Lumix S5 (vs. S1) ISO test shots

[Tom1234]

Advertisement (gone after registration)

Thank you for this S5 post & pictures.  

It seems to show the S5 worthy of Leica M lenses but not perfect near the edges?  

Also the SL2 blacks look slightly, and importantly, better than the S5's on my Benq SW320 with x-Rite calibration.  True to you?

Edited October 19, 2020 by Tom1234
spelling

[Chaemono]

Not sure. The SL2 picture is slightly darker due to faster shutter speed. 

[Advertisement]

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

[Reggie]

Thanks Chaemono for making such a detailed and informative comparison between the Leica and Sigma lenses.

I can see differences between them and I definitely prefer the Leica but at risk of stating the obvious, isn't the principal difference between the images caused by how they handle the out of focus areas? The Leica appears to have a greater DoF which seems odd given the same apertures used. Or if the DoF is indeed the same, the out of focus areas are better focused in the Leica lens than the Sigma which contributes more to the better '3D effect' that the Leica seems to possess.

[Chaemono]

Am 6.11.2020 um 17:25 schrieb Reggie:

Thanks Chaemono for making such a detailed and informative comparison between the Leica and Sigma lenses.

I can see differences between them and I definitely prefer the Leica but at risk of stating the obvious, isn't the principal difference between the images caused by how they handle the out of focus areas? The Leica appears to have a greater DoF which seems odd given the same apertures used. Or if the DoF is indeed the same, the out of focus areas are better focused in the Leica lens than the Sigma which contributes more to the better '3D effect' that the Leica seems to possess.

Yes, some of it is. But the pictures with the Leica lens are brighter in high contrast scenes at the point focus with the same aperture. Look at the motorcycles in #28.

[oculaii]

I'm very surprised by the reviews at Photonstophotos or DX0mark because i have both the S1/S5 and I have come to the same conclusion than Chaemono, the S5 has slightly better noise performance than the S1 up to ISO 12800, pictures are also a bit brighter.

[SrMi]

6 hours ago, oculaii said:

I'm very surprised by the reviews at Photonstophotos or DX0mark because i have both the S1/S5 and I have come to the same conclusion than Chaemono, the S5 has slightly better noise performance than the S1 up to ISO 12800, pictures are also a bit brighter.

PhotonsToPhotos shows a bit better high ISO performance for S5.

[oculaii]

Advertisement (gone after registration)

21 hours ago, SrMi said:

PhotonsToPhotos shows a bit better high ISO performance for S5.

No, your link is for dynamic range. The link for noise is here : Read noise

DXO rates the S1 higher for noise and dynamic range performance. PhotonsToPhotos rates the S5 higher only for dynamic range (your link). They don't agree about dynamic range.
I have not measured the dynamic range between the two cameras but i'm pretty sure the S5 noise performance is slightly better.

[SrMi]

22 minutes ago, oculaii said:

No, your link is for dynamic range. The link for noise is here : Read noise

DXO rates the S1 higher for noise and dynamic range performance. PhotonsToPhotos rates the S5 higher only for dynamic range (your link). They don't agree about dynamic range.
I have not measured the dynamic range between the two cameras but i'm pretty sure the S5 noise performance is slightly better.

Dynamic range is related to noise since it is computed from SNR. However, the Read Noise chart linked has a disclaimer:

"These raw values are not appropriate for comparing camera models because they are not adjusted for gain or area."

However, IIRC, the data should be comparable for the same sensor size and resolution.

The differences between S5 and S1 in PhotonsToPhotos and DxO are minuscule. Therefore, based on charts alone, I would not expect any differences.

[BernardC]

This comes-up a lot in cine circles: there is no good way to measure a sensor's dynamic range, outside of a calibrated lab environment. There probably isn't a good way to do it inside a calibrated lab either, but at least there can be a consistent way.

PtP's methodology is measuring lens flare, monitor calibration, room illumination, room temperature, and firmware optimizations. Theoretically, some of that could even-out with multiple tests from multiple sources, or it could be integrated into the numbers if every test used the same lens/room/monitor. As things stand, their numbers are for entertainment only, and "not appropriate for comparing camera models."

The takeaway is to trust your eyes. All modern sensors come from a handful of billion-dollar fabs, and they all perform more-or-less the same. The real differences, if any, are to be found elsewhere: circuitry, cooling, colour filter arrays, software. Even old-fashioned flocking of the lens cavity. There's a reason why the "Leica look" is maintained over multiple generations of sensors from different manufacturers.

One well-known factor is sensor temperature. You'll get huge variation with a single sensor and a few degrees of temperature shift. That's why cine cameras have giant heat sinks, and fans, and warm-up procedures. Even then, you need to recalibrate them if ambient temperature goes above or below certain values!

[Archiver]

On 11/11/2020 at 3:50 PM, Chaemono said:

Yes, some of it is. But the pictures with the Leica lens are brighter in high contrast scenes at the point focus with the same aperture. Look at the motorcycles in #28.

Just wondering, do you have a M10 or M10-R? I'd like to know if you have any thoughts or comparisons between the M10 and the S5.

I've had my S5 for about six months and found it to be a (gasp) decent alternative to my M9 when used with the right M mount lenses, like the Distagon 35 and Summicron 50 v5. It's not as good in the corners and edges with the Elmarit 28 Asph, or the Biogon 21mm f2.8. But the colours can be just as good as the M9, albeit with a slightly different palette, and of course the dynamic range and high ISO capabilities are better. As the M10 has a more modern sensor, I would be interested in any thoughts or comparisons you might have.

[Chaemono]

vor einer Stunde schrieb Archiver:

Just wondering, do you have a M10 or M10-R? I'd like to know if you have any thoughts or comparisons between the M10 and the S5.

I've had my S5 for about six months and found it to be a (gasp) decent alternative to my M9 when used with the right M mount lenses, like the Distagon 35 and Summicron 50 v5. It's not as good in the corners and edges with the Elmarit 28 Asph, or the Biogon 21mm f2.8. But the colours can be just as good as the M9, albeit with a slightly different palette, and of course the dynamic range and high ISO capabilities are better. As the M10 has a more modern sensor, I would be interested in any thoughts or comparisons you might have.

S5 has a BSI sensor. These capture a lot of light so one needs more negative EV to expose for the highlights than with the M10. I prefer the M10-P between ISO 200-640. After that the S5 pulls ahead for color fidelity. There’s color shift in the corners with the S5 and M lenses, though. See fourth and fifth picture here with a 50mm: https://www.smugmug.com/gallery/n-GChFpc/

[Archiver]

3 hours ago, Chaemono said:

S5 has a BSI sensor. These capture a lot of light so one needs more negative EV to expose for the highlights than with the M10. I prefer the M10-P between ISO 200-640. After that the S5 pulls ahead for color fidelity. There’s color shift in the corners with the S5 and M lenses, though. See fourth and fifth picture here with a 50mm: https://www.smugmug.com/gallery/n-GChFpc/

Thank you for the reply, I appreciate it. Incidentally, I find that the S5 underexposes, especially in bright sunlight, and I have to raise EV by over a stop to make it look 'normal'. Here's an example of such an image which needed EV adjustment in post. Perhaps Panasonic make the S5 underexpose in order to preserve highlights?

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!

I'm really enjoying the S5, and yet I still want the M body experience, so I've been curious about how the sensors perform in relation to each other.

 

Edited January 3, 2022 by Archiver

[Chaemono]

vor 1 Stunde schrieb Archiver:

Thank you for the reply, I appreciate it. Incidentally, I find that the S5 underexposes, especially in bright sunlight, and I have to raise EV by over a stop to make it look 'normal'. Here's an example of such an image which needed EV adjustment in post. Perhaps Panasonic make the S5 underexpose in order to preserve highlights?

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!

I'm really enjoying the S5, and yet I still want the M body experience, so I've been curious about how the sensors perform in relation to each other.

 

I’ll do a high contrast scene, M10-P at base ISO 200 vs S5 at base ISO 100 with the same lens  EV set to 0. It’ll take me a while. I may throw in the α7 III as a bonus.

[jaapv]

2 hours ago, Archiver said:

6 hours ago, Chaemono said:

 

Thank you for the reply, I appreciate it. Incidentally, I find that the S5 underexposes, especially in bright sunlight, and I have to raise EV by over a stop to make it look 'normal'. Here's an example of such an image which needed EV adjustment in post. Perhaps Panasonic make the S5 underexpose in order to preserve highlights?

Not my experience. Sample variation?

[Archiver]

51 minutes ago, jaapv said:

Not my experience. Sample variation?

Interesting, maybe so. Setting the S5 to about +1 EV helps in bright scenes like the one above. Otherwise the image is just too dark.

[jaapv]

In practice I find that “correct “ exposure will vary between -2 and +1.5 on all cameras so I always have the histogram activated in the viewfinder. OTOH pulling up is better than pulling down on a high DR sensor. 

[Archiver]

14 hours ago, jaapv said:

In practice I find that “correct “ exposure will vary between -2 and +1.5 on all cameras so I always have the histogram activated in the viewfinder. OTOH pulling up is better than pulling down on a high DR sensor. 

Good suggestion about histogram, I hardly use it at all. Interestingly, the Panasonic G9 doesn't underexpose like the S5 at all, it seems to have much more accurate matrix metering. Same with all the other Panasonic m43 cameras I have. Only the S5 behaves like that.

[jaapv]

31 minutes ago, Archiver said:

hardly use it at all.

Fairly typical for those who cut their teeth on optical viewfinders. I had to be firm with myself to get into the habit  

[SB4791]

@Archiver I also have an S5 (and an M6 for film) and also find that it seems to regularly underexpose. I’ve always had the histogram visible in the view finder and find myself glancing at it seems much more reliable than the image you see in the view finder.

[Advertisement]

Hallo,

Look here: Leica L-Mount