Are These The Next Canon Cameras To Be Announced? [CR1]

[BillB]

ofc only my personal preference: but i do find cameras with a fixed lens a perversion, especially when it is a prime focal lens and not even a zoom.

"zoom with your feet" = sometimes possible, but often not.
"crop" = hardly ever as good as getting the desired framing "optically" at time of capture; to me really only a last stop measure in "focal length limited situations".

Which lens I would like to use on a small FF mirrorless cam? Any! By mounting them on lens mount.

My lens preferences are
* 3 "as compact as possible" constant f/4 zooms - e.g. EF-X 16-35, 24-70, 50-150 all IS STM of course
* plus a set of 3 ultra-compact primes eg. : EF-X 20/2.8 (my "landscape and indoors pancake"), EF-X 40/2.0 [my "walkaround pancake"], 85/2.4 IS STM [my "portrait ultra-compact"].

Done. Will not buy any other lenses, only rent on the rare occasions when I really need them.

PS: I like what Samyang has started with their new AF lenses. Their AF 24/2.8 is right down my alley in terms of size and price.

Cropping does produce a hit on IQ, but how important is it if you have a couple of dozen megapixels and you are not making prints larger than A3? It is about tradeoffs.

 

[Kit.]

But tethering is not "just a feature in menu". It's an extensive modification of all software parts responsible for hard realtime performance of a camera (which by itself is hard to write).

Without knowledge of Canon source code, I would guesstimate it for at least one man-year to write (and test) for one camera type, and at least one man-month to update it for a new camera model.

But I'm just a programmer, not a project manager, so my estimate is better to be multiplied by Pi.

As a programmer then, you should know that much of that development is valid for a number of different cameras, lowering the cost per product. While each camera is different, they share enough code to spread those costs around. And incremental features cost far less.

That's what I said. Not less than one man-month to adapt the existing tethering implementation code for a new model in the line. Still ~10x cheaper than to develop it from scratch, still not cheap enough to put it into every camera line.

Maybe a lot less than that -- possibly just hours, or even just the addition a few items to static array initialization. If you want to dig in you might look at the source code for this project (I have not).

http://digicamcontrol.com/cameras

That's just a Windows PC client.

 

[Orangutan]

But tethering is not "just a feature in menu". It's an extensive modification of all software parts responsible for hard realtime performance of a camera (which by itself is hard to write).

Without knowledge of Canon source code, I would guesstimate it for at least one man-year to write (and test) for one camera type, and at least one man-month to update it for a new camera model.

But I'm just a programmer, not a project manager, so my estimate is better to be multiplied by Pi.

As a programmer then, you should know that much of that development is valid for a number of different cameras, lowering the cost per product. While each camera is different, they share enough code to spread those costs around. And incremental features cost far less.

That's what I said. Not less than one man-month to adapt the existing tethering implementation code for a new model in the line. Still ~10x cheaper than to develop it from scratch, still not cheap enough to put it into every camera line.

Maybe a lot less than that -- possibly just hours, or even just the addition a few items to static array initialization. If you want to dig in you might look at the source code for this project (I have not).

http://digicamcontrol.com/cameras

That's just a Windows PC client.

Yes, but it gives you a clue regarding the API. If you want info from the other perspective, look at the Magic Lantern source and discussion forums -- that'll tell you how much internal similarity there is from one model to the next. The few times I've looked at ML, most of the discussion is about figuring out how to hook in to the firmware of a new model, not how to get it to work once hooked.

Anyway, you made an off-the-cuff guess as to how much time/effort would be involved, and I pointed you to resources that would give you more information about what would really be involved. It's up to you now to decide whether you want to be informed or not.

 

[3kramd5]

My lens preferences are
* 3 "as compact as possible" constant f/4 zooms - e.g. EF-X 16-35, 24-70, 50-150 all IS STM of course
* plus a set of 3 ultra-compact primes eg. : EF-X 20/2.8 (my "landscape and indoors pancake"), EF-X 40/2.0 [my "walkaround pancake"], 85/2.4 IS STM [my "portrait ultra-compact"].

You could make them even smaller and lighter as fixed focal length, fixed small aperture, fixed focal point non-IS “solid state” lenses to go along with your distaste for moving parts!

Done. Will not buy any other lenses, only rent on the rare occasions when I really need them.

Boy, sounds like a customer I’d want to cater to

 

[3kramd5]

Hi, I a am wondering if it is true that we have to wait at least 1 full year (or longer) before see the new high end Canon flagship top notch camera to replace the EOS 1 DX Mark II

Anyone has any info on this...?
Andrea

Hi Andrea -

Canon could announce a 1Dxiii or 1Dsiv or 1M or whatever today. At this point, predictions are based on past trends, not inside information.

 

[fullstop]

Boy, sounds like a customer I’d want to cater to

hmmm ... so you are saying ... Canon does not need customers willing to purchase 1 Canon EOS FF mirrorless camera + 3 EF-X f/4 zooms + 3 EF-X primes ... and then take some break ... until next body upgrade ...
;D ;D ;D

Well methinks not even Canon would be so stupid and arrogant.

 

[3kramd5]

Boy, sounds like a customer I’d want to cater to

hmmm ... so you are saying ... Canon does not need customers willing to purchase 1 Canon EOS FF mirrorless camera + 3 EF-X f/4 zooms + 3 EF-X primes ... and then take some break ... until next body upgrade ...
;D ;D ;D

Well methinks not even Canon would be so stupid and arrogant.

I missed that it was six lenses, not three.

 

[Keith_Reeder]

hmmm ... so you are saying ... Canon does not need customers willing to purchase 1 Canon EOS FF mirrorless camera + 3 EF-X f/4 zooms + 3 EF-X primes ... and then take some break ... until next body upgrade ...
;D ;D ;D

No, he's just saying that he'd like customers that easy to please...

Well methinks not even Canon would be so stupid and arrogant.

There's very little about Canon's behaviour to suggest that they're either.

Forum members, though? Plenty of evidence.

 

[fullstop]

I missed that it was six lenses, not three.

yep. f/4 for zooms, f/2.8 or a bit faster for primes.

 

[Durf]

Boy, sounds like a customer I’d want to cater to

hmmm ... so you are saying ... Canon does not need customers willing to purchase 1 Canon EOS FF mirrorless camera + 3 EF-X f/4 zooms + 3 EF-X primes ... and then take some break ... until next body upgrade ...
;D ;D ;D

Well methinks not even Canon would be so stupid and arrogant.

Constantly wishing and hoping for things that don't exist can be a bit frustrating and often simply just a waste of time.

It would probably be best to enjoy using and getting the most out of the gear that you do have and possibly working towards upgrades that actually do exist in this reality.

 

[fullstop]

It would probably be best to enjoy using and getting the most out of the gear that you do have

I do.

and possibly working towards upgrades that actually do exist in this reality.

I don't. I boycot them. Until they do.

 

[Kit.]

That's what I said. Not less than one man-month to adapt the existing tethering implementation code for a new model in the line. Still ~10x cheaper than to develop it from scratch, still not cheap enough to put it into every camera line.

Maybe a lot less than that -- possibly just hours, or even just the addition a few items to static array initialization. If you want to dig in you might look at the source code for this project (I have not).

http://digicamcontrol.com/cameras

That's just a Windows PC client.

Yes, but it gives you a clue regarding the API.

If you had a clue about the efforts involved to create a server for such API in hard-realtime environment, you wouldn't try to point me to a Windows PC client.

If you want info from the other perspective, look at the Magic Lantern source and discussion forums -- that'll tell you how much internal similarity there is from one model to the next. The few times I've looked at ML, most of the discussion is about figuring out how to hook in to the firmware of a new model, not how to get it to work once hooked.

For which model exactly did Magic Lantern implement tethering that was absent in the stock firmware?

And Magic Lantern gives you no guarantee at all that what they do won't break things.

Anyway, you made an off-the-cuff guess as to how much time/effort would be involved, and I pointed you to resources that would give you more information about what would really be involved.

They give zero information on the topic.

 

[unfocused]

That's what I said. Not less than one man-month to adapt the existing tethering implementation code for a new model in the line. Still ~10x cheaper than to develop it from scratch, still not cheap enough to put it into every camera line.

Maybe a lot less than that -- possibly just hours, or even just the addition a few items to static array initialization. If you want to dig in you might look at the source code for this project (I have not).

http://digicamcontrol.com/cameras

That's just a Windows PC client.

Yes, but it gives you a clue regarding the API.

If you had a clue about the efforts involved to create a server for such API in hard-realtime environment, you wouldn't try to point me to a Windows PC client.

If you want info from the other perspective, look at the Magic Lantern source and discussion forums -- that'll tell you how much internal similarity there is from one model to the next. The few times I've looked at ML, most of the discussion is about figuring out how to hook in to the firmware of a new model, not how to get it to work once hooked.

For which model exactly did Magic Lantern implement tethering that was absent in the stock firmware?

And Magic Lantern gives you no guarantee at all that what they do won't break things.

Anyway, you made an off-the-cuff guess as to how much time/effort would be involved, and I pointed you to resources that would give you more information about what would really be involved.

They give zero information on the topic.

You are kind of rude considering I think he was trying to be helpful.

 

[Orangutan]

snip

For which model exactly did Magic Lantern implement tethering that was absent in the stock firmware

That's not the question. The question is how hard is it for Canon to make the change, given that they have all the engineering specs and source code. That Windows client and ML firmware might give you important indirect information. I have not looked at their code, but I pointed them out to you because you seemed interested. If you look at their source code you might well find the clues you're looking for. Until then, your claims are just speculation.

My speculation is that they choose their CPUs in part to make model changes easier and more robust.

https://en.wikipedia.org/wiki/DIGIC

 

[stevelee]

Of these three cameras, the one that might interest me is the G7X Mark III. If it is enough of an upgrade for the money, it is likely the next camera I will buy, probably before my next big trip, whenever that might be.

I've been extremely pleased with the G7X II and am in no rush to upgrade it, so if and when for an upgrade can be decided when I have much more information.

I have no need or interest in a new DSLR or mirrorless body any time in the foreseeable future. The new 100-400mm is my new toy, and if I need anything longer, I'll get a 1.4X for it, but don't anticipate that. My next lens purchase will be the 16mm-35mm f/4, since I don't have anything wider for FF than 24mm.

 

[rrcphoto]

everyone has gone crazy over the M5 Mark II and no one has really commented on the fact that this will be Canon's first fully made 1" sensor camera.

Canon making competitive 1" sensors with DPAF could change the narrative for those cameras especially if they manage to keep the costs low especially with Sony's now costing $1200 a pop for a RX100

 

[rrcphoto]

But tethering is not "just a feature in menu". It's an extensive modification of all software parts responsible for hard realtime performance of a camera (which by itself is hard to write).

Without knowledge of Canon source code, I would guesstimate it for at least one man-year to write (and test) for one camera type, and at least one man-month to update it for a new camera model.

But I'm just a programmer, not a project manager, so my estimate is better to be multiplied by Pi.

As a programmer then, you should know that much of that development is valid for a number of different cameras, lowering the cost per product. While each camera is different, they share enough code to spread those costs around. And incremental features cost far less.

That's what I said. Not less than one man-month to adapt the existing tethering implementation code for a new model in the line. Still ~10x cheaper than to develop it from scratch, still not cheap enough to put it into every camera line.

ML did a review of the tethering problem when the M came out. I can't find the thread anymore or the comment, but I recall it being a problem with the legacy code in EOS Utility, that it was built for a 30 fps framerate, and the M's were faster than that in liveview (60 i believe, with the M5 up to 120 fps), and there was a huge complexity for the change, so Canon simply to get the camera out elected to disable tethering in the camera itself was their prognosis.

also at the time ML stated that they would NEVER re-enable a feature that Canon disabled.

 

[fullstop]

yep. Typical ML. No balls. One of the reasons i never cared about ML and don't respect them really. I like the Yongnuos, Meikes and Samyangs much better in that respect. They don't give a damn if some Canon lawyer sends a letter or not.

 

[BillB]

yep. Typical ML. No balls. One of the reasons i never cared about ML and don't respect them really. I like the Yongnuos, Meikes and Samyangs much better in that respect. They don't give a damn if some Canon lawyer sends a letter or not.

And you know all this how?

 

[fullstop]

it is evident.

ML don't touch this [!D eeries cams for example], ML don't do that [re-enabling certain functions when nerfed by Canon], ML knee jerks as soon as Canon all but bats an eye.