Rethinking stacked sensor in om-5

  • Thread starter Thread starter Raist3d
  • Start date Start date
Been saying they need to put the new sensor down the line but I am rethinking this now since stacked sensors are $$$
GH6 sensor is not stacked and it is pretty darn fast. Current generation normal sensors are pretty good in terms of speed so that's the least of worries.
So to me the most important thing is the on-5 sensor if it’s not the stacked one, that I shares the modest image quality improvements of the new sensor and also able to auto iso to 25600
Now you are talking from your heart. See - the important stuff is new UI and new AI, that is, the new processing pipeline.
Maybe read what I wrote first before replying line by line :-)
But all said I rather have them do a pro specced smaller om-5 priced high than a model with some obvious compromises to make it “cheap”

Even if that means a price north of $1,500 USD - say $1600 or even $1799 - make it right make it pro small That would be pretty unique in the market
Unique?
If done right sure. It would be fast, have pretty good image quality in a small size and have a full system of lenses to boot. If this was a weather sealed model they need to re-release their primes as weather sealed (the small ones).
thoughts? What you think is key for the om-5 to come out in todays market competitively - assuming this is an em5 mark iii follow up or even em1 mark iii
Follow the marketing. Only two things are certain about the camera - compact and aimed for outdoors. Implies weather sealing. And a few logical things we can surmise - new interface and new AF. This guarantees that new processor will come as this software can't run on older hardware.

So the only variable here may be the sensor. Rest assured, it certainly won't be worse than EM 1 Mark III, nor will it be slower.

No matter what happens, this new camera will still sit above all existing cameras, though perhaps a tad bit under OM-1, and will likely have price around $1500 or less. Price around $1800 is simply too close to OM-1(and many, many full frame options - Z5, Z6, S5, A7C, A7III, R). The market for an expensive non-stacked MFT body is non-existent as far as photography is concerned (hence Panasonic sells video features only) - especially now that Canon gives AI autofocus in R10 for just over $1000.
Picturing a small body with some OM1 capabilities I would think presents a unique product in the market.
I think you are just overthinking everything. Just chill.
I am quite chill :-). But thanks for your opinion all the same.
We'll get a good camera for a good price in a relatively compact body. If OM-1 didn't radically adopt a new design compared to existing bodies, OM-5 will likely do the same. It may be a better version of EM-5 platform.

Note carefully what Andrea said "I have been told this is not the replacement of E-M5 Mark III". Replacement of E-M5 Mark III is E-M5 Mark IV with internals of E-M1 Mark III. Instead, this is a compact version of OM-1 tech, in a body that probably is a new design.
I can see that as possible and would be great. I think it's not entirely clear what this could mean - seems like several possibilities here.
I personally think that a non-stacked MFT sensor in a DSLR style body ( essentially EM 1 Mark IV) at $1700 or thereabouts has zero market today.
Agreed.
Whatever niche remained was crushed by Canon R7. Either OMDS brings down their AF in a non-stacked sensor for $1200 or they go bust and bankrupt.
I am inclined to think if they don't do that at least in a "small pro" body or "semi enthusiast small" body, they can head the path of trouble, yes. So I guess I sort of agree in a way.

Thanks again.
 
Presuming the OM-5 is still following E-M5 lineage (not a different one where Olympus flips the numbering convention), I don't see OIympus coming up with a completely new sensor. Seems to make more sense to keep with the same sensor like they did for E-M5 III (inheriting E-M1 II/III sensor) and have the volume. Just decontent a bit on the image processor and some of the other features of the camera.
My thought too- by same sensor just to be clear, you mean the OM-1 sensor.
Yes, that's what I meant, although my wording is confusing.
Then presuming they don't kill off the E-PL lineup, that one just inherits the old E-M1 II sensor.
I honestly think they will need better than that, though depends on pricing. Maybe kill the EPL's but do finally a real PenF Mark 2.
I think the E-M1 II sensor is fine and still quite good. I'm pretty sure a lot of people would be quite happy with a GF successor (or even GX successor) if it got the G9 sensor and processor. I don't imagine it'll be a whole lot different for E-PL line. It would allow an inexpensive camera that is still very capable.
Well my concern is that we have seen the 20MP sensor capability in mft for quite a while, the new OM-1 seems to handle better color preservation as ISO climb. Given the competition, it just makes me think they need it to make the camera more competitive.

To put this another way- how long would a "new" 20MP sensor with PDAF on a Pen line would hang around in years further more? Where are the competitors already?
Wasn't the point of this thread that there is concern that a stacked sensor would be too expensive for even a OM5? For E-PL it would be even worse.

If you are instead implying another new lower end non-stacked sensor, I'm not seeing it. Unless Sony already had one in the catalog (like with the IMX269), I'm not seeing it being cost effective to work out yet another sensor, especially given this time Panasonic apparently chose a non-Sony sensor for GH6. The existing IMX270 is going to be the best affordable choice.
 
Presuming the OM-5 is still following E-M5 lineage (not a different one where Olympus flips the numbering convention), I don't see OIympus coming up with a completely new sensor. Seems to make more sense to keep with the same sensor like they did for E-M5 III (inheriting E-M1 II/III sensor) and have the volume. Just decontent a bit on the image processor and some of the other features of the camera.
My thought too- by same sensor just to be clear, you mean the OM-1 sensor.
Yes, that's what I meant, although my wording is confusing.
Then presuming they don't kill off the E-PL lineup, that one just inherits the old E-M1 II sensor.
I honestly think they will need better than that, though depends on pricing. Maybe kill the EPL's but do finally a real PenF Mark 2.
I think the E-M1 II sensor is fine and still quite good. I'm pretty sure a lot of people would be quite happy with a GF successor (or even GX successor) if it got the G9 sensor and processor. I don't imagine it'll be a whole lot different for E-PL line. It would allow an inexpensive camera that is still very capable.
Well my concern is that we have seen the 20MP sensor capability in mft for quite a while, the new OM-1 seems to handle better color preservation as ISO climb. Given the competition, it just makes me think they need it to make the camera more competitive.

To put this another way- how long would a "new" 20MP sensor with PDAF on a Pen line would hang around in years further more? Where are the competitors already?
Wasn't the point of this thread that there is concern that a stacked sensor would be too expensive for even a OM5? For E-PL it would be even worse.
I wasn't saying it needs to use the OM1 sensor then (as I pointed in the original post) but something else, that can still have an image quality benefit to keep those colors better at high iSO like the OM1.
If you are instead implying another new lower end non-stacked sensor, I'm not seeing it.
Certainly that's one possibility. (either, not happening or happening). But maybe it could mean the end of the EPL line and we get to see something more high end like a PenF MKII. I guess that depends how much money really are they making at this point in the market given it seems all data points to higher priced models are the ones to go moving forward (that not just OMDS).
Unless Sony already had one in the catalog (like with the IMX269), I'm not seeing it being cost effective to work out yet another sensor,
Yeah, I was thinking something somewhere in the catalog.
especially given this time Panasonic apparently chose a non-Sony sensor for GH6. The existing IMX270 is going to be the best affordable choice.
Well, was just thinking about improving the IQ in some way. Maybe a tweaked version of that? But I do see your point. I think to me in this market this should happen, but I see your point of maybe what you said is what happens.
 
it started with compact cameras at 20 megapixels it was not otherwise possible
There were multiple generations of 20MP 1" sensors before Sony started introducing stacked versions.
yes but stacked started with the rx100 20 megapixels
What point do you think you're making here?
sony apsc runs ok front illumination and back illuminated full frame run on back illuminated non stacked so only for higher than 24 megapixels full frame stacking has been used
The first stacked sensor available in any MILC was the 24MP A9.
The A9 has a quad bayer cell to have many AF points so actually 96 megapixels
Where did you pull this one from? We have the datasheet for the IMX310, it's definitely not a quad bayer design. Plus it produces PDAF striping against strong backlights like all of Sony's other masked PDAF sensor designs.
From the same datasheet you mention readout mode 1B referring to HDR mode that can only be done with quad bayer it has been commented multiple times on various sony forums
 
it started with compact cameras at 20 megapixels it was not otherwise possible
There were multiple generations of 20MP 1" sensors before Sony started introducing stacked versions.
yes but stacked started with the rx100 20 megapixels
What point do you think you're making here?
sony apsc runs ok front illumination and back illuminated full frame run on back illuminated non stacked so only for higher than 24 megapixels full frame stacking has been used
The first stacked sensor available in any MILC was the 24MP A9.
The A9 has a quad bayer cell to have many AF points so actually 96 megapixels
Where did you pull this one from? We have the datasheet for the IMX310, it's definitely not a quad bayer design. Plus it produces PDAF striping against strong backlights like all of Sony's other masked PDAF sensor designs.
From the same datasheet you mention readout mode 1B referring to HDR mode that can only be done with quad bayer it has been commented multiple times on various sony forums
This is not true. You can make HDR different ways.

4bedff40e39d429ebb884a08bd60e774.jpg.png

This looks more like what Panasonic did with the GH6. By the way, the Sony data sheet of the OM1 sensor clearly specs out it's a quad bayer. Nothing like that here.

--
Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - Apparently Selwyn Duke and not George Orwell
 
Last edited:
it started with compact cameras at 20 megapixels it was not otherwise possible
There were multiple generations of 20MP 1" sensors before Sony started introducing stacked versions.
yes but stacked started with the rx100 20 megapixels
What point do you think you're making here?
sony apsc runs ok front illumination and back illuminated full frame run on back illuminated non stacked so only for higher than 24 megapixels full frame stacking has been used
The first stacked sensor available in any MILC was the 24MP A9.
The A9 has a quad bayer cell to have many AF points so actually 96 megapixels
Where did you pull this one from? We have the datasheet for the IMX310, it's definitely not a quad bayer design. Plus it produces PDAF striping against strong backlights like all of Sony's other masked PDAF sensor designs.
From the same datasheet you mention readout mode 1B referring to HDR mode that can only be done with quad bayer it has been commented multiple times on various sony forums
This is not true. You can make HDR different ways.

4bedff40e39d429ebb884a08bd60e774.jpg.png

This looks more like what Panasonic did with the GH6. By the way, the Sony data sheet of the OM1 sensor clearly specs out it's a quad bayer. Nothing here.

--
Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - Apparently Selwyn Duke and not George Orwell
The sensor did not have PDAF pixels and thats datasheet is doctored there has never been one coming out from sony

and sony didn’t declare the a7s3 quad bayer but it is

the specs of products that are for general use can be trusted but there are none around for any custom ones

There were long discussion as far as the striping was concerned and didn’t lead to anything note tbagyit was not covering the whole sensor so certainly nothing to do with masking so how would this camera have all those AF pixels?



--
instagram http://instagram.com/interceptor121
My flickr sets http://www.flickr.com/photos/interceptor121/
Youtube channel http://www.youtube.com/interceptor121
Underwater Photo and Video Blog http://interceptor121.com
Deer Photography workshops https://interceptor121.com/2021/09/26/2021-22-deer-photography-workshops-in-woburn/
 
for me the E-M5 replacement would ideally be something like the E-M1 Mark three (with any incremental improvements available: i.e, faster HHHR etc) but in the E-M5 mark 2 body.
 
Been saying they need to put the new sensor down the line but I am rethinking this now since stacked sensors are $$$

So to me the most important thing is the on-5 sensor if it’s not the stacked one, that I shares the modest image quality improvements of the new sensor and also able to auto iso to 25600

But all said I rather have them do a pro specced smaller om-5 priced high than a model with some obvious compromises to make it “cheap”

Even if that means a price north of $1,500 USD - say $1600 or even $1799 - make it right make it pro small That would be pretty unique in the market

thoughts? What you think is key for the om-5 to come out in todays market competitively - assuming this is an em5 mark iii follow up or even em1 mark iii
I think it has go one way or the other. Either the same sensor that's in OM1, or a non-stacked, higher resolution BSI sensor to cater to different needs. I don't think 20mp non-stacked sensor will be that appealing, given that it'll be competing with EM1III and EM5III in that price range, not to mention other m43 and APSC offerings in that range. But whether we see higher resolution sensor also depends on how well the current lenses can resolve the additional detail. OM doesn't seem to prioritize spec sheet bragging as much as real world improvements (e.g. they improved IBIS in OM1 even though it makes zero difference in CIPA rating since CIPA does not test for rolls) so I think it would happen only if they believe that there's real image resolution gain from higher res sensor.

There was a low credibility rumor about OM5 having higher resolution sensor. I think it's actually possible they would go for a higher resolution sensor. It will differentiate OM1 and OM5 in a way that will make them more appealing as a set, using one for wildlife and the other for nature
back illuminated sensor read slower so the pixel count is limited like the GH5S
Please explain to me why the Fuji APSC of 26MP then is BSI and 26 MP? That doesn't make sense. Also the Fuji 26MP is a relatively fast sensor (close to EM1.2/.3 readouts) in its class. Of course doesn't compare to the super fast stacked sensors but that's beside the point.

Evidence the Fuji 26MP is BSI and has been that for a while now:

https://www.dpreview.com/news/54716...s-with-a-26mp-x-trans-sensor-and-4k-60p-video
you need stacking to read faster or front illumination
Update: BSI enables actually bigger pixel counts without sacrificing image quality vs some degree of lower pixel counts. No, I am not going to bother looking where this piece of knowledge came from but I am pretty sure you can find it if you look for it. I already gave you evidence.

Hope you now return in kind.
There are back illuminated non stacked sensors typically used for low pixel count cameras like sony And panasonic 24 megapixels
You brought up the GH5s as an example of low pixel count BSI.
no the gh5s is not stacked
those read slow cant do 60 fos without crop
The Fuji 26MP ASPC sensor being used is not slow even if it's not hitting 60fps. It's definitively a step above the average CMOS sensor that was done until that time.
then there sre backilluminated stacked sensor with high pixel count those don’t do equally well in low light and are what goes in full frame high megapixel sensor.
This is a new claim. What you claimed was

"back illuminated sensor read slower so the pixel count is limited like the GH5S

you need stacking to read faster or front illumination"


So looks like the pixel count isn't limited. I gave you quite the example with the sony 61MP sensor. There's others around.
you don’t understand that there are 3 types of construction

front illuminated - sony exmor

back illuminated - sony exmor R

stacked back illuminated in multiple variants - sony exmor rs

Currently the exmor line has very few products it is the oldest line making apsc and MFT sensor with pixel size 3.3-3.97 microns

The exmor r line is the back illuminated line with higher sensitivity it has a wide range of pixel size however those have slow readout the only models that read 60 fps have 12 megapixels or less

Finally the rs line is the new mainstream pixel size up to 2.4 in normal cell and smaller for quad bayer

the sensor in the om-1 has 80 megapixels arranged in 20 megapixel cells

so yes stacking means higher count and smaller pixels
Am not familiar with fuji sony apsc stack uses traditional front illuminated sensors that are still very high performance.
? They are BSI.
nope sony apsc are front illuminated
Why don’t they change to stacked? Not sure but probably dont want to make a 32 megapixel apsc eating into their full frame
? Fuji just announced a 26MP Stacked sensor camera.
sony strategy is to get a winning full frame product and apsc is a commodity product
I was giving you the concrete example of Sony 61MP and Fuji's 26MP BSI sensors.
With more companies using high megapixels high speed full frame also those sensor will become mainstream

For MFT is a bit tricker especially if panasonic no longer uses sony so it will be hard to drive volumes

it would make sense for OMS to move all cameras to the new sensor to reduce purchase price instead kod getting other products
sensor costs is driven by volumes a line only does one process if you make 1 million of a certain product and 10000 of another video on the same line the latter will be significantly more expensive

this is the same reason why you buy a disk and a larger size costs less as raw materials quantity means nothing in the pricing while volumes matter

i think you need to refresh yourself a bit on the technology aspects

so your reasoning om-1 get a new different sensor to have lower cost doesn’t work
So your claim about stacked sensor costing the same as non-stacked is purely based on your speculation and not based on actual sensor pricing information offered by Sony, correct?

Even IF the raw manufacturing costs for the stacked bsi m43 sensor were the same as non-stacked m43 sensor (which is a big if), there's no reason Sony semi conductor would have to offer more desirable product for the same price as the less desirable variant, given the same sensor format. I think it's safe to assume that stacked sensor adds some cost to the manufacturer though the exact amount is still a mystery.
Sony manufacturing is no longer launching any front illuminated sensors since 2016
so everything will be stacked as that is now mainstream.
They seem to be doing BSI's, not necessarily all stacked.
thats the reason why the new mft sensor is stacked not because is more desirable it is simply the only option offered from now on
That doesn't seem to be accurate.
of course it is. The platform is the new mainstream that produces phones and all full frame for Sony and Nikon

Back illuminated is mostly low light security applications low megapixel or monochrome
Of course if you want to buy the old sensor that’s on alibaba at $90-130

sooner or later sony will migrate their apsc to exmor rs and the old exmor will be pronounced dead
At some point maybe that happens, well see.
it is happening hence you got the OM-1 and the DJI drones to use it
 
it started with compact cameras at 20 megapixels it was not otherwise possible
There were multiple generations of 20MP 1" sensors before Sony started introducing stacked versions.
yes but stacked started with the rx100 20 megapixels
What point do you think you're making here?
sony apsc runs ok front illumination and back illuminated full frame run on back illuminated non stacked so only for higher than 24 megapixels full frame stacking has been used
The first stacked sensor available in any MILC was the 24MP A9.
The A9 has a quad bayer cell to have many AF points so actually 96 megapixels
Where did you pull this one from? We have the datasheet for the IMX310, it's definitely not a quad bayer design. Plus it produces PDAF striping against strong backlights like all of Sony's other masked PDAF sensor designs.
From the same datasheet you mention readout mode 1B referring to HDR mode that can only be done with quad bayer it has been commented multiple times on various sony forums
This is not true. You can make HDR different ways.

4bedff40e39d429ebb884a08bd60e774.jpg.png

This looks more like what Panasonic did with the GH6. By the way, the Sony data sheet of the OM1 sensor clearly specs out it's a quad bayer. Nothing here.
The sensor did not have PDAF pixels and thats datasheet is doctored there has never been one coming out from sony
How do we know it was doctored?
and sony didn’t declare the a7s3 quad bayer but it is
That's a fair point. I would still love for you to provide evidence for your original claims though. That does include that BSI / BI designs limit pixel count.
the specs of products that are for general use can be trusted but there are none around for any custom ones

There were long discussion as far as the striping was concerned and didn’t lead to anything note tbagyit was not covering the whole sensor so certainly nothing to do with masking so how would this camera have all those AF pixels?


--
Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - Apparently Selwyn Duke and not George Orwell
 
Been saying they need to put the new sensor down the line but I am rethinking this now since stacked sensors are $$$

So to me the most important thing is the on-5 sensor if it’s not the stacked one, that I shares the modest image quality improvements of the new sensor and also able to auto iso to 25600

But all said I rather have them do a pro specced smaller om-5 priced high than a model with some obvious compromises to make it “cheap”

Even if that means a price north of $1,500 USD - say $1600 or even $1799 - make it right make it pro small That would be pretty unique in the market

thoughts? What you think is key for the om-5 to come out in todays market competitively - assuming this is an em5 mark iii follow up or even em1 mark iii
I think it has go one way or the other. Either the same sensor that's in OM1, or a non-stacked, higher resolution BSI sensor to cater to different needs. I don't think 20mp non-stacked sensor will be that appealing, given that it'll be competing with EM1III and EM5III in that price range, not to mention other m43 and APSC offerings in that range. But whether we see higher resolution sensor also depends on how well the current lenses can resolve the additional detail. OM doesn't seem to prioritize spec sheet bragging as much as real world improvements (e.g. they improved IBIS in OM1 even though it makes zero difference in CIPA rating since CIPA does not test for rolls) so I think it would happen only if they believe that there's real image resolution gain from higher res sensor.

There was a low credibility rumor about OM5 having higher resolution sensor. I think it's actually possible they would go for a higher resolution sensor. It will differentiate OM1 and OM5 in a way that will make them more appealing as a set, using one for wildlife and the other for nature
back illuminated sensor read slower so the pixel count is limited like the GH5S
Please explain to me why the Fuji APSC of 26MP then is BSI and 26 MP? That doesn't make sense. Also the Fuji 26MP is a relatively fast sensor (close to EM1.2/.3 readouts) in its class. Of course doesn't compare to the super fast stacked sensors but that's beside the point.

Evidence the Fuji 26MP is BSI and has been that for a while now:

https://www.dpreview.com/news/54716...s-with-a-26mp-x-trans-sensor-and-4k-60p-video
you need stacking to read faster or front illumination
Update: BSI enables actually bigger pixel counts without sacrificing image quality vs some degree of lower pixel counts. No, I am not going to bother looking where this piece of knowledge came from but I am pretty sure you can find it if you look for it. I already gave you evidence.

Hope you now return in kind.
There are back illuminated non stacked sensors typically used for low pixel count cameras like sony And panasonic 24 megapixels
You brought up the GH5s as an example of low pixel count BSI.
no the gh5s is not stacked
those read slow cant do 60 fos without crop
The Fuji 26MP ASPC sensor being used is not slow even if it's not hitting 60fps. It's definitively a step above the average CMOS sensor that was done until that time.
then there sre backilluminated stacked sensor with high pixel count those don’t do equally well in low light and are what goes in full frame high megapixel sensor.
This is a new claim. What you claimed was

"back illuminated sensor read slower so the pixel count is limited like the GH5S

you need stacking to read faster or front illumination"


So looks like the pixel count isn't limited. I gave you quite the example with the sony 61MP sensor. There's others around.
you don’t understand that there are 3 types of construction

front illuminated - sony exmor

back illuminated - sony exmor R

stacked back illuminated in multiple variants - sony exmor rs

Currently the exmor line has very few products it is the oldest line making apsc and MFT sensor with pixel size 3.3-3.97 microns

The exmor r line is the back illuminated line with higher sensitivity it has a wide range of pixel size however those have slow readout the only models that read 60 fps have 12 megapixels or less

Finally the rs line is the new mainstream pixel size up to 2.4 in normal cell and smaller for quad bayer

the sensor in the om-1 has 80 megapixels arranged in 20 megapixel cells

so yes stacking means higher count and smaller pixels
Am not familiar with fuji sony apsc stack uses traditional front illuminated sensors that are still very high performance.
? They are BSI.
nope sony apsc are front illuminated
Why don’t they change to stacked? Not sure but probably dont want to make a 32 megapixel apsc eating into their full frame
? Fuji just announced a 26MP Stacked sensor camera.
sony strategy is to get a winning full frame product and apsc is a commodity product
I was giving you the concrete example of Sony 61MP and Fuji's 26MP BSI sensors.
With more companies using high megapixels high speed full frame also those sensor will become mainstream

For MFT is a bit tricker especially if panasonic no longer uses sony so it will be hard to drive volumes

it would make sense for OMS to move all cameras to the new sensor to reduce purchase price instead kod getting other products
sensor costs is driven by volumes a line only does one process if you make 1 million of a certain product and 10000 of another video on the same line the latter will be significantly more expensive

this is the same reason why you buy a disk and a larger size costs less as raw materials quantity means nothing in the pricing while volumes matter

i think you need to refresh yourself a bit on the technology aspects

so your reasoning om-1 get a new different sensor to have lower cost doesn’t work
So your claim about stacked sensor costing the same as non-stacked is purely based on your speculation and not based on actual sensor pricing information offered by Sony, correct?

Even IF the raw manufacturing costs for the stacked bsi m43 sensor were the same as non-stacked m43 sensor (which is a big if), there's no reason Sony semi conductor would have to offer more desirable product for the same price as the less desirable variant, given the same sensor format. I think it's safe to assume that stacked sensor adds some cost to the manufacturer though the exact amount is still a mystery.
Sony manufacturing is no longer launching any front illuminated sensors since 2016
so everything will be stacked as that is now mainstream.
They seem to be doing BSI's, not necessarily all stacked.
thats the reason why the new mft sensor is stacked not because is more desirable it is simply the only option offered from now on
That doesn't seem to be accurate.
of course it is. The platform is the new mainstream that produces phones and all full frame for Sony and Nikon
What you mean of course it is- you keep making claims and provide no evidence.
Back illuminated is mostly low light security applications low megapixel or monochrome
Well as I gave you examples there's current BSI/BI cameras from Sony that are being used with a high megapixel count.
Of course if you want to buy the old sensor that’s on alibaba at $90-130

sooner or later sony will migrate their apsc to exmor rs and the old exmor will be pronounced dead
At some point maybe that happens, well see.
it is happening hence you got the OM-1 and the DJI drones to use it
We need to see how "high volume" that is first. Yes, I would expect it will happen over time, just not now. That's pretty. much the gist of the interview on that with Sony.

But I don't know what to tell you- you make claims, provide no evidence (you have been asked to do so several time), seem to ignore claims that have been debunked that you claimed and carry on with what seems making some things up.

I am sorry but if you are going to complain about not providing evidence-which was provided- the last we can expect is that you clearly provide yours and you have provided nothing.

This is my last reply to you in this thread- I can't really take your intent for a discussion seriously particularly when you have misstated several things that were proven wrong and you just ignore that.

--
Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - Apparently Selwyn Duke and not George Orwell
 
Last edited:
it started with compact cameras at 20 megapixels it was not otherwise possible
There were multiple generations of 20MP 1" sensors before Sony started introducing stacked versions.
yes but stacked started with the rx100 20 megapixels
What point do you think you're making here?
sony apsc runs ok front illumination and back illuminated full frame run on back illuminated non stacked so only for higher than 24 megapixels full frame stacking has been used
The first stacked sensor available in any MILC was the 24MP A9.
The A9 has a quad bayer cell to have many AF points so actually 96 megapixels
Where did you pull this one from? We have the datasheet for the IMX310, it's definitely not a quad bayer design. Plus it produces PDAF striping against strong backlights like all of Sony's other masked PDAF sensor designs.
From the same datasheet you mention readout mode 1B referring to HDR mode that can only be done with quad bayer it has been commented multiple times on various sony forums
This is not true. You can make HDR different ways.

4bedff40e39d429ebb884a08bd60e774.jpg.png

This looks more like what Panasonic did with the GH6. By the way, the Sony data sheet of the OM1 sensor clearly specs out it's a quad bayer. Nothing here.
The sensor did not have PDAF pixels and thats datasheet is doctored there has never been one coming out from sony
How do we know it was doctored?
and sony didn’t declare the a7s3 quad bayer but it is
That's a fair point. I would still love for you to provide evidence for your original claims though. That does include that BSI / BI designs limit pixel count.
BI is not the same as stacking and I did not say it physically limits pixel size

because back illuminated sensor read slower sony limited pixel count on those cameras to avoid them being unusable. Readout speed depends on technology used and pixel count

example the S5 that I think you have is 24 megapixels this has been the limit. The full frame readout at 14 bits is 19 fps hence the slow burst speed

it can read at 30 fps but needs to drop to 12 bits

to produce 4K 60 fps the sensor is apsc crop

if you increased the pixel count to 48megapixels you would have a useless camera

here comes stacking originally designed for mobile phones tiny sensor and tiny pixels sorts out speed issues and allows for HDR images, 240 fps slow motion etc

this technology then makes it to cameras together with the quad bayer or octo cells for autofocus which phones have once you get there you have cameras that have no pdaf banding read fast and have cells of c4 microns made by 2x2 smaller pixels array

all of this made possible by mobile phone technology

you can of course produce a high megapixel stacked sensor with PDAF masking but why would you?
the specs of products that are for general use can be trusted but there are none around for any custom ones

There were long discussion as far as the striping was concerned and didn’t lead to anything note tbagyit was not covering the whole sensor so certainly nothing to do with masking so how would this camera have all those AF pixels?
--
Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - Apparently Selwyn Duke and not George Orwell


--
instagram http://instagram.com/interceptor121
My flickr sets http://www.flickr.com/photos/interceptor121/
Youtube channel http://www.youtube.com/interceptor121
Underwater Photo and Video Blog http://interceptor121.com
Deer Photography workshops https://interceptor121.com/2021/09/26/2021-22-deer-photography-workshops-in-woburn/
 
Been saying they need to put the new sensor down the line but I am rethinking this now since stacked sensors are $$$

So to me the most important thing is the on-5 sensor if it’s not the stacked one, that I shares the modest image quality improvements of the new sensor and also able to auto iso to 25600

But all said I rather have them do a pro specced smaller om-5 priced high than a model with some obvious compromises to make it “cheap”

Even if that means a price north of $1,500 USD - say $1600 or even $1799 - make it right make it pro small That would be pretty unique in the market

thoughts? What you think is key for the om-5 to come out in todays market competitively - assuming this is an em5 mark iii follow up or even em1 mark iii
I think it has go one way or the other. Either the same sensor that's in OM1, or a non-stacked, higher resolution BSI sensor to cater to different needs. I don't think 20mp non-stacked sensor will be that appealing, given that it'll be competing with EM1III and EM5III in that price range, not to mention other m43 and APSC offerings in that range. But whether we see higher resolution sensor also depends on how well the current lenses can resolve the additional detail. OM doesn't seem to prioritize spec sheet bragging as much as real world improvements (e.g. they improved IBIS in OM1 even though it makes zero difference in CIPA rating since CIPA does not test for rolls) so I think it would happen only if they believe that there's real image resolution gain from higher res sensor.

There was a low credibility rumor about OM5 having higher resolution sensor. I think it's actually possible they would go for a higher resolution sensor. It will differentiate OM1 and OM5 in a way that will make them more appealing as a set, using one for wildlife and the other for nature
back illuminated sensor read slower so the pixel count is limited like the GH5S
Please explain to me why the Fuji APSC of 26MP then is BSI and 26 MP? That doesn't make sense. Also the Fuji 26MP is a relatively fast sensor (close to EM1.2/.3 readouts) in its class. Of course doesn't compare to the super fast stacked sensors but that's beside the point.

Evidence the Fuji 26MP is BSI and has been that for a while now:

https://www.dpreview.com/news/54716...s-with-a-26mp-x-trans-sensor-and-4k-60p-video
you need stacking to read faster or front illumination
Update: BSI enables actually bigger pixel counts without sacrificing image quality vs some degree of lower pixel counts. No, I am not going to bother looking where this piece of knowledge came from but I am pretty sure you can find it if you look for it. I already gave you evidence.

Hope you now return in kind.
There are back illuminated non stacked sensors typically used for low pixel count cameras like sony And panasonic 24 megapixels
You brought up the GH5s as an example of low pixel count BSI.
no the gh5s is not stacked
those read slow cant do 60 fos without crop
The Fuji 26MP ASPC sensor being used is not slow even if it's not hitting 60fps. It's definitively a step above the average CMOS sensor that was done until that time.
then there sre backilluminated stacked sensor with high pixel count those don’t do equally well in low light and are what goes in full frame high megapixel sensor.
This is a new claim. What you claimed was

"back illuminated sensor read slower so the pixel count is limited like the GH5S

you need stacking to read faster or front illumination"


So looks like the pixel count isn't limited. I gave you quite the example with the sony 61MP sensor. There's others around.
you don’t understand that there are 3 types of construction

front illuminated - sony exmor

back illuminated - sony exmor R

stacked back illuminated in multiple variants - sony exmor rs

Currently the exmor line has very few products it is the oldest line making apsc and MFT sensor with pixel size 3.3-3.97 microns

The exmor r line is the back illuminated line with higher sensitivity it has a wide range of pixel size however those have slow readout the only models that read 60 fps have 12 megapixels or less

Finally the rs line is the new mainstream pixel size up to 2.4 in normal cell and smaller for quad bayer

the sensor in the om-1 has 80 megapixels arranged in 20 megapixel cells

so yes stacking means higher count and smaller pixels
Am not familiar with fuji sony apsc stack uses traditional front illuminated sensors that are still very high performance.
? They are BSI.
nope sony apsc are front illuminated
Why don’t they change to stacked? Not sure but probably dont want to make a 32 megapixel apsc eating into their full frame
? Fuji just announced a 26MP Stacked sensor camera.
sony strategy is to get a winning full frame product and apsc is a commodity product
I was giving you the concrete example of Sony 61MP and Fuji's 26MP BSI sensors.
With more companies using high megapixels high speed full frame also those sensor will become mainstream

For MFT is a bit tricker especially if panasonic no longer uses sony so it will be hard to drive volumes

it would make sense for OMS to move all cameras to the new sensor to reduce purchase price instead kod getting other products
sensor costs is driven by volumes a line only does one process if you make 1 million of a certain product and 10000 of another video on the same line the latter will be significantly more expensive

this is the same reason why you buy a disk and a larger size costs less as raw materials quantity means nothing in the pricing while volumes matter

i think you need to refresh yourself a bit on the technology aspects

so your reasoning om-1 get a new different sensor to have lower cost doesn’t work
So your claim about stacked sensor costing the same as non-stacked is purely based on your speculation and not based on actual sensor pricing information offered by Sony, correct?

Even IF the raw manufacturing costs for the stacked bsi m43 sensor were the same as non-stacked m43 sensor (which is a big if), there's no reason Sony semi conductor would have to offer more desirable product for the same price as the less desirable variant, given the same sensor format. I think it's safe to assume that stacked sensor adds some cost to the manufacturer though the exact amount is still a mystery.
Sony manufacturing is no longer launching any front illuminated sensors since 2016
so everything will be stacked as that is now mainstream.
They seem to be doing BSI's, not necessarily all stacked.
thats the reason why the new mft sensor is stacked not because is more desirable it is simply the only option offered from now on
That doesn't seem to be accurate.
of course it is. The platform is the new mainstream that produces phones and all full frame for Sony and Nikon
What you mean of course it is- you keep making claims and provide no evidence.
Back illuminated is mostly low light security applications low megapixel or monochrome
Well as I gave you examples there's current BSI/BI cameras from Sony that are being used with a high megapixel count.
Of course if you want to buy the old sensor that’s on alibaba at $90-130

sooner or later sony will migrate their apsc to exmor rs and the old exmor will be pronounced dead
At some point maybe that happens, well see.
it is happening hence you got the OM-1 and the DJI drones to use it
We ened to see how "high volume" that is first. Yes, I would expect it will happen over time, just not now. That's pretty. much the gist of the interview on that with Sony.

But I don't know what to tell you- you make claims, provide no evidence (you have been asked to do so several time), seem to ignore claims that have been debunked that you claimed and carry on with what seems making some things up.

I am sorry but if you are going to complain about not providing evidence-which was provided- the last we can expect is that you clearly provide yours and you have provided nothing.

This is my last reply to you in this thread- I can't really take your intent for a discussion seriously particularly when you have misstated several things that were proven wrong and you just ignore that.
 
Thanks for making m point - for providing ZERO evidence for any of your claims. And throwing in non sensical claims like we needed stacked to do 20MP in 1'' sensors (false and proven), that BSI sensors limit pixel counts (false and proven), completely ignoring that sensor size is a big variable in sensor costs in addition to economies of scale- this has been a pretty basic one since forever.

In none of these no corrections, no retractions, no evidence.
 
if you increased the pixel count to 48megapixels you would have a useless camera
IMX492 is a non-stacked sensor, it's already available in some astro cameras.
here comes stacking originally designed for mobile phones tiny sensor and tiny pixels sorts out speed issues and allows for HDR images, 240 fps slow motion etc

this technology then makes it to cameras together with the quad bayer or octo cells for autofocus which phones have once you get there you have cameras that have no pdaf banding read fast and have cells of c4 microns made by 2x2 smaller pixels array

all of this made possible by mobile phone technology

you can of course produce a high megapixel stacked sensor with PDAF masking but why would you?
Because it will read out faster than the equivalent quad bayer setup.
 
Been saying they need to put the new sensor down the line but I am rethinking this now since stacked sensors are $$$

So to me the most important thing is the on-5 sensor if it’s not the stacked one, that I shares the modest image quality improvements of the new sensor and also able to auto iso to 25600

But all said I rather have them do a pro specced smaller om-5 priced high than a model with some obvious compromises to make it “cheap”

Even if that means a price north of $1,500 USD - say $1600 or even $1799 - make it right make it pro small That would be pretty unique in the market

thoughts? What you think is key for the om-5 to come out in todays market competitively - assuming this is an em5 mark iii follow up or even em1 mark iii
Note that the GH-6 does NOT have a stacked sensor, yet its readout speed is extremely high.
Yeah. True.
The E-M5 type of camera arguably doesn’t have to push the extremes of readout speed. This is even more true as the OM-1 already exists, and those who value 120 fps photography are already catered to by OMDS.
To be clear, what I want the most here on an OM-5 is that it keeps the modest gains in ISO/color preservation of the OM1 new sensor.
It would make more sense to create an E-M5 type camera that complements the OM-1 in the market, rather than just being a smaller, straight downgrade.

The OM-1 is targeted at wildlife/birds and to some extent sports. Thus I would target the E-M5 at travel and landscape. Lower weight, not geared towards attaching the longest teles, still weather sealed, higher resolution, compromising on frame rates, and probably grip and battery to save weight.
Maybe. I just think that by going two different sensors OMDS will miss on potential economies of scale. But I'll keep your thoughts as we head into July and see what comes out. I see your point.
The current E-M5 with a prime stays within half a kilo, with the 12-45f4 + 40-150f4 it’s still holds at a kilo with really good image quality through the range. (The 8-25mm and 12-100mm being obvious excellent travelling companions as well.) A camera that blends with these lens options for travel/hiking would be quite attractive. The current E-M5iii already does this however, no need for a new model on that account, so I wish the high resolution sensor rumour will turn out correct. It would justify a new model, and it would occupy its very own niche in the m43 ecosystem and arguably market as a whole, something a cut down OM-1 wouldn’t.
 
it started with compact cameras at 20 megapixels it was not otherwise possible
There were multiple generations of 20MP 1" sensors before Sony started introducing stacked versions.
yes but stacked started with the rx100 20 megapixels
What point do you think you're making here?
sony apsc runs ok front illumination and back illuminated full frame run on back illuminated non stacked so only for higher than 24 megapixels full frame stacking has been used
The first stacked sensor available in any MILC was the 24MP A9.
The A9 has a quad bayer cell to have many AF points so actually 96 megapixels
Where did you pull this one from? We have the datasheet for the IMX310, it's definitely not a quad bayer design. Plus it produces PDAF striping against strong backlights like all of Sony's other masked PDAF sensor designs.
From the same datasheet you mention readout mode 1B referring to HDR mode that can only be done with quad bayer it has been commented multiple times on various sony forums
This is not true. You can make HDR different ways.

4bedff40e39d429ebb884a08bd60e774.jpg.png

This looks more like what Panasonic did with the GH6. By the way, the Sony data sheet of the OM1 sensor clearly specs out it's a quad bayer. Nothing here.
The sensor did not have PDAF pixels and thats datasheet is doctored there has never been one coming out from sony
And you know that it's doctored because...?
and sony didn’t declare the a7s3 quad bayer but it is
Really? [EDIT: Actually looks like I'm out of date on this one and the actual part number is IMX510AQL. Do we have a datasheet from that sensor that doesn't mention quad pixel?]

5413b2c867c04162b298289267ea89e1.jpg
the specs of products that are for general use can be trusted but there are none around for any custom ones

There were long discussion as far as the striping was concerned and didn’t lead to anything note tbagyit was not covering the whole sensor so certainly nothing to do with masking so how would this camera have all those AF pixels?
I'm honestly not sure how to parse this sentence. The A9 shows exactly the same type of PDAF striping against bright light as we've seen in the past from all of Sony's other sensors that use masked-type PDAF pixels. The quad-bayer type sensors, fortunately, don't show any behavior of that sort. Additionally, the A9 is only sensitive to vertical contrast unlike the quad-bayer arrangement. Feel free to actually supply some other evidence that it's a quad-bayer sensor though.
 
Last edited:
Been saying they need to put the new sensor down the line but I am rethinking this now since stacked sensors are $$$

So to me the most important thing is the on-5 sensor if it’s not the stacked one, that I shares the modest image quality improvements of the new sensor and also able to auto iso to 25600

But all said I rather have them do a pro specced smaller om-5 priced high than a model with some obvious compromises to make it “cheap”

Even if that means a price north of $1,500 USD - say $1600 or even $1799 - make it right make it pro small That would be pretty unique in the market

thoughts? What you think is key for the om-5 to come out in todays market competitively - assuming this is an em5 mark iii follow up or even em1 mark iii
Note that the GH-6 does NOT have a stacked sensor, yet its readout speed is extremely high.
Yeah. True.
The E-M5 type of camera arguably doesn’t have to push the extremes of readout speed. This is even more true as the OM-1 already exists, and those who value 120 fps photography are already catered to by OMDS.
To be clear, what I want the most here on an OM-5 is that it keeps the modest gains in ISO/color preservation of the OM1 new sensor.
It would make more sense to create an E-M5 type camera that complements the OM-1 in the market, rather than just being a smaller, straight downgrade.

The OM-1 is targeted at wildlife/birds and to some extent sports. Thus I would target the E-M5 at travel and landscape. Lower weight, not geared towards attaching the longest teles, still weather sealed, higher resolution, compromising on frame rates, and probably grip and battery to save weight.
Maybe. I just think that by going two different sensors OMDS will miss on potential economies of scale. But I'll keep your thoughts as we head into July and see what comes out. I see your point.
Economy of scale is a value, and if it dictates the choice we pretty much know what to expect, except for price.

Personally, I really like how the E-M5 series embody the m43s virtues, so I’ll basically be happy simply if OMDS extends the concept. I won’t actually buy it however unless it brings something to the table that wasn’t in my camera bag already.



If the rumours speak truly we’ll see soon enough.
 
it started with compact cameras at 20 megapixels it was not otherwise possible
There were multiple generations of 20MP 1" sensors before Sony started introducing stacked versions.
yes but stacked started with the rx100 20 megapixels
What point do you think you're making here?
sony apsc runs ok front illumination and back illuminated full frame run on back illuminated non stacked so only for higher than 24 megapixels full frame stacking has been used
The first stacked sensor available in any MILC was the 24MP A9.
The A9 has a quad bayer cell to have many AF points so actually 96 megapixels
Where did you pull this one from? We have the datasheet for the IMX310, it's definitely not a quad bayer design. Plus it produces PDAF striping against strong backlights like all of Sony's other masked PDAF sensor designs.
From the same datasheet you mention readout mode 1B referring to HDR mode that can only be done with quad bayer it has been commented multiple times on various sony forums
This is not true. You can make HDR different ways.

4bedff40e39d429ebb884a08bd60e774.jpg.png

This looks more like what Panasonic did with the GH6. By the way, the Sony data sheet of the OM1 sensor clearly specs out it's a quad bayer. Nothing here.
The sensor did not have PDAF pixels and thats datasheet is doctored there has never been one coming out from sony
How do we know it was doctored?
and sony didn’t declare the a7s3 quad bayer but it is
That's a fair point. I would still love for you to provide evidence for your original claims though. That does include that BSI / BI designs limit pixel count.
BI is not the same as stacking and I did not say it physically limits pixel size
BI is not the same as stacking, I didn't say it was. But you did mention that BI/BSI limits pixel count unless you mistyped. I have quoted this to you several times by now.

BY THE WAY from dpreview's new article on X-H2s:


"The Stacked CMOS chip that enables the X-H2S's high performance for both stills and video is also part of why the camera is so much more expensive than previous X-series models."

"Stacked CMOS sensors tend to be much more expensive to make, but they make possible the high-speed shooting, fast-refreshing viewfinders, advanced AF and high-end video capabilities that the X-H2S offers. This means the X-H2S looks expensive compared with its BSI sister models and predecessor (and, frankly, a bit pricey compared with mid-level full-frame cameras), but promises a level of performance we've only seen from other Stacked CMOS cameras."

because back illuminated sensor read slower sony limited pixel count on those cameras to avoid them being unusable. Readout speed depends on technology used and pixel count

example the S5 that I think you have is 24 megapixels this has been the limit. The full frame readout at 14 bits is 19 fps hence the slow burst speed

it can read at 30 fps but needs to drop to 12 bits

to produce 4K 60 fps the sensor is apsc crop

if you increased the pixel count to 48megapixels you would have a useless camera
I gave you examples of a not useless shipping camera that does not use a stacked sensor. Therefore your claim is wrong.
here comes stacking originally designed for mobile phones tiny sensor and tiny pixels sorts out speed issues and allows for HDR images, 240 fps slow motion etc

this technology then makes it to cameras together with the quad bayer or octo cells for autofocus which phones have once you get there you have cameras that have no pdaf banding read fast and have cells of c4 microns made by 2x2 smaller pixels array

all of this made possible by mobile phone technology

you can of course produce a high megapixel stacked sensor with PDAF masking but why would you?
the specs of products that are for general use can be trusted but there are none around for any custom ones

There were long discussion as far as the striping was concerned and didn’t lead to anything note tbagyit was not covering the whole sensor so certainly nothing to do with masking so how would this camera have all those AF pixels?
--
Raist3d/Ricardo (Photographer, software dev.)- I photograph black cats in coal mines at night...
“The further a society drifts from truth the more it will hate those who speak it.” - Apparently Selwyn Duke and not George Orwell
 

Keyboard shortcuts

Back
Top