Readout speed for GFX 100 II is the same as the 100, 100S!

Thanks Jim for the test.

I wonder how the camera can take 8fps if the readout speed is just 1/6s?
Do you have an idea?
 
This is a shame. Fuji has said the readout speed is twice as fast for the new camera, and my testing doesn't bear that out at all.

https://blog.kasson.com/the-last-word/how-fast-is-the-gfx-100-ii-electronic-shutter/
So what do they mean by claiming that it's twice as fast at read out? Is that in video mode only?
They didn't specify that it was in video mode

a42c80b20b674cc692084ed16bb61c33.jpg
Translated by G from the Japanese website:

13b51e7711d840df9276a2ac96b756d9.jpg.png

Really, I can't imagine Fuji would deliberately lie here, but who knows.

Or is it a matter of the wrong choice of words causing the confusion ?

Perhaps and hopefully future firmware will provide these claimed benefits.

Kind regards.
..............................................................................................................................
Bart
Im not tryng to defend Fuji and strangely I see No lies given they clearly shared during Fuji summit - 100 ii has a new 102mp version ii cmos sensor.

version ii

version ii HS

IMX 461 version II High speed?

I think they are well covered for legal with their choice of words too. Sony can easily help to make an annoucemen w IMX 461 ii HS tomorrow. Would Pentax, Hass, PhaseOne ever use this version ii HS sensor? Ill think they would gladly use v2 given the same pricing as imx461 v1.

Neverthrless - Great effort and still few steps ahead of X2D or PhaseOne - though we will know if this effort to develop imx461 version II HS is a genuine "One small step at a time" contribution to pushing MF sensors boundaries , few years down the road. I am not sure if anyone has ever done that for such a big sensor or in the history of MF sensor? I can only think of Sony recycled same FF sensor e.g a7s a7s ii a7rii a7riii making them faster.

Personally, I would love to see a high speed stacked-BSI 50mp HS 44x33 sensor ☺️
 
Last edited:
This is a shame. Fuji has said the readout speed is twice as fast for the new camera, and my testing doesn't bear that out at all.

https://blog.kasson.com/the-last-word/how-fast-is-the-gfx-100-ii-electronic-shutter/
Thats really a shame. Fuji could have increase 100ii price to $10k -$15k and equip with better, proven premium parts.

what is readout speed btw? lol
How long if takes the camera to read all the image data off the sensor. Faster means less rolling shutter.
Hi,

It is about sensor design and not about premium parts. Sony designs and makes the sensors used by Fujifilm and it seems pretty obvious the sensor used is the same.

Fujifilm probably has made a lot of improvements, with faster readout circuits, probably including a new ASIC. That ASIC would also depend in large part on a third party, including intellectual property and manufacture. That is how things work.

Best regards

Erik
 
Very cool! I would like to try the same with my Sony A1 (even if the scan time has been known for years). Couldn't you just measure the distance between the dark lines and thus calculate the travel over 8.33msec and use that to calculate the scan time of the whole sensor? In that way you wouldn't have to guesstimate the number of bands in the black region? Not that it would make a significant difference of course.
 
Thanks Jim for the test.

I wonder how the camera can take 8fps if the readout speed is just 1/6s?
Do you have an idea?
My guess may be that the camera may switch to a faster readout mode, like 12 bits, when shooting the highest frame rate.

Best regards

Erik
I agree with your guess. My assumption before the release was that the 100 II used the same physical image sensor, plus optimizations or tweaks using a newer faster processor and accessing extended hardware modes. Previous image processors not able to run the sensor in all of its readout modes, I suppose.
 
when using the EVF on the 100 iii does it feel like there is a lag between what you are seeing and what it is capturing to the point it hobbles the camera from being useful for action ? ,,,,,,
I understand its not a Z9 but to a reasonable degree is it pleasurable to use when there is quick panning and zooming ?
is the EVF continuous s viewing or are there highly distracting blackouts .... if so how severe ?
would you say that the camera is really only suitable for landscape . fashion and product ?
What's your agenda behind those questions? What are you trying to say?
 
It would be quite interesting to see Jims test with 8fps continuous shooting. I would expect it to look different. But would a somehow different readout mode affect image quality?

Jim if you could do that, it would be great. But there seems to be so much going on, I could understand if you focus your energy on your own agenda, and not the curiosities of all those forum members ;-)
 
It would be quite interesting to see Jims test with 8fps continuous shooting. I would expect it to look different. But would a somehow different readout mode affect image quality?
Jim if you could do that, it would be great. But there seems to be so much going on, I could understand if you focus your energy on your own agenda, and not the curiosities of all those forum members ;-)
A 12-bit readout will negatively affect IQ at lower ISOs.
 
It would be quite interesting to see Jims test with 8fps continuous shooting. I would expect it to look different. But would a somehow different readout mode affect image quality?
Jim if you could do that, it would be great. But there seems to be so much going on, I could understand if you focus your energy on your own agenda, and not the curiosities of all those forum members ;-)
A 12-bit readout will negatively affect IQ at lower ISOs.
IF Fuji does this with a 12-bit readout mode. We just don't know yet.

I do not know much about sensors. But from a mathematical point of view 12bit vs. 14bit should save 1/7 of data. Not enought to push a sensor with a readout time of 1/6s to 8fps. Or do I miss something?
 
Last edited:
It would be quite interesting to see Jims test with 8fps continuous shooting. I would expect it to look different. But would a somehow different readout mode affect image quality?
Jim if you could do that, it would be great. But there seems to be so much going on, I could understand if you focus your energy on your own agenda, and not the curiosities of all those forum members ;-)
A 12-bit readout will negatively affect IQ at lower ISOs.
IF Fuji does this with a 12-bit readout mode. We just don't know yet.

I do not know much about sensors. But from a mathematical point of view 12bit vs. 14bit should save 1/7 of data. Not enought to push a sensor with a readout time of 1/6s to 8fps. Or do I miss something?
On GFX 100, a 14-bit readout is twice as fast as a 16-bit readout.
 
It would be quite interesting to see Jims test with 8fps continuous shooting. I would expect it to look different. But would a somehow different readout mode affect image quality?
Jim if you could do that, it would be great. But there seems to be so much going on, I could understand if you focus your energy on your own agenda, and not the curiosities of all those forum members ;-)
A 12-bit readout will negatively affect IQ at lower ISOs.
IF Fuji does this with a 12-bit readout mode. We just don't know yet.

I do not know much about sensors. But from a mathematical point of view 12bit vs. 14bit should save 1/7 of data. Not enought to push a sensor with a readout time of 1/6s to 8fps. Or do I miss something?
On GFX 100, a 14-bit readout is twice as fast as a 16-bit readout.
True. I don't know why 2 bits less make it twice as fast, but you're right.
 
It would be quite interesting to see Jims test with 8fps continuous shooting. I would expect it to look different. But would a somehow different readout mode affect image quality?
Jim if you could do that, it would be great. But there seems to be so much going on, I could understand if you focus your energy on your own agenda, and not the curiosities of all those forum members ;-)
A 12-bit readout will negatively affect IQ at lower ISOs.
IF Fuji does this with a 12-bit readout mode. We just don't know yet.

I do not know much about sensors. But from a mathematical point of view 12bit vs. 14bit should save 1/7 of data. Not enought to push a sensor with a readout time of 1/6s to 8fps. Or do I miss something?
14 bits encode 4 times the amount of data compared to 12 bits. So, you can scan 4 times more data with 12 bits than with 14 bits.

But, we could also use 13 bit conversion and put the data into a 14 bit envelope.

One other observation may be that the sensor can probably resolve around 14 bit of information at 80 ISO. But every doubling of ISO is reducing the resolved data by half. Which means that dynamic range drops.

So, once you are around 400 ISO, the amount of data worth sampling drops to about 11.5 bits of data, than rebounds to around 12.5 bits around 500 ISO, due to dual gain.

This article may give some info on Analogue Digital Conversion:


Best regards

Erik
 
This is a shame. Fuji has said the readout speed is twice as fast for the new camera, and my testing doesn't bear that out at all.

https://blog.kasson.com/the-last-word/how-fast-is-the-gfx-100-ii-electronic-shutter/
Thats really a shame. Fuji could have increase 100ii price to $10k -$15k and equip with better, proven premium parts.

what is readout speed btw? lol
How long if takes the camera to read all the image data off the sensor. Faster means less rolling shutter.
I wonder if it's a translation error in the marketing. They probably meant 2x internal bandwidth to benefit FPS, clearing the buffer, processing intense video codecs, etc. because right after that they mention FPS and video. If it had been 2x sensor readout speed, they would have mentioned rolling shutter reduction specifically.
They also said it was a new sensor, and it’s looking more and more like the old sensor with a few tweaks.
Is it a new sensor if it's just a variant of its predecessor? In English there's a big difference between a new sensor and a new version of a sensor.
I think they claim new microlenses as well that improve corner sharpness. That interests me more than the readout speed claims. If the same lens can be tested on the 100S and 100II and compared for corner sharpness that would be nice.
I am under the impression that readout speed is critical in a mirrorless cameras focus system
AF readout rates are usually faster than whole sensor rates.
 
This is a shame. Fuji has said the readout speed is twice as fast for the new camera, and my testing doesn't bear that out at all.

https://blog.kasson.com/the-last-word/how-fast-is-the-gfx-100-ii-electronic-shutter/
Thats really a shame. Fuji could have increase 100ii price to $10k -$15k and equip with better, proven premium parts.

what is readout speed btw? lol
How long if takes the camera to read all the image data off the sensor. Faster means less rolling shutter.
I wonder if it's a translation error in the marketing. They probably meant 2x internal bandwidth to benefit FPS, clearing the buffer, processing intense video codecs, etc. because right after that they mention FPS and video. If it had been 2x sensor readout speed, they would have mentioned rolling shutter reduction specifically.
They also said it was a new sensor, and it’s looking more and more like the old sensor with a few tweaks.
So if the sensor is based on the architect of a Sony A67700/FX30, is it different from the old Gfx100 and gfx100s sensors?

See, “The Fujifilm GFX100 II will come with a new 100 megapixel sensor. Now, in our previous rumors we told you it would be based on the sensor architecture of the Sony A6700/FX30 sensor. Hence a BSI sensor that is faster to read out than previous generation of BSI sensor”- FujiRumors

My understanding from the boards here is the GFX100s sensor is same as Sony A7…. So, how is the Sony A67700/FX30 sensor different from other Sony sensors? If it is different what differences are seen in Sony cameras?

As for 8 fps have people tested this? Is it easy do? Do you need put it in frame rate boost mode and use express card or do you get it without boost and to sd card. The FujiLove podcast interviewee mentions how you get 42 image burst before buffer slow down as compared to 15 image burst slow down with gfx100s.
 
LP, I had not heard this. I have several Godox Speedlites and the latest trigger, so I will check it this morning.

If that is true, Godox will fix it with a firmware update.

No one buys Fuji flashes (although I have one for when I'm just using it on the shoe with no trigger and shooting the kids running around). If you shoot multiple speedlites TTL, you gotta go Godox.
 
Very cool! I would like to try the same with my Sony A1 (even if the scan time has been known for years). Couldn't you just measure the distance between the dark lines and thus calculate the travel over 8.33msec and use that to calculate the scan time of the whole sensor? In that way you wouldn't have to guesstimate the number of bands in the black region? Not that it would make a significant difference of course.
That would work if you measured the cycles precisely enough.
 
Jim, I have not read the responses yet. But that is amazing.

If it is true and not some glitch, you will hear from Fuji either pro or con.

You can be absolutely sure they have read your post.

I just read an article that said it is easy for devices that provide read and write speeds by hooking it up to various USB-C ports to be faulty and provide low numbers (like hooking it up to a USB-C 3.2 Gen 2x2 port (20 Gbps) and it reading less.

I bought one of those devices because I have several ports of various speeds (5, 10, 20 and 40 Gbps, and with cable variations and total wattage loads I never trust what the port is marked as.

That probably has absolutely no bearing on what you measured (I was talking about USB-C and A ports) and I'm certainly not commenting on your test except that if you are right, that is bad news and a big deal.
 
This is a shame. Fuji has said the readout speed is twice as fast for the new camera, and my testing doesn't bear that out at all.

https://blog.kasson.com/the-last-word/how-fast-is-the-gfx-100-ii-electronic-shutter/
Thank you Jim.

So the manual indicating the maximum flash sync speed
The maximum synch speed is not determined by the readout speed, but by the mechanical shutter.
and the Dpreview announcement indicating the same scan speed of the sensor were correct.

Thats a shame that its not a faster sensor, just faster ram.
 

Keyboard shortcuts

Back
Top