Quick Test of Multishot Mode

[johnbuckley]

Advertisement (gone after registration)

It is easy and logical to use it. The setting is at the bottom of the Drive menu screen, and it automatically uses the 2-second delay in shutter activation. It is probably ridiculous to try comparing the 47mp file with the zillion mp file using the down resolution necessary to post on the forum, but here goes. First image is 47mp. Second is the multi shot result.

 

 

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!

[Advertisement]

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

[johnbuckley]

Second one -- the Multi Shot version.

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!

[johnbuckley]

Hard to see the difference at this resolution. Admittedly, hard to see the difference on my Mac.

[douglas fry]

Can you do a similar tight crop on both images? That would help tell any differences and interesting to see.

There was a technical article (which I must try and find again) describing why a mega pixel image downsampled to say a 10x8 print size is better than a more native resolution camera shoot for that size. Better too is a black and white conversion, with say Silver Efex then downsample.

I think it was Luminous Landscape

[douglas fry]

Also how quickly does it take the images in multi shot mode? Would anything with a moving object in the frame be slightly blurred? Cars/people etc

Edited June 19, 2020 by douglas fry

[johnbuckley]

My recollection is it took about 5 seconds, though the metadata says both images were taken at same second.

i can post a crop, but on my computer, I can’t really see a difference so doubt it will be discernible given LUF resolution constraints. Since this image was taken using the 90-280 at 280, focusing on a scene 10 miles away, I think I’ll focus on something much closer, using the 24-90 . That may be a better comparison.

One interesting observation: when pixel peeping samples of the two images in LR, on the 47mp image, I get a much bigger section of the image to look at than on the higher megapixel image.

 

[johnbuckley]

Advertisement (gone after registration)

47mp using the 24-90

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!

[johnbuckley]

187 mp using the 24-90

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!

[johnbuckley]

47 cropped

 

 

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!

Edited June 19, 2020 by johnbuckley
mixed up pictures

[johnbuckley]

187 cropped

 

 

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!

Edited June 19, 2020 by johnbuckley

[SrMi]

Note that post-processing (PS, Nik Software, Topaz) takes much, much longer with a 187Mp file.

Original DNG file: 336Mb

Exported Tiff: 1Gb

Compressed DNG (by LrC): 190Mb

Edited June 19, 2020 by SrMi
Added size

[scott kirkpatrick]

Some nice results so far.  Leica says that they use the electronic shutter only and that means that they can shoot as fast as the chip can complete and stabilize its half-pixel move.  The bulk of the delay is probably processing time.  Here's how to find out -- set up an old fashioned analog stop watch (simulated of course) on a screen.  Point a camera that can focus close on it start the stop watch and see how far the second hand moves in the course of taking the eight shots.  I'll give it a try.  Have to find the right stop watch first.  Olymp;us claims that now you can take a multi-image hi res shot handheld on their latest M4/3 cameras.  I would bet that is how they do it as well, using electronic shutter.

Edited June 19, 2020 by scott kirkpatrick

[Tailwagger]

I posted one of my tests over in the SL2 image thread.  (2k 100% crop immediately follows) Slow, order of 5 seconds or so, but impressive. 

Word to Leica though. I get there's a pandemic on, but thats no reason to go insane. Get rid of the stupid  'Use a stable platform message' that shows up every time you try to adjust framing. How the hell are you supposed to adjust your monoball to frame a shot with you yelling about stability so that you cant actually see what you framing. Get a grip. 

Edited June 19, 2020 by Tailwagger

[scott kirkpatrick]

I bitched about exactly the same piece of idiotic error over-messaging in the other new firmware thread.  You can get around it, but it makes for unnatural handling. 

[scott kirkpatrick]

Here's something interesting.  I found a screen-filling analog online stopwatch and photographed the second hand with multi-shot.  Below, the first shot is the 4X multishot, which is apparently motion-corrected, because the second hand appears sharp even though it is pointing about 0.4 seconds further around than the single frame shot which also results (the second frame below).

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!

U1000014 by scott kirkpatrick, on Flickr

U1000015 by scott kirkpatrick, on Flickr

So the 8 frames are taken in about 0.4 seconds, or about 0.05 (1/20th) sec each.  My actual shutter speed was something like 1/500, so that is a readout speed.  That consistent with estimates that others have made for the electronic shutter sweep speed on the SL2

But you are not done in half a second.  It seemed to me that there was initial processing of about 5 seconds, then some sort of data movement taking 15 seconds more.  During all 20 seconds the screen was black, with some message that I missed on it.  Then it takes another 10-20 seconds to clear the buffer.

 

Edited June 19, 2020 by scott kirkpatrick

[scroy]

Quick test with a stopped clock using 75mm ISO 3200, 1/250s, f/2.0

187 croppped 

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!

47 cropped

[johnbuckley]

Very helpful, Scott. I know it takes about 18-20 seconds from the time the shutter closes to when you get the camera back, but your measurement is really helpful in explaining why the metadata shows both images captured within the same second. 

[Jon Warwick]

Quick test here ....these are 100% crops off what would be a 70" wide print for a printer output at 300dpi. Yes, i like large prints!! 

Both of the native files are therefore resampled larger in Photoshop, ie, the high res image is resampled from its native size of 55" to 70". And the normal image resampled from c 28" to 70".  So quite a brutal test!

The "apparent" difference in resolution is there ....but quite frankly, the SL Summicrons are utterly extraordinary in whatever mode is used.

What is even more apparent (IMHO) is the complete lack of false color and moire ......  

This is clear (on my Photoshop screen at least) from the label on the vitamin bottles ...the high res image has far fewer artefacts and false color in comparison to the normal shot.

I daresay because i have saved these images at the lowest quality JPEGS that some (or perhaps a lot) of this difference is lost, but it gives one a rough idea.

Images below are extreme crops to c 3"x3" for a 300dpi output from these 70" wide files. Taken on an SL2 and SL 50 APO on Gitzo GT1542 and Arca Swiss P0 head.

For me, the "purity" of an image that lacks digital artefacts like false color is what I think is particularly helped by the high res mode. For images of (say) mountain landscapes that have very fine rocks in the distance or glaciers etc can often attract that digital mess, and in that scenario i think high res mode will be superbly useful for me.

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!

 

 

[Advertisement]

Hallo,

Look here: Leica L-Mount

[Stuart Richardson]

Looks much better in that image! I have tried a few where it was worse in some ways (false colors in highlights and aliasing in certain areas of the image). I think the best part of it is that it gives you both versions when you use it, the normal and the resized, so from that perspective it can be even more useful, as if there are problematic artifacts in one version or the other, you have the other version. You do not have to choose and take your chances, in other words.

Edited June 19, 2020 by Stuart Richardson

[Jon Warwick]

10 minutes ago, Stuart Richardson said:

 I think the best part of it is that it gives you both versions when you use it, the normal and the resized, so from that perspective it can be even more useful, as if there are problematic artifacts in one version or the other, you have the other version. You do not have to choose and take your chances, in other words.

Totally agree. I can also imagine possibly incorporating sections of the normal version into the high res version if there are any “problem areas” with the way multishot handles some subjects.