Douglas L wrote:
I just sold my like new A1. The used price for A1 is a steal, IMHO.
In Mark Galer's review, he said the A1 and A1II's sensor read speed is 4X faster than that of the A7V.
Yeah...pretty much. A1/A1II are ~3.8ms and A7V according to DPR is ~15.1ms.
You can use 15.1ms for most birds and not get bad wing distortions. But you will get leaning lines in the background when panning a BIF. The R5 readout is ~16ms and I used it in ES all the time so this A7V will be similar. But it ain't no A1/Z9/R1 type readout where you really don't every have to worry about rolling shutter.
Cliff L. wrote:
Unfortunately, Sony Canada still receives much of their inventory from Sony USA, so the cost of US tariffs is included. They aren't smart enough to use a bonded warehouse to allow the goods to transit to Canada tariff-free, despite most of their retailers asking them to do this...
However, it will be likely be much cheaper in a few months once the initial orders are filled, like most new Sony products.
Still we aren't getting hosed as bad as it could be as the exchange between the CAD and US price for the A7V is ~1.27x. Not the 1.35-1.4x we got before tariffs.
arbitrage wrote:
Yeah...pretty much. A1/A1II are ~3.8ms and A7V according to DPR is ~15.1ms.
You can use 15.1ms for most birds and not get bad wing distortions. But you will get leaning lines in the background when panning a BIF. The R5 readout is ~16ms and I used it in ES all the time so this A7V will be similar. But it ain't no A1/Z9/R1 type readout where you really don't every have to worry about rolling shutter.
I'm not sure what the readout speed of the stacked sensor in the R5 II is, but looking back at my hummingbird images from this past summer, I noticed a few instances of wing distortion, so even a fully stacked sensor is no guarantee of avoid rolling shutter effects. No such issues with the A1, though.
In a video I just watched, it was claimed that the A7V will offer a dynamic range of 16 stops. If this turns out to be true, it would be the first significant improvement in dynamic range since the A7III/A7RIII. A full stop more compared to previous Alpha models – that's a substantial gain. That would put it in the range of common medium format cameras, which also advertise around 16 stops. Add to that the IBIS, which is reportedly much improved... Hmm... if it's all accurate, I might actually pick one up.
Jazzgear296 wrote:
It’s still not true AI in my opinion. In Data Science, this would be equivalent to nothing more than an Edit Check — aka automated execution of some business rules. By definition, true Ai would result in some cameras being at different AI capability stages based on the number of use cases executed with the camera and what it learned from those cases. At the very minimum, there would have to be a data feedback loop from your camera usage patterns back to Sony so that they can retrain the models and update it on some regular basis.
It is AI. Definition being: "The ability of a computer or other machine to perform those activities that are normally thought to require intelligence." => there is no mention of self learning here. You can produce AI without self learning. And Chess with the Alpha-Beta algorithm, before the engines started to use NN/NNUE is one such example, those engines were not learning anything, they were pre-trained, and yet, they're still AI.
Edit: Actually, even those very new Chess engines using NN/NNUE are still not learning anything. The training is done somewhere else, and the engines themselves are just using the trained data, and applying it, just as those Sony cameras are doing with subject recognition.
You're thinking "self learning" and that is a subset of what "AI" is.
hasenbein wrote:
And in the press release the usual bullshit of "AI white balance" and "deep learning".
This camera, like all others, doesn't contain a trace of AI. AI would mean that the camera learns through your shooting experiences, so that each individual copy over time develops individual reactions to shooting situations. Of course this isn't the case, each copy of a certain camera model reacts in the same way, and that reaction doesn't change over time.
Jazzgear296 wrote:
It’s still not true AI in my opinion. In Data Science, this would be equivalent to nothing more than an Edit Check — aka automated execution of some business rules. By definition, true Ai would result in some cameras being at different AI capability stages based on the number of use cases executed with the camera and what it learned from those cases. At the very minimum, there would have to be a data feedback loop from your camera usage patterns back to Sony so that they can retrain the models and update it on some regular basis.
Absolute nonsense.
Lots of people in this thread having very strong opinions on what AI (an ill-defined concept) is or isn't.
The camera probably has some deep learning stuff like classifiers, which are as much AI as anything else.
This idea that you people have that the camera needs to be learning from the customer using it for it to count as AI is entirely hallucinatory.
I know some might be disappointed by the video features, but as someone who primarily just shoots wildlife, this camera is a dream come true. I have an A7RV and didn't want to pay the 8000 CAD for the A1ii. Getting the A7V is a no brainer, pre-capture, no blackout, high DR electronic shutter, 16 stops of DR checks off every wish I had. It even qualified for Sony's student discount when I pre-ordered!
Now comes the question of do I sell my A7RV to buy a GFX100RF as a travel camera...
patotts wrote:
If it had the same body as the a1 II with the updated grip and angled shutter button, I would buy one of these even at MSRP.
Unfortunately it still uses the antiquated body of the A7IV, no raised shutter button
Was considering a backup to A1 but at this time, a second hand A1 makes more sense to me, a strictly bird photographer
Pre-capture is nice to have and maybe new AI chip but the rest still lags behind the A1 IMHO
Dave Sanders wrote:
Yes - I think AI in cameras is really 'pre-AI' as in they used AI to develop the subject recognition and AF algorithms in a lab, but the camera itself doesn't learn and adapt. So it's not AI in the way in which we use the word now with things like LLMs.
Hmm, large language models can't learn and become more intelligent on the fly - they are developed through a training process in the same way the ML algorithms in cameras are trained. Whoever developed the LLM needs to train it with more data for it to become more 'intelligent'.
So in that respect they are identical to the subject detection models used in the cameras.
You folks seem to have your own definition of AI being that it is capable of 'self-learning'. By definition that wouldn't be be ARTIFICIAL intelligence - that would just be INTELLIGENCE.
Artificial intelligence is by definition ARTIFICIAL - i.e. the model is developed through a training process and then deployed for use in a particular situation but the model that is deployed does no have the ability to train itself as it gets used.
deevee wrote:
Unfortunately it still uses the antiquated body of the A7IV, no raised shutter button
Was considering a backup to A1 but at this time, a second hand A1 makes more sense to me, a strictly bird photographer
Pre-capture is nice to have and maybe new AI chip but the rest still lags behind the A1 IMHO
"antiquated body of the A7IV"? What exactly is the problem with the A7IV body?
I would love to know examples of the real world read-out speed difference between the a1 and a7V - for example, the ability to use electronic shutter for a golf swing vs. bird wing vs. soccer game vs. active child. I love my a7RV, but its slow readout prevents electronic shutter on all but the active child example. If the A7V could be solid for a soccer game, usually ok for bird wings, but not a golf swing - count me in.
Cliff L. wrote:
I'm not sure what the readout speed of the stacked sensor in the R5 II is, but looking back at my hummingbird images from this past summer, I noticed a few instances of wing distortion, so even a fully stacked sensor is no guarantee of avoid rolling shutter effects. No such issues with the A1, though.
I will add this to my A1 because - MUCH better IBIS, great rear screen, fixed rolling shutter of the A7r5 (which I tried but rolling shutter in video was a disaster), nicer grip, improved AF, etc.
Overall very good value for $3k.
gocolts wrote:
I see used A1's going for a little over $3k on the B&S boards. I realize there's some new features on the A7V, but for me if I didn't already have an A1 & had $3k I know where my money would go....
If shooting sports I would hands down take a used a1 at $3k over a new a7V. Even without the new AF features, etc.
Douglas L wrote:
I just sold my like new A1. The used price for A1 is a steal, IMHO.
In Mark Galer's review, he said the A1 and A1II's sensor read speed is 4X faster than that of the A7V.
arbitrage wrote:
Yeah...pretty much. A1/A1II are ~3.8ms and A7V according to DPR is ~15.1ms.
You can use 15.1ms for most birds and not get bad wing distortions. But you will get leaning lines in the background when panning a BIF. The R5 readout is ~16ms and I used it in ES all the time so this A7V will be similar. But it ain't no A1/Z9/R1 type readout where you really don't every have to worry about rolling shutter.
Cliff L. wrote:
I'm not sure what the readout speed of the stacked sensor in the R5 II is, but looking back at my hummingbird images from this past summer, I noticed a few instances of wing distortion, so even a fully stacked sensor is no guarantee of avoid rolling shutter effects. No such issues with the A1, though.
R5II is around 1/160, or about 6.25ms. If you pan it really hard, there will be rolling shutter distortion, but from my experience, with most 'normal' sports action, I haven't seen obvious signs of it.
jchapell wrote:
I would love to know examples of the real world read-out speed difference between the a1 and a7V - for example, the ability to use electronic shutter for a golf swing vs. bird wing vs. soccer game vs. active child. I love my a7RV, but its slow readout prevents electronic shutter on all but the active child example. If the A7V could be solid for a soccer game, usually ok for bird wings, but not a golf swing - count me in.
Well, I can tell you that ~15ms readout is OK for American football most of the time. That is roughly the range of the Canon R5 and R6II, which I've used for a couple years to photograph this sport and the only time I have noticed obvious rolling shutter artifacts has been when the ball was kicked at the very moment of contact with the kicker's foot and high speed panning where vertical objects in the background appear to lean at somewhat of an angle. Sometimes if it's a really hard pass, the football might appear a bit more elongated than normal. There are a lot of variables at play but most of the time ~15ms readout speed has been usable. But for round ball and stick sports like soccer, baseball, golf, basketball, volleyball, etc., ~15ms won't be fast enough to minimize rolling shutter distortion when the ball or stick is moving really fast. For slower paced action and lower ability level players, it will probably be fine for sports like soccer, most of the time. Maybe Adobe or someone else will release an AI 'ball de-warping' retouching brush in a future release.
Seems obvious that for those shooting golf swings and bird wings the A1 series is better.
But for the rest…this camera is a great deal. It has most of what the A1 II has for less than half the price. (And a bunch of stuff I wish my A1 had)
And seems to be better than the A7r5 in every way except pixel count.
gocolts wrote:
I see used A1's going for a little over $3k on the B&S boards. I realize there's some new features on the A7V, but for me if I didn't already have an A1 & had $3k I know where my money would go....
This is where I'm at. I already have an a7r V and love it for everything but wildlife (It's ok for that, but a little limited.) I'm not interested in losing resolution for wildlife, so would still rather use what I have. If I do decide to upgrade for wildlife purposes, the a1 makes way more sense to me.
The a7V is a great body for someone who needs a well rounded body and doesn't need a high resolution. If I was shooting general use with some kids soccer games, I'd prefer the a7V to the a7RV for example.
Dual Base ISO @ISO800 and ISO8000 in video (SLOG3) is pretty impressive.
I'm anxiously waiting for my a1 II, I paid for it 3 weeks ago, and it's supposed to be coming in soon... but now it's got me thinking...
