Jump to content
SrMi

Leica CL vs TL2 dynamic range

Recommended Posts

Spend a bit of time working with a 2003 Olympus E-1 ... only 5 Mpixel, ancient CCD Kodak sensor, incredibly slow write times, but capable of spectacular photographs when you get your framing and exposure right! ... and you find out just how flexible and easy a CL or TL2 sensor really is. :D

Share this post


Link to post
Share on other sites
On 12/7/2018 at 9:55 AM, jankap said:

Is that generally true? The effective area of a pixel is a main parameter still.

Jan

Thats one of the myths which always come up. That only is important at base ISO. If you have a large mug (pixel) that can hold 2 cups and a standard cup (smaller pixel). Now you fill in both 1/2 a cup of coffee. In which cup is more coffee? Hence it depends on the exposure and the lenses how much you can make more use of the larger image. 

If you are interested in a profound reading on dynamical range, and noise check out http://www.clarkvision.com/articles/digital.sensor.performance.summary/

Here a quote from Roger that illustrates that

"However, I want to be clear regarding the sensor data: the pixel and its size is only a holder for the photoelectrons. It is the lens that delivers the photons to the sensor. Just because a pixel is larger does not mean that it will produce better or lower noise images if the lens and exposure time does not fill the pixel with enough light (and thus, photoelectrons). If one is working above base ISO, the pixel will not be filled to its full capacity. The larger pixel has the potential to collect more light. But larger pixels see larger angular areas from the lens, so resolve less detail. There is a trade between the lens collecting the light, the focal length spreading out the light, the pixels chopping up the light. and the exposure time limiting the light collected. For the impact on an image of these parameters, the full system must be considered. "

 

 

Share this post


Link to post
Share on other sites

This idea that pixels are cups or buckets which “collect” photons, while illustrative, is really quite misleading.  The reality is that photons aren’t capable of “collection” - that is not a property of light, like saving time in a bottle.  In reality, the sensor site reacts to the light which hits it.  If the light is dim, then the sensitivity of the sensor site needs to be increased; if there is a mixture of different wavelengths (as interpreted from the light coming through the Bayer CFA), then the predominant wavelength with prevail in the final image.

Larger sensor sites will, at an appropriate viewing distance, arguably provide a smoother picture.  There is considerable benefit in larger pixels with greater percentage coverage of the image area (ie, less wastage space between sensor site surface area).  Smaller and more pixels result in greater detail (and less moiré, apparently).  I suspect the issue with the 37.5MP D800 and the A7r was that, while both cameras had more pixels than the 24MP Leica cameras, the Leica sensors had better sensor site design and better coverage (linked elsewhere).

The problem I have with the quote from Clark Vision above is that (1) photons aren’t “collected” and (2) the cell sites don’t create light - they read it.  I agree that the same light hits the sensor (with a given lens), whether the pixels are large and few or small and plentiful, but that is only part of the system - how the light is directed onto the sensor (Leica) and how well the image area is covered (Leica, again) is the primary concern.  For myself, I also don’t need the constraint of lots of smaller pixels to capture motion blur, for resolution I actually don’t need.

Others do.

Edited by IkarusJohn

Share this post


Link to post
Share on other sites

Just to get a little deeper in the weeds in this discussion -- microlenses above the chip and sensor cell surface layout affect the ability of each cell to capture the photons that were meant to be at that point in the image.  As just noted, each photon is then converted in the sensor cell into some number of electrons, usually one electron, sometimes none, sometimes more than one if the sensitivity is turned really high. So it's electrons that they hold.  The electrons are read out as a voltage, which is proportional to the number of electrons.  If the analog to digital circuit that reads this voltage can resolve 14 bits, and the smallest signal of interest is one electron, then the cell can render useful output from at most 2^14 electrons (or 16,384).  While electrons repel each other and won't lie peacefully in a "bucket" there is a depth as well as a surface area to each collecting cell, so the number of electrons it can hold is not simply proportional to the surface area of the cell.  But within a particular generation of technology, making the surface area of a cell bigger or smaller by a factor of two will make the number of electrons it can hold bigger or smaller by about a factor of two. 

Edited by scott kirkpatrick

Share this post


Link to post
Share on other sites

Please don’t be too excited about TL2 :

Please read carefully Bill’s comment in Leica rumors about his own measurements :

- industry worst black level implementation 

- noise reduction kicks in as soon as 400/800 ISO

Looks like the same sensor with different cheating methods to get ahead of the pack

 

Edited by nicci78

Share this post


Link to post
Share on other sites

Interesting; the CL is not mentioned. See post 28, which addresses this point.

 

Actually, there is nothing new here. The raised black point as a method of reducing shot noise has been noted in this forum by Sandy just after the introduction of the M240, back in 2012. It is unclear whether there is a negative impact on image quality in real life. It might just as well improve visual IQ.

I think it is not a bad idea to filter out unwanted random noise. It is an accepted technique in HiFi audio. A cut-off curve at the low end is likely to produce a better image/signal than a curve that flattens out into meaningless noise.

I am quite willing to accept that such a technique is unbeloved by technogeeks as it messes up their pretty graphs, but then, Leica is about the image, not about pleasing the reviewers.

Anyway, using technology with the intention to improve the image can hardly be called cheating, whether one agrees or not.

 

Share this post


Link to post
Share on other sites

Probably different electronic and software implementation of same sensor. 

Just look at the graph, TL2 shows noise reduction kicks in at 800 ISO for sure (little square pointing down)

However the CL graph shows no image treatment before 12800 ISO. From this point a scaling (no more analog gain, exposure is raised by software) is applied (same thing with the SL) 

Edited by nicci78

Share this post


Link to post
Share on other sites

May be the CL as the SL, is targeted to “serious” and traditional user. Who are willing to post-process properly their DNG. 

Whereas the TL2 is intended to be use by beginners, who need excellent jpeg and/or no fuss DNG 

Share this post


Link to post
Share on other sites

Well, the cameras are clearly aimed at a different segment of the market, so it is quite possible that Leica adapted the output to the potential customers. Who do not necessarily need to be beginners. The TL2 is equally useful to an experienced photographer who is looking for a high-level pick-up-and-go camera with a bit of present-day aura.

Share this post


Link to post
Share on other sites
12 hours ago, nicci78 said:

Probably different electronic and software implementation of same sensor. 

Just look at the graph, TL2 shows noise reduction kicks in at 800 ISO for sure (little square pointing down)

However the CL graph shows no image treatment before 12800 ISO. From this point a scaling (no more analog gain, exposure is raised by software) is applied (same thing with the SL) 

Bill Claff writes in Leica rumors: "Yes, a few people have asked about this via email. The TL2 is clearly a different sensor than the CL."

Share this post


Link to post
Share on other sites

He bases this on his finding that the cameras have different raw resolutions.

Share this post


Link to post
Share on other sites
19 minutes ago, jaapv said:

He bases this on his finding that the cameras have different raw resolutions.

.. and black level (according to his email to me).

Why do we assume it is the same sensor?

Share this post


Link to post
Share on other sites

I got the same mail. Which surprises me. The black point is set by the firmware and not by the sensor AFAIK.

We assume it is the same sensor because Leica implied so.

Share this post


Link to post
Share on other sites

Probably pointless to compare then. TL2 is not available at the moment. I was promised one for a weekend once more units come in. 

Share this post


Link to post
Share on other sites

weird, in France, sellers are flooded with TL2, it is almost discounted (-10 to -25%) everywhere except in proper Leica Store. Nobody seems to want one.

One of the largest photo reseller : FNAC refuse to sell it. They are too scared by the previous T and TL experience. Forcing them to fire sale it at -70%. And even at that price the stock lingered a couple of years. 

However the CL is doing OK in France and at the FNAC

Edited by nicci78

Share this post


Link to post
Share on other sites
9 hours ago, jaapv said:

He bases this on his finding that the cameras have different raw resolutions.

raw resolutions does not mean anything at all. A photo sensor is just a cutout of a large round silicon wafer. You can cut it a little bit larger or not. So not a proof at all. 

Share this post


Link to post
Share on other sites
vor 41 Minuten schrieb nicci78:

weird, in France, sellers are flooded with TL2, it is almost discounted (-10 to -25%) everywhere except in proper Leica Store. Nobody seems to want one.

One of the largest photo reseller : FNAC refuse to sell it. They are too scared by the previous T and TL experience. Forcing them to fire sale it at -70%. And even at that price the stock lingered a couple of years. 

However the CL is doing OK in France and at the FNAC

My dealer just doesn’t happen to have any. And there is no continuous production of the TL2. So waiting for the next production run. 

Share this post


Link to post
Share on other sites
50 minutes ago, nicci78 said:

raw resolutions does not mean anything at all. A photo sensor is just a cutout of a large round silicon wafer. You can cut it a little bit larger or not. So not a proof at all. 

It is also not a proof that it is the same sensor. Why would Leica cut the same wafer differently for CL and TL2?

Looking at so many measured differences, it seems more probable that it is different than that it is the same. I could not find any official Leica statement either way. AFAIK, Jono Slack was the first to state that they have the same sensor, but I do not know his source.

Why does this matter? For me, it is interesting to know what is correct. I do not think the sensor is the important differentiator between CL and TL2.

Share this post


Link to post
Share on other sites
1 hour ago, SrMi said:

...

Why does this matter? For me, it is interesting to know what is correct. I do not think the sensor is the important differentiator between CL and TL2.

In other words, it's an academic yearning. Nothing wrong with that. :D 

I agree that the sensor is the least of what I'd be worrying about if I were choosing between the TL2 and the CL. How well does the design—the controls, etc—work for my hands and mind are the big differences to me. 

Share this post


Link to post
Share on other sites

Today I was told by a Leica representative that the CL sensor is not the same as the TL2 sensor and that the CL sensor is the better one.

Share this post


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.

×