Pages: 1 ... 7 8 [9] 10   Go Down

Author Topic: phase versus hassleblad  (Read 47317 times)

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
phase versus hassleblad
« Reply #160 on: March 08, 2010, 08:40:26 pm »

Quote from: Christian Miersch
By now its quite safe to bet the difference must be about useable and not about total DR.

And why would about 5 or 6 stops of DR be "unuseable"?
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

cmi

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 492
phase versus hassleblad
« Reply #161 on: March 08, 2010, 08:53:23 pm »

Quote from: Mark D Segal
And why would about 5 or 6 stops of DR be "unuseable"?

Im was not implying specific numbers.

So far it has been made plausible that measureable DR of current 35mm and MFDB Sensors should be roughly the same. So a difference, if any, must be about useable DR, and this on the other hand has been made plausible too. Definitive proof is missing except for one D3-example. (Did I overlook something?) Exact values for a difference however are not yet clear, but very likely to be under 6 stops.
« Last Edit: March 08, 2010, 09:06:51 pm by Christian Miersch »
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
phase versus hassleblad
« Reply #162 on: March 08, 2010, 09:02:16 pm »

Quote from: Christian Miersch
Im was not implying specific numbers. Just what I said.

Sorry, but this is a bit facile. You're betting the "difference" is between total and useable. It so happens the so-called difference we're talking about in this thread is in the range of 6 to 7 stops. Therefore your statement implies specific numbers whether you meant to or not, and I'm wondering how such a huge amount can be "unuseable". I think it's a fair question.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

BernardLanguillier

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 13983
    • http://www.flickr.com/photos/bernardlanguillier/sets/
phase versus hassleblad
« Reply #163 on: March 08, 2010, 09:03:13 pm »

Quote from: Mark D Segal
As I mentioned further above, it should be ABUNDANTLY CLEAR by now that what's going on here is a comparison of apples and oranges. The methods of measuring DR are differing between the 6-7 crew and the 13+ DxO crew, and this MUST BE driving the differences of outcome.

Yes, so the reasonnable thing to do would be to remove this claim from both the original article and editorial comment wouldn't it?

Cheers,
Bernard

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
phase versus hassleblad
« Reply #164 on: March 08, 2010, 09:07:11 pm »

Quote from: BernardLanguillier
Yes, so the reasonnable thing to do would be to remove this claim from both the original article and editorial comment wouldn't it?

Cheers,
Bernard

Absolutely - if and when my hypothesis has been confirmed.

Cheers,

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

cmi

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 492
phase versus hassleblad
« Reply #165 on: March 08, 2010, 09:08:56 pm »

Quote from: Mark D Segal
Sorry, but this is a bit facile. You're betting the "difference" is between total and useable. It so happens the so-called difference we're talking about in this thread is in the range of 6 to 7 stops. Therefore your statement implies specific numbers whether you meant to or not, and I'm wondering how such a huge amount can be "unuseable". I think it's a fair question.

I agree it could be misunderstood. But I clarified it. I also edited my last post you cited.
Logged

Josh-H

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2079
    • Wild Nature Photo Travel
phase versus hassleblad
« Reply #166 on: March 08, 2010, 09:10:28 pm »

Quote from: BernardLanguillier
Yes, so the reasonnable thing to do would be to remove this claim from both the original article and editorial comment wouldn't it?

Cheers,
Bernard

I agree - And / Or, make available raw files P65+ / D3X / 1DSMK3 as demonstrable evidence of the claimed 6-7 stop advantage of MFDB. This would put the matter to rest one way or the other would it not?
« Last Edit: March 08, 2010, 09:11:31 pm by Josh-H »
Logged
Wild Nature Photo Travel

EricWHiss

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2639
    • Rolleiflex USA
phase versus hassleblad
« Reply #167 on: March 08, 2010, 09:36:05 pm »

Quote from: JeffKohn
If this is true it should be easy to prove by publishing Imatest results to put this matter to rest. I'm not holding my breath, though...
I have done some of these tests and posted the results on this site in the past.  I have only older digital backs so can't comment on the ones in question here but I would not be surprised if there were 3-4 stops difference between the current dslrs and newest digital backs in the measurment important to photographers, and virually no difference in DR in the kind of measurement DXO does. I don't think my imatest license is current so I won't be able to jump in and do them.
« Last Edit: March 08, 2010, 09:39:02 pm by EricWHiss »
Logged
Rolleiflex USA

LKaven

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1060
phase versus hassleblad
« Reply #168 on: March 08, 2010, 10:18:15 pm »

Each time I come upon a post by Michael,  I look for some hint of a claim with any sort of public standards for evaluating its veracity, any sort of common basis for determining its truth or falsity.  But the only basis for claims offered is the Argument From Authority, appearing in different forms in differing postings.  

Then I look at my D3S, with its linear mapped sensor, producing more than 12 good bits of data, at least 10 of which are clean by the strictest standards.  With a linear gamma, this is ten plus stops of dynamic range.  And I think, how do I interpret this when I'm merely offered the argument from authority as a counterargument?  

Luke

Jonathan Ratzlaff

  • Full Member
  • ***
  • Offline Offline
  • Posts: 203
phase versus hassleblad
« Reply #169 on: March 08, 2010, 10:52:40 pm »

After reading through all the pages of discussion I think it boils down to the difference between medium format and 35mm.  A 4x6 print from a medium format negative blows away a 4x6 print from a 35mm negative, even though the dynamic range (read film curve) is identical.  Same goes for a projected slide; medium format blows away the 35mm.  Yet the system resolution stays pretty much the same.

So it is not surprising that you see a difference between a medium format back and a 35mm DSLR.  I am not sure it is dynamic range that makes the difference here.  It is the same difference in tonalty and subtle colour range that makes the difference here.

Having learned to expose using transparency film, my experience is that my dslr has way more dynamic range than transparency film. The only real difference is that the velvia's black has a Dmax of 4 wheras the digital value of black is more or less 0plus sensor noise for that exposure.  

As many have stated, we are dealing with apples and oranges here.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
phase versus hassleblad
« Reply #170 on: March 08, 2010, 11:22:12 pm »

Hi,

Following this discussion I see the ask these questions on the DR issue:

1) Is there any physically feasible explanation for MFDBs having more than two stops of DR? The area of the sensor is about 2.6 larger on the MFDB than on FX this should give a 1-2 step advantage, simple by collecting 2.6 times the photons.

2) If the MFDB has an advantage in DR over full size sensor why doesn't it work well at high ISO? If we assume exposing to the right, underexposure would be equivalent to increasing ISO. So 5 steps of underexposure on a 50 ISO would give 1600 ISO. Does a Phase One P65+ give better ISO 1600 performance than a Nikon 3DX? Does a lesser MFDB than the P65+ give ISO 1600 performance than a Canon 1DsIII. Reason for asking may be that according to the (in) famous DxO numbers both Phase and Nikon seem to have invented some magic sauce pushing the sensors additional two steps. Many readers may be tempted to buy used or simpler backs. Do they have the same magical DR?

3) How much of the perceived difference is not related to sensor but to lens flare? Have MF lenses fewer glass to air surfaces, better coating and internal shielding?

4) Do Phocus and Capture One utilize additional information from the sensor? If we use the MFDB with LR or ACR would we still see the same benefits?

What I'm looking for is a logical explanation on how the large difference in DR would arise.

The comparison by Mark Dubovoy was good read, a great thanks for sharing experience.

Best regards
Erik


Quote from: siba
Mark Dubovoy has done an outstanding, pretty much unbiased test, which is pretty interesting.

And then he writes:

"The dynamic range of a typical high-end professional 35 mm DSLR is around 7 F/stops. Medium Format cameras are closer to 13 F/stops of dynamic range. You can see the difference from 30 feet away in a small print"

You can see the difference from 30 feet! well.....

I'm a phase one user. A professional photographer, who uses the phase one (P 45) for 99% of my work. I also use a canon 5D when light is bad and I'm shooting people, and I want to use the ambient light. I will bet everything I own, and everything my wife owns, that no person on this earth could tell the difference between a small print taken with a high end dslr and a high end medium format digital back from 30 feet (10metres) away.

Any takers?

Stefan Siba
« Last Edit: March 08, 2010, 11:28:11 pm by ErikKaffehr »
Logged
Erik Kaffehr
 

siba

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 61
    • http://www.stefansiba.com
phase versus hassleblad
« Reply #171 on: March 08, 2010, 11:43:18 pm »

Hey Johnathan, heres the original poster - Stefan.

I didn't percieve this thread to go down the comparing medium format versus dslr. That wasn't my original post.
I am a professonal photographer shooting thousands of phase one digital files a month.
I have done so for the last three years or more. I can also tell my MFDB photos from my DSLR photos. After all I shot them.
I hardly ever give any thought to DR. I know my phase one files are beautiful compared to what I was used to before I started using the back.

You say your 6x4 MFDB prints blow away your 6x4 DSLR prints. Well, that's not the point of this argument/discussion

So, if no one minds I'd like to reiterate what my initial point was.

I'll shout it:
IT IS THE 30 FEET away issue that I have. As several posters have thankfully also picked up on, 30 feet is a bloody long way away to see any difference in two prints. Let alone two prints of the same subject taken by the same photographer under the same conditions, processed the same way and printed with the same care.

I am a MFDB user and yet find it surprising that anyone would endorse a comment that claims that from 30 feet away one can tell that a SMALL PRINT was taken using a MDFB

I especially like the duck example. Thank you Dafu post 127

I'm fairly sceptical of the 6 to 7 DR difference, but we can just go round in circles getting smuggish I know better answers from a few people and start going round again. Let it be 6, 7, or call it DR or lenses, or Miriam for that matter. If you guys don't mind I'd like to steer this converastion more towards seeing a print at a certain distance and being able to tell that it is MFDB as opposed to DSLR.

cheers all.

Stefan
Logged

dreed

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1715
phase versus hassleblad
« Reply #172 on: March 09, 2010, 12:05:16 am »

Reading Michael's editorial note at the end of the about f-stops of dynamic range, the question that comes to my mind is - will it always be this way?

Is there something intrinsic to how MF digital backs are made that causes them to have better dynamic range?

If it is simply size, then why haven't we seen a decline in dynamic range as photosites have shrunk in DLSRs?
If this is a result of sensor technology improving, then shouldn't it be possible for Canon/Nikon/Sony to churn out a DSLR with large photosites and thus higher dynamic range?
Or perhaps one might ask, why hasn't Nikon been able to reach a higher dynamic range with their lower MP count modern DSLRs?

If it isn't photosite site, is this the difference between CMOS and CCD?
Or is it something else?

I clicked on through to the cambridge colour link and whilst they discussed what dynamic range was and how it affects pixels, they didn't try to explain why the numbers are what they are.

I'm not yet ready to accept that because the camera is 35mm must mean that it has half the dynamic range of a bigger camera. There's got to be a cause...
Logged

siba

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 61
    • http://www.stefansiba.com
phase versus hassleblad
« Reply #173 on: March 09, 2010, 12:07:21 am »

OK, so now that I think about it, there is also the DR subplot which is at the start of the incriminating sentence from Mark Dubovoy (not Mark Segal   ); which is difficult to dismiss.

The point of Mark Dubovoy's remark was that there is such a HUGE difference between MFDB and DSLR, and that it is the DR that makes this difference tangible, as it were.

But, let's not get too caught up with the DR.

Please let us first refute the notion that one can tell what format camera was used, when looking at a 10x8 print from 30 feet away.

Once we do this I'll be able to sleep at ease, and let anyone who wants to do so, talk about the exact difference between DR of MFDBs and DSLRs to their heart's content.

cheers again

Stefan
« Last Edit: March 09, 2010, 12:37:12 am by siba »
Logged

siba

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 61
    • http://www.stefansiba.com
phase versus hassleblad
« Reply #174 on: March 09, 2010, 12:33:55 am »

The reason I don't want to get caught up in the exact DR difference between a P65+ and a highest end DSLR, is that there may well conceivably be, in some way, a 6 or 7 stop difference.

I can't predict where the proof may come from. I can't imagine where it may come from. But it may be out there. Who knows? Stranger things have happened.

What I am absolutely certain of is that from 30 feet (10 metres), you can not tell with any certainty whether a 10x8 print was taken with a MFDB or a DSLR. And I would claim that given most normal circumstances, you probably wouldn't be able to tell the difference between the same scene taken with a MFDB and an iphone. My iphone would by definition have close to no DR, so the difference would be the whole 13 stops.

I was drinking Sangre de torro 2003 reserva, now I'm on the yellow tail merlot.

goodnight all

Stefan
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
phase versus hassleblad
« Reply #175 on: March 09, 2010, 12:40:52 am »

Hi,

A larger sensor collects more photons and can also store more free electrons. A MFDB sensor (with 1.0 crop factor) has a surface area 2.6 times that of an FX (crop factor 1.0) DSLR sensor. All other aspects kept constant this should increase DR by about one stop.

Usable DR may be more limited by "photon shot noise", random variations of collected photons, than by readout electronics. The engineering definition of DR is log2(Maximum Signal/Signal at SNR==1).

I cannot see any reason (on the sensor side) that an MFDB would have any advantage regarding DR in excess of one stop over an "FX" sensor. This applies to both "engineering DR" and "usable DR". I'm not saying that the advantage does not exist, I'm just saying that I have not seen any feasible explanation for it.

I would not rule out that lens flare may play a role. DSLR lenses often have many air/glass surfaces and may have less baffling than MF-lenses. It is also possible that MFDBs may have more efficient flare reduction in the mirror chamber than DSLRs.

Best regards
Erik


Quote from: dreed
Reading Michael's editorial note at the end of the about f-stops of dynamic range, the question that comes to my mind is - will it always be this way?

Is there something intrinsic to how MF digital backs are made that causes them to have better dynamic range?

If it is simply size, then why haven't we seen a decline in dynamic range as photosites have shrunk in DLSRs?
If this is a result of sensor technology improving, then shouldn't it be possible for Canon/Nikon/Sony to churn out a DSLR with large photosites and thus higher dynamic range?
Or perhaps one might ask, why hasn't Nikon been able to reach a higher dynamic range with their lower MP count modern DSLRs?

If it isn't photosite site, is this the difference between CMOS and CCD?
Or is it something else?

I clicked on through to the cambridge colour link and whilst they discussed what dynamic range was and how it affects pixels, they didn't try to explain why the numbers are what they are.

I'm not yet ready to accept that because the camera is 35mm must mean that it has half the dynamic range of a bigger camera. There's got to be a cause...
« Last Edit: March 09, 2010, 12:49:32 am by ErikKaffehr »
Logged
Erik Kaffehr
 

image66

  • Full Member
  • ***
  • Offline Offline
  • Posts: 136
phase versus hassleblad
« Reply #176 on: March 09, 2010, 12:47:51 am »

So, Michael's "clarification" got me to thinking that just maybe it would be a good idea to test my camera out for what really is the usable dynamic range. For this test I used the ultimate in dogmeat DSLR cameras--the Olympus E-1. This isn't totally fair, though, because this camera sports a Kodak designed and built CCD sensor not too unlike some other larger sensors, but you get the idea.

Zone-10's E-1 Dynamic Range Test

I'm NOT going to tell you how many stops of dynamic range the camera has. I'd like others to tell me based on the images shown. Now, to make things even more ugly for the camera (and limiting) all conversions were done in sRGB and straight as-is by the Olympus converter software. On the second and third pages I do include conversions maximizing highlight and shadow recovery.

I'd think that based on the images in this test, that the E-1 is capable of more than six stops of "usable dynamic range" but I'll let you experts tell me.
Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
phase versus hassleblad
« Reply #177 on: March 09, 2010, 12:51:18 am »

Hi,

No DR actually means that the print would be uniform gray....

Best regards
Erik

Quote from: siba
The reason I don't want to get caught up in the exact DR difference between a P65+ and a highest end DSLR, is that there may well conceivably be, in some way, a 6 or 7 stop difference.

I can't predict where the proof may come from. I can't imagine where it may come from. But it may be out there. Who knows? Stranger things have happened.

What I am absolutely certain of is that from 30 feet (10 metres), you can not tell with any certainty whether a 10x8 print was taken with a MFDB or a DSLR. And I would claim that given most normal circumstances, you probably wouldn't be able to tell the difference between the same scene taken with a MFDB and an iphone. My iphone would by definition have close to no DR, so the difference would be the whole 13 stops.

I was drinking Sangre de torro 2003 reserva, now I'm on the yellow tail merlot.

goodnight all

Stefan
Logged
Erik Kaffehr
 

siba

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 61
    • http://www.stefansiba.com
phase versus hassleblad
« Reply #178 on: March 09, 2010, 12:55:13 am »

sorry, one more point...

I tried this the other day after starting the thread....to put things in perspective, so to speak....

Looking at a 10x8 print from 30 feet is the equivalent in size of looking at your thumbnail at arms length.  Try it.

So, I printed one of my phase files 1x1 cm and one of my DSLR files 1x1 cm and glued them to my thumbnails.

I viewed them at arms length, and my 7 year old daughter who was looking over my shoulder, said to me, "daddy, I can't tell the difference between those two prints on your thumbnails"

Therefore I must come to the conclusion that there is no difference between the DR of my MFDB and my DSLR.

This is purely anecdotal evidence, but I had to agree with my daughter at the time.

goodnight again,

Stefan
« Last Edit: March 09, 2010, 01:21:14 am by siba »
Logged

RomanJohnston

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 72
phase versus hassleblad
« Reply #179 on: March 09, 2010, 01:08:56 am »

Interesting. I have a few questions to ANYONE, including the writer of the article.

1. Has ANYONE really tested CURRENT Dslr's ....lets say the D3X....DIRECTLY against the ....say....P65+ unit to see what TRUE DR numbers are?

2. If so please present your findings and if you have any proof at all, please let them flow.

My hunch is....DSLRs are better than the Mark assumes (as he says others SHOULD not talk if they havent done the tests....then goes on to say that his DSLR data is not from a test like he reccomends, but from heresay from the company.) I think this is why so many people are ticked off. Oh and the 30 feet comment.....ok guys....for somone to say WHOA! look at the DR on that picture....well that is absolutly SILLY. No one knows how much DR is in a shot they know nothing about....come on. Even if it was said....it would be totally silly that the person had all the information to really make a definitive statment about the DR of the picture printed.....no more than a person would be able to comment on noise quality of a camera from a downsized picture posted on the web.

About DXO....well they are off quite a bit, but in all fairness, they are off quite a bit consistantly. As such even if the numbers are off based on the mechanical and not human perception testing procedures, apples are compared to apples with the test. So with DXO numbers being off from real world...I would say that it would be quite silly to assume that the numbers are BETTER on the MF kit, and WORSE on the DSLR kit.....thats just not very logical. Actually that would sound very emotional.....and emotions rarly have any merit when your on the hunt for facts.

So my proposal is, somone who has both cameras.....DO THE TEST.....put this all to rest. Sure the emotional ones will continue to argue over testing procedures, etc.

I know I get about 8 full usable stops of info on my D300 using 14bit RAW files. A bit more than assumed in the article. I expect MF backs to do better than my D300.

I suspect the real numbers with the D3X and MF kit to be more like D3X 10 usable stops, and a solid 12 - 12.5 usable stops of DR for the MF kits.

Anyone want to place any bets?

Roman
« Last Edit: March 09, 2010, 01:14:08 am by RomanJohnston »
Logged
Pages: 1 ... 7 8 [9] 10   Go Up