Sony A55: 16 MP, HD Video and 10 FPS w/AF, Built in IS, GPS, Swivel Screen

Kiwikat wrote in post #10754731
I don't see anything innovative in that list...

It sounds like a good price though!

My comment was not specifically for this camera, but rather the suite of cameras and features (NEX-3/5, NEX-VG10m, A55, Backlit CCD's) that Sony has announced over the last few months. While its debatable how good the cameras are, its hard to deny they are not trying.

I'm a Canon fan, I cannot think of the last "pushing the bounds" thing they did since video in the 5D mark II.

cogliostro wrote in post #10762228
Looking the sample images from a900 and nex 5, i'd say it is plausible though. difficult, but still can be done imo.

The A900 is as good as the 5Dmk2 at iso 400! No it cannot be done. Maybe they could get it as good at the same ISO, but not a 4 stop advantage.

I think that a camera like this would really have to make Canon and Nikon step up their game and stop playing these foolish games with new releases.

The sony has a lot of nice features listed but really interests me is the 10fps with a price point of supposedly less than $1,000. If I wanted to shoot 10 FPS with Canon I am going to spend $2,000 at the least for an older camera, or $5000ish for a 1DIV

Then again I would never buy a Sony camera, or at least not anytime in the near future, but I hope it will make Canon/Nikon get more creative with what they are offering. Each new release from Canon and Nikon is the same camera as before but with one new feature or a button put in a new spot.

RobDickinson wrote in post #10760357
This just sounds absurd.

A crop sensor with 4 stops better sensativity than a sensor 2.5 times larger than it. I'm not sure theres even enough efficiency at 100% to go that far

Not to mention most of the time light will be siphoned off by the pellicle mirror.

Then again, what if the base sensitivity of the sensor itself were higher? Say, base sensitivity is ISO 1600 (as clean as another sensor designed for base ISO 200), and the mirror takes 3 stops of light (all 3 stops of which would go to that impressive-sounding AF system...). When the mirror is down, you get the equivalent of ISO 200 as a base sensitivity, and gain is then applied to get you up to 1600. Above 1600, you have the option to either keep applying gain (and increasing noise) or start flipping the mirror: your ISO 1600 is now just as clean as ISO 200 was, 3200 like 400, 6400 like 800, etc. Given how clean the latest Sony sensors are (K-x, NEX) this could result in a significant performance boost. It's an idea that floated around in my head a lot back when I used 4/3, since they were always a bit behind in the high-ISO game.

The real question as to whether that's feasible or not is whether it's actually possible to make sensors with higher base sensitivities like that. I like to think sensors to this point have been arbitrarily designed to fit a similar low-sensitivity range as we had back in the film era just prior to digital. Having low ISOs available can be very important if you wish to work in normal light with wide apertures or longer exposures. So I don't really know if it can be done or not.

Er1kksen wrote in post #10763030
Not to mention most of the time light will be siphoned off by the pellicle mirror.

Then again, what if the base sensitivity of the sensor itself were higher? Say, base sensitivity is ISO 1600 (as clean as another sensor designed for base ISO 200), and the mirror takes 3 stops of light (all 3 stops of which would go to that impressive-sounding AF system...). When the mirror is down, you get the equivalent of ISO 200 as a base sensitivity, and gain is then applied to get you up to 1600. Above 1600, you have the option to either keep applying gain (and increasing noise) or start flipping the mirror: your ISO 1600 is now just as clean as ISO 200 was, 3200 like 400, 6400 like 800, etc. Given how clean the latest Sony sensors are (K-x, NEX) this could result in a significant performance boost. It's an idea that floated around in my head a lot back when I used 4/3, since they were always a bit behind in the high-ISO game.

The real question as to whether that's feasible or not is whether it's actually possible to make sensors with higher base sensitivities like that. I like to think sensors to this point have been arbitrarily designed to fit a similar low-sensitivity range as we had back in the film era just prior to digital. Having low ISOs available can be very important if you wish to work in normal light with wide apertures or longer exposures. So I don't really know if it can be done or not.

Interesting idea, if it was possible, I think that a lot of professionals would love to have a body such as this. Such as a 5D MK2 HS or 1D MKIV HS(high sensitivity) or something like that. A body they could use specifically for low light. Such as sports, weddings, performing arts, etc. There are a lot of applications where a professional will go to shoot and never go lower than 800 or 1600 ISO. They could have the normal body then the HS body with them, use the normal body when applicable, then the HS when applicable. I think its feasible and if something was made that truly had a clean high ISO I think it would be a big seller.

JoeyBowman wrote in post #10763101
Interesting idea, if it was possible, I think that a lot of professionals would love to have a body such as this. Such as a 5D MK2 HS or 1D MKIV HS(high sensitivity) or something like that. A body they could use specifically for low light. Such as sports, weddings, performing arts, etc. There are a lot of applications where a professional will go to shoot and never go lower than 800 or 1600 ISO. They could have the normal body then the HS body with them, use the normal body when applicable, then the HS when applicable. I think its feasible and if something was made that truly had a clean high ISO I think it would be a big seller.

i believe all we would go for the Hs bodies

tkbslc wrote in post #10762550
The A900 is as good as the 5Dmk2 at iso 400! No it cannot be done. Maybe they could get it as good at the same ISO, but not a 4 stop advantage.

Oops i thought it was the new camera's iso 1600 vs a900 iso 400.
Yeah the crop camera's 6400 that is as clean as the a900 400 is pretty dumb. but sony is always winning the on paper spec tho

Anyone think Canon will cut prices in response?
After all this camera sounds pretty damn good...

rhys216 wrote in post #10767714
Anyone think Canon will cut prices in response?
After all this camera sounds pretty damn good...

The only maker that can move Canon is Nikon and vice versa.

Sony is just too small to have a measurable effect on the plans of Canon I think. While I don't have sales figures, based on my observations of several dSLR heavy locations, I usually see 5 to 10% Sony, of which they are all medium to low end (e.g. Wal-Mart fare) models. Of these, I have never seen one with a lens costing more than a few hundred dollars.

More details released. It's listed as a "SR5" on a Sony rumors site, which means almost certain to be present in the production model. The ISO 6400 noise claim is probably just marketing talk, but the other stuff seems not too far fetched. Sony already has a sub-$800 SLR that can shoot 7 fps (albeit without AF), so 10 fps w/AF in a pellicle mirror design seems very feasible.

Wouldn't the point of a pellicle mirror be to show a TTL view through the viewfinder? Why would they go through the trouble just for the AF? As mentioned earlier, there will be a delay in the display to the EVF, which sort of defeats the point of a high speed camera; you can't shoot the catch that you can't see. And - something I just thought of - you'd still have blackout while the shutter is moving across the sensor to make the exposure, hence defeating the pellicle mirror again.

bjyoder wrote in post #10769637
Wouldn't the point of a pellicle mirror be to show a TTL view through the viewfinder? Why would they go through the trouble just for the AF?

The point of a pellicle mirror is shorter shutter lag, no VF blackout, potentially faster framerates (no longer an issue), and a continuous "feed" to the AF sensor.

In a "normal" mirror system, there is a blackout for both the AF sensor and the VF when the mirror flips up. By leaving the mirror fixed, the AF sensor has continuous information and requires less guesswork in predictive AF (AI Servo).

Since the mirror doesn't have to flip up, this also decreases shutter lag, and the frame rate is theoretically limited by how fast the shutter curtains can be moved, not the mirror. Point is moot, though, since Sony's pellicle design can't do more than 10 fps anyway, which is more than possible with a moving mirror. But a moving-mirror system is limited to about 10fps (I think Nikon got to 12) because the AF sensor isn't given enough time to get new data if the framerate is much faster.

Also, I don't think much more than 10fps is possible even with a pellicle mirror...otherwise Sony would have done it, and Canon wouldn't have stopped with the 1N RS. But they did, because the 1v got 10fps perfectly fine with a normal mirror.

Sure, but 2ms (at 1/500) is better than 80...

The problem is light loss, EVF lag, and the same frame rate as a "normal" mirror.

edit: mirror slap is another problem with "normal" mirrors...but that's not an issue when shooting action.

Sony seems to have stalled in the SLR world, especialy in the pro market.

And nikon shifting their sensors from sony ones will hit them hard.

toxic wrote in post #10770228
The point of a pellicle mirror is shorter shutter lag, no VF blackout, potentially faster framerates (no longer an issue), and a continuous "feed" to the AF sensor.

In a "normal" mirror system, there is a blackout for both the AF sensor and the VF when the mirror flips up. By leaving the mirror fixed, the AF sensor has continuous information and requires less guesswork in predictive AF (AI Servo).

Since the mirror doesn't have to flip up, this also decreases shutter lag, and the frame rate is theoretically limited by how fast the shutter curtains can be moved, not the mirror. Point is moot, though, since Sony's pellicle design can't do more than 10 fps anyway, which is more than possible with a moving mirror. But a moving-mirror system is limited to about 10fps (I think Nikon got to 12) because the AF sensor isn't given enough time to get new data if the framerate is much faster.

Also, I don't think much more than 10fps is possible even with a pellicle mirror...otherwise Sony would have done it, and Canon wouldn't have stopped with the 1N RS. But they did, because the 1v got 10fps perfectly fine with a normal mirror.

Sure, but 2ms (at 1/500) is better than 80...

The problem is light loss, EVF lag, and the same frame rate as a "normal" mirror.

edit: mirror slap is another problem with "normal" mirrors...but that's not an issue when shooting action.

Right, I understand all that, but why bother with a pellicle mirror if they're just going to put in an EVF? In my mind, they're doing all the work for a type of system, then "ruining" it by not taking full advantage of it and giving an optical light path for the VF.

It's all rumor at this point anyway, so no point in trying to figure it all out.