p.3 #1 · Speculation time: Sony a7V body design and...
Ltgk20 wrote:
Getting back to the OP's question, I've seen the rumors that it'll have the same body as the A1, A7Rv, etc. However, thus far in E-mount, I believe when a body change was made, every subsequent camera shared those body changes until new body changes came out. If this is true, then I think it likely the A7V will get the A9iii and A1ii body, likely minus the right control wheels, and possibly the C5.
Anything's possible. The usual Roumors sites have been predicting the new body design for a long time, but recently, it's been said that the RV bodydesign is set. I would have guessed too the new design would be coming, but we'll see soon enough.
p.3 #2 · Speculation time: Sony a7V body design and...
Steve Spencer wrote:
Keep in mind the A9 III is also pushing less data with it 24 MP sensor than the 50 and 60 MP sensors, so it is easier to keep up the VF quality. That and the ultra fast global shutter allow the camera to keep up the viewfinder resolution.
A9iii is pushing 120 * 24 * 14 =40,320 Mb/s
A1ii is pushing 30 * 50 * 12 =18,000 Mb/s
there is no difference in any of the sony cameras as they all use the same processor, the a93 can only shoot a burst for 1.6 secs and takes 12 secs to clear, the a7iv can shoot continously till the card fills at 10 fps. the a1 can only shoot for 5 secs at 30fps till it starts buffereing.
p.3 #4 · Speculation time: Sony a7V body design and...
OK, instead of guessing, I asked Google AI Pro/Gemini 2.5
How is the data read?
The sensor reads all 24.6 MPs simultaneously and instantaneously, not row-by-row. There is no scan time. This simultaneous, instantaneous readout is incredibly efficient. It eliminates the sequential data bottleneck and frees up the processing pipeline. The camera has massive bandwidth available because it's not waiting for a sequential readout process to finish.
Why Resolution Doesn't Drop?
Because the bottleneck is gone, the camera can easily maintain full EVF resolution while still performing 120 AE/AF calculations per second and shooting at up to 120 fps . The data for the high-quality EVF feed is always available instantly, allowing the camera to dedicate enough power to the EVF without having to sacrifice its fidelity.
p.3 #5 · Speculation time: Sony a7V body design and...
A74me wrote:
there is no difference in any of the sony cameras as they all use the same processor, the a93 can only shoot a burst for 1.6 secs and takes 12 secs to clear, the a7iv can shoot continously till the card fills at 10 fps. the a1 can only shoot for 5 secs at 30fps till it starts buffereing.
You can't really compare a7IV's 12-bit compressed RAW at 10 fps to reading a 50 MP sensor in 14-bit RAW (any format) at more fps.
Still, I don't argue that the a7IV doesn't have a great buffer capacity with the right CFExpress-A card being used.
Oct 20, 2025 at 04:09 AM
Steve Spencer Offline • • • • • • • Upload & Sell: On
p.3 #6 · Speculation time: Sony a7V body design and...
patotts wrote:
OK, instead of guessing, I asked Google AI Pro/Gemini 2.5
How is the data read?
The sensor reads all 24.6 MPs simultaneously and instantaneously, not row-by-row. There is no scan time. This simultaneous, instantaneous readout is incredibly efficient. It eliminates the sequential data bottleneck and frees up the processing pipeline. The camera has massive bandwidth available because it's not waiting for a sequential readout process to finish.
Why Resolution Doesn't Drop?
Because the bottleneck is gone, the camera can easily maintain full EVF resolution while still performing 120 AE/AF calculations per second and shooting at up to 120 fps . The data for the high-quality EVF feed is always available instantly, allowing the camera to dedicate enough power to the EVF without having to sacrifice its fidelity....Show more →
You mean you let Google AI Pro/Gemini 2.5 guess. Probably better than my guess, but let's not confuse what AI generates with facts. It does hallucinate and in ways that are hard to predict and sometimes even hard to catch.
p.3 #7 · Speculation time: Sony a7V body design and...
Sure, it hallucinates or fills in the gaps at times, but that also depends on your skills around prompting (before and after getting the first set of answers). In this case, I concluded that part of the info I shared (it wasn't everything the machine spit out) was very reasonable, and certainly as close to the right answer as any of us could come up with. But sure, let's not call it facts.
p.3 #8 · Speculation time: Sony a7V body design and...
A74me wrote:
there is no difference in any of the sony cameras as they all use the same processor, the a93 can only shoot a burst for 1.6 secs and takes 12 secs to clear, the a7iv can shoot continously till the card fills at 10 fps. the a1 can only shoot for 5 secs at 30fps till it starts buffereing.
There’s a difference between the rate at which data is transferred from the sensor to the RAM (buffer) and the rate at which it’s written from the RAM to the CF card. Regarding the data rate that affects the EVF, it’s the former, the EVF video feed never reaches the CF card.
What you are describing the latter.
Oct 20, 2025 at 06:31 AM
Steve Spencer Offline • • • • • • • Upload & Sell: On
p.3 #9 · Speculation time: Sony a7V body design and...
patotts wrote:
Sure, it hallucinates or fills in the gaps at times, but that also depends on your skills around prompting (before and after getting the first set of answers). In this case, I concluded that part of the info I shared (it wasn't everything the machine spit out) was very reasonable, and certainly as close to the right answer as any of us could come up with. But sure, let's not call it facts.
For the record, we don't know what hallucinations depend on or what produces them. It may be that skills around prompting may reduce them, but I don't know that anybody has shown that. It could also be that people thinking they are skillfully prompting AI produces more hallucinations.
My basic point is that although AI is very useful as we integrate it more into our daily lives, I think a healthy does of skepticism about what it produces is also a good thing. Further if people do prompt it in a given way to produce a given result and edit what data they present from what it produces then I think it is helpful to describe both the prompting process and the editing process so others can come to the same conclusion about whether they believe what was produced was reasonable or not.
p.3 #10 · Speculation time: Sony a7V body design and...
patotts wrote:
OK, instead of guessing, I asked Google AI Pro/Gemini 2.5
How is the data read?
The sensor reads all 24.6 MPs simultaneously and instantaneously, not row-by-row. There is no scan time. This simultaneous, instantaneous readout is incredibly efficient. It eliminates the sequential data bottleneck and frees up the processing pipeline. The camera has massive bandwidth available because it's not waiting for a sequential readout process to finish.
Why Resolution Doesn't Drop?
Because the bottleneck is gone, the camera can easily maintain full EVF resolution while still performing 120 AE/AF calculations per second and shooting at up to 120 fps . The data for the high-quality EVF feed is always available instantly, allowing the camera to dedicate enough power to the EVF without having to sacrifice its fidelity....Show more →
The AI is missing the point I think.
I have written code to grab data from cameras. If you did this you will find that some sensors / cameras are able to output multiple data streams simultaneously.
Nikon actually markets this as "dual-stream technology", but its not really that novel. Even some security cameras can do it these days.
The A9iii sensor is capable of the following simultaneously
1. 120fps 24MP 14 bit raw image stream
2. 120hz High Resolution EVF image Stream
The A1ii sensor is capable of the following simultaneously
1. 30fps 50MP 12 bit raw image stream
2. 120hz (i think) Standard Resolution EVF image Stream
The A1ii sensor can output this independently
1. 60hz High Resolution EVF image Stream
If you think about it, it makes sense that the A1 II disables the 60 Hz High-Resolution EVF stream while shooting. How could the camera perform 120 autofocus calculations per second if the input image stream were only 60 Hz? From an engineering standpoint, it wouldn’t make sense to process both a 120 Hz data stream for AF and a separate 60 Hz high-resolution EVF stream simultaneously.
p.3 #11 · Speculation time: Sony a7V body design and...
Steve Spencer wrote:
For the record, we don't know what hallucinations depend on or what produces them. It may be that skills around prompting may reduce them, but I don't know that anybody has shown that. It could also be that people thinking they are skillfully prompting AI produces more hallucinations.
My basic point is that although AI is very useful as we integrate it more into our daily lives, I think a healthy does of skepticism about what it produces is also a good thing. Further if people do prompt it in a given way to produce a given result and edit what data they present from what it produces then I think it is helpful to describe both the prompting process and the editing process so others can come to the same conclusion about whether they believe what was produced was reasonable or not. ...Show more →
I see your points, and they are valid. I'm certainly not saying everything coming out of AI is 100% correct, good, or even bad. Critical thinking should be applied to all aspects of life.
Then again, if we question every answer, thought, idea, and post on forums like FM and ask how they were derived, what the underlying data is, how you chose to structure the presentation of the information, and so forth, we would be here for a while...
On AI hallucinations: they may occur for various reasons like insufficient training data, errors, or biases in the code, etc, etc, but one of the reasons is due to a lack of constraints that limit the possible outcomes, and that is where prompt engineering plays a role. Certainly not the only variable, but not one to ignore either.
I'd love to continue that conversation, but this is probably for a dedicated thread :-)
p.3 #12 · Speculation time: Sony a7V body design and...
aCuria wrote:
The AI is missing the point I think.
I have written code to grab data from cameras. If you did this you will find that some sensors / cameras are able to output multiple data streams simultaneously.
Nikon actually markets this as "dual-stream technology", but its not really that novel. Even some security cameras can do it these days.
The A9iii sensor is capable of the following simultaneously
1. 120fps 24MP 14 bit raw image stream
2. 120hz High Resolution EVF image Stream
The A1ii sensor is capable of the following simultaneously
1. 30fps 50MP 12 bit raw image stream
2. 120hz (i think) Standard Resolution EVF image Stream
The A1ii sensor can output this independently
1. 60hz High Resolution EVF image Stream
If you think about it, it makes sense that the A1 II disables the 60 Hz High-Resolution EVF stream while shooting. How could the camera perform 120 autofocus calculations per second if the input image stream were only 60 Hz? From an engineering standpoint, it wouldn’t make sense to process both a 120 Hz data stream for AF and a separate 60 Hz high-resolution EVF stream simultaneously. ...Show more →
If your are a programmer with hands on experience in writing code in this area, you abviously have far more experience than me on the topic, but if I understood the AI answer correctly, the lower res EVF during AF-C in the a1 II had to do with resource contrains, and to leave enough processing bandwidth to read out the 50mp (stacked BSI CMOS) sensor, it saves on processing power by outputting a lower res EVF image. Is that not correct? It is based on trade-offs that had to be made to achieve 30 fps on a 50mp sensor?!
I could be wrong, just trying to learn and understand...
And yes, Nikon's parallel processing paths is an elegant solution to this problem :-)
p.3 #13 · Speculation time: Sony a7V body design and...
I find all the speculation and speculation quite fascinating, but it will likely remain speculation. My actual question, however, was which Sony cameras besides the A9ii do NOT have this problem. The A7iii, A7iv, A7Rv, A1, and A1ii definitely have it. The question is whether other models don't have it, or whether their owners just haven't noticed it yet. Leica Q3 43 users are having a much worse experience. They report that the EVF becomes extremely grainy when focusing manually, making focusing disastrous.
By the way, the problem can be clearly seen in this video.
?si=uBSQBjGVItcXfKqt
p.3 #14 · Speculation time: Sony a7V body design and...
patotts wrote:
If your are a programmer with hands on experience in writing code in this area, you abviously have far more experience than me on the topic, but if I understood the AI answer correctly, the lower res EVF during AF-C in the a1 II had to do with resource contrains, and to leave enough processing bandwidth to read out the 50mp (stacked BSI CMOS) sensor, it saves on processing power by outputting a lower res EVF image. Is that not correct? It is based on trade-offs that had to be made to achieve 30 fps on a 50mp sensor?!
I could be wrong, just trying to learn and understand...
And yes, Nikon's parallel processing paths is an elegant solution to this problem :-)...Show more →
Nikon uses Sony sensors, and both brands’ cameras appear to feature what’s referred to as “dual-stream technology.” In contrast, Canon’s R5 II and R3 likely lack this capability, as shown by their behavior. They don’t write to the card until the high speed burst is over.
The AI explanation seems incorrect.
The A1/A1 II sensor simply cannot output a low-latency, high-resolution 120 Hz stream, even when the camera isn’t processing anything else. Its “High Resolution” output is limited to 60 Hz.
Because the autofocus system requires an ultra-low-latency stream of at least 120 Hz, the A1 II switches to “Standard Resolution” mode whenever autofocus is active.
p.3 #15 · Speculation time: Sony a7V body design and...
Nifty Fifty wrote:
I find all the speculation and speculation quite fascinating, but it will likely remain speculation. My actual question, however, was which Sony cameras besides the A9ii do NOT have this problem. The A7iii, A7iv, A7Rv, A1, and A1ii definitely have it. The question is whether other models don't have it, or whether their owners just haven't noticed it yet. Leica Q3 43 users are having a much worse experience. They report that the EVF becomes extremely grainy when focusing manually, making focusing disastrous.
By the way, the problem can be clearly seen in this video.
?si=uBSQBjGVItcXfKqt
I don't consider the A1ii's evf situation a "problem".
You can run it on standard resolution (5.7M dots, 0.9x magnification) and you will not see a resolution drop.
The Z9 uses a 3.68M dot, 0.8x magnification EVF
The R5II uses a 5.7M 0.76x magnification EVF
In comparison the A1 / A1ii still has the better EVF even on standard resolution.
p.3 #16 · Speculation time: Sony a7V body design and...
Well, if a viewfinder resolution is advertised to me, I assume it will be maintained even if I prefer AF-C. Incidentally, Sony never states that the standard resolution is 5.7MP; everywhere the viewfinder is listed as 9.5MP, which is why this should be considered the standard. The same applies to the viewfinder of the A7iv, for example. Sony only lists 3.7MP as the standard there, and this standard is not maintained when the shutter button is pressed in AF-C. This is a fact, and no comparisons with other brands change this.
p.3 #17 · Speculation time: Sony a7V body design and...
Nifty Fifty wrote:
Well, if a viewfinder resolution is advertised to me, I assume it will be maintained even if I prefer AF-C. Incidentally, Sony never states that the standard resolution is 5.7MP; everywhere the viewfinder is listed as 9.5MP, which is why this should be considered the standard. The same applies to the viewfinder of the A7iv, for example. Sony only lists 3.7MP as the standard there, and this standard is not maintained when the shutter button is pressed in AF-C. This is a fact, and no comparisons with other brands change this.
I would add to that, for me at least A1's EVF seems too large (0.9x mag) for the standard res (5.7MP). What I mean is that it just looks pretty average with those settings, kinda like when you crop/magnify the image too much...
p.3 #19 · Speculation time: Sony a7V body design and...
BigBabyMoses06 wrote:
Per Sony alpha rumors a few hours ago, looking like a partially stacked sensor.
hmm…. I’ve held on to my a9II partially so I could use the silent shutter under household LED lighting without the banding. (I also like the quieter shutter than that on my a7rIII.) I wonder if this new one might be sufficient to replace my 9II? I never really use the higher FPS.
p.3 #20 · Speculation time: Sony a7V body design and...
BigBabyMoses06 wrote:
Per Sony alpha rumors a few hours ago, looking like a partially stacked sensor.
If they want to get back on top for a best-in-class camera, it needs a partially stack sensor (hey, they sold a 24mp version to Nikon for the Z6 III & ZR...), a7RV flippy screen, hopefully slightly updated, and new a1 II + a9 III body design. Then they can take my money (and oh yes, time for 4k/60p uncropped, of course).
