SL3 Rumors

[Jeff S]

Advertisement (gone after registration)

Stefan discusses the 4 architectural improvements to the sensor at the 20 min mark in this discussion…

He doesn’t, however, say that the electronics have been moved in the same sense (as I understand it) that a BSI design would entail.

Jeff

[Advertisement]

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

[hdmesa]

23 minutes ago, Jeff S said:

Stefan discusses the 4 architectural improvements to the sensor at the 20 min mark in this discussion…

He doesn’t, however, say that the electronics have been moved in the same sense (as I understand it) that a BSI design would entail.

Jeff

Yes, those sound like the improvements made to the last of the Sony FSI sensors, which helped FSI stay competitive with BSI in some ways. Just my opinion that the FSI sensors will one day be regarded the same way we look at CCD sensors like the one in the M9. Compared to the M11 BSI sensor, the M10-R sensor produces a more filmic look to my eye. I wish Leica would produce a budget M11-E using the sensor from the M10-R.

Edited February 17, 2024 by hdmesa

[BernardC]

13 hours ago, hdmesa said:

SL2 is FSI. If the SL2 was BSI, it would be evident from any teardown because the wiring normally on the frontside of the sensor would be moved to the backside of the sensor.

Interesting, do you have a link to this teardown? I know that Chipworks used to publish reports many years ago, but I'm unaware of any recent examples.

[chrismuc]

The 61 MP 3.76 CMOS BSI sensor IMX455 from Sony was initially released in July 2019 in the Sony A7RIV.

The Leica SL2 came out in November 2019 using a 47 MP CMOS FSI sensor from CMOSIS.

So ... we will get now a SL3 with the technology that was already available at the time of the release of the SL2. 

Why that intermediate solution?

The EVF seems also the same. The tiltable screen is very much appreciated, and IMO should have been added already with the SL2. I shoot 90% of my street photography using the back screen tilted in my S1R.

The only thing technology wise today available, not 2019, is a fasted processor and better (AI supported) AF algorithms. 

[SrMi]

15 minutes ago, chrismuc said:

The 61 MP 3.76 CMOS BSI sensor IMX455 from Sony was initially released in July 2019 in the Sony A7RIV.

Yes, and is still state of the art, no other sensor has replaced it for ultimate IQ and resolution.

[chrismuc]

Yes that 3.76 um sensor is excellent, I use it other formats in the X-E4 and GFX100. On pixel level it is 'better' / more flexible in raw adjustments than the CMOSIS in the SL2 and S1R. Looking forward to it in the SL3 and S1R2, and to an CAF quality hopefully not too far behind the A7RV.

[Photoworks]

Advertisement (gone after registration)

15 minutes ago, chrismuc said:

Yes that 3.76 um sensor is excellent, I use it other formats in the X-E4 and GFX100. On pixel level it is 'better' / more flexible in raw adjustments than the CMOSIS in the SL2 and S1R. Looking forward to it in the SL3 and S1R2, and to an CAF quality hopefully not too far behind the A7RV.

I use the gfx100s and a7r4, I can push SL2 and a7r4 more than 100s, and the break color with hue shifts in capture one. I much prefer Leica and Sony files.

[hdmesa]

33 minutes ago, Photoworks said:

I use the gfx100s and a7r4, I can push SL2 and a7r4 more than 100s, and the break color with hue shifts in capture one. I much prefer Leica and Sony files.

That seems very odd. The 100S was the one RAW file I couldn't break in C1, and my edits for IG tend to be a bit over the top. I would check your 100S RAWs in RawDigger to make sure they're not being underexposed. Fujifilm's metering and in-camera histograms might be too conservative. 

SL2 does extremely well at base ISO, where I prefer its output over a lot of other cameras with newer sensor tech.

[hdmesa]

1 hour ago, chrismuc said:

...Why that intermediate solution? ... The EVF seems also the same. The tiltable screen is very much appreciated...

The color and character of the RAW files out of the M11/Q3 are incredible, so expectations should be high for the output of the SL3. The SL3 should be better able to regulate heat versus the M bodies, leading to less noise at higher ISOs. It also lets the processor run faster, which seems in general to be a boost to the image processing. I say that because the files I got from the SL2 were so much nicer for me than those from the Q2, which shares the same sensor.

Perhaps with the EVF they have improved the optics some. At least we're not stuck at 3.XM dot EVFs like Nikon is.

Tilt screen. I hope it's not the recycled two-way tilt from the Q3. Sony finally learned that lesson and has started using 3-way tilt screens. Does Leica think we only take photos in horizontal orientation and don't need a tilt screen when shooting vertically? I'd rather have no tilt screen than the one on the Q3.

[hdmesa]

2 hours ago, BernardC said:

Interesting, do you have a link to this teardown? I know that Chipworks used to publish reports many years ago, but I'm unaware of any recent examples.

I assumed someone had done an SL2 teardown but maybe not. Perhaps that's why no one has definitively written about the SL2 sensor type.

I wonder if you can see the difference between FSI and BSI by looking at the exposed front of the sensor under high power magnification.

Edited February 17, 2024 by hdmesa

[LBJ2]

1 hour ago, hdmesa said:

I assumed someone had done an SL2 teardown but maybe not. Perhaps that's why no one has definitively written about the SL2 sensor type.

I wonder if you can see the difference between FSI and BSI by looking at the exposed front of the sensor under high power magnification.

Viewing other camera tear-downs with the sensor housing intact, I don't see any way to determine BSI or not. Perhaps the cover glass would have to be removed as well, but I do wonder if removed would we be able to see below the microlens layer and/or bayer filter without a microscope anyway? 

For interest here's a look at a damaged sensor and its layers:

https://jmcscientificconsulting.com/microscopy-reflectance-imaging-of-a-modified-camera-sensor/

Non BSI CMOS Sensor Layers:

 

Edited February 17, 2024 by LBJ2

[helged]

On 2/13/2024 at 2:59 PM, LBJ2 said:

 "Sensor readout speed below 17 milliseconds is generally reasonable"

For most types of images, yes. But not for small-ish birds in flight; I have a Canon R5 with a readout time of 16.3 ms, this is insufficient to freeze the wings without deformation. So for this type of scenario, I use the mechanical shutter.

SL2-S has a readout time of about 30 ms (I believe); and here the mechanical shutter is definitely required to avoid deformed wings. It's a visibly clear difference between 30-ish and 16 ms. For small/quick BIF, I would think that sub-10 ms would be good & sufficient.

I guess the readout time of the SL3 sensor can be expected to be around 30 ms? 

Edited February 17, 2024 by helged

[SrMi]

8 minutes ago, helged said:

For most types of images, yes. But not for small-ish birds in flight; I have a Canon R5 with a readout time of 16.3 ms, this is insufficient to freeze the wings without deformation. So for this type of scenario, I use the mechanical shutter.

SL2-S has a readout time of about 30 ms (I believe); and here the mechanical shutter is definitely required to avoid deformed wings. It's a visibly clear difference between 30-ish and 16 ms. For small/quick BIF, I would think that sub-10 ms would be good & sufficient.

I guess the readout time of the SL3 sensor can be expected to be around 30 ms? 

I expect the SL3 sensor's readout time to be the same as a7rIV's: 100ms. If Leica would allow 12-bit mode, the readout would improve to 50ms.

[LBJ2]

29 minutes ago, helged said:

For most types of images, yes. But not for small-ish birds in flight; I have a Canon R5 with a readout time of 16.3 ms, this is insufficient to freeze the wings without deformation. So for this type of scenario, I use the mechanical shutter.

SL2-S has a readout time of about 30 ms (I believe); and here the mechanical shutter is definitely required to avoid deformed wings. It's a visibly clear difference between 30-ish and 16 ms. For small/quick BIF, I would think that sub-10 ms would be good & sufficient.

I guess the readout time of the SL3 sensor can be expected to be around 30 ms? 

Has anyone measured, estimated the Q3 yet ? Not necessarily expecting exact same for SL3, but maybe in the ballpark. 

[SrMi]

29 minutes ago, LBJ2 said:

Has anyone measured, estimated the Q3 yet ? Not necessarily expecting exact same for SL3, but maybe in the ballpark. 

I measured 100ms for Q3, as expected.

[helged]

54 minutes ago, SrMi said:

I measured 100ms for Q3, as expected.

That is a rather long readout time... Hopefully SL3 will be (much) faster, but realistally not given the M11/Q3 type of sensor. 

Edited February 17, 2024 by helged

[Simone_DF]

28 minutes ago, helged said:

Hopefully SL3 will be (much) faster, but realistally not given the M11/Q3 type of sensor. 

Unfortunately it won’t. 

[SrMi]

53 minutes ago, helged said:

That is a rather long readout time... Hopefully SL3 will be (much) faster, but realistally not given the M11/Q3 type of sensor. 

The only way would be to allow 12-bit mode. Otherwise, it is the same as Sony and Sigma 60MP cameras.

[Advertisement]

Hallo,

Look here: Leica L-Mount

[Dr. G]

8 hours ago, SrMi said:

I measured 100ms for Q3, as expected.

FWIW, this was taken with the Q3 with a 1/8000 shutter speed.  A few on the forum say that a global shutter isn't necessary, but faster readout times would have probably negated the issue with the wings here.  So if the Sl3 has the same readout speed we'll have some of the same issues the Q3 has.

 

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!

[SrMi]

31 minutes ago, Dr. G said:

FWIW, this was taken with the Q3 with a 1/8000 shutter speed.  A few on the forum say that a global shutter isn't necessary, but faster readout times would have probably negated the issue with the wings here.  So if the Sl3 has the same readout speed we'll have some of the same issues the Q3 has.

 

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!

Note that SL3 will have a focal plane shutter, allowing 1/8000 with a mechanical shutter.