New DSLR on Tuesday, October 18, 2011. [CR3]

[Meh]

On multiple shots per mirror flip, sometimes I have wondered if it would be nice to have focus bracketing if you make the assumption the tracking focus will not be 100% anyway. Perhaps user settable equivalent to a few units of micro-focus adjust either side of nominal, could be done as the tracking is done to minimise the focus travel distance.

Good idea but I wouldn't suggest "either side of normal" because that would have the focus elements moving back and forth very rapidly and would cause a lot of stops and change of direction which is hard on the motors and gears and defeats the purpose of Ai Servo which is predictive as to the direction of movement.  The small adjustment you suggest for the "multiple frames between mirror flips" could just keep moving the focus elements in the same calculated direction until the next time the mirror flips down and the AF sensor can make a new focus measurement.

[bvukich]

On multiple shots per mirror flip, sometimes I have wondered if it would be nice to have focus bracketing if you make the assumption the tracking focus will not be 100% anyway. Perhaps user settable equivalent to a few units of micro-focus adjust either side of nominal, could be done as the tracking is done to minimise the focus travel distance.

Good idea but I wouldn't suggest "either side of normal" because that would have the focus elements moving back and forth very rapidly and would cause a lot of stops and change of direction which is hard on the motors and gears and defeats the purpose of Ai Servo which is predictive as to the direction of movement.  The small adjustment you suggest for the "multiple frames between mirror flips" could just keep moving the focus elements in the same calculated direction until the next time the mirror flips down and the AF sensor can make a new focus measurement.

That's how I would see it working.  Just keep moving the focus at the calculated rate until the next AF sample, and refine from there.  You're still getting AF samples at 10 per second, and on a fast moving subject AI Servo would have the AF motor moving continuously anyhow.  May as well get a few extra frames in there.

[neuroanatomist]

That's how I would see it working.  Just keep moving the focus at the calculated rate until the next AF sample, and refine from there.  You're still getting AF samples at 10 per second, and on a fast moving subject AI Servo would have the AF motor moving continuously anyhow.  May as well get a few extra frames in there.

Makes sense.  But presumably each exposure would need to end with a shutter curtain - will the shutter reset be sufficiently fast? 

[Ivar]

As for many here, 18MP is beyond me.

I don't believe that Canon is going to introduce another 3rd 1-series cameras, that would be costly to manage, rather I believe the opposite meaning merge (not in the current case though).

Now, one proposed explanation was more than 10fps for the new cam, meaning Canon is not that bad in data processing capabilities. However, 18MP is completely 180-degree turn by Canon, a slap in the face for former users. A new religion must be introduced by marketing.

Trying to rationalize the global picture, less megapixels could explain a new 30+ (only!) 1Ds, which would have been skipped by potential buyers if the 1D was having something like 27MP (upscaled 1D MK4), fullfilling majority of users needs. The key here is to keep enough difference in MP to sell both.

For all high ISO fans - it is not the MP you are after but rather the sensor size, 30MP wouldn't make it worse than 18MP in print using the same technology, rather the opposite, more sensels enable better IQ.

[EYEONE]

On multiple shots per mirror flip, sometimes I have wondered if it would be nice to have focus bracketing if you make the assumption the tracking focus will not be 100% anyway. Perhaps user settable equivalent to a few units of micro-focus adjust either side of nominal, could be done as the tracking is done to minimise the focus travel distance.

Good idea but I wouldn't suggest "either side of normal" because that would have the focus elements moving back and forth very rapidly and would cause a lot of stops and change of direction which is hard on the motors and gears and defeats the purpose of Ai Servo which is predictive as to the direction of movement.  The small adjustment you suggest for the "multiple frames between mirror flips" could just keep moving the focus elements in the same calculated direction until the next time the mirror flips down and the AF sensor can make a new focus measurement.

That's how I would see it working.  Just keep moving the focus at the calculated rate until the next AF sample, and refine from there.  You're still getting AF samples at 10 per second, and on a fast moving subject AI Servo would have the AF motor moving continuously anyhow.  May as well get a few extra frames in there.

Interesting idea. It seems like that would work to me but would the calculation require a bit of processing power by processors that are already being taxed by shooting at 20 or 30fps? Perhaps it is an easier calculation that I think. Something to consider though.

[Ricku]

I can't believe that no one has stepped forth yet and confirmed or denied these rumors..?

Comeon! Theres gotta be lots of people out there who knows what is coming. Lets hear it!

[lol]

Good idea but I wouldn't suggest "either side of normal" because that would have the focus elements moving back and forth very rapidly and would cause a lot of stops and change of direction which is hard on the motors and gears and defeats the purpose of Ai Servo which is predictive as to the direction of movement.  The small adjustment you suggest for the "multiple frames between mirror flips" could just keep moving the focus elements in the same calculated direction until the next time the mirror flips down and the AF sensor can make a new focus measurement.

I'm not sure it would be that big a hit in performance if the bracketing isn't too extreme. Remember I'm only thinking about hedging the bets on relatively small errors, say up to +/- 1 DoF worth or so, not suggesting you do a focus stack! I'm assuming there might be times where having multiple shots at different focus distances would give you a better chance of a good shot than a single conventional track, particularly if relatively shallow DoF is used.

So assuming a 3 shot tracking focus bracket, the 1st shot would under-correct for where the focus thinks the target would be. The 2nd shot would be where the target is expected, and the 3rd shot would go a bit further. Of course the position would need to be based on where the subject is predicted to be at the time of exposure. If the subject isn't changing distance fast, then it should be even easier to move around. If it is moving lens rate limited then you could have custom function options for bracket priority (spend the time to move to position before the shot) or release priority (take the shots even if desired focus point isn't reached) for example.

Stepping back a bit, focus stacking for single-shot AF would be nice too

[iaind]

I can't believe that no one has stepped forth yet and confirmed or denied these rumors..?

Comeon! Theres gotta be lots of people out there who knows what is coming. Lets hear it!

Most are covered by NDA's

Can't wait to get my hands on the 1DsIV 1DV 5DIII and 7DII


 

[Smith]

If Canon is launching a 1D with the rumored specs I suspect that this camera is focused on video. Dual DIGIC V seems like an overkill for a 16-18 megapixel sensor even with high frame rates. A lower resolution sensor would make sense if it has been designed for video applications.

[Ghostdive]

If Canon is launching a 1D with the rumored specs I suspect that this camera is focused on video. Dual DIGIC V seems like an overkill for a 16-18 megapixel sensor even with high frame rates. A lower resolution sensor would make sense if it has been designed for video applications.

I hope its not video. A 1D is maybe within my pricerange. And i'm more for still. If I want video I buy a cam to get the best out of it. Video on DSLR is a nice gimick, but not a must have.

A 1D with arround 21 mpix would be nice.

[ippikiokami]

So can you round up the reasons why you think it's a 1d body release again?

It's just weird to think they would release a new versions of their most updated pro body...

1ds III, 5d2, 7d, 1d IV ... in order of oldest to newest??

Also why do you consider that the best spec list?

I'm not doubting you but I would love some references to where you are getting the info. Even if you say something like one of my most reliable informants... etc

[unfocused]

I feel like I'm the only one who actually read Canon Rumors Guy's post. He gave a CR3 rating to the announcement of a DSLR of some sort and no rating to the spec list.

People are treating the spec list as though it has some validity. Maybe it does, but that was not what the original post indicated.

[K-amps]

I always wondered why they didn't have a drive mode where the mirror stays up for double or triple frames.  Assuming sensor readout the shutter and write to buffer/flash could keep up, you could get 20 or 30 fps, and still get 10 FPS flashing in the viewfinder; which should be enough to track a subject under most circumstances.

+1

+0.5.  I've thought about this, too...but the times when you most want a fast frame rate are with rapidly-moving subjects, and that's when you most need AF between frames.  I wouldn't want 30 fps as in-focus -> a little OOF -> a little more OOF -> in focus -> etc.  Heck, even the S100 I pre-ordered will hit nearly 10 fps (9.6) if it doesn't AF between shots.

I am new to the hobby compared to you guys/pros and have probably only taken about 40k-50k digital pics so far (I have 30k on my PC plus gobs that I throw away).

In those 40-50k, not once do I remember taking a burst with an object moving towards me where I needed AF on every frame. (Not saying others don't need it), but I have never needed it... so I may be in the minority here while I say, I'd rather have 10+fps with 1-3 AF's in between.

Most of my subjects move laterally (requiring little AF adjustment in a burst situation) or are hyper-mobile hills. Now that I have said it... I suspect I am going to regret saying it 

[Gothmoth]

I feel like I'm the only one who actually read Canon Rumors Guy's post.

What camera? No [CR] rating on this yet.
 Everything is pointing to a new 1D series camera, I am getting zero information from sources or from anywhere else on the web about any other camera.

so at least all seems to point to a 1D camera.
that for the guys who ask why we all discuss a new 1D model.

personally i think it will be an FF model that fits between the 7D and 5D.
but maybe thats only because i would like to see it.

as to the spec list... i don´t think it´s written in stone.
 

[Fred_]

Did anyone notice the hunch on the viewfinder in the tease? That means full frame, I'm thinking 1DS, they won't do anything crazily unexpected with the pro segment. Big departures from the present lines are fine for people who like spec sheets, but not for people who depend on this stuff for a living. They have always put a lot of attention in having the user of the previous model be right at home on the new camera.

Also, they have a bit of an arrogance in thinking they can make brilliant small sensors (exhibit a: 7D) with many pixels on them... so why not put some more pixels on some bigger sensors?

Except of course if they pull a full 180° and go the backlit sensors route. But from what I've heard, that can't really be done on larger sensors.