.jpg)
If you own the A7R and would like to ask Sony to review the issue and consider a firmware update that allows for an optional 2 second (approx.) delay after the first curtain closing then please drop them an email:[email protected]
PART I (12/21/13) (Part II below) Here is the discussion of this test on DP Review: http://www.dpreview.com/forums/post/52752354
I decided to test the vibration caused by the shutter release of the A7R, D3 and NEX-7. I used the app called Iseismometer. It’s not as scientific as lasers in the lab but it will have to work due to budget restraints. The Iseismometer app is so sensitive it will respond to a pencil drop 5 feet away on a kitchen table. I’m impressed by the app though an update would be nice. I didn’t set out to get quantitative values my goal was to compare one graphic with another and hopefully get some idea on how the A7R compares to a DSLR and the NEX-7.
The 3 axes oriented in on the mobile phone. Image from Iseismometer.com
Method of test
The iphone was centered on the hot shoe, attached with rubber bands facing up. The X-axis placed in line with the lens, the Y-axis left to right and the Z-axis up and down.
The camera placed on sturdy desk with a tripod plate mounted.
10sec timer
Tested 1/2 second shutter and 1/60 second shutter
10 or more tests were done for each camera. The setup was the same and the location on the desk was the same for all the tests. In reviewing the data I threw out the single worst vibration result for each camera.
.jpg)
The A7R and the D3. Both 1/2 second shutters.
Interpretation of Data
A7R – Have a look at the Z-axis. The vibration begins with the shutter closing, small peaks, then the first long downward peak that extends off the margin is the shutter impact as it fully opens. That is the beginning of the exposure. The X, Y and Z axes, of the 1/2 second exposure consistently show vibrations at the beginning of the exposure then tapering off. This is the critical moment when you would like vibrations to be as close to zero as possible. About half-way through the exposure (1/4 second), vibrations in the A7R settled to near background noise. As expected with the design of the A7R shutter the vibrations are longer in duration than the D3 and NEX-7. As for overall intensity of vibration it appears the A7R and D3 are close to the same, perhaps the A7R is slightly worse (based on the Y-Axis).
D3 – The D3 is a camera with significant mass. My gut says that given two cameras with the same exact mirror slap the camera with greater mass will experience less vibration. However, I’ll leave that for a physics person to prove. One significant finding was during the exposure, the D3 never settles down to background noise. Vibrations are carried through the entire 1/2 second shutter. This is especially evident in the X and Y axes. Perhaps the D3 is not as effective in dissipating the vibration, maybe a momentum thing.
.jpg)
The A7R compared to the NEX-7. Both 1/2 second shutters. NEX-7 Electronic First Curtain is OFF.
NEX-7 – There is no dispute that the NEX-7 wins when it comes to low shutter vibration. The duration is short, the intensity is low and it settles to background noise very quickly during the exposure. The fact that the NEX-7 has the smaller APS-C sensor would bias the comparison because the shutter is presumably smaller. I would imagine the A7 would be a close challenger to the NEX-7. I wish I had one to test – if you have one and live in Maryland USA give me a shout.
Here are all three graphics for 1/60 second shutters. As you can see the A7R has the most vibration in the Z-axis. Probably because the shutter moves up/down whereas the D3 mirror has angular momentum as it flips out of the way.
The D3 with 1/2 second shutter and Mirror Lock-up.
I entered into this experiment hoping I would be dazzled by something. Well I am. When Mirror Lock-up is used in the D3 the results are impressive. Notice that during the 1/2 second exposure period vibration is no more than ground noise. This performance most likely applies to all DSLR’s with Mirror Lock-Up. Hopefully, this article will get you back on track and using Mirror Lock-up.
A7R hand held with 1/30 second shutter.
So I went from being dazzled to being humbled. Good handheld photography takes practice. With shear determination and concentration I was able to improve with each shot. My wife gave it a try before the wine and she did very well, even better than I did. The X-axis reports movement toward and away from the subject. The Y and Z axes are left to right and up and down respectively. The X-axis reports the most movement, it is the hardest to control because the eye cannot see the fore/aft movement as efficiently.
Left graphic A7R is resting on desktop with tripod plate mounted. Right graphic A7R is mounted on tripod, resting on concrete. 1/60 second shutter.
Mounting the A7R on the tripod didn’t dampen the vibrations as I expected. The tripod is carbon fiber with a Really Right Stuff ball head. It appears the Z-axis vertical vibrations are dampened but the left to right vibrations along the Y-axis increase. I think this outcome is realistic given the design of a tripod.
Summary
Duration of vibration is greatest for the A7R over the D3 and NEX-7. This confirms what we already know; the shutter has to close, open, close and then open.
Intensity of vibration in the A7R is about the same as a DSLR but the A7R dissipates the vibration faster than the D3. I am pleased to see that the A7R is not off-the-charts in shutter vibration. Keep in mind this test analyzes the vibration at the hotshoe on the outside of the camera and not the actual sensor. This test is much more accurate than asking my best buddy which one feels like it vibrates the most. Using rubber bands to fasten the iPhone to the camera is admittedly somewhat crude.
PART II (12/23/13)
A possible way to test shutter vibration at the sensor level which ultimately results in image quality is to eliminate the variables by taking an image with shutter vibration and then compare it to an image taken without shutter vibration. Then one could discern how much shutter vibration they are getting at the sensor.
To take the image with NO shutter vibration set the camera up in a dark studio and open the shutter for say 5-10seconds. While the shutter is open and vibration has dampened, fire an off camera strobe(s) to expose the image. This exposure will have no vibration from a shutter release. Then without changing any camera settings except shutter speed, take an image normally using hot lights and compare the two images.
Sony A7R mounted with SEL35mmFE28
Manual focus using 14.4 magnification
All camera settings stay the same except for shutter speed.
Only white balance adjusted, all other settings are out of camera.
Hot lights and strobes are pre-set to give the same exposure.
Camera on tripod is not moved during all images. Really Right Stuff, Carbon Fiber tripod, Really Right Stuff ball head. BTW this is not a rinky-dink tripod, the setup sells for about $1500.
Total of 16 images taken, 8 with strobe and 8 with hot lights.
Sony SEL35mmFE28 @ f/5.6. Out of camera except for white balance adjustment.
The following two images are 200% crops. The first image was taken with a 10 second shutter speed @f/5.6 ISO100. While the shutter was open (about 7 seconds) two strobes were fired exposing the image with no shutter vibration. The second image was taken with hot lights, it is a normal exposure with shutter vibration 1/10 second @f/5.6 ISO100.
In addition to the image taken at 1/10 second I also took images at 1/3, 1/5, 1/20, 1/40, 1/80, 1/160, 1/320. ISO was adjusted by a stop for each to maintain exposure. I wanted to see if there was a particular shutter speed that caused a blurrier image than the others. I could not see any difference in image sharpness based on shutter speed variation.
Strobe exposure no shutter vibration, 200% crop.
Hot light exposure, 1/10sec ISO 100, normal exposure with shutter vibration.
Table Top Resting
How much vibration do you get when your A7R is placed on a firm table top with a 10 second timer? Answer: There is no discernible difference with the 35mm FE 2.8. Compare these images at 400%: No shutter vibration (top), shutter vibration on tripod (bottom L) and shutter vibration (bottom R).
Adobe Camera RAW window with magnifier at 400%. Sony 35mm FE 2.8
Sony A7R mounted with Nikon 85mm f1.4 and Promaster adapter
Manual focus using 14.4 magnification
All camera settings stay the same except for shutter speed.
Only white balance adjusted, all other settings are out of camera.
Hot lights and strobes are pre-set to give the same exposure.
Camera on tripod is not moved during all images. Same Really Right Stuff set up.
Total of 16 images taken, 8 with strobe and 8 with hot lights.The following two images are with the Nikon 85mm f/1.4 and 200% crops. The first image was taken with a 10 second shutter speed @f/5.6 ISO100. While the shutter was open (about 7 seconds) two strobes were fired exposing the image with no shutter vibration. The second image was taken with hot lights, it is a normal exposure with shutter vibration 1/10 second @f/5.6 ISO100.
Shutter opened 10 seconds, dark room, two strobes fired, image created with no shutter vibration. 100% crop
Hot light exposure, 1/10sec ISO 100, normal exposure with shutter vibration. 100% crop
I would have expected the image taken with hot lights with the effects of shutter vibration to be blurry compared to the first image. I’m pleased to see in these two examples, 35mm and 85mm, there is no noticeable blur caused by shutter vibration. The 85mm with promaster adapter is quite heavy, the lens and adapter extend outward and a light touch of the lens shows movement through the viewfinder. I would have guessed that the vibration would have been more evident than in the 35mm.
Below is the lightbox of the various shutter speeds tested with the Nikon 85mm 1.4. Some suspect in other studies that around 1/100 sec has the worst shutter vibration. You be the judge.
An additional comparison above of three images, the strobe image (top) and two normal exposures with shutter vibration (bottom), one on a tripod mount (left) and one on a table top (right). The table top image is something similar to what you’d get if you placed your A7R on a firm table with a 10 second timer and allowed it to take an image. Even resting on the table there is no obvious shutter vibration captured in the image. The magnifier is 400%.
ZOOM LENS Nikon 70-200mm f2.8 on Sony A7R (added 1/4/2014)
The A7R performs well at lower zooms, this section deals with the A7R at higher zooms, 200mm and compares it with the D3 (DSLR).
Cameras: Sony A7R and Nikon D3
Lens: Nikon 70-200mm f/2.8 ED VR, VR-off
Adapters for Sony: Promaster $80 and Novoflex $293
Aperture: f/5.6 for all images
Zoom: 200mm
Shutter Trigger: 10 sec. self timer. Mirror Lock Up (MLU) with cable release
Sensor Orientation: Horizontal (Landscape) and Vertical (Portrait)
No Shutter Shock Images (No SS): 10 sec. shutter in dark room with strobes fired at 7 second mark (allows vibrations to settle.
Tripod and ball head: Really Right Stuff, carbon fiber.
Attachment to tripod: Original tripod collar on Nikon 70-200mm, table top resting had no attachments, 2 points touching – camera base and lens front.
Object Distance: 3m
A7R attached to 70-200mm f/2.8 on a Really Right Stuff Tripod.
Out of Camera image at 200mm.
To view the below image at full size 2000 x 2000 pixels click HERE
To view the image above at full size 2000 x 2000 pixels click HERE
Observations:
Horizontal vs. Vertical orientation
With both cameras the horizontal orientation (landscape) is sharper across all shutter speeds than the vertical orientation (portrait). The blur when the camera is in vertical orientation is most likely due to rotational energy around the post of the ball head and the tripods inability to resist rotational energy.
Shutter vibration is more pronounced with the D3 lending evidence to greater mass of the mirror slap and the vulnerability of the midway coupling of the camera to the tripod.
Promaster vs. Novoflex
Keep in mind the Promaster is $80 and the Novoflex is $293. With respect to vibration control the Novoflex is slightly better than the inexpensive Promaster adapter. Is it worth an additional $213, I would say no if it strains your budget. However, there may be other differences that make the Novoflex more desirable.
Table Top Resting
Table Top Resting setup. The front ball head is one I made in my machine shop.
The Table Top Resting images from the A7R are exceptional. On a quick glance they would be hard to distinguish from the No SS images. They illustrates that the midway coupling of the lens/camera mass to the tripod is not a great design. The midway coupling design is more driven by convenience, quickness of attachment, and maneuverability than camera vibration.
The set up for Table Top Resting images consisted of two tripods, one under the camera base without the camera attached and the other placed under the front of the lens. Consider the analogy of holding a barbell above your head with two hands versus holding the barbell with one hand in the middle.
I didn’t test a bean bag as a resting surface but I would venture to say the results would be as good as Table Top Resting.
D3 with No Mirror Lock Up
The D3 images with mirror slap are unusable. At shutter speeds below 1/100 second it’s a pixel peepers nightmare. As confirmed with the A7R the vertical orientation (portrait) is worse. You can even detect ghosting in the slowest shutter speeds, it indicates a pendulum effect where the camera pauses as it reaches the pinnacle then returns to the resting spot.
It might be interesting to perform Table Top Resting for the D3 to see how much the images are improved.
D3 with Mirror Lock Up (MLU)
Using Mirror Lock Up yields results that rival No SS images at all shutter speeds. When comparing with No MLU it is quite impressive to see how much mirror slap degrades image quality.
A7R vs. D3
Both cameras exhibit shutter shock but the D3, as in many DSLR’s, has Mirror Lock Up. Comparison of the A7R images with the MLU images shows the flaw in the A7R design. In order for the A7R to compete with Mirror Lock Up, it must offer the first curtain closure with a delay before exposure begins.
Until then A7R owners should:
*avoid midway coupling of the lens to the tripod
*use a camera rail with a wedge under the lens for support: http://blog.kasson.com/?p=4558
If a midway coupling to the tripod is necessary then owners should:
*understand that horizontal orientation is better
*stay above 1/100 second shutters
[…] […]
[…] an app that provides three-axis accelerometer readouts using the device built into the phone. https://beforethecoffee.com/sony-a7r-vibration-comparison-with-nikon-d3-and-sony-nex-7/ Although such studies are useful in identifying resonant frequencies and damping factors, and […]
Doing this test with something like a GX-7 would be interesting in comparison as it has a fully electronic shutter option. Other tests have shown that around 1/100 is the worst case scenario for the A7R so that might be another good test.
[…] bonuses create yet another advantage: no added vibrations from extra moving parts. My hands aren’t rock solid steady, so I’m usually limited to […]
I have the nex7, and received the a7r about two weeks ago. Had my reservations on looking at studio images, noticed the blur. Thank you for your article, it convinced me to return the A7r. Disappointing…
[…] The shutter vibration measured by Beforethecoffee […]
Thanks for sharing the data from your testing, much appreciated, i haven’t noticed any issues with my A7r and vibration.
Dear Ferrell,
Thank you for running some interesting and revealing tests. From our experience, it is quite clear, that if your lens were the lighter 70-200 f4 Nikkor, if you also turned your camera vertical, and if you connected the assembly to the tripod via a lens foot (on a collar), and again tested at 1/100th of a second, as we did, then you would not only see vibration quite clearly, but it would almost certainly be striking and quite ruinous to the image quality.
We too have seen many cases where there is either no discernible vibration or so little as to be just a trace which will naturally go unnoticed. The task of a good experimenter is to find out which combinations of lenses and adapters and camera orientations and shutter speeds (etc.) work well and which do not, and when there is an issue, how big it is, and what, if anything can be done about it. Of course, each of us can only test with our own equipment and we’re also naturally most interested in knowing how our own equipment will do, so the old story of the blind men and the elephant is, as usual for these circumstances, very much in play.
If you read my guest article on SonyAlphaRumors carefully http://www.sonyalpharumors.com/the-shutter-vibration-issue-explained-by-joseph-holmes/, you will see that I said absolutely nothing which contradicts your experimental observations. You simply have been lucky to avoid the setups which do reveal the shutter vibration problem quite plainly in captured images. If you were able to read Lloyd Chambers’ Guide to Mirrorless subscription pages, you would see many carefully and thoroughly documented and illustrated examples of A7R images completely ruined by shutter shake, using some of the world’s finest photographic gear, especially some longer Leica R lenses, as well as other results which are either unaffected or only very slightly affected.
In fact, your results with the Nikkor 70-200 f/2.8 at 1/100th and 200 mm and the camera in horizontal position, with the assembly connected via the Mirex foot, do look a little better than I would have guessed that they might, which is great. I do see the tiniest bit of vertical movement in the resulting images, just barely enough to make out shutter shake. The problematic results which we have seen vary up to a great many times worse than this.
If you were mounting your 70-200 lens directly to the camera and connecting the camera directly to the tripod (if that were feasible), you would see even less shutter shake than you did, which would mean none at all, though something like wind would have a much bigger effect on the image by more readily pushing the long lens around. These systems have many variables, and we have taken the time to at least get a rough sense of how each variable contributes to the problem, and we did so before we undertook to make any broad claims about the nature of the elephant.
The A7R most definitely has a serious problem with shutter shake, but the shutter’s shaking has essentially no effect on a great many lenses under a great many circumstances. Rather, as I carefully explained in my article, it is a problem for other combinations of lenses and circumstances: Longer lenses, softer tripod connections by virtue of being further forward of the sensor, vertical worse than horizontal, and so on. Don’t doubt that this problem is very real. If this problem had affected many shorter lens and natively mounted lenses and so on, the camera would never have shipped in the first place. The good news is that attaching a dead weight to the base of the camera of about 24 ounces so far has been seen to reduce shake in the image by over 90% in every test result I have seen so far. So if Sony does put in the pause between the two movements of the first curtain in a usable way, and we have the weight for those times when it’s needed, we may have a perfectly workable solution, despite not having the elegance of electronic first curtain.
Sincerely,
Joseph Holmes
Dear Ferrell:
Thanks for sharing your results. Below is a response I posted on your article on the SAR web site late last night. I thought your readers might like to see it as well. I don’t doubt your results and your work was well thought out. Here is what I wrote on SAR.
“I have been working with Joe Holmes over the past 2 weeks on the shutter vibration issue. I am the guy who supplied the images showing the shutter shake in Joe’s article. ( and no I did not kick the tripod!). Today I spent 4 hours inside an aircraft hanger (no wind, no shake) testing my A7r with a wide variety of lenses and adapters. The reality is that the vibration is dependent upon a number of variables: mass of lens, orientation of camera body, how you anchor the camera to the tripod( I.e.camera body, lens ring, adapter foot), and finally shutter speed. It is not simple. I would encourage each person to test their various set ups, lenses, and a variety of shutter speeds. Is there shutter shake…. YES! But not noticeable in many circumstances, and very detrimental in others. I have plenty of well done tests to show it is so, and so do many others. But I know which of my lenses/adapter combinations are problematic, and at what shutter speeds. The counter weight on the bottom of the camera is a must in certain situations. For the Canon TSE 17 and 24mm with a Metabones attached to the tripod head via the Metabones foot….the images are excellent at any shutter speed ( in both vertical and horizontal orientation)… If you only saw those tests you would say ” shutter shake is not an issue”. As a professional photographer I demand the most out of my equipment. I am often called to print 40 x 50 inch prints and I want to deliver the best I can. It is part of what makes me a professional. Each person has to determine their threshold of acceptability.
I think some readers need to re-read Joe’s well written article. I know Joe personally and he is one of the most meticulous and knowledgable people I have meet in 40 years of making images. Do I think there is a fundamental problem with this camera design? Yes. Am I going to send the camera back? No. Why? Because I have spent the time and effort to know what it is capable of and how I will use it. Am I selling my 800e, or P45? No, they each serve a specific purpose. I think the A7r is a good camera that needs some improvement”
Thanks for the interesting results! They weren’t surprising though. Going to try attaching the dead weight at the bottom using a 300mm adapted Nikon lens if it trully does fix the problem. It is though a very light camera. Definatly not made for any huge telephoto lens. Portability does have a downside. I wonder how much the battery grip helps towards this problem. Seeing it’s only 8.8oz, maybe not so much.
Just realized it you use the battery grip with two batteries, problem solved. Clever Sony to do that. Without the battery grip though any small lens (18-55mm) isn’t going to be that huge of a problem.
I would be careful about attaching a dead weight to the bottom of the camera. It could add unnecessary stress to the camera mount. Also, I would want to be sure that I’m doing it because it corrects a camera problem and not an adapter, tripod mount or collar problem. We are talking about shutter vibrations that are very small. I doubt the tolerances of tripod collars, mounts and adapters are good enough to eliminate them as problematic. Just put you entire rig together, zoom in and touch the front of the lens, what happens. Yea it moves, something is giving? What is it and why? – Ferrell
[…] [Merged] Ad I don't know if this has been posted before but here you are anyway…. Sony A7R vibration comparison with Nikon D3 and Sony NEX-7 Play nicely […]
Thanks Ferrell for doing some extensive testing. I shot with the Sony A7R and haven’t noticed any problems with shutter vibration.
I have produced consistently sharp results using the Sigma 105mm F2.8 EX DG OS HSM Macro Lens with the LAEA4 Sony A-E mount adapter.
Dav – Glad to hear your results but some are getting less sharp results which are most likely due to “rigging” of the camera through adapters and collars. -Ferrell
The mass of the iPhone will modify the vibrationpattern, so I would not base to much on those measurements. (Modify: Improve/worsen, it can go both ways).
However you comparison of strobe-exposed pictures and normal shutter-exposed is nice. Thank you very much for the effort.
Best wishes for 2014
I agree Claus, it is not hard core science. However, it does support what we know about the mechanics of the A7R shutter, the EFCS (nex-7) and Mirror lock up etc. Also, it does support what we “feel” when we are holding the camera and the shutter is released. -Ferrell
I would guess that the next version A7R will have Electronic front curtain shutter. I’m sure some Sony engineers are reading this and scratching their head ” we got to fix this ”
The present owners of the A7R are hoping for a firmware update that adds the option for a delay after the first curtain closing. -Ferrell
You did a better job of studying and presenting the vibration issue on A7R than Diglloyd. That is why he is attacking your work. As an engineer, I think your graphs and photo comparisons worth 1000 times the Diglloyd’s nice but empty websites and photos.
Keep up the good work.
Thank you Tuong. I added Sony’s email at the top to voice concerns about the shutter vibration. Sony: [email protected] -Ferrell
Grat information – thank you for the post!
On many of my cameras including the D800E, NEX6, Mamiya 7II I am able to hand-hold down to 1/15sec with consistent results. On the A7r I struggle even at 1/100sec with any of the 50mm lenses I tested (55 1.8, Leica lux 50, Nokton 50, Summitar 50, etc). At times, I see a ghost double image that has the characteristics of the pendulum effect described in the above article. Do you or anyone else have better results hand holding the A7 camera?
I never bother to carry a tripod but I use good technique to get the best results possible while hand holding. I would switch to lower resolution any time of the day if I can get the same results I get with all the other cameras I use.
BTW, I am not sure why diglloyd would have an issue with this article. I find it very complete and interesting – keep them coming…
Hi,
Thank you for stopping by and I’d say 1/15 sec handholding is impressive. The A7R exposure is further along (time wise) in the vibration/sound than other shutters so I wonder if YOUR “steady moment” is out of sync with the A7R. Just a thought on finding an explanation for struggles at 50mm. For some high-tech reading about the A7R shutter Jim Kasson is doing some interesting research.
http://blog.kasson.com/?p=4459
-Ferrell
@Mike Shultz
@Joseph Holmes
Maybe try putting a “Black Rapid FastenR” in you A7r’s 1/4 20. Then source a small bungy cord and fasten it between the FastenR and tripod leg (rather than lugging weights around). Filmmakers are well aware of vibration issues caused by a cameras film transport system. A filmmaker will hang a lens off of the cameras lens mount to avoid just these vibration issues (even with digital). Sony allude to this mounting method in their description of the internal design of the A7r. I personally wouldn’t have an issue hanging the Canon 70-200 f/2.8 off a tripod mounted A7r body.
Ok, shooting as a filmaker negates the portrait orientation of the camera but, aren’t we
[…] over a wide range of focal lengths and shutter speeds. I was by no means alone; Lloyd Chambers, Ferrell McCullough and many others also discovered and documented the problem. I spent a lot of time trying to pin it […]
[…] Forum Article Joseph Holmes Article To complete your review, you might also read the article at Before the Coffee, which refutes the claim that such a problem exists at […]
[…] https://beforethecoffee.com/sony-a7r-vibration-comparison-with-nikon-d3-and-sony-nex-7/ […]
I work around the problem by using shutter priority(1/250)and auto ISO with my A7r (FE55mm) Not an elegant solution, but the photos are SHARP.
Good thing I have other FF cameras that don’t have this shutter vibration problem.
I’ve rented the A7R twice and noticed hand-held shots not being sharp but tripod mounted ones to be great. Still, I decided to purchase one as I was blown away by the pictures that were sharp as well as thinking that perhaps my handling skills could be better.
2 months later I’m noticing a drastic improvement in the number of successful shots. I believe I’m getting the hang of it now. Ok; the L-plate and Sony Battery grip really work wonders.
The usage of electronic first curtain is worse at high speed:
http://www.street-photo.fr/fr/forum/topic?id=927&p=1#p10559
[…] Sony A7R Shutter ~~VIBRATION~~ » Before The Coffee […]