Leica Q: Missing the Decisive Moment

[arnklit]

Advertisement (gone after registration)

Missing the Decisive Moment

 

1932, Paris, near the Gare Saint-Lazare the great Henri Cariter Bresson approaches a flooded construction site enclosed by an iron fence. He quickly realizes the potential of the situation and as a man dressed in black leaps over the ladder lying on the ground trying to avoid getting his feet wet, he presses the shutter just a fraction of a second before the heel of the man’s right shoe hits the water. He just made what was arguably going to be his most iconic street photograph ever. If, instead of using one of his beloved mechanical Leicas, he had used a Leica Q, looking though its fantastic EVF, he would immediately have avoided the parallax issue that caused him to include a part of the fence through which he was photographing and which forced him to apply a slight crop. The last glimpse he would see in the viewfinder was in fact the much desired decisive moment, but what he would find when hitting the PLAY button would be a picture with a big splash of the man landing in the puddle of water. Not because his reflexes were not quick enough, not because of any delay in the class-leading shutter of the “Q”, but because of a more serious quirk of the Leica Q which under certain circumstances makes it almost useless as a tool for capturing the “decisive moment”.

It turns out that under certain circumstances the delay with which the real world is presented through the EVF (or the LCD on the back of the camera) makes it almost impossible to catch the right moment. By the time you see something happening in the viewfinder the world has moved on and what you get in the final picture is not what you had expected.

Try this: set your camera for a typical daylight street scene on a sunny day using the “sunny sixteen” method (F16, Shutter = 1/ISO), so for example F16, 1/500th, ISO 500. Now watch someone walking while half-pressing the shutter and then press through completely just as the subjects front foot is just about to hit the ground (as in HCB’s picture). What you get is a rather unpleasant front leg with a bent knee. If you repeat the exercise looking at your subject directly (ie, not through the viewfinder) you should be able to get exactly the moment you want.

 

To be more explicit, and in order for anybody to easily repeat the experiment, I’ve used a metronome app running on an iPAD as a continuously moving subject.

In the following video I’ve set the speed of the metronome to 248 bpm running on an iPad set at maximum luminosity. The Leica Q is set to F16, 1/125th, ISO 50000 and the rear LCD screen and the iPad are recorded with an iPhone 6. I apologize for the quality of the video, but I think it is absolutely clear that when the shutter is half-pressed (which closes down the diaphragm) the delay becomes immediately visible. Also notice that if the shutter is not half-pressed there is no delay. In theory, therefore, it would be possible to follow the action without half-pressing and then doing a full press at the right time, but often you want to half-press to make sure all of the normal chores of autofocus, metering etc. are taken care of so as not to waste further time.

The problem seems to be intimately related to the F-stop, the higher the F the worse. At F below 4-5.6 it is hardly noticeable.

 

https://youtu.be/EPTyJrfnkgs

 

In this second video I’m showing a Fuji X100T in the same conditions and as you can see there is no noticeable delay. Also the Fuji X-T1 is without delay in the same conditions, which leads me to believe the problem to be specific to the “Q”.

 

https://youtu.be/8GoNQXW0Ke0

 

I hope the design engineers of Leica get a chance to read this post and perhaps to repeat these experiments for themselves in the hope that they may come up with a solution in a future firmware update.

 

On a side note I wonder if the new SL has the same problems. Since it is marketed as an action camera I think it would be a real deal breaker if it did.

 

Aside from the problem described I would like to express my full appreciation of the camera. For reportage/street photography with its excellent intuitive handling, super silent shutter,  unquestionable image quality, buttery smooth manual focus, lightning fast autofocus and general unobtrusiveness the Leica Q is really hard to beat.

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[soizic]

Thanks, very interesting 

[biglouis]

Thank you. That is an interesting 'feature' I was not aware of. The comparison with the Fuji is interesting. I think you should email Leica and point them at your video and ask for their comments.

 

LouisB

[SrMi]

In a quick test, I was not able to reproduce the large lag. I assume that was because my 'test' had much more light involved. I believe that half-pressing the shutter closes down the iris, and much less light reaches the sensor. In most cameras the EVF/LCD has much more lag when there is not much light.

 

I do not know if Fuji X100T also closes the iris with half-press.

[jaapv]

I have worked with cameras with far more lag (like the Digilux2). The key word, in my experience, is: anticipate.

[Manoleica]

I have worked with cameras with far more lag (like the Digilux2). The key word, in my experience, is: anticipate.

Know your equipment -- Anticipate -- 100%..

[Manoleica]

Advertisement (gone after registration)

FWIW --

D-Lux4   --- Anticipate..""

 

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[ricky1981]

I think what is happening there is the EVF is trying to give you a 'properly exposed' view but since it stops down and you are shooting in low-ish light it needs a larger 'shutter speed' hence the lag. I don't really see this as an issue, it's just a limitation of physics and current technology, it's very common for any EVF to lag or become less useful in low light. That's why some still insist on using OVFs instead or use the EVF for framing but use their eye to judge when to shoot.

[Bill Livingston]

This is going back to the original posters point...

 

...Which is exactly what HCB would have done in that same scenario...  just taken the photo and then fortuitously found something special once developed.

 

Take the photo first... if you get a chance to refine focus, composition or anything else, then thats your second and subsequent shots... just don't miss that first one!

 

There is absolutely no way that someone can come across that scene and been able to fire the shutter just as the foot touches the puddle... not at that speed... if you read the actual story, in the book, it was an opportunistic photo that didn't reveal itself until it was developed*.

 

For some reason, everyone seems to think that great photographers take great images all the time and every one of them is down to skill and technique.

 

The reality is they are able to know where an opportunity will present itself and then anticipate the action or the composition they had in mind for that photographic solution... thats the raw material... and then make sensible and interesting choices once they go through their negatives.

 

Then, they, or with their printer, make choices in cropping, dodging and burning... choice of paper, print exposure... in fact pretty much everything we do today with digital cameras and Lightroom...

 

Pre-visualisation, Anticipation, Composition, Selection... from then on, its what you do with the image and the choices you make in either a darkroom or Lightroom... same as we all do.

 

The best photographers have a feel for an image and the message they wish to convey and have an instinct in getting the right raw material in the first place... its an art, not a science. 

 

It remains an art all the way through... even if the process feels 'scientific', its the choices made at each stage, even the capture stage - although that is the point where you have the least time and instinct therefore plays the greater part - that render it an art.

 

It is completely irrelevant whether the image is digital or film, whether it is perfectly exposed or not, whether it is in focus, or not...

 

Disagree...? Then look at the HCB image being discussed. We know it is film, but is it sharp and perfectly exposed? No... not even close and a LOT of work was done in the printing stage to complete the image as we now know it.

 

I think it worth looking through a photojournalists contact prints and look at the notes, framing etc written on them... in fact, any decent photographer in the past will have taken thousands and thousands of photos during their lifetime... how many of those individual images are truly memorable... six? ten? of their ENTIRE life's work...?

 

*(in this case, it was NOT the case, HCB just looked through the fence saw something potentially interesting and then poked his camera through... if I remember the story correctly, he wasn't even looking through the viewfinder!)

Edited January 4, 2016 by Bill Livingston

[seabea23]

Good points Bill, the Thames & Hudson book Magnum Contact Sheets is a great resorce to peruse.

[matlep]

How come the Iso is set to ISO 50.000 on the Leica and only ISO 4000 on the Fuji? If conditions where the same then all setting would have to match to make it a fair comparison. Right?

Edited January 4, 2016 by matlep

[NERICSSON]

One could also wonder if Bresson had the shutter half pressed in order to activate AF on his Leica... ;-)

[Bill Livingston]

One could also wonder if Bresson had the shutter half pressed in order to activate AF on his Leica... ;-)

 

Clearly not if you know the photo..!

[Ecaton]

Clearly not if you know the photo..!

 

...did he even lift the camera to the eye to frame?

[matlep]

So this was just a trolling thread then?

[Bill Livingston]

Seems so... I need to learn not to rise to them... Maybe that should be another of my new year resolutions..

[NERICSSON]

Coming back to the technical isseue, EVF lag, it would only be an issue if it is used with tracking AF at large f-numbers? Otherwise you would use MF or just press fully down and trust AF at the decisive moment

Edited January 5, 2016 by NERICSSON

[millsart]

The Q does 11 FPS....just spray and pray as soon as you see a kid who might jump into a muddle LOL

[Advertisement]

Hallo,

Look here: Leica Q

[arnklit]

Thanks everyone for your kind feedback, I'm sorry not to have come back earlier.

I agree with you that the perfect camera does not exist and we must therefore adapt to what our cameras offer us. In the case of the Leica Q the obvious solution is to learn to shoot from the hip, or look with one eye in the EVF and the other directly at the subject, or use the snappy 10 fps shooting mode (although with the shallow buffer it takes a long time even with a fast SD card before it gets ready for more action).

I was just trying to document the delay and hope for an explanation. What annoys me a bit is that the delay is very variable and therefore it is difficult to compensate for by anticipation. It can be a bit hit and miss. It is quite unlike the zero delay of classical optical viewfinders and the reactivity of an old M6 or any high-grade DSLR is easily compensated for. I again stress the fact that the shutter of the Leica Q is lightning fast in all conditions, it's what you see through the EVF/LCD that is delayed in certain conditions.

I think what is happening is that when operating in low light, which happens when stopped down to F16 during a half-press, the EVF/LCD starts to average the frames to keep the noise down. I believe the EVF/LCD has a refresh rate of 30 frames/sec and in fact filming the LCD with a high frame-rate video camera seems to show an update only every 4-5 frames. (the subject in this case is the stopwatch of an iPad). Averaging of 4 to 5 frames out of 30 does in fact correspond to the delay of about 100-200ms as observed in the worst-case as shown in my previous video of the metronome. Strangely enough I never experienced this with the two Fujis.

However I guess all this is probably only something that hits my own curiosity. Since the Q is such a fast camera in all other aspects, and I found that I was getting unexpected results in some cases I tried to figure out why and narrowed it down to the delay in the EVF/LCD and thought it would be worthy of your attention. No trolling intended.

 

Happy shooting.

[matlep]

Thanks everyone for your kind feedback, I'm sorry not to have come back earlier.

I agree with you that the perfect camera does not exist and we must therefore adapt to what our cameras offer us. In the case of the Leica Q the obvious solution is to learn to shoot from the hip, or look with one eye in the EVF and the other directly at the subject, or use the snappy 10 fps shooting mode (although with the shallow buffer it takes a long time even with a fast SD card before it gets ready for more action).

I was just trying to document the delay and hope for an explanation. What annoys me a bit is that the delay is very variable and therefore it is difficult to compensate for by anticipation. It can be a bit hit and miss. It is quite unlike the zero delay of classical optical viewfinders and the reactivity of an old M6 or any high-grade DSLR is easily compensated for. I again stress the fact that the shutter of the Leica Q is lightning fast in all conditions, it's what you see through the EVF/LCD that is delayed in certain conditions.

I think what is happening is that when operating in low light, which happens when stopped down to F16 during a half-press, the EVF/LCD starts to average the frames to keep the noise down. I believe the EVF/LCD has a refresh rate of 30 frames/sec and in fact filming the LCD with a high frame-rate video camera seems to show an update only every 4-5 frames. (the subject in this case is the stopwatch of an iPad). Averaging of 4 to 5 frames out of 30 does in fact correspond to the delay of about 100-200ms as observed in the worst-case as shown in my previous video of the metronome. Strangely enough I never experienced this with the two Fujis.

However I guess all this is probably only something that hits my own curiosity. Since the Q is such a fast camera in all other aspects, and I found that I was getting unexpected results in some cases I tried to figure out why and narrowed it down to the delay in the EVF/LCD and thought it would be worthy of your attention. No trolling intended.

 

Happy shooting.

 

Ok. But you still have not explained why the cameras you are comparing have different settings.