Jump to content

SL2-S


nicci78

Recommended Posts

Advertisement (gone after registration)

2 hours ago, Slender said:

I am not sure the SL2 uses line skipping, but pixel mixing... that can explain the alliasing issues in some FF modes.. which also give us this wonderful sharpness in the image. In s35 the rolling shutter falls below 10ms and is very good. I am still getting more jobs in still than moving image atm, but pairing/downgrading to SL2-s could be an interesting options for video work indeed.

Yes, in FF the SL2 uses pixel mixing.   Sensor readout is 1/60 th second (16.6 ms).  In crop mode (s35) it is less than 1/100 th s = less than 10 ms.         (It makes good use of the Maestro III)

The Sony a9 II is faster, but this is probably not the new sensor for the SL2-S , or is it ?  The predicted price is higher than the a9, so they could maybe afford a high end sensor in the SL2-S....  But Sony will maybe not give it away, unless they have something new that is better.

It is a BSI sensor, does this mean it is from Sony ? Or do others also produce BSI sensors ?

Edited by caissa
Link to post
Share on other sites

15 minutes ago, caissa said:

Yes, in FF the SL2 uses pixel mixing.   Sensor readout in 1/60 th second (16.6 ms).  In crop mode (s35) it is less than 1/100 th s = less than 10 ms.         (It makes good use of the Maestro III)

I bet it's exactly the same as the S1R, so pixel binning, it does not matter how they are naming it.
Probably same 1.09x crop, same rolling shutter, except the S1R is more crippled for video, regarding no Log and more compressed codecs.
The S1/S5 sensor is simply a better fit for video as you can see in this comparison (except the 1.5x crop 4k60p versus 1.09x crop, but worse quality)
 

 

Link to post
Share on other sites

1 minute ago, caissa said:

There is a description in the web. It is definitely pixel mixing for the SL2 in FF mode.

It is in a video magazine (I think from UK), from Nov 2019. I try to find it again.

It does not matter, if the quality, rolling shutter, 1.09x crop is exactly the same as the S1R, it is the same thing, e.g. pixel binning with a puffed-up name. Only the codecs and the presence of Log which are different. It does not mean that the SL2 is not a quantum leap over previous Leicas for video.

But not compared to other mirrorless cameras, and the SL2-S is more up-to-date while being cheaper.

Link to post
Share on other sites

And no, it is not the same as the S1R. (Why do you say that ? You did not verify.) Just the sensor resolution is the same. The SL2 is much faster. Why else offers the S1R 9 FPS and the SL2 20 FPS.   Check the specs. Also the video specs are clearly different. (Too complicated to write down here. Simply compare the official spec sheets).

Edited by caissa
Link to post
Share on other sites

I found the description again, It was in the fdtimes.(= film and digital times) In the article below (Report on Leica SL2), in the section  Leica SL2 for Cine . (Pixel mixing    and following lines).

https://www.fdtimes.com/2019/11/06/leica-sl2/

The explanations are written by Stephan Schulz. Who if not he should know the details better ?

Edited by caissa
Link to post
Share on other sites

Advertisement (gone after registration)

2 hours ago, padam said:

It does not matter, if the quality, rolling shutter, 1.09x crop is exactly the same as the S1R, it is the same thing, e.g. pixel binning with a puffed-up name. Only the codecs and the presence of Log which are different. It does not mean that the SL2 is not a quantum leap over previous Leicas for video.

A couple of comments. Isn't the "1.09x crop" just the difference between the 3:2 (1.5:1) still ratio and 16:9 (1.77:1) video ratio? If so, it's hardly worth mentioning, given that every "full frame 35mm" still sensor has a 3:2 ratio.

"Pixel Binning" was term used a lot during the HD days. It's not as accurate here, since there is a necessary conversion between RGGB Bayer data and RGB video. You can call it over-sampling instead; each RGB pixel uses colour information from 4 pixels instead of 3. It's really nothing new, or special. The Arriflex Alexa did this from day 1, as have many other high-end video cameras. You need the additional samples in order to get clean colour information for green-screen/blue-screen/compositing work. I am just nit-picking, of course, but "pixel binning" has negative connotations, which aren't accurate here.

  • Like 1
Link to post
Share on other sites

5 hours ago, Stuart Richardson said:

I find it interesting at all the criticism directed towards the camera. Leica is doing what every single other company has done, which is to sell a lower res version of the same body to market to people who: 1. Do not care about resolution. 2. Prefer better high ISO performance. 3. Want better video. 4. Want to pay as little as possible for the modern tech. If you don't want to do that, you can get the more expensive higher resolution model, which in this case is the SL2. They are doing literally the exact same thing as Sony, Panasonic, Canon and Nikon. Clearly this is a strategy that works...you save on all the tooling and R&D of creating a completely different body, and you satisfy a greater proportion of users. If you introduce the more expensive one first, you get more buyers at a higher profit as well, as you sell cameras to the people who would want that body, but also to the people who would be satisfied with the cheaper one, but prefer to buy the newer model immediately. Those same people might still buy the lower rez body later, meaning you get two sales. Once you are done satisfying demand for the higher rez body, the lower cost one comes out and you get a wider group of buyers as well as some of the initial group who might have wanted it as a backup or for the low light/video abilities. It seems like this is one of those cases where everybody wins...buyers get more choice and the company gets more sales for minimum additional effort.

The negative is not about this offer itself but the features offered and the price ratio between High Res vs Low Res one. Yes it has been a nice model used by others. Nikon Z: 3300Z7(or 3000Z72)/2000Z6, Canon R 3900R5/2500R6, Sony A 3500A7R4/2000A73. Panasonic 3500S1R/ 2500S1or2000S5.  Here with Leica is 6000/5000.

If there is new features such as smaller size, flip screen, it would be justified. Just saying. As a SL2 owner, I am happy to see this happen so I don't need to struggle to pick which one, it is a very easy decision for me, TBH.  

Edited by ZHNL
Link to post
Share on other sites

1 hour ago, BernardC said:

A couple of comments. Isn't the "1.09x crop" just the difference between the 3:2 (1.5:1) still ratio and 16:9 (1.77:1) video ratio? If so, it's hardly worth mentioning, given that every "full frame 35mm" still sensor has a 3:2 ratio.

"Pixel Binning" was term used a lot during the HD days. It's not as accurate here, since there is a necessary conversion between RGGB Bayer data and RGB video. You can call it over-sampling instead; each RGB pixel uses colour information from 4 pixels instead of 3. It's really nothing new, or special. The Arriflex Alexa did this from day 1, as have many other high-end video cameras. You need the additional samples in order to get clean colour information for green-screen/blue-screen/compositing work. I am just nit-picking, of course, but "pixel binning" has negative connotations, which aren't accurate here.

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!



Of course not, no crop means full width of the sensor with whatever aspect ratio the camera is recording at. (For instance some Canon cameras use Cinema 4k so for 16:9 one has to crop in slightly from the full width to achieve that)
This is the S1R, the SL2 seems to be a mystery as even full reviews don't seem to talk about it but I don't see a lot of vignetting so I assume it is slightly cropped in.
I can't find anything reliable about it other than it can record in C4K as well as UHD, I would tend to think in the first case it would be the red area except very slightly cropped in top and bottom.

In any case, the camera throws away a handful amount of data, and you can see the effect of that in the previous video.
It's not bad overall, but there is aliasing, more noise and less dynamic range (does not tolerate under- or overexposure), compared to using all the available pixels to record video, which is just a little bit slower readout, but it can provide a fairly clean video image even at ISO 25600.
Some may ask, is that actually useful?
The answer is a definite yes. It does provide the benefit of using more depth-of-field in low-light if necessary, not forced to using apertures f/2 or lower, which is fine for nailing one single image, but pretty shallow for FF video at close range, where it might constantly drift in and out of focus. Same with vignetting or uneven bokeh in the background, appealing for stills but it may be a bit distracting for video as cinema lenses are known to have a very even rendering across the frame (and of course as I wrote the 1.4-1.5x 1:1 crop (C4K/UHD) might have external ProRes RAW video recording enabled later on, which is another potentially big benefit).
 

OK finally found it, as expected it is the same 1.09x crop as the S1R.

The SL2-S will not have any crop on the sensor width (and it does actually matter if one is shooting with primes, which is a likely scenario with Leica lenses).

Edited by padam
Link to post
Share on other sites

On 10/29/2020 at 9:26 AM, BernardC said:

Leica probably knows exactly what market share they want, and I would be surprised if their ideal share made them into a mainstream player. An important part of that strategy is not trying to be everything to everybody. There's an oft-reoccurring thread here in which people demand (demand!) that Leica release super-telephotos, the like of which only Nikon and Canon have ever been able to sell. You'll read fortellings of doom if Leica doesn't design/productise/manufacture/sell these lenses right away.

I'm sure it happens in every brand forum: "I absolutely need a feature that is offered by a different brand. My brand should drop everything and implement it right now!" There is no perfect camera.

If you want some reach, put a Hasselblad HC 300 on a Leica S. There's a reason I sold my Canon big white lenses.

Link to post
Share on other sites

On 12/5/2020 at 10:17 AM, LocalHero1953 said:

That is the voice of reason, and I wouldn't criticise it - it's what I aspire to. But all the fans of the S system give as their primary justification the gorgeous image quality. One day I'm just going to have to borrow or rent one to try for myself.

I bought an S 007 new at the end of 2016. I have the 24, 45, 70CS, 100 and 180. Also via the S to H adaptor, I have the HC 35, 50-110, and 300. I have an SL with 16-35, 24-90, and 90-280. I have a holdover Otus 55mm that I use on the SL with a Novoflex adaptor. I sold 3 Canons- 2 5DSR's and 1 5D4 plus eleven of Canon's best lenses for the SL and 3 lenses. That was a good decision. Regarding the S, it easily best the SL and Canons. The 5DSR with the Otus attached had more resolution and produced a nice look, but you would likely choose the S over anything mentioned above when comparing images side-by-side. 

  • Like 1
Link to post
Share on other sites

And Leica Rumors wrongly predicted a size difference.  This is the third time they messed up on predicting a compact Leica, having done it twice regarding the phantom C/M. And they (Peter) also twice wrongly predicted the timing of the SL launch, eventually being almost a year early. Not hard to outperform their reporting the last couple of years.

Jeff

Link to post
Share on other sites

3 hours ago, padam said:

OK finally found it, as expected it is the same 1.09x crop as the S1R.

Thanks. In that case it's done so they can get an even number of pixels for over-sampling.

According to specifications, the SL2 has 8368 pixels horizontally. UHD is 3840 pixels across. Double that (=7680), and multiply by 1.09. The actual crop ratio is 1.0896 (if the pixel numbers are correct), which is close enough to 1.09.

In other words, they over-sample by a ratio of 2:1 to get UHD out of the 47MP sensor. Which is exactly what anyone would do with the same sensor, since oversampling 2.17916667:1 would give you more aliasing. Try resizing an image by 8.96% in an image editor to get an idea of what would happen.

Link to post
Share on other sites

36 minutes ago, Steven said:

18 days before the rumour websites ! What else do you guys want to know ? 😜😂😂

If I recall correctly, you stated specifically who your source was, didn't you? People at Leica read these threads, so if you want to protect your source (or at least not cause trouble for them), as well as keep getting information, you might have to be a bit more discreet. Then again, maybe I misremembered what you said!

Link to post
Share on other sites

6 minutes ago, Jeff S said:

Leica Rumors wrongly predicted a size difference.

They translated a Japanese web site. I guess Japanese uses similar phrasing for MP count and for LxWxH dimensions. Not a big deal, especially since they almost always mention their source and provide a link.

They often keep Chinese or Japanese phrasings, which makes it clear that they are providing a machine translation. Anyone who wants to read the source in its original language is free to do so.

Link to post
Share on other sites

3 minutes ago, Stuart Richardson said:

If I recall correctly, you stated specifically who your source was, didn't you? People at Leica read these threads, so if you want to protect your source (or at least not cause trouble for them), as well as keep getting information, you might have to be a bit more discreet. Then again, maybe I misremembered what you said!

You remember correctly. My source didn’t mind sharing the information though. He didn’t seem to think it was highly classified. It seemed rather well know among the Leica team. 
 

So for those of you interested, what he insisted on the most when I spoke with him was the variety of codecs that will be offer with the sl2s. It seemed so impressive that I would be shocked if they announced external raw. 
 

the real mistery to me will be wether or not they will add the same AF improvements as the s5. I would bet that they won’t. Not at launch at least. 

  • Thanks 2
Link to post
Share on other sites

SL2 had several promised functions that did not appear until firmware 20, about 6 months into the product life.  This is getting to be a pattern -- get the product out, state the goals, and then work over time to deliver all of them, but not all at once...  I'll bet some of the SL2-S video facilities take their time arriving.

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...