Jump to content

Recommended Posts

Advertisement (gone after registration)

Really nice.  It will be interesting to see comparative images made with both methods from the same camera.  

From a logical perspective, the the high res, in-camera image should reveal considerably more color detail and perhaps even greater overall detail, without the potential jaggies caused by software algorithms.  That said, I've begun to use Photo Zoom in place of ON1 for up sizing SL files ... it appears to upsize without jaggies.

I'll have to wait until my pre-ordered S1R arrives to make that comparison.

Edited by ron777
Link to post
Share on other sites

2 hours ago, ron777 said:

ON1 and similar upsizing software essentially clone existing data by virtue of their proprietary algorithms, the in-camera technology described above, however, derives a larger file from the data accumulated via multiple image acquisitions, the result of which should be far superior to that produced by upsizing software.  Whether or not that holds true remains to be seen, but logically, it should bear fruit. 

Up-rezzing and up-sizing software uses a process called "interpolation" where it creates a matrix of new, empty pixels around each existing pixel and uses intelligent algorithms to fill the empty pixels with data that 'fits' based on the existing data in neighbouring pixels.  This is an iterative process, ie it uses 'compare and repeat' cycles to achieve the best fit and yes it does create new data to fill the empty pixels in the matrices.  This used to be a bit clunky owing to the computing power available so jaggies and other data errors would appear but the massive increase in sheer computing power that's available today, largely driven by the computer games industry, allows a massive increase in the number of iterative cycles which has driven the data error rate close to the notional limit or entropy.

I work with systems that transfer Terabytes of data per second over a single optical fibre link of, say 150 km, where data error correction software is crucial to preventing systems clogging up with 'bad' data and slowing down or killing systems such as millions of Visa transactions per second.  None of us would want our bank cards wrongly charged with someone else's purchase owing to incorrect data and it's error correction software working in the background that prevents this from happening.  The increase in proficiency of this software even over the past 5 years has been simply amazing and, for example, is what is enabling NFC (Near Field Communications) technology - 'touch card' systems - and the proliferation of places that offer this convenience seemingly instantly.

The point is that these massively powerful error correction software algorithms are also at the heart of interpolation and up-rezzing software so I'm also interested to know whether pixel-shifting technology is as far ahead as we might suppose and whether it produces far superior images because, after all, the same process is going on at the heart of pixel shifting software, ie iterative compare and repeat cycles albeit from the other direction: an overabundance of data versus created data; both boil down to which data to pick as the best 'fit'.  While I'm happy to suppose that pixel-shifting should be superior, up-rezzing software is catching up up fast and not all vendors software is the same.  It's likely to be an interesting comparison. Either way, we the photographers are the winners through increased options. 😊

Pete.

  • Like 4
Link to post
Share on other sites

So far i have been skeptical of up-sizing images except for one image that presented no great challenge due to lack of detail. But that is today; what comes tomorrow is still in my skeptical bin.

Here is one image which is likely impossible to view. Up-sampled from 35mm 2475 film.

 

  • Like 1
Link to post
Share on other sites

Only very few reports available right now, but the Lumix G9 is already longer available and is a “close relative” of the S1R. So here from a Lumix expert his experiences with G9 high res photography.  That should give a rough idea of the way it will work on S1/R.

https://www.lumixgexperience.panasonic.co.uk/learn/expert-advice/working-in-high-resolution-mode-with-the-lumix-g9/#.XIWchRoo-hA

Edited by caissa
  • Thanks 1
Link to post
Share on other sites

10 hours ago, caissa said:

Only very few reports available right now, but the Lumix G9 is already longer available and is a “close relative” of the S1R. So here from a Lumix expert his experiences with G9 high res photography.  That should give a rough idea of the way it will work on S1/R.

https://www.lumixgexperience.panasonic.co.uk/learn/expert-advice/working-in-high-resolution-mode-with-the-lumix-g9/#.XIWchRoo-hA

mmmmm ...... quite a lot of restrictions on the settings that can be used on the G9 ..... (ISO, longest exposure, aperture) which would make it a bit limited for landscape if these are similar in the S1R. I recall a recent video where they took images and used the function to give a water smoothing effect by having motion blur active..... instead of taking a long exposure. 

Edited by thighslapper
Link to post
Share on other sites

Also diglloyd has made his reflections regarding use of adapted lenses on the Lumix S cameras. And interestingly he thinks that S1 and S1R and SL have a "huge lead" over CaNikon mirrorless. Because of the native SL lenses and because there are many adapters for many other brands.  https://diglloyd.com/blog/2019/20190201_1144-PanasonicS1R-LMount.html
He also mentions the Leica M to SL adapter (from Leica) and automatically assumes that all M lenses can be used. He did not explicitly test them, but we can safely (?) assume that the lenses working on the X1D will also work well on the S1R. 

This is not a proof, but an indication that the S1R will not be too bad with M lenses.

Edited by caissa
Link to post
Share on other sites

42 minutes ago, caissa said:

Also diglloyd has made his reflections regarding use of adapted lenses on the Lumix S cameras. And interestingly he thinks that S1 and S1R and SL have a "huge lead" over CaNikon mirrorless. Because of the native SL lenses and because there are many adapters for many other brands.  https://diglloyd.com/blog/2019/20190201_1144-PanasonicS1R-LMount.html
He also mentions the Leica M to SL adapter (from Leica) and automatically assumes that all M lenses can be used. He did not explicitly test them, but we can safely (?) assume that the lenses working on the X1D will also work well on the S1R. 

This is not a proof, but an indication that the S1R will not be too bad with M lenses.

What you mean when you say lenses working on the X1D will work on the S1R? you mean M lenses working fine on the X1D?

There are not a lot of M lenses that work fine on the X1D

 

Edited by Fedro
adding context
Link to post
Share on other sites

On 3/10/2019 at 3:54 PM, farnz said:

Up-rezzing and up-sizing software uses a process called "interpolation" where it creates a matrix of new, empty pixels around each existing pixel and uses intelligent algorithms to fill the empty pixels with data that 'fits' based on the existing data in neighbouring pixels.  This is an iterative process, ie it uses 'compare and repeat' cycles to achieve the best fit and yes it does create new data to fill the empty pixels in the matrices.  This used to be a bit clunky owing to the computing power available so jaggies and other data errors would appear but the massive increase in sheer computing power that's available today, largely driven by the computer games industry, allows a massive increase in the number of iterative cycles which has driven the data error rate close to the notional limit or entropy.

I work with systems that transfer Terabytes of data per second over a single optical fibre link of, say 150 km, where data error correction software is crucial to preventing systems clogging up with 'bad' data and slowing down or killing systems such as millions of Visa transactions per second.  None of us would want our bank cards wrongly charged with someone else's purchase owing to incorrect data and it's error correction software working in the background that prevents this from happening.  The increase in proficiency of this software even over the past 5 years has been simply amazing and, for example, is what is enabling NFC (Near Field Communications) technology - 'touch card' systems - and the proliferation of places that offer this convenience seemingly instantly.

The point is that these massively powerful error correction software algorithms are also at the heart of interpolation and up-rezzing software so I'm also interested to know whether pixel-shifting technology is as far ahead as we might suppose and whether it produces far superior images because, after all, the same process is going on at the heart of pixel shifting software, ie iterative compare and repeat cycles albeit from the other direction: an overabundance of data versus created data; both boil down to which data to pick as the best 'fit'.  While I'm happy to suppose that pixel-shifting should be superior, up-rezzing software is catching up up fast and not all vendors software is the same.  It's likely to be an interesting comparison. Either way, we the photographers are the winners through increased options. 😊

Pete.

Partial response maximum likelihood algorithms have been used in hard disk drives like forever.

Link to post
Share on other sites

Maybe everybody has already seen these great images, but anyway, just in case I add here the link to the Lumix Ambassador gallery (here Ken Duncan, but there are others as well).  https://www.panasonic.com/global/consumer/lumix/ambassadors/photo/ken-duncan.html

It’s not all full frame, but many are also shot with G9 (e.g. the hippos, where a greater distance is probably safer).

Edited by caissa
Link to post
Share on other sites

In some shops there are already some S1 or S1R cameras “on display”. Probably for demo purposes not for sale. Maybe somebody could check his neighbourhood shops and at the same time take a critical wide-angle M lens with adapter with him. The WATE is easy to carry - though it is usually a well-behaved lens.

Actually I have already seen a few WATE images on S1R (only jpgs, no raws of course, I remember a big blueish building with glass front), and they looked very good (unsurprisingly).

I live in rural surroundings, no shops close by (my preferred shop actually surprisingly went bankrupt a few months ago). Maybe somebody from London or NewYork or Paris or Berlin or Moscow or Tokyo has a better selection of shops ?

Edited by caissa
Link to post
Share on other sites

2 hours ago, Joakim said:

There are no indications that M lenses will work well on S1/S1R yet, only speculations and that is quite different 

Of course there are indications which M lenses are likely to work well on the S1R. (And a possible strategy is to prefer these lenses.) But of course there is also still a lot to learn about this topic. And especially one thing is still open, if a future SL2 will have the same sensor and the same (or only similar) thickness of the cover glass and similar (equal) aspherical prisms (or whatever the correct technical term is). For that we need to wait for the specialists - at least I would not know how to measure it. Maybe somebody knows if e.g. Reid Reviews has this on their agenda ?

Edited by caissa
Link to post
Share on other sites

The user manuals for the S1/S1R are now available which might help the "backup camera" decision.

On the Panasonic EU site here select AV Products, Digital Still Camera and then DCS1E or DCS1RE. Click View for the language of choice, click I Accept in the subsequent pop-up and finally click View  to get to the PDF. Crikey, all that clicking to see the manual - I hope the camera menus are simpler! 🤣

  • Haha 2
Link to post
Share on other sites

42 minutes ago, Bob Andersson said:

The user manuals for the S1/S1R are now available which might help the "backup camera" decision.

On the Panasonic EU site here select AV Products, Digital Still Camera and then DCS1E or DCS1RE. Click View for the language of choice, click I Accept in the subsequent pop-up and finally click View  to get to the PDF. Crikey, all that clicking to see the manual - I hope the camera menus are simpler! 🤣

520 pages. ohh...

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

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