Pixel Perfect wrote:
Well Sony uses a 0.18micron fab process. This means the line traces for the back-side circuitry is much smaller. Sony has managed to keep the entire readout, ADC-pipeline, NR etc on chip. Canon has to do a lot of this off chip and Sony's entire pipeline after the initial ADC is digital. The very short pipeline means less chance for noise to be introduced. The Sony has low ISO DR advantage due to the fact it's read noise is about 10x less than Canons' comparing D800 and 5D III. DR is the maximum signal divided by the noise floor in a pixel at each ISO and at low ISO the Canon noise floor is much higher. At higher ISO the read noise improves quickly for Canon and this is why at about ISO 800 or so the DR of the 5D III is similar to D800 and pulls ahead at higher ISOs....Show more →
So in essence a smaller fab process enables a better design for the back-end electronics. That makes sense. Thanks.
dolina wrote:
There area already prosumer cameras that have 4K resolution video like the Sony A7s, Panasonic GH4 and FZ1000 to name a few. So why shouldnt a $1800 Canon SLR not have it as well?
I can see them waiting a bit to see how the 4K market plays out. They aren't the only ones sitting back and waiting. Apple has yet to release a 4K monitor which would kind of be nice to go with the new Mac Pro's.
Thanks for that. Canon Japan has more details: link.
Yes, it's obviously an 8-channel read-out with two 4-channel A/D front-end chips (AFEs) .
Basically, it's the same architecture as any other Canon - with presumably the same DR characteristics.
I doubt we'll see ISO improvements either (vs the 70D).
So, that's the " newly developed" sensor that we are getting.
jorkata wrote:
Thanks for that. Canon Japan has more details: link.
Yes, it's obviously an 8-channel read-out with two 4-channel A/D front-end chips (AFEs) .
Basically, it's the same architecture as any other Canon - with presumably the same DR characteristics.
I doubt we'll see ISO improvements either (vs the 70D).
So, that's the " newly developed" sensor that we are getting.
And now there is more and more talk building that they probably won't even put 4k into even the 5D4 (or zebras, focus peaking, etc. etc.). It seems that everyone who has talked to Canon reps has gotten a really negative feeling from them.
kezeka wrote:
Thank you for the excellent explanation. What resources are you using for this information? I know more than a few people besides myself who would like to read more.
Basically just googling, checking the technical discussions in forums, a wee bit of of my own limited knowledge. I used to have an excellent link that explains the Sony Exmor technology that I'll try and track down. The stuff about the line process is from that article I quoted from earllier
jorkata wrote:
Thanks for that. Canon Japan has more details: link.
Yes, it's obviously an 8-channel read-out with two 4-channel A/D front-end chips (AFEs) .
Basically, it's the same architecture as any other Canon - with presumably the same DR characteristics.
I doubt we'll see ISO improvements either (vs the 70D).
So, that's the " newly developed" sensor that we are getting.
Only possible hope for improved DR is a lower read noise but not sure if that will eventuate and I'd only expect a fraction of an EV.
Really a "NEW" sensor, or just the old 70D "redesigned? As i read the sentence after their claim of "new" and they are still being pretty vague with their choice of words. The old saying of , Smoke and Mirrors comes to mind.
"20.2 Megapixel CMOS Sensor
A brand new 20.2 Megapixel sensor helps record high- resolution image files with stunning detail. Redesigned specifically for the EOS 7D Mark II, "
From what I know: QE and readnoise are fine, the problem is the "downstream noise". Canon tranfers the charges analog to an off-sensor analog-to-digital converter, while on all other modern image sensors this conversion happens directly on the sensor-chip ...
phibes wrote:
From what I know: QE and readnoise are fine, the problem is the "downstream noise". Canon tranfers the charges analog to an off-sensor analog-to-digital converter, while on all other modern image sensors this conversion happens directly on the sensor-chip ...
The sensorgen.info QE and read noise numbers are estimated from raw files, so they include the downstream noise. They do not state read noise at low ISO though.
I could be wrong, but I believe that this "new" sensor is basically a 70d sensor with very minor hardware changes and the differences we might be getting in IQ will mostly derive from software, not hardware.
Case in point: 70d has better IQ than 7d in jpeg, not so much (much smaller difference) in raw. I think the same will happen with the 7DMII.
Also, better high iso will likely come at the expense of detail/sharpness, that is to say that it will probably be just in-camera NR and not actual better hardware performance.
I wish I am wrong, because I am interested in this camera and apart from the sensor, I see a lot of goodies thrown in there and from the spec sheet, it has improved in many areas.
The AF does seem impressive (at least in specs) and the 10fps surprised me as well, since this has previously been exclusively series 1 territory.
I would have liked a fully articulated touch screen, but I guess that would be asking for too much, since we got 10fps and a theoretically very good AF, among other goodies.
But I do fear that regarding IQ, we won't be getting much better than the 70d.
Regarding 4k video, my $299 cell phone has it... need I say more?
Here is what Canon Europe says about the 7D2 sensor.
"New photodiodes with improved transistors inside each pixel offer high sensitivity, while gapless microlenses offer superior light-gathering ability for enhanced image quality in low light. Canon’s CMOS technology integrates advanced noise reduction circuitry at each pixel site, delivering virtually noise-free images and a wide native ISO speed range of 100-16,000 (expandable to up to 51,200). A high-speed 14-bit signal conversion combines with two “DIGIC 6” image processors for smoother tonal gradations and natural colours. An eight-channel read-out also supports high speed shooting of up to 10 frames per second (fps) at full resolution."
Jeff, after the 50D sensor propaganda, I swore to myself to never believe anything canon says about their products.
I will wait for testing and real world results from fellow photographers.
retrofocus wrote:
If they moved to a smaller process technology, we don't see it yet in the recent announcements from Canon. Canon was caught in a bad surprise when the D800(E) was released about the same time as the 5D MkIII. At the time, Canon mgmt must have decided to stick to the 24 MP FF sensor for a while not expecting others like Sony and Nikon would catch up as quickly and release new sensor-based products. One year later Canon's CEO was replaced without given reason to the media - I am sure it was at least partially related to the misjudgment of the demand in high MP/high DR sensors. If we assume that Canon since started at some point 2013/2014 to build new fabrication lines and investigate new sensor technology, they are likely not there yet. It is a tough catch-up game now for Canon....Show more →
Excellent post. So much opinion and speculation wrapped into one paragraph.
I'm still sticking to my opinion that the masses don't know about DR and don't really care. The current cameras are so good that most people wouldn't be able to tell the difference between a lowly canon FF and an Exmoor. Most people can't tell the difference between a crop and FF sensor either.
jctriguy wrote:
Excellent post. So much opinion and speculation wrapped into one paragraph.
I especially liked the part about the Canon CEO being terminated because of low ISO DR. I had no idea the Canon board of directors were all landscapers. Who would have thought.
ggreene wrote:
I especially liked the part about the Canon CEO being terminated because of low ISO DR. I had no idea the Canon board of directors were all landscapers. Who would have thought.
Interesting thread on the Nikon forum about everyone wanting to jump ship to the 7d since Nikon hasn't released a d400 yet...grass is always greener...
phibes wrote:
From what I know: QE and readnoise are fine, the problem is the "downstream noise". Canon tranfers the charges analog to an off-sensor analog-to-digital converter, while on all other modern image sensors this conversion happens directly on the sensor-chip ...
Read noise is not fine on the 5D III and if you go to Clarkvision you'll see it's around 35e- at ISO 100, compared to the D800's 4e-. But indeed Canon does most stuff off-sensor which also contributes to downstream noise. In fact the 5D III has worse read noise at low ISO than the 5D II, but better at high ISO, though still not as good as 1D IV, 1D X or 6D.
Nokia Lumia Denim firmware (coming soon) will record 4K video on several Nokia windows phones with the PureView. This highlights (Panasonic $800 4k camera quoted below) that some of this geek/numbers stuff is blown out of proportion. Numbers are just numbers in this case, not an end result.
The sensor is just one component of the camera ecosystem. The 7D2 is going to kill in terms of sports/action/reach for enthusiasts (the ones that spend $). Price, lens selection and the entire content in the 7D2 will guarantee this.
I personally don't like Canon/Apple/etc "incremental update" (anti-consumer) strategy but they add the technology that makes sense to keep sales going first and foremost at a time that is financially sensible to the company. I don't use Apple for the reason there are (much) better alternates. Not so with Canon. So I understand that if Canon feels a better sensor can help (sales and future products with a real road map) more later they will add it later not today. Again, please show where these great sensors are helping Sony DSLR or Nikon or Panasonic sales & revenue.
popinvasion wrote:
Especially when panasonic now has an $800 camera that can shoot 4k video.
