Jump to content

  • Log in with Facebook Log in with Twitter Log In with Google      Sign In   
  • Create Account

Welcome to Fuji X Forum!

Welcome to the Fuji X Forum, the world's largest online community of Fujifilm X camera users!

Creating a forum account is FREE! or CLICK HERE TO LOG IN

Registration enables you to post in the community, and be notified to replies.

Other benefits: sharing photos, creating a photo gallery, get answers to technical questions, assistance with trouble-shooting, elgibility for contests, and more!

Registering is easy and simple. Become a part of the forum by creating an account. You can also log in using Facebook, Twitter or Google.

We take pride in being the friendliest photo forum on the net.

Come on in join the fun!

Note: after registering if you don't see the validation email, check your spam folder.
Guest Message by DevFuse
 

Photo
- - - - -

Shutter Lag - Verdict


  • You cannot start a new topic
  • Please log in to reply
23 replies to this topic

#1 markopa

markopa
  • Premium Member
  • 487 posts
  • Local time: 03:27 PM

Posted 03 June 2012 - 03:47 PM

As announced I've tried to measure the shutter lag. This could be done in two ways without being intrusive into the camera:

a) Each camera action consumes battery power. That said one could measure the current drain from the battery. The current waveform directly follows the camera actions. However this approach is more com0plicated as you have to solder few wires t othe battery plus teh current waveform could be smoothed by the internal electronics capacitors and power managment circuitry.

B) The second method is by recording the sounds. Every mechanical action of the camera emits sound. By recording the sound and measuring the intervals one could figure out what the camera is doing. And that was the method for measuring shutter lag.

Below is a typical audio footprint of the cpro1 taking a shot. Camera setting were:

- MF, prefocusued
- Manual expsure 1/4s
- EVF
- ISO1600 (manual)
- Latest camera and lens firmware, 35mm lens

Posted Image

So let examine the audio fingerprint:

The part highlighted by the red circle is the X-Pro1 taking an exposure:

Posted Image

Taking into account the sampling frequency of 44100 samples per second we can observe the exposure time:

Posted Image

The grayed region is approx 11132 samples which leads to (1/44100)*11132 = 252ms. So we are pretty accurate as our manual exposure time was set to 1/4s.

Moving further we can measure the part before the exposure - the shutter lag:

Posted Image

Again, the grayed region is 2132 samples which yields to a shutter lag of 48ms. Hum. Not as declared, but not so bad.

So why I'm missing shots if the lag is only 48ms ? Lets examine again the first audio fingerprint. We can notice tow small dots after the exposure has been taken. Intrigued by this I started to observe what the X-Pro1 is doing in that moment ? Yes, those are the blades returning after you depress the button.

Posted Image

We can observe this movement after the first and second exposure. After each exposure I was again half pressing the shutter button preparing the camera for the next shot. So far so good. But, hey ? What happening after the third exposure ? The blades were not returned into the typical position, but just half of the operation was performed by the xpor1. Whoups ? :blink:

Please Login or Register to see this Hidden Content



Bankg ! Pressing the button in such situation the X-Pro1 has to first move the blade to its final position and than take the shot. Let's measure the "lag":

Please Login or Register to see this Hidden Content



There are 9924 samples grayed which leads to 225ms. Ajajaj ! :huh:

Verdict:

- The shutter lag is contained below 50ms
- For some reason after you take one shot you don't allow the xpor1 to take its time and settle down, but you half press the shutter button as soon as you can an anomaly happens which results in a consequent lag of 225ms after you press the button all the way down. An eternity.
- Not confirmed, but it looks we have also a ~100ms of the delay in the EVF. I can't measure this, but is there for sure !
- If you have noticed we have about 1 to 3s between each shot. Pretty slow.

Fuji :angry: ?! Its time to release a firmware fix for this as I'm missing decisive moments and my kinds are growing !!!

#2 flysurfer

flysurfer
  • Premium Member
  • 3,077 posts
  • Local time: 04:27 PM
  • LocationNUE

Posted 03 June 2012 - 04:34 PM

Could it be an unintended side effect of the now much slower aperture dance? Aperture adjustmend used to be faster (and noisier) before the "hysteria" surrounding "aperture chatter" hit the net and prompted Fuji to roll out a (maybe too) quick fix. So is is this "fix" actually a new bug?



#3 EvilTed

EvilTed
  • New Member
  • 21 posts
  • Local time: 02:27 PM

Posted 03 June 2012 - 06:41 PM

Could it be an unintended side effect of the now much slower aperture dance? Aperture adjustmend used to be faster (and noisier) before the "hysteria" surrounding "aperture chatter" hit the net and prompted Fuji to roll out a (maybe too) quick fix. So is is this "fix" actually a new bug?


I've noticed this too.
If you prime the lens by half-pressing the shutter button, sometimes it just goes off and takes the picture?

ET

#4 flysurfer

flysurfer
  • Premium Member
  • 3,077 posts
  • Local time: 04:27 PM
  • LocationNUE

Posted 03 June 2012 - 07:51 PM

I made some unscientific tests involving a "fast" 18mm version 1.00 lens and a "slow" 18mm version 1.01 lens and could not notice any difference in shutter response once the camera was primed. That said, both lenses need some time after taking a shot and immediately half-pressing the shutter button to re-prime the camera. However, I can't notice any differences between both lenses with this respect.

Using an adapted M mount lens reduces the "prime time" between shots, meaning you can take several single shots in a row more quickly.

All three tests involved MF and manual exposure mode.

I any case, once the camera was primed (= green ready light next to the VIEW MODE button), there was no significant shutter lag.



#5 Arjay

Arjay

    Admin

  • Admin
  • 10,444 posts
  • Local time: 04:27 PM
  • LocationMunich, Germany

Posted 03 June 2012 - 11:47 PM

Thank you for your thorough tests, Markopa!

- For some reason after you take one shot you don't allow the xpor1 to take its time and settle down, but you half press the shutter button as soon as you can an anomaly happens which results in a consequent lag of 225ms after you press the button all the way down. An eternity.

I have observed this after-shot lag effect on the X100 too. It doesn't stop me from getting the 'decisive moment' on the first go, but it certainly does for consecutive shots.

- Not confirmed, but it looks we have also a ~100ms of the delay in the EVF. I can't measure this, but is there for sure !

It's very strange that Fuji gave up one of the cooler features of the X100 here: On the X100, the video frame rate is boosted considerably when you press the shutter button to the first pressure point, effectively at least halving EVF lag. Why on earth did Fuji do away with this feature on the X-Pro1?

Fuji :angry: ?! Its time to release a firmware fix for this as I'm missing decisive moments and my kinds are growing !!!

Hm - we don't know yet if this after-shot lag is due to a software or hardware deficiency. If it were due to hardware, we'd have to wait for an X-Pro1 Mk II. :(

#6 Machspeed

Machspeed
  • New Member
  • 12 posts
  • Local time: 05:27 PM

Posted 04 June 2012 - 12:50 AM

Thank you markopa for this well-thought test!
At least it somewhat proves that the shutter lag is a common issue with these cameras and not some statistically insignificant occurrences.



#7 WillemH

WillemH
  • New Member
  • 3 posts
  • Local time: 03:27 PM

Posted 04 June 2012 - 01:22 AM

What a great job. Have the fujiguys given a reaction yet?



#8 Ivophoto

Ivophoto
  • Member
  • 214 posts
  • Local time: 04:27 PM
  • LocationFlanders

Posted 04 June 2012 - 01:23 AM

I didn't need this test to know. Just using the cam was enough. But this test proves things. It surely explains the inconsistency I'm talking about, and that is the most irritating thing. Especially if you are in snipe mode. I didn't understand why I was not able to anticipate on the lag, the inconsistency is the reason....



#9 markopa

markopa
  • Topic Starter
  • Premium Member
  • 487 posts
  • Local time: 03:27 PM

Posted 04 June 2012 - 01:32 AM

What a great job. Have the fujiguys given a reaction yet?


Thanks !

As for now they didn't. At least I'm not seeing a firmware release :) Joking....

Let's hope they notice this and possibly fix it.

#10 Machspeed

Machspeed
  • New Member
  • 12 posts
  • Local time: 05:27 PM

Posted 04 June 2012 - 02:05 AM

markopa, in your wave graphs, can you identify the moment the shutter release button is fully pressed? Not the moment the camera begins the exposure but the actual moment you are achieving full shutter release button press.
I guess not (because it's not a very audible event).



#11 markopa

markopa
  • Topic Starter
  • Premium Member
  • 487 posts
  • Local time: 03:27 PM

Posted 04 June 2012 - 02:16 AM

markopa, in your wave graphs, can you identify the moment the shutter release button is fully pressed? Not the moment the camera begins the exposure but the actual moment you are achieving full shutter release button press.
I guess not (because it's not a very audible event).


You've yourself partly answered. If fact is hardly audible. If you take a closest look at the last wave you can notice something is happening before the camera tries to move the blades.

However I didn't wanted to take also this into account as I think is really minimal. Seems the lag itself is not a problem. The problem seems to be inconsistency. On the other side the lag is not within specifications given by Fuji - correct me if I'm wrong.

[BTW: I'm really tempted to prove/measure how the battery management is bad designed (and not worth the ~1500€ the camera costs) as the battery indicator doesn't work as it should be. But let's us leave this for another thread.]

#12 Machspeed

Machspeed
  • New Member
  • 12 posts
  • Local time: 05:27 PM

Posted 04 June 2012 - 02:47 AM

Mike Kobal just posted on the shutter lag in his blog

Please Login or Register to see this Hidden Content

.

"Shutter lag in manual focus mode: this problem could easily get fixed with the next firmware upgrade.
Even in mf and full manual exposure mode, Fuji lenses assume a resting position until the shutter is half pressed. This causes significant shutter lag. Take the Sony Nex 7 for example, it lacks a manual focusing/DOF scale, however, once you focus, set aperture and shutter speed, the lens stays in the prefocused position and holds the selected aperture value. There is practically no shutter lag with this setting. Once I spot a subject or scene, I half press and hold the shutter button until I am ready to take the picture. This eliminates shutter lag completely, if I miss the shot I can’t blame the camera. "

#13 Sergio

Sergio
  • Member
  • 82 posts
  • Local time: 09:27 AM

Posted 05 June 2012 - 10:15 AM

On the other side the lag is not within specifications given by Fuji - correct me if I'm wrong.


The published shutter lag is 0.05 seconds, which seems to be in line with your test results. This is from the X-Pro brochure.
X100 is 0.01 😔 but Leica M9 is 0.08 😉

#14 flysurfer

flysurfer
  • Premium Member
  • 3,077 posts
  • Local time: 04:27 PM
  • LocationNUE

Posted 05 June 2012 - 10:22 AM

Leaf shutters vs. focal plane shutters. M9 has the additional handicap of a larger shutter for a larger sensor, so no wonder it is even slower.



#15 jknights

jknights

    Moderator

  • Moderator
  • 7,595 posts
  • Local time: 04:27 PM
  • LocationOliva, Valencia, Spain

Posted 05 June 2012 - 10:43 AM

Thank you markopa for this well-thought test!
At least it somewhat proves that the shutter lag is a common issue with these cameras and not some statistically insignificant occurrences.


Yes indeed the use of Audacity to capture the camera sounds and to analyse these results is very good.

Thanks Mark. Much kudos.

Is it possible to measure the AF performance in a similar/some way?