Does the S1Rii have pixel binning like Leica Q3?

[dirk]

Hi,

I saw in a German review, that there is the possibility with the Lumix S1Rii to choose between different MP resolution.



How does this work? Is this for jpeg only or also for RAW files?

Is this the same way like with the Leica Q3? Full 24x36mm sensor size but less pixel by combining them like in a smartphone?

I saw this setting in no other review, but I watched too many anyway

Maybe @Richard Wong can check this?

 

[Richard Wong]

Pretty sure this is jpg only like previous Lumix cameras. But I can double check later and let you know

 

[dirk]

Pretty sure this is jpg only like previous Lumix cameras. But I can double check later and let you know

Yes please, thank you.

And ask them how it works and whether it works like with the Q3.

If it would be real pixel binning like in the Q3, this would be a game changer for me.

 

[GeorgeHudetz]

Yes please, thank you.

And ask them how it works and whether it works like with the Q3.

If it would be real pixel binning like in the Q3, this would be a game changer for me.

Yes indeed - mark me as interested as well. This could be very useful for astro.

 

[Richard Wong]

Just checked. it's for JPG only.
I'm curious, since both of you don't shoot sports, what makes small RAW so special/attractive?

btw, I've received quite a number of test requests/things want me to test from some of you guys but mostly from Youtube/FB/IG, I will try my best to do all for you guys!
But I do still have 2 more small S1RII videos I want to create and a few other reviews need to publish this month so I'm sorry If I couldn't get around to do some of the tests you guys have requested.

 

[GeorgeHudetz]

For me, it's astro. If the binning does indeed increase the DR by one stop, that could be very valuable for astro. Theoretically, it should be similar to using a lens that is one stop faster - which is to say, it should cut the exposure time in half to reach a desired noise level. That's important when you are shivering in the cold, LOL.

I don't fully understand the mathematics of binning, but as I have seen it explained it, it's different than simply reducing the size of the image. There is a YT channel called "mathphotographer" who did a deep dive into the binning on the SL3 and he seems to validate that it does increase the DR of the raw file as compared to the full 60 MP image.

 

[GeorgeHudetz]

For me, it's astro. If the binning does indeed increase the DR by one stop, that could be very valuable for astro. Theoretically, it should be similar to using a lens that is one stop faster - which is to say, it should cut the exposure time in half to reach a desired noise level. That's important when you are shivering in the cold, LOL.

I don't fully understand the mathematics of binning, but as I have seen it explained it, it's different than simply reducing the size of the image. There is a YT channel called "mathphotographer" who did a deep dive into the binning on the SL3 and he seems to validate that it does increase the DR of the raw file as compared to the full 60 MP image.

And, I will admit that I don't know if the binning in the SL3 would be effective for astro in the way I describe above. It's just an assumption on my part that needs to be validated.

 

[xaviergut]

btw, I've received quite a number of test requests/things want me to test from some of you guys but mostly from Youtube/FB/IG, I will try my best to do all for you guys!

Richard, you are doing a lot of work and we benefit from it.

Have you got a Patreon account, to be able to support you in the fantastic work you do analyzing cameras, lenses and photo and videographic equipment in general?

I have seen content creators with fewer subscribers than you on YouTube who have a community of fans on Patreon who collaborate with them financially.

 

[Richard Wong]

For me, it's astro. If the binning does indeed increase the DR by one stop, that could be very valuable for astro. Theoretically, it should be similar to using a lens that is one stop faster - which is to say, it should cut the exposure time in half to reach a desired noise level. That's important when you are shivering in the cold, LOL.

I don't fully understand the mathematics of binning, but as I have seen it explained it, it's different than simply reducing the size of the image. There is a YT channel called "mathphotographer" who did a deep dive into the binning on the SL3 and he seems to validate that it does increase the DR of the raw file as compared to the full 60 MP image.

I haven't test the Q3/SL3's small RAW, but when I previously tested the small RAW output from Nikon cameras, I saw no difference in terms of image quality compared to resizing a full size raw ourselves in Photoshop.

The only way I can think of that would give us better DR would be if the sensor has a different readout mode that exposes some pixels shorter than the normal pixels.
Some quad Bayer sensor allows that, for example the Kandao QooCam3 Ultra has a special mode that increases dynamic range (preserve more highlight) by doing exactly that when you choose the lower resolution output. However since some pixels are now exposing at shorter interval, the trade-off is that lowlight performance is worse than the normal mode.
Still it is a pretty fascinating trick.

Does the Leica Q3/SL3 sensor have this kind of trick?
If not, the smaller RAW should be just summing/averaging the value from 4 or more pixels into one. So mathematically, it should be the same as doing downsizing in Photoshop as long as you choose the correct re-sampling algorithm.

That's my understanding at least but I would love to learn more if anyone know more about it

 

[Richard Wong]

Richard, you are doing a lot of work and we benefit from it.

Have you got a Patreon account, to be able to support you in the fantastic work you do analyzing cameras, lenses and photo and videographic equipment in general?

I have seen content creators with fewer subscribers than you on YouTube who have a community of fans on Patreon who collaborate with them financially.

hey @xaviergut I'm glad to hear you found my videos useful
Regarding Patreon, no, i don't have one. While I'm definitely not against the idea of having more money, it is not my intention to start my YouTube channel. Frankly, I feel uncomfortable asking my viewers to donate money to me. I also turned down pretty much all the sponsor deals as I really don't want to add ads in my video as well.
Maybe one day I will change, but right now I feel a lot happier when I heard my reviews help you guys make the right purchase decision.

 

[dirk]

Just checked. it's for JPG only.

Thank you. Only jpeg. That is a bummer.

I'm curious, since both of you don't shoot sports, what makes small RAW so special/attractive?

For me there are 2 reasons:

1. Higher DR. I did not really test it on my Leica Q3 43 yet, but Leica claims that you get a higher DR. Nowadays DR is already very good, therefore for my kind of photography this would not be very important. But for others that might be an argument to buy the camera for sure.

2. Simple smaller file sizes, less storage space, faster writing times. Makes my life easier. I do not need 44MP in 98% of my photography. 24MP is enough for me. I shoot always Jpeg & Raw at the same time. But this takes a lot of space, which is unnecessary for me.

I have on my Leica Q3 43 always 33MP activated. Only if I switch i.e to 70mm, I use a different file size. Or if there is a tremendous good looking scene. For sport it is the additional zoom range I could get with a small telezoom and the S1Rii.

Do not underestimate file sizes. The main reason why I never bought the Lumix G9ii and stick with my G9 and Oly EM 1.3 and why I do almost never use my XT5 is the file size. With the XT5 you have even 70 MP!

If you need more ideas what to check with Panasonic for future S1Rii update short videos, like your writing speed video:

1. How exactly does this smaller file size works technically and influence in reality DR etc. Why not in RAW too?

2. EVF. Does resolution or fps changes, if you press the shutter (see Sony A7Rv)

3. Will this very slow writing time with CF express and SD card be improved by Panasonic?

4. Can Panasonic add in a future FW update in the menu the option to have i.e. only 20 fps, but writing on the card starts already from the beginning, so your fps drops not that much/ not to zero, once the buffer is full?

5. Overheating. Every reviewer has different claims, depending on whether they use CF express or SD or both, 8k, 4k and the different fps, temperature outside and their FW version etc pp.

It would help the marketing of the S1Rii, if there would be new tests with the latest FW in all kind of possible settings-combinations to have a reliable guide when overheating starts really.

 

[Richard Wong]

Thanks Dirk! I can answer some of your questions:
1. I believe the smaller JPG is just resized (downsized) JPG. I don't believe changing in RAW size affects dynamic range at all. Smaller raw may have better signal to noise ratio at PIXEL LEVEL, but you could achieve the same if you resize your full size raw to smaller raw. As I said in one of the post above, I previously tested some RAW file from Nikon cameras and I found zero difference between a full size raw and a small raw in dynamic range or image quality, once I normalise to the same size. So unless the sensor has a special High dynamic range readout mode, I still don't believe raw size would affect dynamic range.
2. Can't comment on FPS, but I don't see any resolution change at all, no matter half or full press and I tried different burst speed.
3. I don't know, pretty sure Panasonic won't answer even if I ask . But as I said in my review, I believe the bottleneck is the processing when you shooting in SH mode with a fast card. I noticed similar behaviour when testing GH7, G9II...etc so my GUESS is no. But we will have to see.
4. I've asked. I believe 20fps or even 15fps would benefit a lot of users. But Panasonic won't answer about future firmware. Re fps drops, unless the current processing speed is improved in the future, otherwise lowering to 20fps would only allow you to shoot longer before you fill up the buffer. Once the buffer is full, you still need to wait same time before you can capture the next photo, and the rate you can shoot should be identical as shooting at 20fps wouldn't reduce processing/write time for each photo.
5. There are just too many variables that would affect the overheating test results. I found that when doing the overheating test in my studio room, even when I was testing on same day.. since two consecutive tests would be at least 2-3 hours apart (test is like 1-2 hour, + at least 1 hour to let the camera to cool down) the difference in room temperature is enough to make two consecutive identical test give me quite different results. So that's probably why all reviewers would give you different results.
But I can tell you I can shoot 1 hour+ 10bit 4K video easily indoors even though it is still summer here in NZ, and the S1RII overheats easier than the S5II and G9II. Also when I did some side by side rolling shutter test with a Sony A1 this weekend, the Sony A1 overheated much faster than the S1RII when both shooting 8k video.

 

[RBecker]

Hello Richard,
Thanks for your work! I appreciate also your task, to not use adds in your videos. Hope you can go on with this…

One question to the write speed bottleneck of the S1RII. I saw your video testing all the different options for memory cards. I hope that Panasonic did not use just the old interface from the S5II and connected a CFE slot to a limited transfer rate. Then FW could help, otherwise not.
Have you tested the 40 picture/sec with saving to an external SSD? That should be the fastest transfer possible. Would be very intersting what the camera can do with a SSD…

 

[RBecker]

Additional question:
All the old Lumix cameras offered a feature to transfer stills to a PC with a shared folder via WLAN. Tethering without cable and without any need of an additional app or software. I used this in the studio. Unfortunately this feature is deleted now in Europe with the last FW updated from January from all the actual models (the S5IIx also lost the live streaming feature).

Is your S1RII still offering the feature to transfer pictures without cable to a PC? You are far from Europe… ;-) Or maybe Panasonic has decided to stop this function with new models everywhere?

 

[xaviergut]

hey @xaviergut I'm glad to hear you found my videos useful
Regarding Patreon, no, i don't have one. While I'm definitely not against the idea of having more money, it is not my intention to start my YouTube channel. Frankly, I feel uncomfortable asking my viewers to donate money to me. I also turned down pretty much all the sponsor deals as I really don't want to add ads in my video as well.
Maybe one day I will change, but right now I feel a lot happier when I heard my reviews help you guys make the right purchase decision.

I just can say thank you, Richard!

 

[RBecker]

For me, it's astro. If the binning does indeed increase the DR by one stop, that could be very valuable for astro. Theoretically, it should be similar to using a lens that is one stop faster - which is to say, it should cut the exposure time in half to reach a desired noise level. That's important when you are shivering in the cold, LOL.

I don't fully understand the mathematics of binning, but as I have seen it explained it, it's different than simply reducing the size of the image. There is a YT channel called "mathphotographer" who did a deep dive into the binning on the SL3 and he seems to validate that it does increase the DR of the raw file as compared to the full 60 MP image.

Binning was „invented“ with astrophotography in the 90s to raise sensitivy. I did my diploma at physics at that time with photometry of stars using first commercial CCD cameras from that time (ST6, ST7 if someones remembering these…).
Every pixel is like a barrel collecting photons and with binning during the read out all the collected photons where put together in a bigger 2x2 or 3x3 barrel (to keep it simple). With this trick the sensitivity could be raised and also the signal to noise ratio was better.
BUT: these cameras where all monochrome! With todays bayer sensors and using a real 2x2 binning you would mix 2green, 1red and 1blue pixel from the matrix. Could be also possible but requires a totally different processing! These 4 pixels would create a real RGB pixel by loosing resolution and creating high demand on memory for this pixel. I think thats the reason why nobody think about it.
But Canon offers some years mRAWs with reduced resultion, but I habe no idea how these are created.

 

[Quentinquirelino]

What could be interesting for wildlife shooters, would be a cropped RAW mode. We know that the sensor can read much faster, when it reads not the whole sensor. A 1.4x or APS-C crop should bring the rolling shutter way down to about 10ms or faster and there would be still around 20 Mpix left. That would be comparable to a OM-1 in terms of read out speed and resolution, but the sensor area would be still larger. The gain in reach would be welcome for such types of photography as well.

 

[GeorgeHudetz]

Thanks everyone for the replies. @RBecker that is - I think - exactly what the SL3 does. Again "mathphotographer" on YT does a deepdive on the SL3 binning feature and does show that the 33 MP raw file has 1 stop more DR than the 60 MP raw file. If you ever have the time to watch it, I'd be curious if you think his methodology is good.

And @Richard Wong yes, Leica does seem to have this trick.

If I were to ask for two astro enhancements to the S1RII, it would be:

1. Let us go up to 4 (or even 8?) minute exposure times, rather than the current limit of 60 seconds, and
2. Bring in binning for a super-high DR 22.5 MP raw file.

I also wonder if the fan can be used for LE stills to keep the sensor cooler. Not if it actually cools the sensor or the processor, however.

 

[RBecker]

.

If I were to ask for two astro enhancements to the S1RII, it would be:

1. Let us go up to 4 (or even 8?) minute exposure times, rather than the current limit of 60 seconds, and
2. Bring in binning for a super-high DR 22.5 MP raw file.

I use a modified S5 for astrophotography and the 24MP sensor is doing very well, especially in terms of dark current. So with stacked pictures I can do already extreme editing in post…

My first wish for an enhancement is an astrotracer like Pentax provides.
To get exposure times up to some minutes without mount just using a tripod would be awesome for travelling.
Todays cpu in the cameras are good enough to calculate the necessary movement for the IBIS from one or some testpictures and this process is free from patents. The only problem that you need some firmware programmers who understand the logic of the necessary routines and the decision that this would be a cool feature…

 

[RBecker]

Thanks everyone for the replies. @RBecker that is - I think - exactly what the SL3 does. Again "mathphotographer" on YT does a deepdive on the SL3 binning feature and does show that the 33 MP raw file has 1 stop more DR than the 60 MP raw file. If you ever have the time to watch it, I'd be curious if you think his methodology is good..

I havn‘t look the video yet, but I will do!
In general: If you reduce the resolution from 61MP to 33MP by interpolation you will reduce noise and by reducing noise you gain DR. The same happens if you stack multiple pictures like doing with HighRes mode or in Post. That is why I use HighRes with my OM1 and the S1R already some time (if possible), because the RAWs are much better for editing in post. It is a real push for city nightscapes or architecture, especially for the smaller MFT sensor!
It would be interesting to get info from a programmer who works and decodes raw files. There are free raw tools and Leica uses DNG, so the logic should be freely documented. I raw just from smaller resolution alone seems not simple…