Ad
Ad
Ad
Poll
Question: Dof and DB
Yes - 3 (75%)
- 1 (25%)
Total Voters: 4

Pages: [1]   Bottom of Page
Print
Author Topic: Hyperfocal and digital backs  (Read 1263 times)
esox
Full Member
***
Offline Offline

Posts: 117


« on: January 20, 2013, 07:21:03 PM »
ReplyReply

I'm looking forward to switch to digital MF (I'm on hassy with film now) and I'm a bit confused about circle of confusion, dof...

Let's speak about a 6 microns pixel digital back wich means a circle of confusion of about 10 microns. So we have at f8 (wich is the best opening usually regardons sharpness) an hyperfocal distance of ... 80m ! Wich means that in ordre to have sharpness to infinity an object has to be positionned at at least 40m... So everything closer than 40m won't be sharp. And also if one make fucus closerer than 80m there wońt be sharpness at infinite...

AM I wrong ?

If not, that is very difficult to be sharp at infinite no ?

But why choisi ne a cof of 0.010 when human eye cańt do Moore than 30l/mm wich is a cof of 0.03 ?
« Last Edit: January 20, 2013, 07:51:30 PM by esox » Logged
Graham Welland
Sr. Member
****
Offline Offline

Posts: 610


« Reply #1 on: January 20, 2013, 08:20:28 PM »
ReplyReply

You need to decide what you will be looking at. 100% pixel peep on screen or images printed on paper seen from 'normal' distances ... those choices will determine what real CoC you will use in your DoF calculations.

If you do a search on this forum you'll see literally reams & reams of DoF/CoC discussions, arguments and hissy fits.
« Last Edit: January 20, 2013, 10:30:42 PM by Graham Welland » Logged

Graham
FredBGG
Sr. Member
****
Offline Offline

Posts: 1651


« Reply #2 on: January 20, 2013, 08:52:13 PM »
ReplyReply

The circle of confusion is more dependent on lens and aperture than film or digital of about the same capture format size.

However digital will give you more enlarging ability compared to most film for overall detail.

However one thing to be considered is a thing known as focus buzz. When shooting very shallow depth of field
because film is thicker than the active part of the digital sensor film can be a bit more forgiving.

When it comes to landscape and larger depth of field I would say that an important factor to consider is tilt shift ability.
Tilt will give you more focus control.

As for what circle of confusion is acceptable is more dependent on enlargement size, viewing distance, subject matter and viewer subjectivity.

Something you can also explore with digital is focus stacking. A technique where several images with different focus distance
are stacked together and blended to produce an artificially high depth of field.

Also another thing that is important in my opinion is how pleasant and unobtrusive slightly out of focus elements are.
One thing to keep in mind is how fewer aperture iris blades effect this. Fewer blades tend to produce choppy or chatter of sorts in the transition to out of focus
when stopped down. Sometimes this chatter can clash with subject texture. In general its best to go for lenses with more iris blades.
Logged
Graham Welland
Sr. Member
****
Offline Offline

Posts: 610


« Reply #3 on: January 20, 2013, 10:30:09 PM »
ReplyReply

You might find this thread of interest as it gets to the point pretty quickly.
http://www.getdpi.com/forum/medium-format-systems-digital-backs/35964-what-c-c-use-iq180-dof-calculation.html

Determine your own acceptable hyperfocal settings based on trial (as Doug describes in the thread above) and use that to decide acceptable infinity / near focus ranges and use those. If you are going to use your monitor at 100% pixel peeping then you're going to be setting yourself up with a very small CoC if approximately 1.4 - 2x sensor pitch and worrying that this results in a much smaller DoF than you are used to. Make some large prints and you'll get a lot more relaxed about what you see on screen vs what you can ACTUALLY see on paper.
Logged

Graham
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7333


WWW
« Reply #4 on: January 20, 2013, 11:05:35 PM »
ReplyReply

Hi,

All that depends on how much you enlarge.

Sharpening plays an important role.

This page demonstrates effect of defocus and diffraction: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/49-dof-in-digital-pictures?start=1

And here I look a little bit into sharpening: http://echophoto.dnsalias.net/ekr/index.php/photoarticles/49-dof-in-digital-pictures?start=2

Best regards
Erik


I'm looking forward to switch to digital MF (I'm on hassy with film now) and I'm a bit confused about circle of confusion, dof...

Let's speak about a 6 microns pixel digital back wich means a circle of confusion of about 10 microns. So we have at f8 (wich is the best opening usually regardons sharpness) an hyperfocal distance of ... 80m ! Wich means that in ordre to have sharpness to infinity an object has to be positionned at at least 40m... So everything closer than 40m won't be sharp. And also if one make fucus closerer than 80m there wońt be sharpness at infinite...

AM I wrong ?

If not, that is very difficult to be sharp at infinite no ?

But why choisi ne a cof of 0.010 when human eye cańt do Moore than 30l/mm wich is a cof of 0.03 ?
Logged

Gel
Full Member
***
Offline Offline

Posts: 159


Excuse me while I bust a cap


« Reply #5 on: January 21, 2013, 04:26:54 AM »
ReplyReply

When I can't be bothered, and there's enough light it's a case of 80mm lens, TV mode, Manual focus to infinity, F/8

Point and shoot.
Logged

Chris Giles Photography
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3585


« Reply #6 on: January 21, 2013, 05:23:51 AM »
ReplyReply

I'm looking forward to switch to digital MF (I'm on hassy with film now) and I'm a bit confused about circle of confusion, dof...

Let's speak about a 6 microns pixel digital back wich means a circle of confusion of about 10 microns. So we have at f8 (wich is the best opening usually regardons sharpness) an hyperfocal distance of ... 80m ! Wich means that in ordre to have sharpness to infinity an object has to be positionned at at least 40m... So everything closer than 40m won't be sharp. And also if one make fucus closerer than 80m there wońt be sharpness at infinite...

AM I wrong ?

No, you are correct for images that need uncompromized sharpness.

Quote
But why choisi ne a cof of 0.010 when human eye cańt do Moore than 30l/mm wich is a cof of 0.03 ?

That means that you can magnify the image from sensor size to something larger for viewing. BTW, human vision cannot resolve features as small as you suggest. Normal vision resolves features in the order of 10-16 pixels/mm (0.1 - 0.0625 mm), which means that you can magnify some 6 - 10 times, when you need that uncompromized sharpness (there will be no visible difference in sharpness between the focal plane and the edges of the DOF zone). However, sharpness doesn't suddenly stop at the edge of the hyperfocal distance, it gradually starts to fall off, and when you increase your viewing distance more than normal reading distance, nearer objects will also appear to remain sharp to our vision.

You should choose a CoC criterion, also based on the final viewing distance. When you view large format output from e.g. 4x the normal reading distance, you can also multiply your critical CoC of 0.010 to 0.040 mm.

Cheers,
Bart
« Last Edit: January 21, 2013, 05:27:56 AM by BartvanderWolf » Logged
esox
Full Member
***
Offline Offline

Posts: 117


« Reply #7 on: January 21, 2013, 10:32:20 AM »
ReplyReply

Thanks a lot for your answers and sorry for the weird words that apeared in my message, I wrote it on the iPad and that guy not only speaks french but decides by himself to correct english words into the "nearest" french word... Kind of a circle of confusion in itself !

So this is what I thought. And if the picture is dedicated to big enlargemets one can choose a bigger coc (phonetically those last words sounds... well...  Grin). Do that mean that for a picture dedicated to screen viewing - i.e. let say 1200x800@96dpi - the coc would be 0.010 ?
Logged
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3585


« Reply #8 on: January 21, 2013, 12:01:24 PM »
ReplyReply

Do that mean that for a picture dedicated to screen viewing - i.e. let say 1200x800@96dpi - the coc would be 0.010 ?

Yes, if you look at the image at 100% zoom. If on the other hand you downsample the image to fit/fill full screen, then the CoC criterion to calculate the hyperfocal distance (HFD) increases with the same (reciprocal) factor as you downsample with (1/6x smaller image, 6x larger CoC for HFD calculation).

Cheers,
Bart
Logged
esox
Full Member
***
Offline Offline

Posts: 117


« Reply #9 on: January 21, 2013, 02:48:52 PM »
ReplyReply

Clear
Logged
jonathan.lipkin
Full Member
***
Offline Offline

Posts: 148



« Reply #10 on: January 21, 2013, 07:10:00 PM »
ReplyReply

I have a Hasselblad H3Dii39. I was told by Hasselblad that a rough COC of .04mm was the appropriate measure. Of course, a more nuanced answer would take into account print size and other factors above my pay grade.
Logged
Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad