CCD vs CMOS, again

boojum said:
DOS will NEVER DIE? Are there other operating systems? Will they run on my XT? What's going on here? Did I miss something?

.
Click to expand...

Not for me. However- My code will not run on the XT, I use the PharLap DOS-Extender that puts the I5 in 32-bit mode and lets me use 4GBytes of memory and the 512GByte disk. It is Fast.

The KAF-18500 in the M9 and M Monochrom dates back to 2008. Revisions are always made, and the primary improvements are reductions in noise. The uniformity of the CCD in the M Monochrom is amazing, true scientific instrument quality. No surprise, these detectors were sold through DigiKey and other electronic shops to the scientific market. The "button dance" puts the cameras into a true "raw mode", but unlike the M8: the raw files are in DNG format and easily viewed in Lightroom and Photoshop. CMOS Back-Side-Illuminated cut the gap for collecting light at sharp angles. I'm surprised the M10 series used the older FSI CMOS detectors.
 
Sonnar Brian said:
Not for me. However- My code will not run on the XT, I use the PharLap DOS-Extender that puts the I5 in 32-bit mode and lets me use 4GBytes of memory and the 512GByte disk. It is Fast.

The KAF-18500 in the M9 and M Monochrom dates back to 2008. Revisions are always made, and the primary improvements are reductions in noise. The uniformity of the CCD in the M Monochrom is amazing, true scientific instrument quality. No surprise, these detectors were sold through DigiKey and other electronic shops to the scientific market. The "button dance" puts the cameras into a true "raw mode", but unlike the M8: the raw files are in DNG format and easily viewed in Lightroom and Photoshop. CMOS Back-Side-Illuminated cut the gap for collecting light at sharp angles. I'm surprised the M10 series used the older FSI CMOS detectors.
Click to expand...

The M9 I have has a Leica installed new sensor and board. From that I would assume I have a lower noise KAF-18500 than the one when new. I am not sure the "button dance" will reveal the upgrade install date. I can say that I am pleased, very pleased with the sensor images.

It is disappointing that while the best sensors were available they were not always used. We both know the "Faster, Better, Cheaper" mantra. Two out of three ain't bad and it was the "Better" that got dropped.

You must have a bunch of old DOS programs. Fast DOS is an oxymoron. ;o)
 
Boy, this thread took a weird turn somewhere south of Panasonic.

One thing I was going to add to the mix was that every review I read that compared the Panasonic and Leica equivalent models usually came with the caveat that the Leica models had better software, and that was the main difference between them and Panasonic. I could never get as good an output from my LX100 MkII as Rob-F did on his Digilux2 due to the image smearing that seemed to be inherited from the TZ-3 I owned many years before.

PF
 
ptpdprinter said:
I have no idea why you brought up the mic/amp thing or this Osborne guy. What does any of that have to do with Panasonic making the Leica Digilux 2?
Click to expand...

I think the analogy of mic > amp pairs well with sensor > software. Each have seen a move from analog(ue) to digital and have their adherents. I had hoped it would help explain. Perhaps I was wrong.

Osborne does extensive reviews and has a good one on Panasonic Lumix vs Leica SL which points out there are differences he ascribes to software. Likewise the differences between the Leica Digilux and its corresponding Panasonic Lumix. In each case it is more than just the sensor, it is what happens to the output of the sensor. Just as you can have two identical mics feeding two different amps and get different results.

But, back to the thread, I believe that CCD sensors have better color and imaging. This is only from my experience and observations. No double-blind tests. I could do M9 - M240 tests as other than sensors they are still pretty similar. There are those smarter and better educated in this who agree.
 
boojum said:
I could do M9 - M240 tests as other than sensors they are still pretty similar. There are those smarter and better educated in this who agree.
Click to expand...

Shoot both cameras with lens detection off, and shoot the M9 using uncompressed DNG. Use a range of lenses from wide-angle to telephoto. Shoot a fast lens wide-open. Without in-camera corrections for lens type, will show how the CMOSIS sensor does with handling light at angles.
 
Sonnar Brian said:
Shoot both cameras with lens detection off, and shoot the M9 using uncompressed DNG. Use a range of lenses from wide-angle to telephoto. Shoot a fast lens wide-open. Without in-camera corrections for lens type, will show how the CMOSIS sensor does with handling light at angles.
Click to expand...

You got it, Gunney.
 
boojum said:
I think the analogy of mic > amp pairs well with sensor > software. Each have seen a move from analog(ue) to digital and have their adherents. I had hoped it would help explain. Perhaps I was wrong.

Osborne does extensive reviews and has a good one on Panasonic Lumix vs Leica SL which points out there are differences he ascribes to software. Likewise the differences between the Leica Digilux and its corresponding Panasonic Lumix. In each case it is more than just the sensor, it is what happens to the output of the sensor. Just as you can have two identical mics feeding two different amps and get different results.

But, back to the thread, I believe that CCD sensors have better color and imaging. This is only from my experience and observations. No double-blind tests. I could do M9 - M240 tests as other than sensors they are still pretty similar. There are those smarter and better educated in this who agree.
Click to expand...

I'll just repeat what i said in post #2: I don't know about "special magic", but if you like the images the CCD sensors in your M8.2 and M9 render, then it makes sense to continue to use them.

And I hope you enjoy using your Pixii II, and like the images it produces with its CMOS sensor. Maybe Matt Osborne will come out with some of his special "Leica look" presets for the Pixii.
 
ptpdprinter said:
I'll just repeat what i said in post #2: I don't know about "special magic", but if you like the images the CCD sensors in your M8.2 and M9 render, then it makes sense to continue to use them.

And I hope you enjoy using your Pixii II, and like the images it produces with its CMOS sensor. Maybe Matt Osborne will come out with some of his special "Leica look" presets for the Pixii.
Click to expand...

I appreciate your graciousness.
 
boojum said:
...there is a perceived higher image quality to the CCD and the CMOS is always playing catch up.I do not understand the hows or whys of sensors but I am convinced, perhaps wrongly, that the CCD generally returns a better image than the CMOS. ....
Click to expand...

I agree completely. The evidence some Leica owners prefer images rendering from their cameras with CCD sensors is irrefutable.

The issue is - does this have anything to do with CCD photodiode array technology alone?

Nikon owners never, ever miss the the aesthetics of images from cameras with CCD sensors - D100, D100X and D200 CCD. Are we to believe a Kodak CCD photodiode in the sensor array is radically different than a CCD photodiode in the Nikon sensor arrays? Or is it more likely the other components in the sensor assembly as well as differences in proprietary demosaicking algorithms play a significant role? There were many more Nikon CCD camera users than Leica users. Surely some of them would notice a loss of image rendering aesthetics when they upgraded to Nikon CMOS cameras.


If I read: "No there camera renders images as well as my Leica M9", I would have no comment. I would think to myself - it's happy they enjoy their M9.
 
We finally got a day of sunshine here, all day long! Woohoo!!! Last Saturday was the first day this year I did not heat the house. Anyway, I ran parallel tests between CCD and CMOS sensors, M9 and M240. Each instance the files are uncompressed DNG's with no lens correction at ISO 200 with 28, 35, 50 and 85mm lenses. The 85 is a Jupiter, the rest are old LTM Canons. I will find an internet repository for the DNG's. Here is a JPG from each, the M9 and the M240.

What do you think? Both were converted in Darktable and resized in GIMP, both taken with the Canon 35mm LTM f/2.0 at f/5.6.


Click image for larger version Name: Pic#1.jpg Views: 0 Size: 420.2 KB ID: 4791106


Click image for larger version Name: Pic#2.jpg Views: 0 Size: 409.2 KB ID: 4791107​
 
I am uploading twelve files from the M9 and twelve files from the M240 to a Proton"Cloud Drive." Each is ISO 200, no lens correction, uncompressed DNG, -1/3 EV, taken with a 28mm Canon LTM f/2.8, Canon 35mm LTM f/2.0, Canon 50mm LTM f/1.8 and Jupiter 85mm f/2.0. The lens opening progression for each camera, each lens, was f/2.8, f/5.6, / f/11. Four lenses, three exposures, twelve images. I'll post the links as soon as the files are uploaded.
 
OK, links:

<snip>

OK, the files are available for DL only. So do that if you wish and I will try and find another way to post them for viewing.

Removed these links in favor of the Google Drive links.
 
boojum said:
...

What do you think? ...
Click to expand...

I think the exposure for the bottom (second image) image was slightly lower than the first image.

In the bottom image:
  • The left window is darker.
  • The siding above the porch is darker.
  • The clouds on the horizon are visible (assuming they were physically present when the first image was made).
I propose over exposure (photo diode full-well capacity was exceeded) for a small percentage of the blue channel pixels is consistent with the loss of horizon cloud detail in the top image.

I assume the shutter time settings were manually set to be equal. If so, this small level of over exposure could be due to small differences in the real, but unknown, shutter times between the two cameras. If the metering systems chose the shutter times, then small exposure differences are to be expected.

Otherwise, on my calibrated McBook Pro Retna display ,using Safari (i.e. gamma-corrected CSS colors) these images have essentially identical hue renderings.
 
The biggest difference for me is in the sky rendering. Either one is believable, but I would say the first picture is more so. In fact, it exactly matches the north sky I'm looking at out of my window right now.
 
You must log in or register to reply here.
Back
Top Bottom