Ad
Ad
Ad
Pages: [1] 2 »   Bottom of Page
Print
Author Topic: D800 mirror vibration - myth buster surprise  (Read 9821 times)
kramer11x
Newbie
*
Offline Offline

Posts: 7


« on: April 25, 2013, 07:39:01 PM »
ReplyReply

Since getting a D800 I have been chasing the mirror vibration loss of resolution.  Mostly shooting test charts under controlled studio continuous lighting conditions.   I have used many different tripod/head combinations trying to get the most rigid configuration assuming that stopping the camera from any motion was the solution.  Yesterday I was using a manfrotto 405 head with Hejnar adapter plate and large 80mm clamp on a 057 CF tripod. Assuming that this very rigid and mythically best dampening tripod system would improve the loss of resolution due to mirror movement.   Still the affects of camera motion were evident at 100% or more examination.  This is the most rigid combination I have access to now.  There was no expected improved dampening of the mirror vibration at slower (1/60 - 1/8) shutter speeds compared to smaller aluminum tripods with small ball heads.  For some reason my thought process changed to try something else to dampen rather than rigidly preventing the mirror induced camera motion.  I think reading an article once about how NASA handles camera dampening problems.  So I took off the massive rigid metal to metal Hejnar and reinstalled the standard 410PL ( RC-4) plate with a small 50mm clamp.  This is a typical manfrotto plate with the rubber pad which was between the plate and the mounted clamp.  The same much maligned rubber pads stated as the cause of camera movement in many forums for many years.  Indeed if you grab the camera mounted on one of these pads and apply torque you can feel the camera move.  BUT

Surprise.  There was a small but definite improvement in vibration dampening as evidenced by examination at 100-400% in photoshop.  So I continued to look at the best conditions for slow shutter speeds.  Another surprise.  The 0-1-2-3 second delayed release sequence showed that the 1 second delay was better than no delay as expected, but also better than the 2 and 3 second delays.  Much longer Mup delays were the best but only very slightly better than the 1 second delay and detectable only at 400%.

Bottom line for me.  Use 1 second delayed release on the D800 when on a tripod with or WITHOUT a cable release.

Since the D800 seems to be very susceptible to this mirror problem because of the higher resolution I was wondering if anyone else has seen the same thing.

NASA uses something called "sorbothane".  Thin pads to mount cameras and other instruments affected by vibration and it seems that their testing shows that it works.  And they have a much bigger budget than most of us.  Maybe they were on to something that has been ignored for years due to the negative rubber pad dogma.  Look in any large studio and there is a great deal of manfrotto and other camera mounting devices with some sort of rubber pad.  HMMMM.

My guess (I am not a physicist) is that it is not the mirror slap at the top of the movement that is setting up a vibration.  But rather the "equal and opposite" reaction of the camera body to the swing of the mirror.  Newton's third law for rotational systems.  And this reactive movement only lasts for the duration of the mirror movement and is dampened by the rubber mounting pad at the 1 second delay period.   Doesn't explain the 1 second delay being better than 2 and 3.  Possible something to do with the period of the mirror action, body reaction oscillations.

Would love to hear from others who have tested for this.  Any real experts out there?
Logged
MichaelEzra
Sr. Member
****
Offline Offline

Posts: 659



WWW
« Reply #1 on: April 25, 2013, 08:53:51 PM »
ReplyReply

If you are saying that it is not a slap but the rotational movement of the mirror that is responsible for camera shake, then it would be due to the law of conservation of the angular momentum, which would require camera to rotate in the direction opposite to the mirror.

The effect of the slap would be subject to the third law of Newton, just as in case of the opposing forces during dampening of camera movement.

One method to improve stability would be to make the camera extremely heavier than the (fast moving) mirror.
In this case the angle of camera rotation (and any other further displacement) would be decreased.

At some point the energy of all of these movements is transferred into heat.
Therefore, another option I see is to dissipate the energy of camera disturbance very quickly, much faster than the duration of exposure.
I am curious, whether material used by NASA is targeting this effect.
This would likely also help dampening due to the vertical movement of the shutter.
Logged

BernardLanguillier
Sr. Member
****
Online Online

Posts: 8343



WWW
« Reply #2 on: April 25, 2013, 09:01:21 PM »
ReplyReply

Any reason why you are not using mirror lock up?  Huh

Cheers,
Bernard
Logged

A few images online here!
kramer11x
Newbie
*
Offline Offline

Posts: 7


« Reply #3 on: April 25, 2013, 09:20:11 PM »
ReplyReply

If I understand your logic then the rubber pad could be acting like a spring, absorbing the energy of the camera motion, dissipating some as heat, and then returning the rest to the camera.  This would repeat cyclically until all the energy was dissipated.  If the period of the oscillation were just right this could explain why the 1 second delayed release was better than the 0, 2 and 3 second release delays.  And Mirror up with a 10+ second release was the best of all.  The real mystery to me was why the 1 is better than the 2 or 3 second delays.
Logged
MichaelEzra
Sr. Member
****
Offline Offline

Posts: 659



WWW
« Reply #4 on: April 25, 2013, 09:21:07 PM »
ReplyReply

For the many-frames panoramic stitching I find the 1-2 seconds delay much easier to use than the mirror lockup, especially when it is *cold*, as it was this January in the Grand Canyon -16F.
At some point even while using gloves, it is difficult enough to press the shutter button even once (that is, after you find it), and mirror lockup requires pressing it twice.
Disclaimer: I forgot the cable release at home on that trip:)
Logged

kramer11x
Newbie
*
Offline Offline

Posts: 7


« Reply #5 on: April 25, 2013, 09:21:58 PM »
ReplyReply

Mirror up worked as expected so I was not puzzled as I was by the 1 sec delayed release being better than the 0, 2, or 3.
Logged
Glenn NK
Sr. Member
****
Offline Offline

Posts: 308


« Reply #6 on: April 26, 2013, 12:49:32 AM »
ReplyReply

For what it's worth, it is common practice to insert 1/2 inch thick neoprene/rubber pads between a diesel-powered backup generator in a building and a concrete floor to reduce the transmission of vibrations to the floor (which results in audible sound).

These generators weigh anywhere from 2,000 to 10,000 lbs (905 kg to 4,599 kg); larger ones are usually located outdoors.

Glenn

« Last Edit: April 26, 2013, 12:55:07 AM by Glenn NK » Logged

Economics:  the study of achieving infinite growth with finite resources
MrSmith
Sr. Member
****
Offline Offline

Posts: 918



WWW
« Reply #7 on: April 26, 2013, 01:28:33 AM »
ReplyReply

Does the Nikon have a silent shutter setting?
Not done any analytical testing but I know with the 5dIII you feel less through your hands when shooting at slower speeds with long lenses and I guess this would translate to less mirror slap on a tripod.
Logged
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3865


« Reply #8 on: April 26, 2013, 01:56:27 AM »
ReplyReply

Since getting a D800 I have been chasing the mirror vibration loss of resolution.  Mostly shooting test charts under controlled studio continuous lighting conditions.   I have used many different tripod/head combinations trying to get the most rigid configuration assuming that stopping the camera from any motion was the solution.  Yesterday I was using a manfrotto 405 head with Hejnar adapter plate and large 80mm clamp on a 057 CF tripod. Assuming that this very rigid and mythically best dampening tripod system would improve the loss of resolution due to mirror movement.   Still the affects of camera motion were evident at 100% or more examination.

Hi,

Do you have any quantification (sample image / measured angular displacement over time) as to the severity of what you observed?

Cheers,
Bart
Logged
design_freak
Sr. Member
****
Offline Offline

Posts: 1074



« Reply #9 on: April 26, 2013, 04:09:40 AM »
ReplyReply

I think that this simple and ingenious idea could solve this problem:
http://www.youtube.com/watch?feature=player_embedded&v=KvV7XGqK8sc#!
POLAND TECHNOLOGY  Cool
http://epar.pl/index.php?pl_nagrody-dla-epar,106#.UXpEAytHBE8

Only if it's in anyone's interest? For example Car manufacturers do not want to protect your car, but at best the passengers ...
Do you see any manufacturer who has balls? And risked research in this direction?
Logged

Best regards,
DF

-------------------------------------------
WORK HARD AND BE NICE TO PEOPLE
-------------------------------------------
RFPhotography
Guest
« Reply #10 on: April 26, 2013, 07:46:54 AM »
ReplyReply

Still the affects of camera motion were evident at 100% or more examination. 

There was a small but definite improvement in vibration dampening as evidenced by examination at 100-400%

I know I'm going to get slammed for this, but I don't care.  What are the results like on real subjects viewed under 'normal' conditions?  How does it appear in a, for example, 16x20 print framed and viewed normally (i.e., not with the nose pressed up against the glass)?  Or even a 20x30 print? 

Quote
Does the Nikon have a silent shutter setting?

Yes, it does.
Logged
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3865


« Reply #11 on: April 26, 2013, 07:56:14 AM »
ReplyReply

What are the results like on real subjects viewed under 'normal' conditions?

Hi Bob,

For some people making huge enlargements is 'normal' conditions, and they most certainly want to avoid camera shake.

For those only interested in small size output, most of the cameras used by LuLa members are overkill (as far as number of Megapixels are concerned) anyway. They should concentrate on down-sampling correctly.

Cheers,
Bart
Logged
RFPhotography
Guest
« Reply #12 on: April 26, 2013, 10:23:21 AM »
ReplyReply

Very true, Bart.  I'd have the same question though.  If making big prints, how do they look under 'normal conditions'.  A 5' wide print isn't going to be viewed from 12".  Huge billboards often look like nothing but pixels up close but from a distance are fine. 
Logged
Kirk Gittings
Sr. Member
****
Offline Offline

Posts: 1551


WWW
« Reply #13 on: April 26, 2013, 10:43:21 AM »
ReplyReply

I know I'm going to get slammed for this, but I don't care.  What are the results like on real subjects viewed under 'normal' conditions?  How does it appear in a, for example, 16x20 print framed and viewed normally (i.e., not with the nose pressed up against the glass)?  Or even a 20x30 print?  

Yes, it does.

I have always found this a funny concept, which is why I mainly shoot 4x5 for my personal work. How do you prevent people from looking at your prints so close and having them look like crap? Or do you post a disclaimer?
Logged

Thanks,
Kirk

Kirk Gittings
Architecture and Landscape Photography
WWW.GITTINGSPHOTO.COM

LIGHT+SPACE+STRUCTURE (blog)
RFPhotography
Guest
« Reply #14 on: April 26, 2013, 12:24:35 PM »
ReplyReply

Kirk, the implication that anything other than made with a 4x5 will look like crap up close is a bit silly. 

I don't 'prevent' people from looking closely.  But the simple fact is that most people don't.  The simple fact is that the concept of 'normal viewing distance' has developed for a reason. 

I ask about 'real' images because test charts at 300% or 400% aren't overly indicative of much.  People in images move.  Trees, leaves, grass, move.  At 300% or 400% even a tiny amount of movement will likely be picked up.  Will it be evident in a print?  Even quite a large print?  Call me crazy but I'm one of those photographers who still believe that the ultimate presentation of a photograph is a print, not something on a computer screen. 

Blow a film image up far enough and you get nothing but a mess of grain up close, even at reasonable viewing distances in some cases.  We didn't object to it then the way we do now.  Medium format cameras have far bigger mirrors and much more significant mirror slap than a 35mm format DSLR. 
Logged
theguywitha645d
Sr. Member
****
Offline Offline

Posts: 970


« Reply #15 on: April 26, 2013, 02:27:29 PM »
ReplyReply

Funny, I shoot with a 645D and D800E. I am not having any problems with vibration caused by the mirror. I am not sure if you have to go to a 400% viewing condition you are actually seeing mirror vibration.

BTW, if the images is sharp at 100% with a D800, 40" wide print should remain acceptably sharp to viewer at a viewing distance of about 9", assuming you can focus that close.
Logged
David Watson
Sr. Member
****
Offline Offline

Posts: 396


WWW
« Reply #16 on: April 26, 2013, 02:58:29 PM »
ReplyReply

Since getting a D800 I have been chasing the mirror vibration loss of resolution.  Mostly shooting test charts under controlled studio continuous lighting conditions.   I have used many different tripod/head combinations trying to get the most rigid configuration assuming that stopping the camera from any motion was the solution.  Yesterday I was using a manfrotto 405 head with Hejnar adapter plate and large 80mm clamp on a 057 CF tripod. Assuming that this very rigid and mythically best dampening tripod system would improve the loss of resolution due to mirror movement.   Still the affects of camera motion were evident at 100% or more examination.  This is the most rigid combination I have access to now.  There was no expected improved dampening of the mirror vibration at slower (1/60 - 1/8) shutter speeds compared to smaller aluminum tripods with small ball heads.  For some reason my thought process changed to try something else to dampen rather than rigidly preventing the mirror induced camera motion.  I think reading an article once about how NASA handles camera dampening problems.  So I took off the massive rigid metal to metal Hejnar and reinstalled the standard 410PL ( RC-4) plate with a small 50mm clamp.  This is a typical manfrotto plate with the rubber pad which was between the plate and the mounted clamp.  The same much maligned rubber pads stated as the cause of camera movement in many forums for many years.  Indeed if you grab the camera mounted on one of these pads and apply torque you can feel the camera move.  BUT

Surprise.  There was a small but definite improvement in vibration dampening as evidenced by examination at 100-400% in photoshop.  So I continued to look at the best conditions for slow shutter speeds.  Another surprise.  The 0-1-2-3 second delayed release sequence showed that the 1 second delay was better than no delay as expected, but also better than the 2 and 3 second delays.  Much longer Mup delays were the best but only very slightly better than the 1 second delay and detectable only at 400%.

Bottom line for me.  Use 1 second delayed release on the D800 when on a tripod with or WITHOUT a cable release.

Since the D800 seems to be very susceptible to this mirror problem because of the higher resolution I was wondering if anyone else has seen the same thing.

NASA uses something called "sorbothane".  Thin pads to mount cameras and other instruments affected by vibration and it seems that their testing shows that it works.  And they have a much bigger budget than most of us.  Maybe they were on to something that has been ignored for years due to the negative rubber pad dogma.  Look in any large studio and there is a great deal of manfrotto and other camera mounting devices with some sort of rubber pad.  HMMMM.

My guess (I am not a physicist) is that it is not the mirror slap at the top of the movement that is setting up a vibration.  But rather the "equal and opposite" reaction of the camera body to the swing of the mirror.  Newton's third law for rotational systems.  And this reactive movement only lasts for the duration of the mirror movement and is dampened by the rubber mounting pad at the 1 second delay period.   Doesn't explain the 1 second delay being better than 2 and 3.  Possible something to do with the period of the mirror action, body reaction oscillations.

Would love to hear from others who have tested for this.  Any real experts out there?

I have a D800E which I use hand held with either a 35mm f1.4 or an 85mm f1.4.  If I want ultimate print quality however I use a Hasselblad H4D-60 with its very nice leaf shutter lenses and mirror up tethered - no problem.  Oh I guess MFD isn't dead after all.

LOL
« Last Edit: April 28, 2013, 11:43:07 AM by David Watson » Logged

David Watson ARPS
mcbroomf
Sr. Member
****
Offline Offline

Posts: 423


WWW
« Reply #17 on: April 26, 2013, 05:37:45 PM »
ReplyReply

Mirror up worked as expected so I was not puzzled as I was by the 1 sec delayed release being better than the 0, 2, or 3.
If you use liveview in combo with MLU then the mirror is up before you start the exposure
Logged

Mike Broomfield
Website
leeonmaui
Full Member
***
Offline Offline

Posts: 213


« Reply #18 on: April 27, 2013, 12:37:47 AM »
ReplyReply

Aloha,

I have had similar problems when butterflies are around, the vibration caused by their wings as they move through the air plays havoc with the overall sharpness of my images...
Logged
elf
Full Member
***
Offline Offline

Posts: 232


« Reply #19 on: April 27, 2013, 01:06:40 AM »
ReplyReply

Sorbothane works well for dampening vibrations, but it needs to be tuned to the object needing the dampening.  It also won't dampen all frequencies, so it may or may not help with shutter vibration.  My studio setup for macro has the camera sitting on an 80 pound block of granite which in turn sits on 4 sorbothane feet which are matched to the weight.  Even more effective was adding a 5 pound sandbag on top of the camera.  If you really want to eliminate shutter vibration get a Canon as they have an electronic first curtain shutter Roll Eyes
Logged
Pages: [1] 2 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad