Buffering

[drolfe]

Advertisement (gone after registration)

Can someone explain to me why the M9 buffering is so poor? My M9 stops shooting after just a few frames, while my Nikon slams along with no worries.

Drolfe

[Gregm61]

If that's what you need, the Nikon is what you should be using, not a 7 year-old under-developed for 2009, manual focusing rangefinder.

 

Resort to the M9 when you are only interested in taking pictures, one at a time.

[Advertisement]

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

[jdlaing]

Leica M is, and always was, NOT a run and gun camera.

[jaapv]

The key words are "slams along" I guess

[drolfe]

All good answers, and accurate.  But not exactly what I was getting at. 

 

I am a newspaper shooter, and during my 40-year career have always shot Leica M and  Nikon  side by side.  By "slamming away," I was thinking of sports sequences, fast-moving news events, and so on. 

 

For most sports, the DSLR is always my tool of choice.  When I can work in close and take my time, I use the Leica M9.  Particularly for very wide lenses; 21mm, 15mm, and such.  The Leica is not so good at football.

 

But I use the Leica for its size, unobtrusiveness, and simple operation.  It is not intimidating to most people, not like having a massive DSLR with a zoom pointed at them.

 

So my question is simply a technical one; why is the Leica buffering so much slower than DSLRs, and are the latest models any better?

 

Thanks for your responses.

[Luke_Miller]

The M-240 M-P and the M-246 have 2 GB buffers, an increase of 1 GB over that the M-240. I believe the M-262 has the same 1 GB buffer. I can't find data on the M9, but I suspect it has a smaller (than 1 GB) buffer.

Edited November 29, 2016 by Luke_Miller

[-Lss-]

Advertisement (gone after registration)

So my question is simply a technical one; why is the Leica buffering so much slower than DSLRs, and are the latest models any better?

 

 

The buffer is small. It's a design choice, or perhaps rather accurately a cost-cutting measure. As far as I know (meaning I am too lazy to google for the specs right now), the M9 has the same buffer as the M8. The problem is that the file size is significantly larger on the M9, which means it takes fewer frames to fill a buffer of equal size.

 

People complain even about the M8 buffer. As I never ran into practical issues with it myself, I made a simple test some time ago to simulate very fast real-world shooting. The experiment consisted of two parts.

 

In the first part of the experiment, I started a stop watch, and began shooting mostly individual frames using continuous drive and DNG only. By individual frames I mean that I did not keep the shutter button pressed down, but rather simulated fast consecutive individual exposures. For this reason I used continuous drive to err on the side of faster shooting. I kept my eye on the stop watch, and stopped shooting when it was approaching 15 seconds. I stopped the watch at 14.851s. Between exposures 6 and 7, I paused to "consider the composition". Again, this was done to simulate a real-world scenario that makes sense for myself. Omitting the hour and the minute, the DNG files show the following times:

 

Image 01: 34 seconds

Image 02: 34 seconds

Image 03: 35 seconds

Image 04: 36 seconds

Image 05: 37 seconds

Image 06: 38 seconds

Image 07: 42 seconds

Image 08: 43 seconds

Image 09: 44 seconds

Image 10: 45 seconds

Image 11: 46 seconds

Image 12: 47 seconds

Image 13: 47 seconds

 

There were no delays whatsoever during this part of the experiment. The camera shot every frame either according to my command or the continuous drive performance (2 fps for 10 shots, I believe). I did not time how long it took for the buffer to clear. Note that this time affects your chimping, not your shooting.

 

In the second part of the experiment, I put the camera again on continuous drive with DNG only and let it fire away for 1 minute. It achieved a total of 32 exposures. The first 12 exposures were fast, then the camera got sluggish for the remaining 20 exposures.

[Gregm61]

Does using compressed DNG vs. non-compressed alter the speed at all? I've never been in such a fast-paced shooting frenzy I've managed to make myself have to wait between frames with my M9-P.

[John Ricard]

The buffer is better on the M240, but still not enough for most of my shooting needs.   People often act as if a larger buffer is necessary only for sport shooters.  Personally, I find I hit the buffer if I'm just shooting 2 people talking, and I need to ensure that I have a few images where both faces look right while speaking.  The M-P version of the M240 has an adequate buffer in my opinion, but that camera is hard to find used.  Expect to pay over $7,000 for that one.  You might find a used M240 for half that price.   

 

The crazy thing about the M240 that no one ever talks about is that the buffer gets worse as you increase ISO.  At 800 and below you'll get about 7 quick shots before you can't take another image.  At ISO 1250, those same 7 shots will take much longer.  Even the M-P slows down the fps once you get above ISO 800.  Very frustrating.

 

I'm confident the next M body will have a large buffer, since Leica now has 2 full frame bodies that can shoot virtually unlimited frames.  Already sold my M240 in anticipation of that one.  Using the M9 these days -even though it has a truly unbearable buffer most of the time.

[John Ricard]

Somewhere in this video, I did a side by side comparison of the buffers of the M9 vs the M240: 

[Paul J]

Make sure you have compressed DNG set.

 

I generally find the M9 buffer workable, but yes, limited.

[tobey bilek]

Maybe not cost but space constraint ?  

[pico]

Maybe not cost but space constraint ?  

 

That's interesting, Toby. What space are you referring to, please.

 

Aside: I am perfectly happy with the M9's time-to-process possibly because my expectations are from Leica's film era.

 

Best,

Pico