Fuji and Panasonic develop organic sensor with 26.2 stops of dynamic range

[o hey tyler]

lol wut

Fuji and Panasonic's New Organic Sensor Boasts Insane 26.2 Stop Dynamic Range

 

[Garbz]

Sorry I can't take an article seriously which says going from 46dB to 88dB is a "doubling" of performance. I understand not everyone is sciency, but this is a site which has some experience with cameras so they should also know that 26EV is not double 15EV. Not to mention that their example images even if they are a mockup don't even remotely represent the level of performance described by the numbers. ... assuming the numbers are in any way real.

I'll believe it when I see it.

 

[o hey tyler]

Sorry I can't take an article seriously which says going from 46dB to 88dB is a "doubling" of performance. I understand not everyone is sciency, but this is a site which has some experience with cameras so they should also know that 26EV is not double 15EV. Not to mention that their example images even if they are a mockup don't even remotely represent the level of performance described by the numbers. ... assuming the numbers are in any way real.

I'll believe it when I see it.

It will be interesting to see where the tech goes.

 

[Derrel]

Sorry I can't take an article seriously which says going from 46dB to 88dB is a "doubling" of performance. I understand not everyone is sciency, but this is a site which has some experience with cameras so they should also know that 26EV is not double 15EV. Not to mention that their example images even if they are a mockup don't even remotely represent the level of performance described by the numbers. ... assuming the numbers are in any way real.

I'll believe it when I see it.

The new process was shown yesterday in Japan at the VLSI Technology Show in Japan, and is supposed to be presented in the USA by Fuji and Panasonic on June 15 at the International Image Sensor Workshop. The drawing you are mocking is direct from the Fuji/Panasonic joint press release. Perhaps you could send the engineers at Fuji and Panasonic your thoughts on their new process.

Fujifilm and Panasonic collaborate on 'industry-leading' organic/CMOS sensor: Digital Photography Review

 

[Garbz]

The drawing you are mocking is direct from the Fuji/Panasonic joint press release. Perhaps you could send the engineers at Fuji and Panasonic your thoughts on their new process.

Oh Derrel please re-read my post. I'm not mocking any engineer. Quite the opposite. I'm mocking non-engineers trying to write technical articles. Maybe the writers can talk to to Fuji and Panasonic in joint and confirm for us whether they really have doubled the performance or increased the performance by a factor of 128x as their numbers would suggest before they write the article.

The dpreview article you linked was much better.

 

[timor]

Now there will be no excuses for blown highlights. If Fuji or Panasonic going to make FF sensor in this technology ?

 

[o hey tyler]

If [sic] Fuji or Panasonic going to make FF sensor in this technology ?

I wouldn't hold my breath... But who knows.

 

[amolitor]

I don't get it. 88dB S/N does sound like about 2x the range of 46dB S/N. I know audio, not imaging, though.

 

[dxqcanada]

Cameras five times more sensitive to light? An ultrasensitive molybdenum-based image sensor developed

 

[Garbz]

I don't get it. 88dB S/N does sound like about 2x the range of 46dB S/N. I know audio, not imaging, though.

Decibels are logarithmic. For every 6dB increase you end up with double the dynamic range. Same in audio.
16bits provides 96dB of dynamic range and 65536 possible loudness values for a given sample.
24bits provides 144dB of dynamic range. Looks like a 33% increase only except that you get 16777216 possible loudness values for a given sample or a 256x increase.

Something is amiss.

 

[amolitor]

Right. And 46dB is is about 7.6 bits, 88dB is about 14.6, which looks like generally about twice as many bits. There's a basic problem in that "double the performance" is pretty much completely undefined, but twice the bit depth sounds like a reasonable definition to me.