Where is the EOS 5D Mark IV? [CR2]
- Thread starter Canon Rumors
- Start date
cnardo said:
Lots of times.
When you're shooting small objects in a big blue sky, why record all that blue sky? It wastes space and fills your buffer. Plus, on other cameras that have this feature, they can run at faster frame rates in crop mode because of fewer pixels to process.
Upvote
0
cnardo said:
When you don't need the full frame. When you need a faster frame rate. When you need a deeper buffer. When you want to save space.
Upvote
0
“Imagination is more important than knowledge. For knowledge is limited to all we now know and understand, while imagination embraces the entire world, and all there ever will be to know and understand.” - Albert Einstein ;Dajfotofilmagem said:
Upvote
0
pedro said:
That is a misunderstanding. Any camera can be updated to support "native" ISO's higher than their predecessor's maximums. That actually has nothing to do with IQ, though. It's just a configurable setting, that's all. The sensor would actually have to gather more light to make the prior "highest" setting better. Cameras are already at 60% Q.E. At best, we'll be able to get 40% more, but it is unlikely we'll see much more than 80-90% with consumer-grade sensors. That means we can get reasonably expect 20% more Q.E. over what we have now...quite some time down the road off into the future (Canon was improving Q.E. by 2-3% per generation for a long time there.)
I could see Canon going to 64% Q.E. tops with the next generation, although I kind of doubt it will be that high...it'll probably stay around 59-60%. If Canon starts strengthening the CFA again, that will reduce EQE, which will affect the total amount of light reaching the photodiode. IQE may reamain the same or higher...but it'll still be working with less light...so in the grand scheme of things, we could see WORSE ISO 25600 if the 5D IV got 51200.
Assuming Canon came right out with a 20% improvement over current Q.E. and the 5D IV hit with 80%, we still wouldn't see a factor-of-two reduction in noise at ISO 25600. We would see about a 25% improvement at best...and that is barely going to be noticeable under scrutinization.
Increases in maximum selectable ISO don't really have any bearing on whether that actually means less noise. Increases in Q.E. can mean improvements in noise...however there is both IQE (Internal Q.E.) an EQE (External Q.E.) The measurements we usually see give us IQE, which is just the conversion rate of photons incident on the photodiode. An increase in IQE paired with a reduction in EQE can result in worse overall performance. If we see increases in both EQE and IQE, or an increase in EQE at the same IQE, then we can see real improvements.
A truly significant increase in EQE could be made by ditching color filters entirely and moving to an MCS design, where we simply split the light via certain diffraction effects, but gather 100% of it. That kind of technology already exists...it would just need to be employed.
Upvote
0
QE = Quantum Efficiency
MCS = Micro Color Splitting
QE is the rate of conversion of incident photons to electrons in the photodiode. Very few cameras are actually true photon detectors, in that they detect every single photon and release charge as a result. Most have some lower conversion rate, like 41% or 56% or 65%.
MCS is a technology that was developed by Panasonic. It replaces color filters, which block certain frequencies of light and pass others (therefor reducing the amount of light reaching the photodiodes) with a special kind of color splitter. The color splitter deflects some light to neighboring pixels, and passes the rest. That results in pixels that have W+R and W-R color (white plus red and white minus red). That means 100% of the light is sent to the photodiodes, where color filters will block 60-70% of the light. The Q.E. of each photodiode in an MCS sensor could be the same as that in a Bayer CFA (color filter array) sensor, but the MCS sensor is going to be more sensitive overall.
MCS = Micro Color Splitting
QE is the rate of conversion of incident photons to electrons in the photodiode. Very few cameras are actually true photon detectors, in that they detect every single photon and release charge as a result. Most have some lower conversion rate, like 41% or 56% or 65%.
MCS is a technology that was developed by Panasonic. It replaces color filters, which block certain frequencies of light and pass others (therefor reducing the amount of light reaching the photodiodes) with a special kind of color splitter. The color splitter deflects some light to neighboring pixels, and passes the rest. That results in pixels that have W+R and W-R color (white plus red and white minus red). That means 100% of the light is sent to the photodiodes, where color filters will block 60-70% of the light. The Q.E. of each photodiode in an MCS sensor could be the same as that in a Bayer CFA (color filter array) sensor, but the MCS sensor is going to be more sensitive overall.
Upvote
0
jrista said:
Thank you, jrista, for taking your time to explain this technical aspect to me. Lacking any technical knowledge in the aforementioned area, board members like you really contribute a lot to the better understanding of folk like me. Let's hope that Canon will start soon to walk that road as outlined in your reply.
Upvote
0
@jrista:
I've been searching about MCS tech, and wow... it's quite a step ahead...!
Panasonic in FEb 2013:
http://www.diginfo.tv/v/13-0021-r-en.php
http://news.panasonic.com/press/news/official.data/data.dir/2013/02/en130204-6/en130204-6.html
As I do not now much about this materia, did Canon mention anything in the past two years? Are they working on it, or is it panasonic tech achievement only?
So, given that Canon might be developing a similar product, how long is the wait for the next revoluationary EOS Camera then, which will mention MCS in its sensor specs...? Is there hope for an outing within the next 4 years?
uh, and it was shortly discussed here...now I got a new abreviation to jump at in camera fora ;-)
http://www.canonrumors.com/forum/index.php?topic=12783.0
The most intresting statement is this one here:
Can this technology be implemented right away?
Yes, says Panasonic. The “micro color splitters” can replace all color filters in current sensors and they support both CCD and CMOS sensors. Furthermore, the new sensors can be manufactured using conventional semiconductor techniques and cheap, inorganic materials.
http://camyx.com/news/2013/02/panasonic-sensor-doubles-low-light-image-quality/
sample showing improvements
http://cdn.camyx.com/wp-content/uploads/2013/02/panasonic-sensor-double-low-light-image-quality.jpg
I've been searching about MCS tech, and wow... it's quite a step ahead...!
Panasonic in FEb 2013:
http://www.diginfo.tv/v/13-0021-r-en.php
http://news.panasonic.com/press/news/official.data/data.dir/2013/02/en130204-6/en130204-6.html
As I do not now much about this materia, did Canon mention anything in the past two years? Are they working on it, or is it panasonic tech achievement only?
So, given that Canon might be developing a similar product, how long is the wait for the next revoluationary EOS Camera then, which will mention MCS in its sensor specs...? Is there hope for an outing within the next 4 years?
uh, and it was shortly discussed here...now I got a new abreviation to jump at in camera fora ;-)
http://www.canonrumors.com/forum/index.php?topic=12783.0
The most intresting statement is this one here:
Can this technology be implemented right away?
Yes, says Panasonic. The “micro color splitters” can replace all color filters in current sensors and they support both CCD and CMOS sensors. Furthermore, the new sensors can be manufactured using conventional semiconductor techniques and cheap, inorganic materials.
http://camyx.com/news/2013/02/panasonic-sensor-doubles-low-light-image-quality/
sample showing improvements
http://cdn.camyx.com/wp-content/uploads/2013/02/panasonic-sensor-double-low-light-image-quality.jpg
Upvote
0
Glad to help, Pedro. 
Regarding Canon's technology. Canon has been investing in layered sensor technology. In regards to light preservation, layered sensors, if designed properly, certainly have a leg up on bayer sensors. You gather the full constituent of light at each pixel.
In regards to whether you preserve 100% of it like MCS, I don't know. Canon has invested some R&D in anti-reflective technologies with their layered sensors. They have both nanocoatings on the lower-layer photodiodes (red specifically, maybe green) as well as anti-reflective coatings on the bottom of each photodiode. That tells me that Canon is still losing light in some ways with their layered sensor designs. I don't know how much...I suspect the losses are primarily to heat, and nanocoatings on the photodiodes should certainly help in that area.
I have not read anything that indicates Canon is developing MCS of their own. I was kind of more hoping Canon might just buy Panasonic's technology...maybe buy their whole sensor patent portfolio.
I don't think that is going to happen any time soon, though. I think Omnivision is on the ball with similar technologies as well. Omnivision is probably one of the biggest dangers to Canon. They have already developed their own form of DPAF, so Canon will not be the sole company to have such technology in the not too distant future. I think more of Omnivision's customers are video based, where DPAF truly shines. There is at least one other company developing similar technology as well.
Canon has some key innovations in the image sensor arena, but the hounds are just about ready to bite down on Canon's ass and shred. We'll have to see if Canon brings their lucrative technology to market fast enough or not. Maybe with the 5D IV...I certainly hope.
Regarding Canon's technology. Canon has been investing in layered sensor technology. In regards to light preservation, layered sensors, if designed properly, certainly have a leg up on bayer sensors. You gather the full constituent of light at each pixel.
In regards to whether you preserve 100% of it like MCS, I don't know. Canon has invested some R&D in anti-reflective technologies with their layered sensors. They have both nanocoatings on the lower-layer photodiodes (red specifically, maybe green) as well as anti-reflective coatings on the bottom of each photodiode. That tells me that Canon is still losing light in some ways with their layered sensor designs. I don't know how much...I suspect the losses are primarily to heat, and nanocoatings on the photodiodes should certainly help in that area.
I have not read anything that indicates Canon is developing MCS of their own. I was kind of more hoping Canon might just buy Panasonic's technology...maybe buy their whole sensor patent portfolio.
I don't think that is going to happen any time soon, though. I think Omnivision is on the ball with similar technologies as well. Omnivision is probably one of the biggest dangers to Canon. They have already developed their own form of DPAF, so Canon will not be the sole company to have such technology in the not too distant future. I think more of Omnivision's customers are video based, where DPAF truly shines. There is at least one other company developing similar technology as well. Canon has some key innovations in the image sensor arena, but the hounds are just about ready to bite down on Canon's ass and shred.
We'll have to see if Canon brings their lucrative technology to market fast enough or not. Maybe with the 5D IV...I certainly hope.
Upvote
0
