SL2 v. SL2-S –Interesting Video

[John Smith]

Advertisement (gone after registration)

Interesting video about megapixels and low-light performance.

 

[Advertisement]

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

[Anthonylars]

I watched this video. This is really interesting. I own the SL2-S, and have rented/used the SL2. Just from a perception standpoint, it just seems the SL2-S files are MUCH more clean than what I got with the SL2. I also own the Q2 & the same applies. In low light I can use a faster shutter speed & get the same light/quality image than on the Q2 & SL2. 

Curious to other's thoughts on this. 

[Steve Miller]

2 minutes ago, Anthonylars said:

I watched this video. This is really interesting. I own the SL2-S, and have rented/used the SL2. Just from a perception standpoint, it just seems the SL2-S files are MUCH more clean than what I got with the SL2. I also own the Q2 & the same applies. In low light I can use a faster shutter speed & get the same light/quality image than on the Q2 & SL2. 

Curious to other's thoughts on this. 

I don’t know about the SL2, but I can definitely say that I’ve never had a camera like the SL2-S when it comes to shooting in low light. I had the Q2 and found that above ISO 3200, I would need to convert the image to B&W because the color noise was very unpleasant. I don’t think twice about shooting the SL2-S at ISO 12500 and even 16000 and keeping the image in color. The noise level in the SL2-S files is just so low and the quality of the noise is night and day (literally, one might say) compared to any other camera I’ve used.

I haven’t tried the SL2 but was told by two different salespeople at Leica stores that the SL2-S was a good 1-2 stops better in terms of low light noise performance versus its 47 MP big brother. Comparing it to my (now former) Q2, I definitely see at least two stops difference. FYI, I ended up with the SL2-S after I bought a used M10 and was blown away by its low light performance, which I’d put pretty much on par with the SL2-S. I loved that camera so much that I ended up selling the Q2 and bought the SL2-S to get the same sensor performance but in an autofocus body…plus the ability to shoot manual focus M mount lenses on it with focus peaking.

Steve

P.S. I haven’t watched the video yet, but assume the argument made is that if you downsample the larger 61 MP files (or 47 MP files in the case of the SL2), the noise cleans up so well that you’re better off going with the larger MP sensor. Just a guess.

[LBJ2]

33 minutes ago, Anthonylars said:

I watched this video. This is really interesting. I own the SL2-S, and have rented/used the SL2. Just from a perception standpoint, it just seems the SL2-S files are MUCH more clean than what I got with the SL2. I also own the Q2 & the same applies. In low light I can use a faster shutter speed & get the same light/quality image than on the Q2 & SL2. 

Curious to other's thoughts on this. 

What you are seeing is the result of the BSI sensor in the SL2-S. 

[hansvons]

Good guess, @Steve Miller.

The video is one of the better YouTube takes on photography. They compared a 12 MP Sensor to a 60 MP sensor, took a snap at ISO 6400 and printed the files professionally in a fairly large format. 

In the end, we are talking about uprezing noise and detail. From that perspective the high-resolving sensor with a lesser noise/signal ratio did better than the low-resolving video-optimised sensor.

What they were missing out is the saturation in the shadows which often less resolving sensor are better at in lowlight situations. That particular image didn’t carry much colour. A more colourful photograph would have been more challenging. 
And finally, if there is no uprez happening, things will again be different, especially if one shoots in environments that require more extreme ISOs.

 

 

[hoolyproductions]

Not really sure what to make of this. The results (in those specific test conditions) do speak for themselves, but still I am left with doubts (all the other considerations affecting image quality, plus my own experience in severely low light).

To oversimplify my doubts, they started by explaining well that the camera industry used to promote this 'myth' that smaller sensors have better high ISO by comparing results at 100% magnification at all sensor sizes. They concluded by saying this is now 'fixed' by showing results at 100% for the smaller sensors and then downsampled to equivalent size for larger sensors. This seems a 'better' than the former way of doing it indeed, but means that we end up pixel peeping the smaller sensor at 100% and still end up comparing apples with oranges?

It seems like quite an oversimplification that a smaller sensor is not (ever) going to give better noise performance (on the final image) and only has a 'place' for video or for photographers that  prefer smaller file sizes? (which was basically  the main conclusion)

From my own experience, my digital camera journey was along the lines D700 > Sony APS-C (portability, but never happy with IQ in low light or horrific menu) > Sony A7 (back to full frame, more portable than dSLR, did not like the menu/ergonomics) > 50mm Summilux on my A7 was my 'gateway drug' to Leica > M6, M240, M10 as main cameras for a few years > added a CL for portability (finding APS-C more acceptable for low light due to more modern sensor, but still 'at the limit' for my type of low light shooting, not on the level of my prior full frame cameras) and, most recently selling my M10 after obtaining an SL2-S.

I've only had the SL2-s for a few weeks but am astonished by the files at 12500 ISO. With prior full frame cameras I have limited to 6400 and even then, depending on the nature of the light, got mixed results (sometimes great, often not). I shot all of the above cameras many times in low light in a local music bar and for me the difference in that setting is 'night and day' (although the D700 was pretty good, particularly given how old it is now)

Well, however 'black and white' this all is I'm still delighted with the SL2-s and the lenses I chose for it. Curious how the discussion unfolds

 

 

 

 

 

[beewee]

Advertisement (gone after registration)

The video brings up some valid points but it took things to the extreme in the sense that 12MP vs 60 MP is 5x the number of pixels and more than 2.23x the pixel density. In comparison, the SL2 vs SL2-S is only 40% increase in pixel density. To this end, as far as 24MP vs 47MP goes, the increase in actual increase in resolution (pixels per area) from SL2-S to SL2 is not that significant as compared to the improvement in dynamic range from SL2 to SL2-S which is on the order of 2-3 stops (i.e. 4x to 8x improvement). That being said, unless you’re pushing the SL2 ISO to the point where dynamic range and noise become noticeable or significant (i.e. ISO 6400+).

[jrp]

The answer is that “it depends”. If you are producing electronic output at full size you are probably not going to notice the difference even if you have an 85 inch tv unless you press your nose against the glass (but why would you?). If you are always shooting at high iso then the lower megapixel camera will suit you better.  If you are printing big, then the higher megapixel camera (with noise machine learned noise reduction, if you need it) will best the smaller megapixel camera uprezzed. 

If you are a pro, you will probably take both with you for a shoot. If not, you will end up leaving one, or other at home. 

[beewee]

3 hours ago, jrp said:

The answer is that “it depends”.

Yup. There’s really no simple answer. It’s very much use case dependent.

I recall also, for many bill board applications, contrary to what many people would think, the resolution requirements are actually fairly low. A 24 MP resolution would be plenty for large billboards or something you might put on the wall of a building. The reason is because for this application, even though print size is huge, the audience would never look at it up close. It’s designed mainly to be viewed from a far distance in which it’s about the sample angular resolution as a 4x6 or 8x10 print viewed at arms length.

Edited August 17, 2021 by beewee

[Jeff S]

3 hours ago, beewee said:

I recall also, for many bill board applications, contrary to what many people would think, the resolution requirements are actually fairly low. A 24 MP resolution would be plenty for large billboards or something you might put on the wall of a building. The reason is because for this application, even though print size is huge, the audience would never look at it up close. It’s designed mainly to be viewed from a far distance in which it’s about the sample angular resolution as a 4x6 or 8x10 print viewed at arms length.

24MP could be far more than needed…

https://fstoppers.com/originals/how-many-megapixels-do-you-need-print-billboard-220239
 

Jeff

[dgramagra]

This is interest Japanese Issue.

https://dc.watch.impress.co.jp/docs/review/comparison/1337319.html

 

Check it 

[hoolyproductions]

Thanks dgramagra, that translated well on google and some interesting points there!