Pages: 1 ... 5 6 [7] 8   Go Down

Author Topic: 8x10/MFDB Comparison  (Read 70193 times)

JeffKohn

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1668
    • http://jeffk-photo.typepad.com
Re: 8x10/MFDB Comparison
« Reply #120 on: August 29, 2010, 01:44:22 pm »

Someone said that 40 Mpx is equivalent to 5 * 4, so you would need 80 Mpx for 10 * 8
Actually 160mp, since you're quadrupling the surface area.

Quote
The old yardstick was that you could enlarge from film up to 10* the neg size, and the new yardstick is 360 original camera pixels per print inch...
You're the only one I see regularly advocating 360 original camera pixels per inch, I think most would agree that's overkill. By your two yardsticks, a 10x enlargement of 35mm film (to 14.1x9.4") would yield the same print quality as a 15.4mp digital capture printed at the same size. You don't seriously believe that, do you?
Logged
Jeff Kohn
[url=http://ww

feppe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2906
  • Oh this shows up in here!
    • Harri Jahkola Photography
Re: 8x10/MFDB Comparison
« Reply #121 on: August 29, 2010, 02:25:30 pm »

The old yardstick was that you could enlarge from film up to 10* the neg size, and the new yardstick is 360 original camera pixels per print inch...

I'm afraid that's not enough. According to the LL LR3 tutorial you need 720dpi to yield good results.

[Yes, this is a joke. Yes, Mr Schewe makes a case for 720ppi in some cases, but I doubt he or anyone else seriously considers even 360ppi a "new yardstick" for prints that are viewed at from reasonable distances. In fact, he used to advocate anything above 180ppi and still seems to stick to it. If one subscribes to the notion that "viewing distance is only limited by the size of one's nose," that's a whole different story.]
« Last Edit: August 29, 2010, 02:27:13 pm by feppe »
Logged

Dick Roadnight

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1730
Re: 8x10/MFDB Comparison
« Reply #122 on: August 29, 2010, 02:26:26 pm »

Actually 160mp, since you're quadrupling the surface area.
Yes, Thanks
Quote
You're the only one I see regularly advocating 360 original camera pixels per inch, I think most would agree that's overkill. By your two yardsticks, a 10x enlargement of 35mm film (to 14.1x9.4") would yield the same print quality as a 15.4mp digital capture printed at the same size. You don't seriously believe that, do you?
I appreciate that 360 original camera pixels is "excellent" and anything higher is OTT... and I expect that I might be making some prints at 180.

35mm film looks rough at greater than A4, as does the output of my 12Mpx P&S, but the mega-pixels from a multi-shot MFD...

Not all films are equal, and not all Megapixels are equal.
Logged
Hasselblad H4, Sinar P3 monorail view camera, Schneider Apo-digitar lenses

Dick Roadnight

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1730
Re: 8x10/MFDB Comparison
« Reply #123 on: August 29, 2010, 02:36:56 pm »

I'm afraid that's not enough. According to the LL LR3 tutorial you need 720dpi to yield good results.

If one subscribes to the notion that "viewing distance is only limited by the size of one's nose," that's a whole different story.
If your eyes do not have bellows built-in, and you are over 25, viewing distance is limited by the power of you reading glasses.

I do not want to print pictures for people to ignore them and walk past them... I want them to take in the whole picture, and then perhaps be drawn in to the picture to look at the detail.

The "old master" paintings did this about half a millennium ago, but photographers have tried to re-write the rules.
Logged
Hasselblad H4, Sinar P3 monorail view camera, Schneider Apo-digitar lenses

Gigi

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 548
    • some work
Re: 8x10/MFDB Comparison
« Reply #124 on: August 29, 2010, 07:26:21 pm »

How long is a piece of string?
What film are you using?

Someone said that 40 Mpx is equivalent to 5 * 4, so you would need 160.

The old yardstick was that you could enlarge from film up to 10* the neg size, and the new yardstick is 360 original camera pixels per print inch...

So for film you get 2.25" * 10 = 22.5 inches from  2.25" 120 MF roll film,...

and form a 60Mpx H4D-60 or P65 + you get 9,000 pixels/360 = 25 inches,,,

So, using the 360 original camera pixels per print inch yardstick, silicon is little higher res than film.

If you use 180 as your yardstick, 645 MDF is equivalent to 5 * 4.

If you use the right yardstick, or compare to low-res high speed sheet film, 60Mpx is equivalent to  10 * 8".

...like I said, how long is your piece of string?

Short and sweet. Put in your own numbers, change them as you believe, but its great to have a simple "rule of thumb" to allow comparisons.....
Logged
Geoff

Dan Wells

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1044
Re: 8x10/MFDB Comparison
« Reply #125 on: September 04, 2010, 01:32:13 am »

Another issue is the quality of scans... 8x10 drum scans are extremely expensive (and slow). Anything short of a drum scanner (and perhaps some of the Creo or Scitex "super flatbeds") will not produce good results. I was amazed recently (I've been shooting digital for years) when I had to deal with a bunch of slide scans (35mm) made on various Nikon Coolscans. 35mm average-quality slide film is supposed to be about 6mp equivalent, and it might well have been in terms of resolution (the files were 10-15 mp in dimension, but a scan will always be softer than native digital of the same resolution). What shocked me was the lack of color accuracy and dynamic range. I've, admittedly, been spoiled by shooting the D3x (11 stops of DR, at least approaching MF quality), but these weren't even in the same ballpark. These were files that often had only 4-5 stops of DR and glaring color problems. I'll be the first to admit that a good scanner and a good operator could have done much better than this garbage, but that is another variable to add into any film/digital comparison - there is an art to good scans, and people who can produce them are rare. In addition to the perfect capture, you need a perfect scan, and those are tough to come by...
Logged

JonathanBenoit

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 414
Re: 8x10/MFDB Comparison
« Reply #126 on: September 04, 2010, 08:33:07 am »

Who says the D3x has 11 stops of dynamic range.. poppycock! more like 8 stops
Logged

Dan Wells

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1044
Re: 8x10/MFDB Comparison
« Reply #127 on: September 04, 2010, 04:14:38 pm »

I've gotten a heck of a lot more than 8 stops out of the D3x! I don't believe DxOMark's nearly 13 stops, but I don't believe 8 stops either... I'd say (from 25,000+ shots on my D3x) that 11 is a pretty good estimate. It really reaches deep into the shadows in 14-bit mode, while holding highlight detail as well. Of course, the first and last of those stops are equivalent to Zones II and VIII - texture, but not full detail. From prints I've seen from Hasselblad and Phase systems, the D3x is comparable (not better, but also not worse) in printed DR. Of course, I haven't made those Hassy/Phase prints myself, and I don't know exactly how they were made (other than that they were large prints being used as manufacturer samples at a show, so one would hope they had done a decent job). Any digital system in this range is going to have MUCH more DR than an average film scan, which was my initial point (substitute your favorite MF system for my D3x and the point is exactly the same)... If it's even possible to get equivalent DR out of any film, it would have to be a slow B+W film, and a fairly special scan. Drum scans of 8x10 range from $100 on up, and the equipment is even more expensive than a P65+ (not to mention difficult to use).
Logged

ondebanks

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 858
Re: 8x10/MFDB Comparison
« Reply #128 on: September 04, 2010, 06:40:56 pm »

Who says the D3x has 11 stops of dynamic range.. poppycock! more like 8 stops

Jonathan, you cannot issue such a searing statement without backing it up with evidence. We await your reply...
Logged

JonathanBenoit

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 414
Re: 8x10/MFDB Comparison
« Reply #129 on: September 04, 2010, 06:54:13 pm »

well, maybe not 8, but not 11 either. Prove to me that it's 11. It goes both ways.
Logged

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: 8x10/MFDB Comparison
« Reply #130 on: September 04, 2010, 07:09:12 pm »

well, maybe not 8, but not 11 either. Prove to me that it's 11. It goes both ways.

Correct, not 11 but 13.65  :D

Proof is here (click the dynamic range 'button'), unless you want to debate the proof ... Okay, I'll help you, engineering DR may be different from practical DR, but that's for you to substantiate, since it goes both ways.

Cheers,
Bart
« Last Edit: September 04, 2010, 07:49:17 pm by BartvanderWolf »
Logged
== If you do what you did, you'll get what you got. ==

JonathanBenoit

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 414
Re: 8x10/MFDB Comparison
« Reply #131 on: September 04, 2010, 07:32:25 pm »

DxO isn't proof. I'm not going to debate this.
They claim the d3x has more dynamic range than a phase one P65 ..
Logged

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: 8x10/MFDB Comparison
« Reply #132 on: September 04, 2010, 07:54:22 pm »

DxO isn't proof. I'm not going to debate this.
They claim the d3x has more dynamic range than a phase one P65 ..

Burying one's head in the sand never helps (unless when seeking a disconnect from reality) ... ;)

Cheers,
Bart
Logged
== If you do what you did, you'll get what you got. ==

JonathanBenoit

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 414
Re: 8x10/MFDB Comparison
« Reply #133 on: September 04, 2010, 08:01:33 pm »

Bart,

I am 100% certain that the dynamic range in files from a nikon are much lower than that of a full frame MF back, and even a back with less than 22mp. If you want to jump on dxo's marketing train, go for it. Do a proper test with real images. When people talk about dxo, I give them a link to dpreview. Here it is ... http://www.dpreview.com
Logged

feppe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2906
  • Oh this shows up in here!
    • Harri Jahkola Photography
Re: 8x10/MFDB Comparison
« Reply #134 on: September 04, 2010, 09:03:22 pm »

Bart,

I am 100% certain that the dynamic range in files from a nikon are much lower than that of a full frame MF back, and even a back with less than 22mp. If you want to jump on dxo's marketing train, go for it. Do a proper test with real images. When people talk about dxo, I give them a link to dpreview. Here it is ... http://www.dpreview.com

So I take it that we're supposed to take your word for it... because it's your word? And we're supposed to ignore DXOMark, despite the fact they they publish their methodology.

Seriously.

Bart_van_der_Wolf

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 8913
Re: 8x10/MFDB Comparison
« Reply #135 on: September 04, 2010, 09:38:48 pm »

Bart,

I am 100% certain that the dynamic range in files from a nikon are much lower than that of a full frame MF back, and even a back with less than 22mp.

Even a back with less than 22MP? Especially easy with such a back and appropriate low noise read-out circuitry.

Quote
If you want to jump on dxo's marketing train, go for it.

I don't want to, but it happens (?) to be quite well in line with what one can determine for oneself. I did a test for the lowely Canon 1Ds3 (I know, not MF but you are questioning the DxO stats without any proof), and found their tests to be closely aligned with my results (my 11.3 stops, vs DxO's 11.97 stops, within 2/3rds of a stop similar for 2 different copies of a model), so I have no reason to assume the rest is bogus, do you?

Quote
Do a proper test with real images.

I do all the time, and find the tests to be a reliable predictor of my own results.

Quote
When people talk about dxo, I give them a link to dpreview. Here it is ... http://www.dpreview.com

I'm not sure which part of DPR you are specifically referring to, but it's in general not a site I'd refer to for fundamental proof of anything but mass hysteria (although I do enjoy the contributions of a select few of the contributors).

Cheers,
Bart
Logged
== If you do what you did, you'll get what you got. ==

JonathanBenoit

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 414
Re: 8x10/MFDB Comparison
« Reply #136 on: September 04, 2010, 11:39:33 pm »

I am very close to deleting my luminous landscape account because of the previous two posters. Waste of my time.
Logged

PaulSchneider

  • Full Member
  • ***
  • Offline Offline
  • Posts: 115
Re: 8x10/MFDB Comparison
« Reply #137 on: September 05, 2010, 12:15:29 am »

kindergarten
Logged

deejjjaaaa

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1170
Re: 8x10/MFDB Comparison
« Reply #138 on: September 05, 2010, 01:43:49 am »

I am 100% certain that the dynamic range in files from a nikon are much lower than that of a full frame MF back

that will be no less than infamous "6 stops of DR"  ::) ?
Logged

feppe

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2906
  • Oh this shows up in here!
    • Harri Jahkola Photography
Re: 8x10/MFDB Comparison
« Reply #139 on: September 05, 2010, 04:45:27 am »

I am very close to deleting my luminous landscape account because of the previous two posters. Waste of my time.

"Extraordinary claims require extraordinary evidence"
- Carl Sagan

You're not the first one to claim incredible DR advantage for MFDBs. It seems you can't back your claims with verifiable data, and dismiss well-documented and falsifiable analysis which disagrees with your views with mere handwaving. That's just not good enough.
Pages: 1 ... 5 6 [7] 8   Go Up