1Dx M2 Sensor Resolution - Back of envelop estimate

As well as the physical mirror speed, does the auto aperture close down speed also have to be figured into the equation of how many FPS can be achieved? I assume (I don't know) that the diaphragm cycles between each shoot and that's another thing that gets disabled if you use the current 1DX at 14FPS?
 
Upvote 0
motion folks: can anyone help me with a comparison to i.e. Red Dragon or Warrior or C300?

struggling to grasp just how big the difference is.

I have worked with Dragon footage and own a 5d mkiii and I am keenly aware of the difference in footage quality but I don't have the numbers.
 
Upvote 0
dilbert said:
rs said:
dilbert said:
Read the 1DX people.

At 14fps on the 1DX, the mirror gets locked up. That should tell you all that you need to know.

No-one on here has suggested anything but 12 fps with mirror movement on the 1D X. The previous generation with its smaller sensor ran at 10 fps. We're debating possibilities for the next generation.

I think it is safe to say that if moving to 14fps means that the shutter stays open then there's an engineering limit somewhere between 12fps and 14fps that makes it undesirable to move the mirror.

Are there any audio recordings of the 1DIV and 1DX that people have put through analysis software to show how long it takes the mirror to move, etc?

The defining limit on the fps will be the shutter lag, which is 55ms for the 1DX (1000/55 =~ 18) but it can be shortened to 36ms.

To get 12fps, the shutter speed needs to be 1/1000 or faster - i.e. 1ms or less is used for the shutter.

At 12fps, each shot can take 83ms. If 1ms is the allowed time for the exposure (1/1000) then 82ms to setup and restore the image box. With a shutter lag of 55ms, that suggests the restoration of the mirror is 28ms. If the mirror moves the same speed in both directions then it takes 28ms to move the mirror up and another 8ms before the camera is ready to take a photo.

Now at 14fps, each shot has 71ms to complete. That's less time than it takes to do complete the image box setup and restore - hence the mirror stays up.

So the question about "more FPS" is dependent on shutter lag - can Canon reduce the shutter lag from 55ms?

And yes, it is possible to reduce the shutter lag to 36mm but Canon don't guarantee that however we can take that to mean that the camera allows 19ms for the lens to get ready however with the mirror locked up, that doesn't need to be repeated.

http://www.imaging-resource.com/PRODS/E1DMK4/E1DMK4A6.HTM

http://www.imaging-resource.com/PRODS/canon-1dx/canon-1dxA6.HTM

From 1DIV -> 1DX, shutter lag went from 49ms to 55ms (larger mirror?) for prefocused shots and from 88ms to 108ms for continuous AF (i.e in continuous AF mode, you won't be able to shoot more than 10fps?)

My reading of all this is that there are physical limits that Canon is near (or has found) that mean 12fps is the limit with a moving mirror.

https://www.youtube.com/embed/CmjeCchGRQo

10 fps was the limit back in 2009 when they announced the 1D mk IV. 12 fps was the limit back in 2011 with the 1D X. It's now 4 years on, so why do the same limits apply? To add some more credit to this, I'll quote what I said earlier on:

rs said:
As for 12 FPS, Canon have managed that reliably with the 1D X. And since then they have introduced the Mirror Vibration Control System on the 5DS(r) and 7D mk II. This tech allows the mirror to softly stop at each end of travel instead of slapping into the end stops, allowing for less vibration and quicker settling for viewfinder/AF purposes. This should allow them to push the envelope a little further on their flagship product without breaking a sweat. I imagine 14 FPS could be a given for them, and they could well pull one out of the bag and push it even further.
 
Upvote 0
dilbert said:
I think it is safe to say that if moving to 14fps means that the shutter stays open then there's an engineering limit somewhere between 12fps and 14fps that makes it undesirable to move the mirror.

I think it is safe to say that you don't know what the H-E-double-hockey-sticks you're talking about. The shutter does not remain open when shooting at 14 fps on the 1D X.

The remainder of your post contains a plethora of factual errors and unwarranted assumptions. Typical dilbert.
 
Upvote 0
neuroanatomist said:
takesome1 said:
Canon was promoting frame grabs from video with the 1Dx.
With the next version that is what will likely be pushed even more with 4K video.
Increasing shutter speed is a redundant increase in technology if Canon believes that technology is headed in that direction.

So you're saying the 1D X II will have DPAF and it will be as good as dedicated phase AF for subject tracking?

I didn't say that. Then again what does the future hold in technology?

I will say it is my speculation that we will not see a substantial frame rate increase out of the next body.

Another reason I believe this.
Geeks and gearheads like the fastest, but is a frame rate over 12 fps necessary for most of the photog's and pro's that will buy the 1Dx II? Some will want it, but will Canon think it is worth while to invest the R&D for the mechanical system to do this for those that do? I can only think of a hand full of scenarios that I would want more and in some of those situations no current AF system keeps up. We could ask those on this forum if they need the increased frame rate, I think many here would but it wouldn't be a fair sampling.
 
Upvote 0
takesome1 said:
Canon was promoting frame grabs from video with the 1Dx.
With the next version that is what will likely be pushed even more with 4K video.
Increasing shutter speed is a redundant increase in technology if Canon believes that technology is headed in that direction.
For that reason I think you see mp increase, ISO and DR performance increase if possible and all balanced to the abilities of the latest greatest processors. 24-28mp should be possible, keep in mind the speed and depth of the buffer can be just as important as frame rate.

There are disadvantages to using video for stills capture vs genuine stills. Most are pretty big to overcome, and the only solution is to shoot stills:

  • video fires off a frame to a precise timeline. Stills can fire off a frame when the conditions are right, e.g. shutter speed, aperture, ISO, anti flicker can all delay the next frame, but for the right reasons
  • shutter speed
  • metering per frame
  • AF
  • rolling shutter
  • tracking action
 
Upvote 0
takesome1 said:
Another reason I believe this.
Geeks and gearheads like the fastest, but is a frame rate over 12 fps necessary for most of the photog's and pro's that will buy the 1Dx II? Some will want it, but will Canon think it is worth while to invest the R&D for the mechanical system to do this for those that do? I can only think of a hand full of scenarios that I would want more and in some of those situations no current AF system keeps up. We could ask those on this forum if they need the increased frame rate, I think many here would but it wouldn't be a fair sampling.

Many users of the 1D X shoot things that move fast, where catching the key moment or peak of action is critical. I've shot action where the frame rate was 12 fps...and I've shot action where the frame rate was as fast as my thumb could push the lever between shots. More is better.
 
Upvote 0
rs said:
takesome1 said:
Canon was promoting frame grabs from video with the 1Dx.
With the next version that is what will likely be pushed even more with 4K video.
Increasing shutter speed is a redundant increase in technology if Canon believes that technology is headed in that direction.
For that reason I think you see mp increase, ISO and DR performance increase if possible and all balanced to the abilities of the latest greatest processors. 24-28mp should be possible, keep in mind the speed and depth of the buffer can be just as important as frame rate.

There are disadvantages to using video for stills capture vs genuine stills. Most are pretty big to overcome, and the only solution is to shoot stills:

  • video fires off a frame to a precise timeline. Stills can fire off a frame when the conditions are right, e.g. shutter speed, aperture, ISO, anti flicker can all delay the next frame, but for the right reasons
  • shutter speed
  • metering per frame
  • AF
  • rolling shutter
  • tracking action

I wonder if Canon knows that those problems may be to big to overcome, otherwise the may be wasting R&D money on this instead of important things like DR.
I tried to find the article Canon put out when the 1Dx was released. There was a video they released with it. Apparently it a very popular article as it doesn't pop up easily on searches.
 
Upvote 0
takesome1 said:
Geeks and gearheads like the fastest, but is a frame rate over 12 fps necessary for most of the photog's and pro's that will buy the 1Dx II? Some will want it, but will Canon think it is worth while to invest the R&D for the mechanical system to do this for those that do? I can only think of a hand full of scenarios that I would want more and in some of those situations no current AF system keeps up. We could ask those on this forum if they need the increased frame rate, I think many here would but it wouldn't be a fair sampling.

Eh, yeah I think it would be desirable, especially for sports. Sports photos are often most dramatic at very specific instants and more frames means more chances to catch those moments. Wildlife is often the same and while actual paid professional wildlife photographers are rarer than professional astronauts there are gazillions of amateurs willing to spend tons of money to get those dramatic moments. I think there is definitely a market for boosts to burst speeds.
 
Upvote 0
neuroanatomist said:
dilbert said:
I think it is safe to say that if moving to 14fps means that the shutter stays open then there's an engineering limit somewhere between 12fps and 14fps that makes it undesirable to move the mirror.

I think it is safe to say that you don't know what the H-E-double-hockey-sticks you're talking about. The shutter does not remain open when shooting at 14 fps on the 1D X.

The remainder of your post contains a plethora of factual errors and unwarranted assumptions. Typical dilbert.

A couple of questions

1. at 14 FPS does the mirror go up and down? what is difference between 12FPS and 14FPS (and I don't mean 2 FPS).

2. how does canon video work? Is the mirror up all the time?
 
Upvote 0
RGF said:
neuroanatomist said:
dilbert said:
I think it is safe to say that if moving to 14fps means that the shutter stays open then there's an engineering limit somewhere between 12fps and 14fps that makes it undesirable to move the mirror.

I think it is safe to say that you don't know what the H-E-double-hockey-sticks you're talking about. The shutter does not remain open when shooting at 14 fps on the 1D X.

The remainder of your post contains a plethora of factual errors and unwarranted assumptions. Typical dilbert.

A couple of questions

1. at 14 FPS does the mirror go up and down? what is difference between 12FPS and 14FPS (and I don't mean 2 FPS).

2. how does canon video work? Is the mirror up all the time?

1. The mirror is locked up at 14 fps (not at 12). The shutter opens and closes with each exposure at 12 and 14 fps. The mirror being locked up precludes AF between shots, and you're also limited to jpg only.

2. The mirror is locked up during video shooting.
 
Upvote 0
dilbert said:
RGF said:
....
A couple of questions

1. at 14 FPS does the mirror go up and down? what is difference between 12FPS and 14FPS (and I don't mean 2 FPS).

No, it does not.

Also, for 12fps, you need to shoot at 1/1000 or faster.

From the 1DX manual (page 116 Drove: Selecting the Drive Mode):
The maximum continuous shooting speed of approx. 12 shots/sec is attained under the following conditions*: At ISO 25600 or lower, 1/1000sec. or faster shutter speed, and at the maximum aperture (varies depending on the lens). The continuous shooting speed may be slower depending on the ISO speed, shutter speed, aperture, subject conditions, brightness, lens type, flash use, etc.

Is my 1DX broken since I'm using 1/500 at the 12fps?
 
Upvote 0
filluppa said:
neuroanatomist said:
RGF said:
Wonder why the ISO limit?

Higher ISO means more noise which means larger files. From TDP, a studio scene that yields a 24 MB file at ISO 100 is 32 MB at ISO 25600 and 37 MB at ISO 102400.

higher ISO means different signal to noise ratio, not more noise

If a higher ISO makes the S/N ratio lower, how doesn't that raise the noise proportion of the signal?
 
Upvote 0
filluppa said:
the noise is lower at higher iso, higher iso means not more noise. as you can se below, but the signal decreases
so it is wrong to suggest that noise increases with higher ISO

It is wrong to suggest that read noise is the only component of image noise.

Do you know why a scene shot at ISO 100 that yields a 24 MB file yields a file that's 75% larger (41 MB) when shot at ISO 204800? The short answer is more noise. The long answer is...also more noise, albeit with more details and explanation because it's the long answer.
 
Upvote 0
neuroanatomist said:
RGF said:
Wonder why the ISO limit?

Higher ISO means more noise which means larger files. From TDP, a studio scene that yields a 24 MB file at ISO 100 is 32 MB at ISO 25600 and 37 MB at ISO 102400.

Is the time limited by read time from the sensor? Does the write time to memory really matter? I thought memory writes were well in excess of 1GB/Sec (perhaps 10 GB/sec). could that be a bottle neck?
 
Upvote 0
dilbert said:
RGF said:
Thanks everyone for the replies.

Wonder why the ISO limit?

Canon's explanation:
At these high ISO speeds, the system requires slightly more time for noise reduction and the maximum available fps rate slows to 10 fps (even when set to 12 fps).

Reference:
http://www.learn.usa.canon.com/resources/articles/2013/eos1dx_high_speed_shooting.htmlp

Consider that (for example) taking 1ms rather than .5ms to process the image for noise would be enough to slow the shooting speed down.

Feel free to come up with your own reason if you don't like Canon's.

So if I turn off noise reduction I can hit 12 FPS? Need to try that when I get home.
 
Upvote 0
filluppa said:
neuroanatomist said:
filluppa said:
the noise is lower at higher iso, higher iso means not more noise. as you can se below, but the signal decreases
so it is wrong to suggest that noise increases with higher ISO

It is wrong to suggest that read noise is the only component of image noise.

Do you know why a scene shot at ISO 100 that yields a 24 MB file yields a file that's 75% larger (41 MB) when shot at ISO 204800? The short answer is more noise. The long answer is...also more noise, albeit with more details and explanation because it's the long answer.

no
the short answer is more amplification of a given signal,and a different signal/noise ration the noise does not increase with higher iso

Oh, I see. So, more amplification of something does not mean more ... So, (5 x 25600 widgets) is not more than (5 x 100 widgets). Yeah, that makes so much sense it must be true.

An image shot at higher ISO has more noise that the same image shot at lower ISO. You can continue to disagree, but you'll only make yourself look foolish (as usual...at least until you disappear again).
 
Upvote 0