Pages: 1 2 [3] 4 5 ... 21   Go Down

Author Topic: Is the difference of DR on MFDB vs 35mm dslr discernible on print?  (Read 87761 times)

evgeny

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 495
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #40 on: September 13, 2009, 10:04:35 am »

I shot with a Nikon F6/17-35/70-200 until 2007 and was pleased with my images, especially when compared to a friend's Nikon D200/18-200.

Then, my Contax 645 beats my Nikon F6. The 6x4.5 format was a huge step forward.

A half year later I added a digital back to Contax, sold all film gear, including the Nikon 9000 scanner, and stopped to spend long hours and days in film development and scanning.

Regarding money, well, the MFDB is more expensive than film, but I don't know Photoshop enough to create a WOW with a cheaper gear.
Logged

Jim2

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 96
    • http://
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #41 on: September 13, 2009, 10:05:15 am »

I just had a quick look at www.dxomark.com and it seems that Canon 1ds3 beats Phaseone P65 in all respects (I'm using the camera comparison)

http://www.dxomark.com/index.php/eng/Image...%29/Phase%20One

I am wondering whether the better printed images from MFDB is due to higher DPI it afforded from having a higher resolution? - Assuming of course that it is 'better' - I am just saying here because I have not compared one myself. Very few people here argued that dslr produces better prints?
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #42 on: September 13, 2009, 10:28:04 am »

Hi,

Resolution, microcontrast and acutance are not part of DxO mark. DxO mark is about SNR (Signal to Noise Ratio) and DR (Dynamic Range). They present all their data on six tabs (in comparison mode) and nine tabs if you look at a single model. There are two sets of plots, one for "actual pixels" and the other for "print". The rationale behind this is here:

http://www.dxomark.com/index.php/eng/Insig...s-offsets-noise!

My suggestion is that Phase One P65+ comes to advantage for the fullowing reasons:

1) Phase One owners probably use Capture One which can utilize Phase One proprietary informition
2) Shear number of pixels
3) Higher MTF of lens at equivalent lp/mm (sorry for engineering speak)
4) No optical low pass filter

So DxO-mark is just about sensor signals, not about what is in front of the sensor or the actual processing on raw data.

Hope this helps...

Best regards
Erik


Quote from: Jim2
I just had a quick look at www.dxomark.com and it seems that Canon 1ds3 beats Phaseone P65 in all respects (I'm using the camera comparison)

http://www.dxomark.com/index.php/eng/Image...%29/Phase%20One

I am wondering whether the better printed images from MFDB is due to higher DPI it afforded from having a higher resolution? - Assuming of course that it is 'better' - I am just saying here because I have not compared one myself. Very few people here argued that dslr produces better prints?
Logged
Erik Kaffehr
 

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #43 on: September 13, 2009, 10:31:50 am »

Quote from: ErikKaffehr
Hi Ray,

DR is nothing you can see in a print or, to be more exact, prints can reproduce about seven seven steps of DR. It would be possible to map a large DR. like 10 steps, to print range but the image would be boring and flat. DR is something we can use in raw-conversion and post processing to achieve attractive results.

D-max on papers is about 2.15 - 2.35. D-min on glossy paper is about 0.05. So DR is about 2.2, to get it in step you simply divide by 0.3 .

Best regards
Erik


Hi Erik,
Have I been imprecise in my expression? I've amended my statement as follows. Is this better?  

Quote
However, the situation is different when comparing equal size images or prints. The P65+ image then appears to have better SNR, tonal range and color sensitivity, but oddly enough not better DR. The DR of the D3X, as represented even on an 8x12" print, is still about 2/3rds of a stop better than the downsampled P65+ image, excluding enhanced results from proprietary software.

Obviously a wide dynamic range capability allows one to capture more detail in the shadows. However it's interesting that DXO tests indicate that the DR of the D3X is so much greater than the P65+ at the pixel level, that even when both images are downsampled to the quite small file size sufficient for an 8x12" print at 300dpi, the D3X would apparently still show more detail in the deepest shadows of a scene of wide brightness range, assuming both images have been correctly exposed to the right.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #44 on: September 13, 2009, 10:42:30 am »

Quote from: ErikKaffehr
My suggestion is that Phase One P65+ comes to advantage for the fullowing reasons:


3) Higher MTF of lens at equivalent lp/mm (sorry for engineering speak)


Erik,
Point #3 doesn't seem right. The pixel density of the D3X is the same as that of the P65+. I doubt whether MF lenses in general would have a higher MTF at equivalent lp/mm than the best Nikon glass, considering the larger image circle required of MF lenses. But the lack of an AA filter might give the impression of greater lp/mm resolution.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #45 on: September 13, 2009, 11:16:55 am »

Quote from: Jim2
I just had a quick look at www.dxomark.com and it seems that Canon 1ds3 beats Phaseone P65 in all respects (I'm using the camera comparison)

http://www.dxomark.com/index.php/eng/Image...%29/Phase%20One

I am wondering whether the better printed images from MFDB is due to higher DPI it afforded from having a higher resolution? - Assuming of course that it is 'better' - I am just saying here because I have not compared one myself. Very few people here argued that dslr produces better prints?

DSLRs tend to have better performance than MFDBs at high ISO. However, at base ISO the Ids3 appears to be about equal to the P65+, at the pixel level only. When both images are downsampled to 8mp, the P65+ is clearly better in all departments. I imagine the P65+ would also be better in all departments when compared to a 1Ds3 image upsampled to 60mp.
Logged

PHOTO ZARA

  • Newbie
  • *
  • Offline Offline
  • Posts: 39
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #46 on: September 13, 2009, 11:29:10 am »

Quote from: Ray
DSLRs tend to have better performance than MFDBs at high ISO. However, at base ISO the Ids3 appears to be about equal to the P65+, at the pixel level only. When both images are downsampled to 8mp, the P65+ is clearly better in all departments. I imagine the P65+ would also be better in all departments when compared to a 1Ds3 image upsampled to 60mp.

are you suggesting then 35mm FF 1ds markXX with native 60mp would be surely equal or better to the current 645 FF P65+ at any size?

(both taken at optimum quality exif.)
« Last Edit: September 13, 2009, 11:31:09 am by PHOTO ZARA »
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #47 on: September 13, 2009, 11:36:58 am »

Ray,

A 645 image needs less enlargement than an 24x36 image, so smaller frequency needs to utilized to achieve a similar resolution in print, so if you need 40 lp/mm on sensor for a certain print on 24x36 you would need less, like 30 or 20 lp/mm, on a bigger format. When I wrote equivalent lp/mm i meant just this factor. I'd also suggest that the new lenses designed for digital photography may be quite good.

Best regards
Erik

Quote from: Ray
Erik,
Point #3 doesn't seem right. The pixel density of the D3X is the same as that of the P65+. I doubt whether MF lenses in general would have a higher MTF at equivalent lp/mm than the best Nikon glass, considering the larger image circle required of MF lenses. But the lack of an AA filter might give the impression of greater lp/mm resolution.
Logged
Erik Kaffehr
 

gwhitf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 855
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #48 on: September 13, 2009, 03:08:45 pm »

Quote from: Jim2
I guess this is all technical and at the end of the day a nicely captured shots with 1ds3 would be quite nice too and I know a few professional landscape photographers who sell their photos for a living use 5d2.

To Jim2:

This forum can get very heavy on Theory, but the last time I looked, it was named Photography, not Pixel Theory. (Maybe there should be another forum for that). So before you go drop thirty or forty grand on a MF system, based on Theory Feedback here, I'd advise you to call CaptureIntegration and RENT a system. It might be the cheapest money you ever spent.

Take your 1ds3 and that rental system, and GO SHOOT REAL PHOTOGRAPHS, and then come back and make real Epson prints, and then lay them out on a table, and do not get out a magnifying glass, (unless your potential clients do). Act like a normal human being, and simply look at the prints. Do not put on a Lab Coat. Just stand there and act like a real human being. Do not consult an MTF Chart. Just be a Photographer, and stand there and compare the prints.

Then ask yourself again the original question that you asked here, and then see if you think it's worth writing a check for forty grand.

Many of the so-called science experts here have never purchased a MF system, let alone used one hard, yet they seem quite content advising you on a sizable purchase. Theory is fine, but your potential clients aren't writing you a check for Theory; they're writing you a check for Ink On Paper.

Only if you do this can you truly answer this question for yourself.

And, if you do this test, make sure you bracket, so that you get within three stops.
Logged

asf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 510
    • http://www.adamfriedberg.com
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #49 on: September 13, 2009, 03:20:00 pm »

I own and use regularly both types of systems you are considering

Take GWHITF's advice and don't buy anything until you compare. Rental units can be shipped to you.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #50 on: September 13, 2009, 05:39:21 pm »

Sorry Ray,

My comment was not personally ment, it was more intended as a response to the original posting and I happened to reply to your posting.

On the DR issue, I don't know. There had been several postings indicating very high DR values for the D3X.

I'm not really sure that DR is really higher for MF backs than for DSLRs. Sometimes the eyes don't give the same information as measurements. I'd suggest that the observer says "I prefer the image on the left" but cannot explain in exact terms. So it's perhaps called DR, sometimes Microcontrast or even tonality.

Best regards
Erik

Quote from: Ray
Hi Erik,
Have I been imprecise in my expression? I've amended my statement as follows. Is this better?  



Obviously a wide dynamic range capability allows one to capture more detail in the shadows. However it's interesting that DXO tests indicate that the DR of the D3X is so much greater than the P65+ at the pixel level, that even when both images are downsampled to the quite small file size sufficient for an 8x12" print at 300dpi, the D3X would apparently still show more detail in the deepest shadows of a scene of wide brightness range, assuming both images have been correctly exposed to the right.
Logged
Erik Kaffehr
 

tho_mas

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1799
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #51 on: September 13, 2009, 06:00:33 pm »

Quote from: GLuijk
If the propietary software by Hasselblad performs noise reduction, no fair DR comparision can be done. We are talking about sensor's DR, not about propietary software abilities in noise reduction.
so you think there is no noise reduction in Canon and Nikon RAW files? AFAIK CMOS sensors perform NR on chip before a RAW file is produced. How would you compare those files with MFD files that need their respective software to perform a similar "clean" stage of the image (RAW)?
IMHO in this regard DX0 is pointless and for me personally everything that counts is how good is a given image at base ISO after the "best" possible processing (either way if on chip or with the manufacturers software).
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #52 on: September 13, 2009, 07:56:44 pm »

Quote from: PHOTO ZARA
are you suggesting then 35mm FF 1ds markXX with native 60mp would be surely equal or better to the current 645 FF P65+ at any size?

(both taken at optimum quality exif.)

Not without a huge leap in technology. If the new 18mp Canon 7D were full frame, it would be only 45mp. The current flagship DSLRs from Canon and Nikon have around the same pixel size as the P65+, so it's not totally surprising that performance at the pixel level is as good (in the case of the 1Ds3) or better (in the case of the D3X).

However, in practice such performance figures can be misleading. What they mean is, if you were to take shots with both cameras using lenses of the same focal length, and then crop the P65+ image to the same FoV as the D3X image, the D3X image would likely be better, on balance, with regard to SNR, DR, tonality and color sensitivity. Whether or not one image is sharper than the other would depend on the quality of lenses used and the effect of the D3X's AA filter.

In practice, you might rarely do this. I certainly wouldn't because I use mostly zoom lenses. However, if I were to rely upon best quality primes outside of a studio setting, I can envisage that I would sometimes want to crop heavily.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #53 on: September 13, 2009, 08:06:23 pm »

Quote from: ErikKaffehr
A 645 image needs less enlargement than an 24x36 image, so smaller frequency needs to utilized to achieve a similar resolution in print, so if you need 40 lp/mm on sensor for a certain print on 24x36 you would need less, like 30 or 20 lp/mm, on a bigger format. When I wrote equivalent lp/mm i meant just this factor. I'd also suggest that the new lenses designed for digital photography may be quite good.


I see! In this situation I think it would be better to talk about LW/PH (line widths per picture height) since lp/mm usually refers to the absolute resolution.
Logged

Ray

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 10365
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #54 on: September 13, 2009, 08:53:05 pm »

Quote from: ErikKaffehr
Sorry Ray,

My comment was not personally ment, it was more intended as a response to the original posting and I happened to reply to your posting.

On the DR issue, I don't know. There had been several postings indicating very high DR values for the D3X.

I'm not really sure that DR is really higher for MF backs than for DSLRs. Sometimes the eyes don't give the same information as measurements. I'd suggest that the observer says "I prefer the image on the left" but cannot explain in exact terms. So it's perhaps called DR, sometimes Microcontrast or even tonality.

Best regards
Erik


No offense taken, Erik. I've always associated lack of DR of the film or sensor with lack of detail in the shadows. I infer from the DXO test results that an 8x12 print from the D3X would reveal more detail in the shadows, despite the obvious advantage of the larger file size of the P65+. The difference on the DXOMark site is about 2/3rds of a stop. I interpret that as meaning, if you want the P65+ shot to have equal shadow detail on the 8x12" print, you would have to overexpose the P65 shot by 2/3rds of a stop and blow the highlights. It is assumed that DR comparisons are only meaningful when shooting scenes of a high brightness range, such as this one below, taken with the 5D.

It would be interesting if Bernard could do a comparison with his hired P65+ in order to see if there is a noticeable difference at the 8x12" print size and if this is really what DXO mean.

[attachment=16555:Moonlit_night.jpg]

Logged

marcs

  • Newbie
  • *
  • Offline Offline
  • Posts: 43
    • http://
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #55 on: September 13, 2009, 09:32:18 pm »

DR is only one small piece of the pie, so to speak.  And is so overrated.

Having printed many, many large prints (20x30 inch and up) from P45+/P65+, D3X and MF film, I "feel" the results are completely relative.  Though I can say, without a doubt, that nicely executed 4x5s (drum scanned, Inkjet printed) look by far the most pleasing from a tonality and texture perspective (to me and many of my cronies).

The many subtle nuances of film capture, all difficult to articulate, preclude many fine artists from going exclusively digital.  Of course, commercial concerns are entirely another thing.

Regards, MS
Logged

Professional

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 309
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #56 on: September 13, 2009, 09:37:00 pm »

Quote from: Jim2
I just had a quick look at www.dxomark.com and it seems that Canon 1ds3 beats Phaseone P65 in all respects (I'm using the camera comparison)

http://www.dxomark.com/index.php/eng/Image...%29/Phase%20One

I am wondering whether the better printed images from MFDB is due to higher DPI it afforded from having a higher resolution? - Assuming of course that it is 'better' - I am just saying here because I have not compared one myself. Very few people here argued that dslr produces better prints?

Long time ago i did a quick simple test between my 1Ds3 and H3DII-39, the H3DII-39 was the winner in all aspect until ISO400, so for quality and clarity and sharpness and DR and colors up to ISO400 my H3DII blow away my 1Ds3 so easy, i will not wonder that H3DII will blow away even Nikon D3X, and that i say Phase One is better than H3DII, so i will be very happy to get P65+ and never look back to Canon/Nikon for same type of work i will use MFDB for, even my kids portraits from H3DII was all amazing wonderful over all my old shots taken by 1Ds3/1D3/1Ds2/5D, even my friend 5D2, now i want to see a real D3X shots to see what is there that making it the best [ISO400 and lower].
Logged

EricWHiss

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2639
    • Rolleiflex USA
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #57 on: September 14, 2009, 12:55:56 am »

Quote from: GLuijk
I am sorry Eric, but this discussion was already hold in the forum. If the propietary software by Hasselblad performs noise reduction, no fair DR comparision can be done. We are talking about sensor's DR, not about propietary software abilities in noise reduction.

Regards.


As I wrote earlier - you have my respect for your "zero noise" work.  Nothing however that you write in this thread adds to that respect.  If you are not willing to use the proprietary software such as capture one for phase backs or phocus or flexcolor for hasselblad backs in your tests then you will certainly be missing something.  What's unfair about using the software a digital back maker supply's with their own backs?   I don't think there's a fair or unfair since no contest was declared or rules stated by anyone but you.  There certainly is informed and uniformed however, and I consider those that have shot extensively with both cameras for jobs and real image production to be informed and all those that post conjecture and theory but maybe only downloaded a few files that they had no control or knowledge of how they were shot to be uninformed.  I see that every time someone asks a sincere question like the OP did here, a bunch of people jump in and drag this into a needless argument about how DSLR's are just as good or better.  What a waste!  

With regard to the idea that someone needs to prove these differences by supplying you RAW files for the differences to exist - this is simply a flawed argument.    No one owes anything here - its all free.   I did my own tests for my own work.  I know what my cameras and backs will do and in what situations which gear will be most appropriate for my own shooting.   Everyone owes it to themselves to do their own testing.  


Logged
Rolleiflex USA

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #58 on: September 14, 2009, 02:09:59 am »

Hi Eric,

Many photographers, amateurs or professionals, preferably use one tool over another. There can be many good reasons for that, one is that many of us prefer to have a parametric workflow. The approach taken by Guillermo or DxO-mark is perfectly valid in my view. It's about the the raw data.

I'd also say that this discussion is  bit off topic.

Best regards
Erik


Quote from: EricWHiss
As I wrote earlier - you have my respect for your "zero noise" work.  Nothing however that you write in this thread adds to that respect.  If you are not willing to use the proprietary software such as capture one for phase backs or phocus or flexcolor for hasselblad backs in your tests then you will certainly be missing something.  What's unfair about using the software a digital back maker supply's with their own backs?   I don't think there's a fair or unfair since no contest was declared or rules stated by anyone but you.  There certainly is informed and uniformed however, and I consider those that have shot extensively with both cameras for jobs and real image production to be informed and all those that post conjecture and theory but maybe only downloaded a few files that they had no control or knowledge of how they were shot to be uninformed.  I see that every time someone asks a sincere question like the OP did here, a bunch of people jump in and drag this into a needless argument about how DSLR's are just as good or better.  What a waste!  

With regard to the idea that someone needs to prove these differences by supplying you RAW files for the differences to exist - this is simply a flawed argument.    No one owes anything here - its all free.   I did my own tests for my own work.  I know what my cameras and backs will do and in what situations which gear will be most appropriate for my own shooting.   Everyone owes it to themselves to do their own testing.
« Last Edit: September 14, 2009, 02:11:48 am by ErikKaffehr »
Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Is the difference of DR on MFDB vs 35mm dslr discernible on print?
« Reply #59 on: September 14, 2009, 02:20:02 am »

Hi!

1) DxO mark is essentially about noise related issues
2) If you really check DxO-mark there is a "screen" and a "print" option, the difference is that "print" option normalizes noise a standard resolution. Screen corresponds to actual pixels whereas "print" corresponds to what would be seen in a print.
3) DxO-mark is based on the "print" option as DxO regards this to be the relevant parameter

DxO mark has nothing to do with sharpness, resolution and so on.


Best regards
Erik


Quote from: Professional
Long time ago i did a quick simple test between my 1Ds3 and H3DII-39, the H3DII-39 was the winner in all aspect until ISO400, so for quality and clarity and sharpness and DR and colors up to ISO400 my H3DII blow away my 1Ds3 so easy, i will not wonder that H3DII will blow away even Nikon D3X, and that i say Phase One is better than H3DII, so i will be very happy to get P65+ and never look back to Canon/Nikon for same type of work i will use MFDB for, even my kids portraits from H3DII was all amazing wonderful over all my old shots taken by 1Ds3/1D3/1Ds2/5D, even my friend 5D2, now i want to see a real D3X shots to see what is there that making it the best [ISO400 and lower].
Logged
Erik Kaffehr
 
Pages: 1 2 [3] 4 5 ... 21   Go Up