Sony DSC-R1 - Focus speed and response a weakness?

No, it's not. The shutter can't fire until the mirror is raised. If
you don't have a mirror, you don't have to raise it.
The mirror raising and shutter opening is done almost at the same
time, producing a single noise.
They are separate events (hence mirror lockup), so in that pedantic case, Joe is correct, but you are correct that the I-R "lag" measurements include both times.
"Its shutter lag is an astonishing 0.007 second (7 milliseconds) in
that mode"
Heh you forgot one very important fact about that number. It's
PRE-FOCUSED. So not representable of a real world situation.
The one thing you can say about the real world, is that YMMV. I know I try to prefocus as much as I can. Of course any number this low becomes somewhat moot as human response time begins to dominate (e.g. you still have to anticipate or "lead" the action), but it's onviously important as Nikon/Canon spent a lot of money trying to minimize this time for their pro DSLRs.
"We found some of its ergonomics decidedly awkward, and both timing
performance and high-ISO noise levels aren't up to the best of the
digital SLRs currently on the market."

Notice the word "timing".
Actually, in this context of "timing" refers to all of the timing measurements and not just the "lag" times you are trying to emphasize.

--
Erik
 
"Its shutter lag is an astonishing 0.007 second (7 milliseconds) in
that mode"
I wonder what the "blackout" time is? In this case it would be the time to readout and reset the sensor. 10 million pixels but what clock rate and how many channels? I remember the surreal "stop motion" effect my old D7i had it it's 7fps UHS mode. It was almost impossible to follow a moving object because of it.

--
Erik
 
The mirror raising and shutter opening is done almost at the same
time, producing a single noise. And no matter how much you try to
argue semantics, it doesn't change the fact that the R1 is slower
than DSLRs. The numbers speak for themselves.
Yes, they do. 7mS shutter lag if prefocused.
So in one single situation that's not very likely to occur, it's fast. In all other cases it's slower.
From the same
imaging resource review that you like to quote:

"Its shutter lag is an astonishing 0.007 second (7 milliseconds) in
that mode"
Heh you forgot one very important fact about that number. It's
PRE-FOCUSED. So not representable of a real world situation.
It's the only difference between the two examples. As you yourself
pointed out, its meter, focus, and fire.
What do you mean? The time for the whole thing is still around 0.48-0.77 seconds.
Raising the mirror takes longer than the update
cycle of an EVF. You're having a lot of trouble with a very, very
simple concept.
You are just in complete denial, aren't you?
Sorry, that's my line. Read the script.
You haven't answered regarding the processing lag yet.
No, the total
delay of using an EVF or LCD is not the refresh rate of the
display. It's the refresh rate PLUS the delay created by the
electronic interpretation of the light on the sensor. It's you who
are having a lot of trouble with this concept.
Poor, poor dinosaur. Did you notice it's getting cold outside?
Hmm here you can't claim to just answer in kind. You answer a real argument with an insult (that, to be fair, I found amusing). Are you going to keep doing that or maybe come up with a counter argument? Again, refresh rate is not equal to the total delay when using an EVF or LCD. It is just the rate at which the image on the EVF or LCD is updated and does not reflect the time it takes for an image to be processed from light on the sensor, through the PCB and finally shown on the LCD/EVF. Are you claiming this processing is instantaneous?
Heh, yes apparently you are. Did you even read Jared Hunter's post?
You are confusing refresh rate with lag. The two have to be added
together. First you have the lag from when the light hits the
sensor, is registered in the electronics, the information altered
and manipulated and finally ready to display on the EVF. On top of
this you have the update/refresh rate of the display.
Did you ever read anything, at all, ever?
Again, I'm wowed by your powers of argumentation. Try reading my
answer again, and maybe you'll see the light.
Again, it's your rude argument, and you've again stepped on my line
for the reply.
My rude argument?

I wrote this:
Did you even read Jared Hunter's post?
You are confusing refresh rate with lag. The two have to be added
together. First you have the lag from when the light hits the
sensor, is registered in the electronics, the information altered
and manipulated and finally ready to display on the EVF. On top of
this you have the update/refresh rate of the display.

Which you answer with this:

"Did you ever read anything, at all, ever?"

I think that speaks for itself.
I'll provide you with another quote from the conclusion:

"We found some of its ergonomics decidedly awkward, and both timing
performance and high-ISO noise levels aren't up to the best of the
digital SLRs currently on the market."

Notice the word "timing".
Note the word "best". I have one of them, a $5000 Nikon D2X.
Good for you, great camera. However I don't see how it should have any relevance as to whether you're right or wrong.
 
I'll provide you with another quote from the conclusion:

"We found some of its ergonomics decidedly awkward, and both timing
performance and high-ISO noise levels aren't up to the best of the
digital SLRs currently on the market."

Notice the word "timing".
Note the word "best". I have one of them, a $5000 Nikon D2X.
Forgot to say, the word best doesn't necessarily mean the high end DSLRs in this case, as both the timing and noise performance is better on the 350D.
 
Erik Magnuson wrote:
snip
Well if you read the whol paragraph it says "At 1.27 seconds (best
case) between frames, single-shot cycle times are reasonably fast
compared to other all-in-one digital cameras but once again slow
relative to d-SLRs." so the 0.33 - 0.75 numbers are obviously not
shot to shot time. I think that it's you who need to read the
paragraph (and the table above it) more carefully for context.

--
Erik
The point is that It's irrelevant to my comments which have nothing to do with shot to shot, but have to do with the EVF's affecting or not affecting the ability to get a quick autofocus and capture.

Lin
 
The reason for the slow response is not the lens on the R1, it's the contrast AF, which is slower and less accurate than phase change AF.
where do you get this info from, i didn't see any mention of af system type in phils preview, maybe i missed it.
--
Paul A Roberts
 
The time from half pressed to picture in the best prosumers are under 0.1 seconds (.007 seconds if you can believe that.) dSLR's are around 0.1-0.2 seconds or so. Although dSLRs are certainly faster on the average, prosumers are catching up. Sony has been a leader in focusing speed, and there is no reason to believe they didn't pay a lot of attention to this issue. The 828 is already about the same speed as the 350! The R1 just might beat a low end dSLR in focusing and delay times. Most of today's prosumer cameras beat the 350 in prefocused response time, based on a random sample below. I believe this to be due to electronics beating the rise/fall time of the mirror. The best prosumers are coming closer to the focusing time of a dSLR. Yes, a high end dSLR is probably a lot faster overall, but that's an apple to oranges comparison. Imagine what this comparision will be like 5 years from now!

Using the 350 as a baseline:

Shutter lag, full autofocus 0.24 with kit lens
Shutter lag, manual focus 0.168
Shutter lag, prefocus 0.095

The Canon G1
Shutter lag, full autofocus 0.75
Shutter lag, manual focus 0.48
Shutter lag, prefocus 0.12

The FZ20
Shutter lag, full autofocus 0.53 at wide angle
Shutter lag, manual focus 0.080
Shutter lag, prefocus 0.038

The FinePix F10
Shutter lag, full autofocus 0.29
Shutter lag, prefocus 0.014

The Sony 828 (now two years old!)
Shutter lag, full autofocus 0.25 at wide angle
Shutter lag, continuous autofocus 0.42
Shutter lag, manual focus 0.20
Shutter lag, prefocus 0.009
 
but have to do with the EVF's affecting or
not affecting the ability to get a quick autofocus and capture.
Actually, EVFs may have an effect - a lot depends on the implementation. For example, many Sony CCDs have a special high-speed scanning mode that just scans the center area (but at up to 120 FPS!) to get the maximum possible data for AF. However, while the sensor is in this mode, it obviously cannot be also feeding the EVF a full frame at the same time. This sometimes shows up as a slight jump or freeze in the EVF when the camera is focusing. It's possible to focus using the normal 30 FPS mode that feeds the EVF, but with less speed and accuracy.

Then there is the issue of keeping the EVF frame rate at 30 FPS in low light. Obviously, this means the minimum CCD exposure during framing is at least 1/30 of a second. This means you either have to increase the gain (noise), decrease the frame rate or both. (Or play the trick of switching the image to B&W to hide the color noise.) Obviously decreasing the frame rate increases the latency.

And then there is the issue of blackout time. Again, if the sensor is only capable of 4 FPS full-frame, that implies that it takes at least 250ms to read and reset the sensor before it can begin to feed the EVF again.

So just as there are a few advantages (7ms minimum latency), there are some speed disadvantages as well w.r.t focus and capture.

--
Erik
 
where do you get this info from, i didn't see any mention of af
system type in phils preview, maybe i missed it.
It's an assumption based on the 2.4mm lens-to-sensor clearance. There is obviously not enough room between them to put the optical parts needed for phase detection.

--
Erik
 
The time from half pressed to picture in the best prosumers are
under 0.1 seconds (.007 seconds if you can believe that.) dSLR's
are around 0.1-0.2 seconds or so.
Yep. The advantage of no mirror lag.
there is no reason to believe they
didn't pay a lot of attention to this issue.
I think the point is that the shallower DOF with the larger sensor makes the problem harder. Note how the AF times for non-DLSRs get much slower as you increase the zoom.
The R1 just might beat a low end
dSLR in focusing and delay times.
The preproduction R1 tested by I-R appears to be noticably slower than both the 350 and the 828. It could be just preproduction issues, but they might have a tough time equaling the 828.
Shutter lag, full autofocus 0.24 with kit lens
0.20 at 55mm. Odd how it got a little faster at the longer setting.
The Sony 828 (now two years old!)
Shutter lag, full autofocus 0.25 at wide angle
and 0.69 at full telephoto. The R1 time for full telephoto is 0.74s. Interesting enough, both of these are about 71mm actual focal length at full telephoto.
Shutter lag, manual focus 0.20
Shutter lag, prefocus 0.009
Now why should these be any different? The only thing that comes to mind is the time to meter the exposure although I would be surprised if that takes 0.19 seconds.

--
Erik
 
This is an utter prejudice against anything without a an optical VF. If auto focus works well an EVF is just good. Even the sharpest eye in the best of the best of the optical VF can not be sure of the perfect focus in a camera with 10 MPx sensor. The resolution it delivers is too high for the naked eye.

--
Andre
 
This very inexpensive lens costs almost 1/3 of the R1.
--
Andre
If you're talking about the 50mm f1.8, it costs $100. The R1 will cost $1000. Is this 1/3?

PS: Even if it did cost 1/3 of the R1, so what? It can do something that's very useful, but quite impossible with the R1. I'd say that's worth a price premium, imaginary or not...
 
Yes, once they can build multi MP EVF and remove the lag it'll be the future. However that doesn't change what you can get today...
 

Keyboard shortcuts

Back
Top