Pages: 1 2 3 [4] 5 6 ... 10   Go Down

Author Topic: phase versus hassleblad  (Read 47318 times)

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
phase versus hassleblad
« Reply #60 on: March 07, 2010, 02:43:58 pm »

Hi,

My view is that DxO does unbiased and scientifically sound work. I certainly feel they deserve a lot of respect for publishing their results, no one else is doing a similar effort. So I'm much against DxO bashing.

The way I see it sometimes the DxO numbers agree with our expectations and than we say that they are great, sometimes they don't fit our expectations than we say they are not so great.

What folks need to realize that photography is about perception and perception is not really the same as physics. There is always a possibility that measurements don't measure the right parameters the right way, or that they are not easy to interpret. There is always a temptation to condense all findings into a single "figure of merit", in my view it's simply a stupid approach.

My view on this discussion may be:

- Some of Mark Dubovoy's observations regarding DR need some elaboration or evidence. They certainly seem to contradict basic science. I may be wrong on this but would really like to see some explanation, proof or evidence.
- It's not possible to reproduce a DR in excess of 7 to eight step in a print without manipulating tonality. So if we discuss prints, we also discuss processing.

Best regards
Erik





Quote from: fredjeang
Jeff,
I really like your landscape photographs and I think you are a knowledgable photographer. That is why your post surprises me. How do you give that much credit to DxO and not the same credit to photographers who works daily with MFD ?  
But then I also see that you work with D3x, so is it that you are really serious in your post or in a way you defend your gear?
Because then, how can you explains the enormous amount of top photographers that are working in fashion, landscape, arquitecture and fine arts who only work with MFD or LF ?
I'd like to see a real world comparaison between the Nikon and some MFD, with texture clothes, complex patterns, difficult light etc...to see if this DxO is trustable in the real world. Numbers are one thing, and DxO is NOT absolute mathematics calculations but just a way of testing, reality is another story.
In my understanding of course and with all my respect.

Cheers.

Fred.
Logged
Erik Kaffehr
 

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
phase versus hassleblad
« Reply #61 on: March 07, 2010, 02:58:41 pm »

Quote from: Ray
Mark, Mark, Mark,

You seldom print larger than 13"x19" yet you've invested in an MFDB system??? What are we to think?

Are you the ultimate pixel-peeper?  

Here's what Mark Dubovoy wrote about pixel peepers, to quote:



Now to contrast this statement, I would like to describe my own experiences as regards pixel-peeping.

As one gets older, one tends to become more long-sighted. That is, reading requires spectacles, but the full moon on a clear night might still appear detailed.

So it was with me. About 16 years ago, reading became a bit of a strain, and the optician prescribed spectacles with a +1x magnification.

16 years later, those same spectacles required for ease of reading are now perfect for long distances, such as trees on the horizon or the moon on a clear night.

I don't need the spectacles to get around, drive a car etc. If I wasn't into Photography and wasn't concerned about resolution, I just wouldn't bother. But I know through experimentation, if I want to appreciate the maximum detail of a distant scene, I need spectacles with a +1x magnification.

If I want to pixel-peep two images on my monitor at 100%, or 200% or even 400%, I need spectacles with a +2.5 to +3x magnification.

If I want to see the maximum detail in a high quality HD source on my 65" plasma TV, I need spectacles of +1.25x magnification, and I need to sit no further than 2.4 metres (or 7ft 10in) from the screen.

I mention this just to demonstrate that I have no reason to believe that my eyesight is deficient when using the appropriate specatcles.

If one examines Mark Dubovoy's statement about pixel peeping more carefully, he seems to be implying that the print from the MFDB file is so significantly better than any crop at any enlargement viewed on the monitor, that pixel peeping is not relevant, partly because an inkjet printer has higher resolution than a monitor, and also because some printers have a slightly wider color gamut than even a good monitor.

I don't find these reasons at all credible in light of my own experiences. If the print is better than the view on the monitor (in terms of resolution at the appropriate magnification on the monitor), or in terms of color gamut which the monitor cannot display, then such differences are of a pixel-peeping magnitude.

Furthermore, Mark Segal, any 'additional' qualities you see in your 13"x19" print from an MFDB file (as opposed to a 35mm file), viewed from the distance you would read a book, will disappear from even a slightly greater distance, like 600mm, never mind 2.4 metres.

Ray,

While I could have much to tell you, for the moment I won't comment further on this post, except to make one point and to ask you a question. The point is that I'm not a wealthy person who can afford to spend this kind of money frivolously and I wasn't either an idiot or on drugs when I bought that system. The question is: have you personally had a recent opportunity to compare similar images, having high scene DR, as rendered by a high-end DSLR and a Phase-1 P40+ or P65+ back?

Mark
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
phase versus hassleblad
« Reply #62 on: March 07, 2010, 02:58:52 pm »

Hi,

I do admit, my statement was meant to be a bit provocative!

What I mean is essentially that a print has different qualities. Tonality is something can be seen at long distance. You don't need to put the print under your noose to see that it has dense black and bright highlights. These to parameter actually define the DR of a print. The next set of parameters are tonal differentiations. None of these parameters is really dependent on viewing distance, if you have normal vision (except cataract).

Closer up and with good vision fine detail contrast, noise and other factors show up.

So, it's perfectly possible to tell apart a print with deficiency in DR (defined as black to white ratio) at long distance. Other deficiencies are only visible at short distance.

Best regards
Erik

 

Quote from: siba
And Erik, you have made the same sort of sweeping comment that has made me come in once again. According to you a "blind man" can tell the difference between two small prints at thirty feet. That the issue is just being able to discern between black and white in the two prints.
Logged
Erik Kaffehr
 

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
phase versus hassleblad
« Reply #63 on: March 07, 2010, 03:02:54 pm »

Quote from: BernardLanguillier
Hum... Mark, have you have heard the equivalent speech done by Sony/Nikon engineers? I believe that they also describe themselves as being the best in the world at what they do. In the end these are just claims.

Figures do tell the whole story, especially when they are measured on an actual device just like DR can be. We are not talking about doggy predictions resulting from a potentially flawed theoretical model, we are talking about hard facts here.

Cheers,
Bernard

I wasn't talking about "claims", I was talking about technical explanations. And no, the kind of figures being bandied around in this thread definitely DO NOT tell the whole story. And facts are no harder then the definitions and methodologies used to establish them, which in this case obviously vary.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

JeffKohn

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1668
    • http://jeffk-photo.typepad.com
phase versus hassleblad
« Reply #64 on: March 07, 2010, 03:03:11 pm »

Quote
What folks need to realize that photography is about perception and perception is not really the same as physics. There is always a possibility that measurements don't measure the right parameters the right way, or that they are not easy to interpret. There is always a temptation to condense all findings into a single "figure of merit", in my view it's simply a stupid approach.
I can agree with that. In the case of DxOMark, the overall "Sensor Mark" score strikes me as pretty silly, and of very little use. However drilling down into the individual test results can be useful, if you understand what they're testing.

I think the DR advantage of MFDB's over DSLR's was considerably larger in the past. And it's also a fairly easy concept to understand; if the difference is large enough it will be readily apparent in images, and of course there's the whole 16-bit versus 12-bit (whether the MF backs were truly using a full 16-bits is debatable, but that's the sort of technical detail that's easily ignored).

So the "DR advantage" of MF became part of the conventional wisdom, and was oft-repeated when the merits of MF came up in discussion/debate. But technology doesn't stand still, and the small-format CMOS sensors have improved more than their FF-CCD counterparts, so the gap has closed with regard to DR. Some people don't recognize this for whatever reason, and still bring out the "DR advantage" as if it were one of the main reasons to prefer MF; in some cases even suggesting that 2- and 3-generation old MF sensors are still vastly superior in DR to today's latest DSLR's.
« Last Edit: March 07, 2010, 03:34:48 pm by JeffKohn »
Logged
Jeff Kohn
[url=http://ww

JeffKohn

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1668
    • http://jeffk-photo.typepad.com
phase versus hassleblad
« Reply #65 on: March 07, 2010, 03:09:12 pm »

Quote from: ErikKaffehr
What I mean is essentially that a print has different qualities. Tonality is something can be seen at long distance. You don't need to put the print under your noose to see that it has dense black and bright highlights. These to parameter actually define the DR of a print. The next set of parameters are tonal differentiations. None of these parameters is really dependent on viewing distance, if you have normal vision (except cataract).
Differentiating pure black and paper white is not hard from a distance, I agree. Subtle tonal differentiations? I'm not so sure about that one, you have to be close enough to distinguish the regions the tones are distributed over. I think really subtle tonality would require closer examination than can be achieved from across a room.


Quote
So, it's perfectly possible to tell apart a print with deficiency in DR (defined as black to white ratio) at long distance. Other deficiencies are only visible at short distance.
All well in good if you're comparing the white and black points of various papers; but not so much if you're interested in the DR of cameras.
Logged
Jeff Kohn
[url=http://ww

Mosccol

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 88
phase versus hassleblad
« Reply #66 on: March 07, 2010, 03:10:27 pm »

Quote from: michael
Please see my Editor's Note just added to Mark's review. It will hopefully shed some light on the DR question.

Michael

Thanks Michael for this agnostic update. The next question of course is "so what of the human eye?"

If I understand correctly, although humans have a theoretical DR of about 24 stops accounting for pupil size adjustment etc. our instantaneous DR is about 10-14 stops (I picked this page for reference, but there seems to be many others). If this is correct, then it means that a correctly exposed MFB photograph will match the human eye at the time a photograph is taken (10-14 stops v. 12-13 stops).

Now as an impecunious amateur, I can only dream of ever owning a $20k MFB, so can you pro users out there tell me: how far from 'the real thing' is the range of MFB? - particularly for outdoor photography.

Cheers

François
Logged
Canon 7 & L Glass + iPhone 6 ... Where n

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
phase versus hassleblad
« Reply #67 on: March 07, 2010, 03:13:09 pm »

Quote from: BernardLanguillier
Unless you have done a thorough measurement of the DR of your back vs that of high end DSLRs it would appear that you have believed the numbers proposed by the vendor of your back at least in absolute terms if not in relative terms compared to your previous camera. If not you may want to either present your test results, or phrase your position as an impression.


Cheers,
Bernard

Bernard,

Between DxO and Phase-1 Inc., there is a difference of numbers amounting to 0.7 of DR. One says it is 13, the other says it is 13.7. Are they measuring the same way under the same conditions, I have no idea. But the DR is way up there. As for the DSLRs, two different approaches to measurement produce very different numbers. Is this unusual?  - not in the least. So we need to be careful about putting blind faith in numbers. I have not measured the DR using any of these approaches from either my Canon 1Ds3, or my Phase P40+. I have photographed the same scenes with both cameras under as near identical conditions as I could, and I can tell which has higher DR, better shadow detail and better resolution straight out of the box. You'll have to take my word for it - no sense dumbing it all down into JPEGs. As I told Ray, I don't THINK I was drugged when I bought the system. But I still like my Canon 1Ds3 a lot and I still use it a lot. It makes great images, as well it should for what it cost, but it isn't a Phase-1 P40+, thats all.

Mark
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Graeme Nattress

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
    • http://www.nattress.com
phase versus hassleblad
« Reply #68 on: March 07, 2010, 03:14:32 pm »

I have not seen any evidence to suggest that a medium format back has twice as many stops DR than a top end 35mm DSLR. When it comes to DR on sensors - 1 stop is a pretty good difference - 6 or 7 is absurdly different. It's not like there's this magic sensor different for MF that only MF manufactures have and the guys that make 35mm sensors don't. Point the cameras at a decent chart, and I have an 18 stop backlit chart sitting right next to me now and let's see it.

Graeme
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
phase versus hassleblad
« Reply #69 on: March 07, 2010, 03:32:10 pm »

Quote from: Graeme Nattress
I have not seen any evidence to suggest that a medium format back has twice as many stops DR than a top end 35mm DSLR. When it comes to DR on sensors - 1 stop is a pretty good difference - 6 or 7 is absurdly different. It's not like there's this magic sensor different for MF that only MF manufactures have and the guys that make 35mm sensors don't. Point the cameras at a decent chart, and I have an 18 stop backlit chart sitting right next to me now and let's see it.

Graeme

Graeme, this is exactly what makes me think that the ways in which DR is defined and measured differ between various organizations. Perhaps there is one right way of doing it, I don't pretend to know, but it seems pretty clear that what's going on here is most likely a comparison between apples and oranges. I like your idea of pointing a high-end DSLR and a Phase back at your chart and measuring DR using exactly the same method in exactly the same conditions. As Jeff Kohn said, it is the relative results which mattes, not the absolute, and that would probably settle it in a relative sense. If you don't mind, should I have an opportuity to get over to Ottawa, I shall bring the gear, look you up and we can do it. I'd love to see the outcome, whatever it reveals.

Mark
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

PierreVandevenne

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 512
    • http://www.datarescue.com/life
phase versus hassleblad
« Reply #70 on: March 07, 2010, 03:41:36 pm »

Quote from: ErikKaffehr
My view is that DxO does unbiased and scientifically sound work. I certainly feel they deserve a lot of respect for publishing their results, no one else is doing a similar effort. So I'm much against DxO bashing.

Yes, indeed.

Whenever I have been able to compare DxO numbers and the results of calculations based on CCD/CMOS sensors data sheets, I've found them to be in good agreement. It is certainly possible to design a system poorly and worsen the noise part of the equation. But, on the other side, it is impossible to improve on the best result a sensor can deliver. There's nothing magical about larger sensors well depth (especially if they have a pitch that"s similar to their smaller siblings) and even if the cabling/amplifiers/DACs used in MFDB had the "magical" properties of very high end audio gear, they couldn't improve results by 6 stops.  That's very obvious when those sensors are used for science and SNR or DR are really supposed to have a standard meaning. Let's not forget that improving things by a factor of "1 stop" is roughly equivalent to doubling the performance of the system, "6 stops" improve the performance of a system by a factor of 64... That's _really_ a lot.




Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
phase versus hassleblad
« Reply #71 on: March 07, 2010, 03:52:42 pm »

Mark,

I really appreciate that approach!

Best regards
Erik

Quote from: Mark D Segal
...
If you don't mind, should I have an opportuity to get over to Ottawa, I shall bring the gear, look you up and we can do it. I'd love to see the outcome, whatever it reveals.

Mark
Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
phase versus hassleblad
« Reply #72 on: March 07, 2010, 03:57:12 pm »

Hi,

The original statement was by Mark Dubovoy, claiming that the differences were visible at 30 feet on a small print. At that distance tonal differences would be visible on a large scale. The original statement was about DR on paper and a small print.

Thanks for your thoughtful comments!

Best regards
Erik


Quote from: JeffKohn
Differentiating pure black and paper white is not hard from a distance, I agree. Subtle tonal differentiations? I'm not so sure about that one, you have to be close enough to distinguish the regions the tones are distributed over. I think really subtle tonality would require closer examination than can be achieved from across a room.


All well in good if you're comparing the white and black points of various papers; but not so much if you're interested in the DR of cameras.
Logged
Erik Kaffehr
 

Alan Goldhammer

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 4344
    • A Goldhammer Photography
phase versus hassleblad
« Reply #73 on: March 07, 2010, 04:32:36 pm »

I just did part of the experiment.  I have an 8x10 silver gelatin Ansel Adams Merced River, Autumn hanging on my wall (not an original print by Adams but by Alan Ross from the original 8x10 negative  (see: this link).  I also have a very good print from the Ansel Adams calendar.  At 30 feet, they pretty much look alike.    Back in the old film days when I was learning the Zone system, we were instructed to do the simple experiment of taking a picture of a uniformly colored surface with some texture, exposing it for all the Zones.  The negatives were developed all under the same conditions and prints were made using exactly the same enlarger exposure and development.  You end up with a series of prints of all Zones and the texture disappears at some point before you get to absolute white and black.  It's of course dependent on the film of choice.  I think Adams goes over all this in "The Negative" which unfortunately is over at a friend's house.  From what DxO, Norman Koren and others have tested, digital SLRs have a greater dynamic range than some kinds of film but do have to be treated differently (see: Koren's good explanation)
Logged

fredjeang

  • Guest
phase versus hassleblad
« Reply #74 on: March 07, 2010, 04:45:04 pm »

Quote from: Mark D Segal
Graeme, this is exactly what makes me think that the ways in which DR is defined and measured differ between various organizations. Perhaps there is one right way of doing it, I don't pretend to know, but it seems pretty clear that what's going on here is most likely a comparison between apples and oranges. I like your idea of pointing a high-end DSLR and a Phase back at your chart and measuring DR using exactly the same method in exactly the same conditions. As Jeff Kohn said, it is the relative results which mattes, not the absolute, and that would probably settle it in a relative sense. If you don't mind, should I have an opportuity to get over to Ottawa, I shall bring the gear, look you up and we can do it. I'd love to see the outcome, whatever it reveals.

Mark
That's great.
Hope you guys will do it, and of course published the results here  

Regards,

Fred.
Logged

Graeme Nattress

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
    • http://www.nattress.com
phase versus hassleblad
« Reply #75 on: March 07, 2010, 05:09:19 pm »

Quote from: Mark D Segal
Graeme, this is exactly what makes me think that the ways in which DR is defined and measured differ between various organizations. Perhaps there is one right way of doing it, I don't pretend to know, but it seems pretty clear that what's going on here is most likely a comparison between apples and oranges. I like your idea of pointing a high-end DSLR and a Phase back at your chart and measuring DR using exactly the same method in exactly the same conditions. As Jeff Kohn said, it is the relative results which mattes, not the absolute, and that would probably settle it in a relative sense. If you don't mind, should I have an opportuity to get over to Ottawa, I shall bring the gear, look you up and we can do it. I'd love to see the outcome, whatever it reveals.

Mark


At least if the images are posted and the raws made available, people with different methodologies of DR measurement can take that data and do to it whatever they want. Those who work visually can push and pull the files in Photoshop, those who put numbers to it can do so, but with the cameras shooting exactly the same chart you can make valid comparisons. Feel free to come by and shoot charts with me.

Graeme
Logged

Josh-H

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2079
    • Wild Nature Photo Travel
phase versus hassleblad
« Reply #76 on: March 07, 2010, 05:13:57 pm »

Quote
Unfortunately DxO Labs has complicated matters. Their DxO Mark camera tests have become widely read by photographers around the world, and are quoted as gospel by many. Regrettably there are concerns that I, and others have expressed about some aspects of DXO's tests, and I have been in communication with them over this in the past.

The only aspect that is relevant to this discussion is with regard to dynamic range. The standard definition is, as mentioned above, how many F stops above and below middle gray can be recorded while delivering full texture and detail. The DXO definition, according to their web site, is the range between zero signal to noise and full saturation of the sensor.

This approach is not inherently flawed, it's just that it does not take into account the linear nature of sensors. It is therefore not a particularly relevant way of measuring DR to a photographer, as opposed to an engineer.

Tests done by me in the past, as well as others more recently, show that DSLRs have a dynamic range of 6–7 stops while top medium format backs are in the 12-13 stop range. This is using the common definition of DR as mentioned above.

The exact numbers are open to some debate because of the subjective aspect of the test, but usually a one stop differential is the most seen between testers or test runs. But, if someone is using a different set of criteria for their tests then obviously results can differ much more widely.

The question that this raises in my mind (as I see it)  is why are DXO measuring DR by taking the difference between zero signal to noise and full saturation of the sensor? And how does this really correlate to how many F stops above and below middle gray can be recorded with full texture and detail? And further.. why is the difference in DR so great using the two different measurement techniques? And why is is DXO's methodology not relevant in the real world? (as this seems to be what is being implied).

It seems odd to me that DXO with all their scientific efforts would measure DR in a method that did not take into account the linear nature of the sensor. Its not that I am a DXO pundit - far from it. I just want to fully understand why different measuring methods are being used (as this is clearly the reason for the differences) and which (if any) actually correlates to real world results.



Logged
Wild Nature Photo Travel

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
phase versus hassleblad
« Reply #77 on: March 07, 2010, 05:20:50 pm »

Quote from: Graeme Nattress
At least if the images are posted and the raws made available, people with different methodologies of DR measurement can take that data and do to it whatever they want. Those who work visually can push and pull the files in Photoshop, those who put numbers to it can do so, but with the cameras shooting exactly the same chart you can make valid comparisons. Feel free to come by and shoot charts with me.

Graeme

Thanks Graeme. I shall definitely keep this invitation in mind. It won't be in the very near future because I am in Toronto and you in Ottawa, but I shall ping you when I see an opportunity to get up there, and if you're avaiable we can do that.

Mark
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

Graeme Nattress

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 584
    • http://www.nattress.com
phase versus hassleblad
« Reply #78 on: March 07, 2010, 05:43:34 pm »

Quote from: Josh-H
The question that this raises in my mind (as I see it)  is why are DXO measuring DR by taking the difference between zero signal to noise and full saturation of the sensor? And how does this really correlate to how many F stops above and below middle gray can be recorded with full texture and detail? And further.. why is the difference in DR so great using the two different measurement techniques? And why is is DXO's methodology not relevant in the real world? (as this seems to be what is being implied).

It seems odd to me that DXO with all their scientific efforts would measure DR in a method that did not take into account the linear nature of the sensor. Its not that I am a DXO pundit - far from it. I just want to fully understand why different measuring methods are being used (as this is clearly the reason for the differences) and which (if any) actually correlates to real world results.

http://www.dxomark.com/index.php/eng/Techn.../Noise-protocol shows how DXO do it - with grey targets over a 13 stop range. I actually think that's too small a range - they should be going much further from practical experience, and that's why I've just sourced an 18 stop chart rather than the 13.66 stop one I was using. Because sensors are linear and we know  noise is the limit of dynamic range and that the brighter the tones the less noise they contribute to the final image, measuring SNR is a completely valid way of determining absolute dynamic range.

Graeme
Logged

deejjjaaaa

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1170
phase versus hassleblad
« Reply #79 on: March 07, 2010, 06:21:35 pm »

Quote from: Mark D Segal
I have photographed the same scenes with both cameras under as near identical conditions as I could, and I can tell which has higher DR, better shadow detail and better resolution straight out of the box. You'll have to take my word for it - no sense dumbing it all down into JPEGs.
may we have just your raw files instead of your word ? please don't tell that you are not keeping them
Logged
Pages: 1 2 3 [4] 5 6 ... 10   Go Up