSLRs, but it remains to be seen if their pixels are any better than those of the top dSLRs, such as the Nikon D3x. The former do lack low pass filters (probably since such filters would be impractical for large sensors) and their images do appear sharper, but suffer from aliasing artifacts. The blurring of low pass filters on dSLRs can largely be overcome by deconvolution sharpening as Diglloyd has shown on his web site.

[bjanes]

"MFDBs do have more pixels than dSLRs, but it remains to be seen if their pixels are any better than those of the top dSLRs, such as the Nikon D3x"

No.....That is a fact 16 bit, larger 9 micron photosite...etc  They are better at base ISO question is ...do you really care ?  D3X is an awesome tool for the right job, just as a MFDB is DEPENDING ON THE JOB  .... sheesh

Dennis,

You should check your facts and understand some theory before posting. For example: the Phase 1 P65+, usually considered the top dog in MFDB, has a pixel size of 6 um. The D3x, has a pixel size of 5.9 um. The Phase 1 has a 16 bit ADC, but that does not mean that the sensor has enough DR to make full use of that bit depth. The excess bits serve to quantify noise. See Emil Martinec. The Nikon D3 has 8.4 um pixels, but the D3x bests it in DR. Quantum efficiency, fill factor, and micro-lens technology, and better electronics give the D3x the edge in DR. The D3 has a 14 bit ADC, but that is mainly a marketing ploy, since it does not have 14 stops of DR.

[eronald]

bjanes,

 All the fun of forum posting is that you get readers more qualified than you to fact check for free, kind of like a journal review, but less unpleasant

 That aside, if you can explain this graph (D3, D3s, D3x) please do
http://www.dxomark.com/index.php/en/Camera-Sensor/Compare-sensors/(appareil1)/485|0/(appareil2)/628|0/(appareil3)/438|0/(onglet)/0/(brand)/Nikon/(brand2)/Nikon/(brand3)/Nikon

Edmund

Dennis,

You should check your facts and understand some theory before posting. For example: the Phase 1 P65+, usually considered the top dog in MFDB, has a pixel size of 6 um. The D3x, has a pixel size of 5.9 um. The Phase 1 has a 16 bit ADC, but that does not mean that the sensor has enough DR to make full use of that bit depth. The excess bits serve to quantify noise. See Emil Martinec. The Nikon D3 has 8.4 um pixels, but the D3x bests it in DR. Quantum efficiency, fill factor, and micro-lens technology, and better electronics give the D3x the edge in DR. The D3 has a 14 bit ADC, but that is mainly a marketing ploy, since it does not have 14 stops of DR.

[cyberean]

photography by numbers ... is bestest

[bradleygibson]

For those who don't understand (or want to understand) the science behind the tools they're using, you are welcome to continue using your favorite tools that get the job done in just the way you like.

But please don't try to imply that because a scientific test doesn't have your particular favorite camera come out where you felt it should that the entire idea of objective measurements is somehow useless.  We all know that optics and software processing tools have a huge impact on final image quality.  These DxO tests don't measure these.  That alone can explain many of the perceived discrepancies, so let's not exaggerate the relevance of what they do test.

If you know what you're looking for, this type of objective measurement helps to cut through the marketing BS and can tell you what's really going on--if you're interested.

Edmund--not sure what you are looking for an explanation of with the three Nikons you linked to...

[Joe Behar]

11 pages, 218 posts and almost 6200 views.....

I guess that old guy I met when I was much younger really did have it right when he told me;

"Kid, opinions are like a$$holes...most everybody has one, and they all stink"

[cyberean]

photography by science ... is almost as good

[Dennis Carbo]

Dennis,

You should check your facts and understand some theory before posting. For example: the Phase 1 P65+, usually considered the top dog in MFDB, has a pixel size of 6 um. The D3x, has a pixel size of 5.9 um. The Phase 1 has a 16 bit ADC, but that does not mean that the sensor has enough DR to make full use of that bit depth. The excess bits serve to quantify noise. See Emil Martinec. The Nikon D3 has 8.4 um pixels, but the D3x bests it in DR. Quantum efficiency, fill factor, and micro-lens technology, and better electronics give the D3x the edge in DR. The D3 has a 14 bit ADC, but that is mainly a marketing ploy, since it does not have 14 stops of DR.

Ok Bjanes,

What facts do I need to check ? The back I own has a pixel size of 9 um and is 16 Bit.  I actually use a camera to make a living and it is my sole source of income, you can throw all the numbers and theory at it you want - I find a MFDB file takes many times the abuse in PP that a file from a DSLR does... thats all Im sayin ......I dont care what the literature says, I have actually used both and speak from real world experience. I dont think someone who has never USED a MFDB is qualified to make any meaningful comparison, many of the posters in this thread have never even used one and rely on sales literature to throw out opinions - to me that is just silly.  Not sure if you have used one or not - no disrespect intended

Happy Shooting !  (With what ever you use ! )

 

[bjanes]

 All the fun of forum posting is that you get readers more qualified than you to fact check for free, kind of like a journal review, but less unpleasant

 That aside, if you can explain this graph (D3, D3s, D3x) please do
http://www.dxomark.com/index.php/en/Camera-Sensor/Compare-sensors/(appareil1)/485|0/(appareil2)/628|0/(appareil3)/438|0/(onglet)/0/(brand)/Nikon/(brand2)/Nikon/(brand3)/Nikon

Edmund,

I am not sure what you want explained or if I am more qualified than you (doubtful), but I do have a couple of things to suggest. Emil Martinec has made some interesting observations regarding sensor DR vs camera DR that are worth looking at, and I will add a few comments of my own.

With regard to the DR of the D3 and D3x, it is instructive to look at the DR vs ISO plots published by DXO. The D3x plot is nearly linear, with DR dropping by 1 stop for each doubling of ISO from the base ISO and upwards. However, the D3 demonstrates this linearity only for higher ISOs, and exhibits flattening towards base ISO. This indicates that the D3 has higher read noise at base ISO. This is likely due to inferior electronics (amplifiers and ADC). The D3 is geared for speed and it is difficult to design an ADC with low noise at high readout rates. Also, the D3x is a newer generation sensor. Pixel size has little to do with DR when the the overall sizes of the sensors are the same and one uses the same print or viewing size: the large and small pixel cameras both collect the same number of photons if other things are equal.

As Emil has pointed out, the graphs are important when one is shooting with limited light and exposure is limited by shutter speed and f/stop considerations. With the D3, a higher ISO is desirable because of reduced read noise. When the graph becomes linear, increasing ISO will not improve read noise and it is advisable to increase exposure in the raw converter rather than increasing the ISO on the camera. In this case the histogram on the camera will be to the left, but headroom is preserved. With many MFDBs one can merely shoot at base ISO and increase exposure with the raw converter. Indeed, I understand this is what is done with Hasselblad MFDBs and possibly some others.

[eronald]

Bjanes,

Thanks for your comments. I'm pretty superficial these days, as we have all noticed

I would agree that it's clear that the D3x and D3 higher Iso's are basically software (left shift). I don't think the low ISO flattening is a problem, I think its is probably a design feature. What feature would be interesting to know. Last not least is the case of the D3s; I think here we are seeing the additional effect of some software processing in the High ISO regime.

I would agree that the D3x is newer tech than the D3, if those are unit-cell DR estimates. But with those virtually coinciding lines, one can wonder why the D3x cannot be pushed by software to the same ISO as the original D3 with identical results.

 This is just from a cursory inspection, I'm not very technical these days. A week in the library could change that

Edmund

Edmund,

I am not sure what you want explained or if I am more qualified than you (doubtful), but I do have a couple of things to suggest. Emil Martinec has made some interesting observations regarding sensor DR vs camera DR that are worth looking at, and I will add a few comments of my own.

With regard to the DR of the D3 and D3x, it is instructive to look at the DR vs ISO plots published by DXO. The D3x plot is nearly linear, with DR dropping by 1 stop for each doubling of ISO from the base ISO and upwards. However, the D3 demonstrates this linearity only for higher ISOs, and exhibits flattening towards base ISO. This indicates that the D3 has higher read noise at base ISO. This is likely due to inferior electronics (amplifiers and ADC). The D3 is geared for speed and it is difficult to design an ADC with low noise at high readout rates. Also, the D3x is a newer generation sensor. Pixel size has little to do with DR when the the overall sizes of the sensors are the same and one uses the same print or viewing size: the large and small pixel cameras both collect the same number of photons if other things are equal.

As Emil has pointed out, the graphs are important when one is shooting with limited light and exposure is limited by shutter speed and f/stop considerations. With the D3, a higher ISO is desirable because of reduced read noise. When the graph becomes linear, increasing ISO will not improve read noise and it is advisable to increase exposure in the raw converter rather than increasing the ISO on the camera. In this case the histogram on the camera will be to the left, but headroom is preserved. With many MFDBs one can merely shoot at base ISO and increase exposure with the raw converter. Indeed, I understand this is what is done with Hasselblad MFDBs and possibly some others.

[bjanes]

Ok Bjanes,

What facts do I need to check ? The back I own has a pixel size of 9 um and is 16 Bit.  I actually use a camera to make a living and it is my sole source of income, you can throw all the numbers and theory at it you want - I find a MFDB file takes many times the abuse in PP that a file from a DSLR does... thats all Im sayin ......I dont care what the literature says, I have actually used both and speak from real world experience. I dont think someone who has never USED a MFDB is qualified to make any meaningful comparison, many of the posters in this thread have never even used one and rely on sales literature to throw out opinions - to me that is just silly.  Not sure if you have used one or not - no disrespect intended

If the sensor sizes of two cameras are the same and other things are equal, pixel size has little to do with DR, as I pointed out earlier and has been confirmed by DXO and other experts. 16 bits does not ensure high DR. What DSLRs have you used for your comparisons, the D3x or some obsolete cropped sensor model? As I said, I have not even seen or held a MFDB, but the laws of physics still apply. I am relying on science, not sales brochures or exaggerated and subjective opinions of some.

I'm glad that your camera is meeting your needs for your type of photography. MFDBs do have definite advantages, but let's cut out the hype and overstatements.

Regards,

Bill

[Doug Peterson]

MFDBs do have more pixels than dSLRs, but it remains to be seen if their pixels are any better than those of the top dSLRs, such as the Nikon D3x. The former do lack low pass filters (probably since such filters would be impractical for large sensors) and their images do appear sharper, but suffer from aliasing artifacts. The blurring of low pass filters on dSLRs can largely be overcome by deconvolution sharpening as Diglloyd has shown on his web site.

Just FYI the Mamiya ZD had an optional AA filter - it was not a popular option.

"Largely Overcome" is of course a relative statement. Certainly some of the softening effect caused by the loss of data via low pass filtering can be offset through meticulous post-processing.

[pcunite]

Let me just say that I could not be more pleased, operationally and financially speaking, with my Canons 1D. Hold the camera and it fits like a glove, AF that is fast, put it on a tripod and the LiveView is a pleasure to use, lenses for almost every need, and ISO to suit any situation.

Leaving megapixels out of the discussion, what would make someone like me want to move to MFD? I can answer that. The DOF look, the different color (out of the bag anyway, and different not better from what I've seen), and greater ease of retouching (sharper files at 100%).

What keeps me away? Firstly, no LiveView, secondly cost, thirdly, no customer complaints.

[bradleygibson]

photography by science ... is almost as good

[Audii-Dudii]

11 pages, 218 posts and almost 6200 views.....

I guess that old guy I met when I was much younger really did have it right when he told me;

"Kid, opinions are like a$$holes...most everybody has one, and they all stink"

...and perhaps most importantly, nobody has much use for anybody else's!