Sequence of events when you shoot in Live View?

[Big Ga]

Ok, so I have a D5300, but I suspect this applies to a whole bunch of nikon cameras.

Lets say I'm looking through the (optical) viewfinder. The shutter is closed, the mirror is down. I press the shutter release to take a shot.

The mirror flips up, the shutter opens, the exposure is made, the shutter closes and then the mirror flips down again. Simple.

But

When I'm in Live view and looking at the rear screen, what EXACTLY happens.

I am assuming that:

The mirror is already flipped up and the shutter is obviously open so the sensor is 'seeing' the scene and transferring the image to the rear LCD.

What EXACTLY is the sequence of events when I now press the shutter button to take the shot.

(and why does it take such a relatively long time to do it?)

 

[qunamax]

My best guess is it goes from liveview mode to actual record image mode, closes the shutter, opens it at the selected speed, captures the image, closes the shutter to finalize the exposure, opens the shutter again to go back to liveview mode.

 

[Big Ga]

My best guess is it goes from liveview mode to actual record image mode, closes the shutter, opens it at the selected speed, captures the image, closes the shutter to finalize the exposure, opens the shutter again to go back to liveview mode.

Ok, my fault for not being clear enough in the original post as to what information I was trying to glean.

I would imagine that when in live view, the sequence of events goes something like this:

The mirror is already up. The shutter is open - the sensor is exposed

You press the shutter release all the way down to take the picture

(The mirror stays up all the time)

The shutter closes to cover the sensor. The sensor is 'reset', ready to accept light.

There MIGHT be a delay at this point? is there ??

At some time now, I'm assuming a conventional mechanical shutter based exposure takes place. I.e. the shutter blades open (or a slit moves across the sensor), exposing it. The shutter blades then close, completely covering the sensor. The sensor data is read off.

There MIGHT be a delay at this point? is there ??

Then the shutter can open again, the sensor goes into electronic shuttering mode, effectively reading live view video type data off to provide the view on the rear LCD.

Now knowing how fast the camera can take shots in continuous shooting mode, it SHOULD be able to do the above really, really fast. Even faster because there is now no mirror movement at all involved. But it doesn't do it fast. Or at least it doesn't seem to. it seems to be a fairly slow, laborious process in comparison to what it could be and I'm trying to understand why.

It seems that there is some additional delay in there somewhere. Can anyone tell me what is going on?

I'm pretty sure this is a known issue, because when the D800 first came out, I noticed exactly the same type of laboured shooting in live view, and I have a feeling that there were discussions on the matter, however I didn't follow them at the time.

And to explain why I'm interested in this, the other day when I was checking out the focus on my D5300, I took a number of different shots, on a (steady) tripod (using the mirror up delay mode on), first shot would have been using 'normal' PDAF and I was looking through the viewfinder, a second shot straight afterwards was using live view based CDAF. When I compared the images afterwards even though I was initially checking for focus, I noticed that even if the focus was good, pretty much all of the live view shots were a little crisper and sharper than the corresponding OVF based ones. Hence I'm wondering why, and if the slow operation of the live view mode and a possible extra settling delay deliberately introduce somewhere, might be responsible.

Of course it might just be that the 1 sec exposure delay mode of the D5300 isn't long enough for all vibrations to subside. Not sure... but knowledge is power and all that, hence why I'm trying to understand properly what might be going on.

Cheers

G.

PS - if anyone can point me to any other threads or info on previous cameras such as the D800 that have this similar laborious LV shooting, that might help!

 

[Mako2011]

My best guess is it goes from liveview mode to actual record image mode, closes the shutter, opens it at the selected speed, captures the image, closes the shutter to finalize the exposure, opens the shutter again to go back to liveview mode.

Ok, my fault for not being clear enough in the original post as to what information I was trying to glean.

I would imagine that when in live view, the sequence of events goes something like this:

The mirror is already up. The shutter is open - the sensor is exposed

You press the shutter release all the way down to take the picture

(The mirror stays up all the time)

That's not always the case and depends on the model.

The shutter closes to cover the sensor. The sensor is 'reset', ready to accept light.

That;'s usually true...unless using flash. Sometimes the mirror will drop so the flash exposure can be metered (shutter closes at the same time) then the mirror rises again for the shutter to open and expose the scene to the sensor. Look though the viewfinder while taking Live-View shots with and without flash. It will be obvious when the mirror flips back down to meter the scene

There MIGHT be a delay at this point? is there ??

Depends and other setting and model.

At some time now, I'm assuming a conventional mechanical shutter based exposure takes place. I.e. the shutter blades open (or a slit moves across the sensor), exposing it. The shutter blades then close, completely covering the sensor. The sensor data is read off.

There MIGHT be a delay at this point? is there ??

Depends on model and settings

Then the shutter can open again, the sensor goes into electronic shuttering mode, effectively reading live view video type data off to provide the view on the rear LCD.

Basically true

Now knowing how fast the camera can take shots in continuous shooting mode, it SHOULD be able to do the above really, really fast. Even faster because there is now no mirror movement at all involved. But it doesn't do it fast. Or at least it doesn't seem to. it seems to be a fairly slow, laborious process in comparison to what it could be and I'm trying to understand why.

Part of it is the focusing is slow in Live-View and if you are in focus priority, things can slow down

It seems that there is some additional delay in there somewhere. Can anyone tell me what is going on?

I'm pretty sure this is a known issue, because when the D800 first came out, I noticed exactly the same type of laboured shooting in live view, and I have a feeling that there were discussions on the matter, however I didn't follow them at the time.

D800 basically works the same as the D5500 regards when the mirror flips in Live-View

And to explain why I'm interested in this, the other day when I was checking out the focus on my D5300, I took a number of different shots, on a (steady) tripod (using the mirror up delay mode on), first shot would have been using 'normal' PDAF and I was looking through the viewfinder, a second shot straight afterwards was using live view based CDAF. When I compared the images afterwards even though I was initially checking for focus, I noticed that even if the focus was good, pretty much all of the live view shots were a little crisper and sharper than the corresponding OVF based ones.

Likely because focus wasn't actually perfect in PDAF or you were in a shutter speed range where mirror slap is an issue. Was VR on? Also, there is a built in bias in Live-View that can affect expose when shooting very bright scenes

Hence I'm wondering why, and if the slow operation of the live view mode and a possible extra settling delay deliberately introduce somewhere, might be responsible.

Unlikely

Of course it might just be that the 1 sec exposure delay mode of the D5300 isn't long enough for all vibrations to subside.

Certainly can be part of the issue when mirror slap is a major factor and when using tripod

 

[beatlephile]

Like others have said this behavior varies by model. Some shutter mechanisms are linked to the mirror mechanism and requires a full cycle of both in order to recock the shutter. Since they are linked then the mirror must cycle down and back up even when shooting in live view so a subsequent shot can be taken.

Other models have a decoupled mirror mechanism that allows the shutter to fire and recock without moving the mirror. For example, Canon Rebels (XXXD and XXXXD) require a full cycle of the mirror during live view shooting while XXD and XD lines don't. Correct me if I am wrong but the same holds true as you go up the Nikon line. I'll have to get out my D7200 and D750 and test.

 

[Mako2011]

Like others have said this behavior varies by model. Some shutter mechanisms are linked to the mirror mechanism and requires a full cycle of both in order to recock the shutter.

Not the case with Nikon. With Nikon they are separate...it's just a matter of whether or not the RGB meter needs to see the scene for proper metering. When it does (usually just in the case of flash), the mirror is lowered so the RGB meter/sensor can image the scene.

Since they are linked then the mirror must cycle down and back up even when shooting in live view so a subsequent shot can be taken.

Not the case with the D5500...as the mirror does not cycle in Live-view when not using flash

Other models have a decoupled mirror mechanism that allows the shutter to fire and recock without moving the mirror.

That's also the case most Nikon DSLR models...the shutter mechanism in not mechanically linked, as far as I know to the mirror mechanism. That's why you get a different sequence action when shooting with flash

 

[fieldfest]

Loading…

photo.stackexchange.com

 

[Big Ga]

(The mirror stays up all the time)

That's not always the case and depends on the model.

The shutter closes to cover the sensor. The sensor is 'reset', ready to accept light.

That;'s usually true...unless using flash.

Ok, yes, I can see that the shutter might need to drop then. I had my D5300 on the desk and I peeped into the mirror box without a lens on, took a shot and I the mirror seemed to stay up. That was not using flash though.

Sometimes the mirror will drop so the flash exposure can be metered (shutter closes at the same time) then the mirror rises again for the shutter to open and expose the scene to the sensor. Look though the viewfinder while taking Live-View shots with and without flash. It will be obvious when the mirror flips back down to meter the scene

There MIGHT be a delay at this point? is there ??

Depends and other setting and model.

Interesting. Do you have more information on this?

At some time now, I'm assuming a conventional mechanical shutter based exposure takes place. I.e. the shutter blades open (or a slit moves across the sensor), exposing it. The shutter blades then close, completely covering the sensor. The sensor data is read off.

There MIGHT be a delay at this point? is there ??

Depends on model and settings

Again. Interesting, and do you have more info on what models do what? I did a bit of digging after my initial post on the slow operation of the D800 and to be honest, it was fairly inconclusive (although that might just be the fact that I hadn't found the right threads!). One theory for the known and recognised slow operation of the D800 in LV is that the camera does not read the data off the sensor into its RAM buffer before then writing onto the memory card - it goes directly to the memory card, and since the memory cards are relatively slow, this causes the delay. But as I said, this seemed to be a guess and I couldn't find anything confirming this is actually the case.

Now knowing how fast the camera can take shots in continuous shooting mode, it SHOULD be able to do the above really, really fast. Even faster because there is now no mirror movement at all involved. But it doesn't do it fast. Or at least it doesn't seem to. it seems to be a fairly slow, laborious process in comparison to what it could be and I'm trying to understand why.

Part of it is the focusing is slow in Live-View and if you are in focus priority, things can slow down

I was locked down on a tripod and would have pre-focused. Focusing is not part of the delay I'm talking about here.

It seems that there is some additional delay in there somewhere. Can anyone tell me what is going on?

I'm pretty sure this is a known issue, because when the D800 first came out, I noticed exactly the same type of laboured shooting in live view, and I have a feeling that there were discussions on the matter, however I didn't follow them at the time.

D800 basically works the same as the D5500 regards when the mirror flips in Live-View

And to explain why I'm interested in this, the other day when I was checking out the focus on my D5300, I took a number of different shots, on a (steady) tripod (using the mirror up delay mode on), first shot would have been using 'normal' PDAF and I was looking through the viewfinder, a second shot straight afterwards was using live view based CDAF. When I compared the images afterwards even though I was initially checking for focus, I noticed that even if the focus was good, pretty much all of the live view shots were a little crisper and sharper than the corresponding OVF based ones.

Likely because focus wasn't actually perfect in PDAF or you were in a shutter speed range where mirror slap is an issue. Was VR on?

In this situation, I was specifically testing for focus and in the cases I mentioned where the Live View shots were marginally sharper than the conventional OVF ones, I was sure focus wasn't the issue as I could see the focus plane. As I was on a tripod, VR was off ;-)

Mirror slap could well have been an issue, but as I also had the exposure delay mode on, this is worrying if its the cause, as it means the delay isn't long enough, even with a fairly small lens like the 18-140 (which I was using at this time).

Also, there is a built in bias in Live-View that can affect expose when shooting very bright scenes

That's also interesting and I wasn't aware of it. Got any more info on this or what would I search for specifically?

Hence I'm wondering why, and if the slow operation of the live view mode and a possible extra settling delay deliberately introduce somewhere, might be responsible.

Unlikely

It is indeed looking unlikely (although not impossible). I wondered because in my experience using the MFT mirrorless system, Oly have implemented a number of systems where there is a delay in the cycle of events, specifically it seems to avoid shock issues. Sometimes this delay can be set, sometimes its 'baked in' and hidden and people only found out about it when they analysed sound recordings of the shutter operation! If I was a nikon designer, knew there was an issue with say shutter shock, but also knew that someone using the live view mode was probably not concerned with ultimate shooting speed, I'd be tempted to stick in some kind of delay in there to give you the best IQ possible!

Of course it might just be that the 1 sec exposure delay mode of the D5300 isn't long enough for all vibrations to subside.

Certainly can be part of the issue when mirror slap is a major factor and when using tripod

Yes. There may be two totally separate issues going on here. The slow operation using LV (which may be a data writing/no buffer issue) and the sharper images in LV (which may just be a mirror slap thing). Have to be careful I don't do a 2+2=5

But at the moment, using what I'd consider my 'normal' (and fairly careful) techniques for tripod shooting, it looks as if I can get sharper shots when I'm using Live view, which itself is interesting.

 

[Mako2011]

Now knowing how fast the camera can take shots in continuous shooting mode, it SHOULD be able to do the above really, really fast. Even faster because there is now no mirror movement at all involved. But it doesn't do it fast. Or at least it doesn't seem to. it seems to be a fairly slow, laborious process in comparison to what it could be and I'm trying to understand why.

Part of it is the focusing is slow in Live-View and if you are in focus priority, things can slow down

I was locked down on a tripod and would have pre-focused. Focusing is not part of the delay I'm talking about here.

Don't forget that in Live-view the shutter must first close to reset the sensor for the shot, then open again. That doesn't happen in PDAF shots and simply flipping up the mirror is faster than first closing the shutter and resetting the sensor. That extra delay needed to reset the sensor may be what you are noticing.

Also, there is a built in bias in Live-View that can affect expose when shooting very bright scenes

That's also interesting and I wasn't aware of it. Got any more info on this or what would I search for specifically?

Live-View exposure is biased slightly to produce an image similar to what you see on the LCD. This is most obvious on bright sunny days and can result in a live-View image that is slightly overexposed compared to the same viewfinder shot. The manual does not explain it well, but it is hinted at in some Nikon manuals when is says...

(D7000 page 53) "Depending on the scene, exposure may differ from that which would be obtained when live view is not used. Metering in live view is adjusted to suit the live view display producing photographs with exposure close to what is seen in the monitor"

Of course it might just be that the 1 sec exposure delay mode of the D5300 isn't long enough for all vibrations to subside.

Certainly can be part of the issue when mirror slap is a major factor and when using tripod

Yes. There may be two totally separate issues going on here. The slow operation using LV (which may be a data writing/no buffer issue) and the sharper images in LV (which may just be a mirror slap thing). Have to be careful I don't do a 2+2=5

But at the moment, using what I'd consider my 'normal' (and fairly careful) techniques for tripod shooting, it looks as if I can get sharper shots when I'm using Live view, which itself is interesting.

Try using fast shutter speeds (greater than 1/250s) to rule out mirror slap and shutter vibration/shock. All that is left then is focus differences and exposure/brightness.

 

[tcg550]

The Googles found this from Nikon.

Loading…

www.nikonusa.com

 

[BirgerH]

Like others have said this behavior varies by model. Some shutter mechanisms are linked to the mirror mechanism and requires a full cycle of both in order to recock the shutter.

Not the case with Nikon. With Nikon they are separate...it's just a matter of whether or not the RGB meter needs to see the scene for proper metering. When it does (usually just in the case of flash), the mirror is lowered so the RGB meter/sensor can image the scene.

Curious, Mako:

My D90 requires a full cycle (the first LV-Nikon) of the mirror going up and down for each image. Even in burst-mode, it flips up and down for each image

So what you are saying, is that my D90 meters with the RGB-sensor in LV too - while my D7000 uses the image-sensor for metering?

I really thought, that the D90 used the image-sensor for metering, just that the engineers haven't separated the shutter-mechanism and the mirror-mechanism (yet).

Since they are linked then the mirror must cycle down and back up even when shooting in live view so a subsequent shot can be taken.

Not the case with the D5500...as the mirror does not cycle in Live-view when not using flash

Other models have a decoupled mirror mechanism that allows the shutter to fire and recock without moving the mirror.

That's also the case most Nikon DSLR models...the shutter mechanism in not mechanically linked, as far as I know to the mirror mechanism. That's why you get a different sequence action when shooting with flash

--
My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

BirgerH.

 

[Mako2011]

Like others have said this behavior varies by model. Some shutter mechanisms are linked to the mirror mechanism and requires a full cycle of both in order to recock the shutter.

Not the case with Nikon. With Nikon they are separate...it's just a matter of whether or not the RGB meter needs to see the scene for proper metering. When it does (usually just in the case of flash), the mirror is lowered so the RGB meter/sensor can image the scene.

Curious, Mako:

My D90 requires a full cycle (the first LV-Nikon) of the mirror going up and down for each image. Even in burst-mode, it flips up and down for each image

So what you are saying, is that my D90 meters with the RGB-sensor in LV too - while my D7000 uses the image-sensor for metering?

It may, but researching, the D90 may indeed have the mirror tied to cocking the shutter. Looks like D90 was the last body that works as beatlephile suggested above. But it may also be that the D90 is using PDAF during the Live-View exposure:

Loading…

www.nikonusa.com

I really thought, that the D90 used the image-sensor for metering, just that the engineers haven't separated the shutter-mechanism and the mirror-mechanism (yet).

Possible with the D90, and then the tech got better. Hard to say without opening up your D90

 

[BirgerH]

Like others have said this behavior varies by model. Some shutter mechanisms are linked to the mirror mechanism and requires a full cycle of both in order to recock the shutter.

Not the case with Nikon. With Nikon they are separate...it's just a matter of whether or not the RGB meter needs to see the scene for proper metering. When it does (usually just in the case of flash), the mirror is lowered so the RGB meter/sensor can image the scene.

Curious, Mako:

My D90 requires a full cycle (the first LV-Nikon) of the mirror going up and down for each image. Even in burst-mode, it flips up and down for each image

So what you are saying, is that my D90 meters with the RGB-sensor in LV too - while my D7000 uses the image-sensor for metering?

It may, but researching, the D90 may indeed have the mirror tied to cocking the shutter. Looks like D90 was the last body that works as beatlephile suggested above. But it may also be that the D90 is using PDAF during the Live-View exposure:

http://www.nikonusa.com/en/learn-and-explore/article/ftlzi4ra/live-view-shooting-modes.html

I really thought, that the D90 used the image-sensor for metering, just that the engineers haven't separated the shutter-mechanism and the mirror-mechanism (yet).

Possible with the D90, and then the tech got better. Hard to say without opening up your D90

No - not my D90 - it's not that necessary for me to know

 

[qunamax]

My best guess is it goes from liveview mode to actual record image mode, closes the shutter, opens it at the selected speed, captures the image, closes the shutter to finalize the exposure, opens the shutter again to go back to liveview mode.

Ok, my fault for not being clear enough in the original post as to what information I was trying to glean.

I would imagine that when in live view, the sequence of events goes something like this:

The mirror is already up. The shutter is open - the sensor is exposed

You press the shutter release all the way down to take the picture

(The mirror stays up all the time)

The shutter closes to cover the sensor. The sensor is 'reset', ready to accept light.

There MIGHT be a delay at this point? is there ??

At some time now, I'm assuming a conventional mechanical shutter based exposure takes place. I.e. the shutter blades open (or a slit moves across the sensor), exposing it. The shutter blades then close, completely covering the sensor. The sensor data is read off.

There MIGHT be a delay at this point? is there ??

Then the shutter can open again, the sensor goes into electronic shuttering mode, effectively reading live view video type data off to provide the view on the rear LCD.

Now knowing how fast the camera can take shots in continuous shooting mode, it SHOULD be able to do the above really, really fast. Even faster because there is now no mirror movement at all involved. But it doesn't do it fast. Or at least it doesn't seem to. it seems to be a fairly slow, laborious process in comparison to what it could be and I'm trying to understand why.

It seems that there is some additional delay in there somewhere. Can anyone tell me what is going on?

I'm pretty sure this is a known issue, because when the D800 first came out, I noticed exactly the same type of laboured shooting in live view, and I have a feeling that there were discussions on the matter, however I didn't follow them at the time.

And to explain why I'm interested in this, the other day when I was checking out the focus on my D5300, I took a number of different shots, on a (steady) tripod (using the mirror up delay mode on), first shot would have been using 'normal' PDAF and I was looking through the viewfinder, a second shot straight afterwards was using live view based CDAF. When I compared the images afterwards even though I was initially checking for focus, I noticed that even if the focus was good, pretty much all of the live view shots were a little crisper and sharper than the corresponding OVF based ones. Hence I'm wondering why, and if the slow operation of the live view mode and a possible extra settling delay deliberately introduce somewhere, might be responsible.

Of course it might just be that the 1 sec exposure delay mode of the D5300 isn't long enough for all vibrations to subside. Not sure... but knowledge is power and all that, hence why I'm trying to understand properly what might be going on.

Cheers

G.

PS - if anyone can point me to any other threads or info on previous cameras such as the D800 that have this similar laborious LV shooting, that might help!

I don't think so, delay is only for the first exposure when shooting in burst.
You can hear the shutter go click-click when liveview goes off and on, it's just a slow mechanical-software process to go from electronic shutter (liveview) to mechanical (shooting time).

 

[Big Ga]

I don't think so, delay is only for the first exposure when shooting in burst.

Interesting. I tried burst mode when in LV and after the first one, it does indeed rattle through the subsequent ones at what seems like full speed. It doesn't look as if its waiting for the card write between each shot.

However, there is a tad more of a delay at the start than in between the burst shots, and then it looks like it doesn't return control (or a live view display) to you until its written ALL the data to the card.

Hmmmm.

You can hear the shutter go click-click when liveview goes off and on, it's just a slow mechanical-software process to go from electronic shutter (liveview) to mechanical (shooting time).

Not really sure what you mean here.

 

[qunamax]

I don't think so, delay is only for the first exposure when shooting in burst.

Interesting. I tried burst mode when in LV and after the first one, it does indeed rattle through the subsequent ones at what seems like full speed. It doesn't look as if its waiting for the card write between each shot.

However, there is a tad more of a delay at the start than in between the burst shots, and then it looks like it doesn't return control (or a live view display) to you until its written ALL the data to the card.

Hmmmm.

You can hear the shutter go click-click when liveview goes off and on, it's just a slow mechanical-software process to go from electronic shutter (liveview) to mechanical (shooting time).

Not really sure what you mean here.

It doesn't return to normal operation quickly because liveview is software intensive, while using viewfinder/mirror you can capture the next shoot as soon as the buffer clears for at least one exposure more to fit in.

I mean, the process of transition from liveview (which uses electronic shutter to capture and display image from sensor) to shooting time capture (which uses mechanical shutter) and all the software processing used in the process (to engage and disengage liveview).

 

[Big Ga]

I don't think so, delay is only for the first exposure when shooting in burst.

Interesting. I tried burst mode when in LV and after the first one, it does indeed rattle through the subsequent ones at what seems like full speed. It doesn't look as if its waiting for the card write between each shot.

However, there is a tad more of a delay at the start than in between the burst shots, and then it looks like it doesn't return control (or a live view display) to you until its written ALL the data to the card.

Hmmmm.

You can hear the shutter go click-click when liveview goes off and on, it's just a slow mechanical-software process to go from electronic shutter (liveview) to mechanical (shooting time).

Not really sure what you mean here.

It doesn't return to normal operation quickly because liveview is software intensive, while using viewfinder/mirror you can capture the next shoot as soon as the buffer clears for at least one exposure more to fit in.

I mean, the process of transition from liveview (which uses electronic shutter to capture and display image from sensor) to shooting time capture (which uses mechanical shutter) and all the software processing used in the process (to engage and disengage liveview).

I'm still not sure why you think the delay is because there is some very intensive software processing going on. Anything LV based is effectively real time! or at least as fast as the FPS refresh on the rear LCD. What exactly is causing a delay in simply transitioning from one mode to the other? (apart from a mechanical mirror flipping up, but in the above discussion, its already up so can be ignored)

 

[qunamax]

I don't think so, delay is only for the first exposure when shooting in burst.

Interesting. I tried burst mode when in LV and after the first one, it does indeed rattle through the subsequent ones at what seems like full speed. It doesn't look as if its waiting for the card write between each shot.

However, there is a tad more of a delay at the start than in between the burst shots, and then it looks like it doesn't return control (or a live view display) to you until its written ALL the data to the card.

Hmmmm.

You can hear the shutter go click-click when liveview goes off and on, it's just a slow mechanical-software process to go from electronic shutter (liveview) to mechanical (shooting time).

Not really sure what you mean here.

It doesn't return to normal operation quickly because liveview is software intensive, while using viewfinder/mirror you can capture the next shoot as soon as the buffer clears for at least one exposure more to fit in.

I mean, the process of transition from liveview (which uses electronic shutter to capture and display image from sensor) to shooting time capture (which uses mechanical shutter) and all the software processing used in the process (to engage and disengage liveview).

I'm still not sure why you think the delay is because there is some very intensive software processing going on. Anything LV based is effectively real time! or at least as fast as the FPS refresh on the rear LCD. What exactly is causing a delay in simply transitioning from one mode to the other? (apart from a mechanical mirror flipping up, but in the above discussion, its already up so can be ignored)

Not really, I think the delay is caused by the combination of both mechanical (close shutter, open shutter) and software operation (turn off electronic shutter, turn on electronic shutter and live feed).

Once again:

0. (mirror raised all the time)

1. Shutter is open - electronic shutter and live feed are being displayed - sensor exposed all the time.

2. CPU turns off electronic shutter, turns off live feed, flushes the sensor (camera cpus are dedicated and fast for the job they do but ain't lightspeed fast, that's why they constantly put out new models)

3. Mechanical shutter closes to assume the start positions for shooting time

4. Mechanical shutter opens to expose the sensor. It takes the shot.

Steps 2, 3 and 4 are what makes the delay.

5. You take a burst of shots - it flies trough multiple exposures (because camera does not have to take steps 2, 3 and 4 again).

6. Mechanical shutter closes for the last exposure - opens again for live feed, cpu turns on the electrnoci shutter and live feed (after it's done processing the buffer which is cpu and bandwight intesive task)

(Nikon) Liveview made a hybrid system out of the 'all mechanical' dslr, as an add-in compromise technology, it's expected not to work very smoothly and with some drawbacks (aperture control comes to mind).

Of course, all this might not be right, it's just my understanding.