Oscilloscope Zoom Out Quirk

[rsjsouza]

No its classic chinese out of the box thinking... you leave it like that in the first place and set your zoom in to where you want it.. works fine in all modes, stop, single, trigger, auto

If Rigol, Owon and Uni-T did the same, I would agree with you there, but to me it looks like a classic second Tier brand copying thinking. I suspect LeCroy did that originally and Siglent followed suit.

(of course, you can always argue that the opposite is true as well: the other brands copied the top tiers Tek or HPAK)

And that's the thing. For a scope not to offer the full memory for a single capture even when deliberately selected (without using a convoluted dual timebase system) is just silly.
The user needs to be able to chose how that memory is used, and Siglent/Lecroy/Picoscope are seemingly the only ones crippled in this regard.

I think that is exactly the crux of the matter. Since the manual memory depth setting is not translated to the capture itself, why have it in the first place? It seems a waste of memory when the user deliberately told the scope to use it.

BTW, the choice of doing the "Zoom In" approach does not immediately mean there is something wrong with it - after all, it is a design choice.

A question for anyone with a scope that has the 'history' mode:

Does changing the settings between triggers, e.g. changing vertical or horizontal scale, or enabling an additional channel, clear the history buffer? If not, do such changes get reflected in the on-screen display of the settings, when reviewing history?

If I understood it correctly, you mean in Run mode, right? If so, the buffer is discarded when enabling channels on the Rigol DS4014 and I would expect so, given the memory is split between pairs of channels.
In stop mode you can change the vertical and horizontal settings without a problem.

[nfmax]

A question for anyone with a scope that has the 'history' mode:

Does changing the settings between triggers, e.g. changing vertical or horizontal scale, or enabling an additional channel, clear the history buffer? If not, do such changes get reflected in the on-screen display of the settings, when reviewing history?

If I understood it correctly, you mean in Run mode, right? If so, the buffer is discarded when enabling channels on the Rigol DS4014 and I would expect so, given the memory is split between pairs of channels.
In stop mode you can change the vertical and horizontal settings without a problem.

Presumably you have to be in Stop mode anyway to look at the history?

[tv84]

I think they had already developed the whole architecture when they decided to put a history button... 

"Uhmm, what about putting a history button in this free space over here?"

Why if it's always in history mode? ?   

Siglent's 200 Mpts: "We have plenty of storage space in this server. But it's all tapes..."

[JPortici]

Also the way LeCroy has always operated and the way Picoscopes have always operated.
Why do i have to keep telling people this?

That doesn't make it right.
Here is the thing. If you want to look at a short timebase signal then you'll set a short timebase, that's normal oscilloscope usage.
Even on a Siglent, Lecroy or Picoscope, you aren't going to go through to convoluted process of setting a slow timebase and then chosing the zoom mode and then setting the secondary timebase just for regular use.
The point is that when using the scope the normal way, you can't zoom out if you need to.

I removed the message because i though about it twice and didn't want to have this argument again.
There is nothing wrong in one or the other. Both have edge cases. End of story.

It doesn't make it wrong either.
It doesn't make it some weird ass chinese blah blah  (not sure if i have understood that post correctly though)
Granted i've been using scope for just 10 years and had the chance to use everything available with reasonable price, but i like the lecroy way better, about many things. Don't care about zoomout because i can always reproduce the shot (don't care about other what if i had zoom out bullshit, i can guarantee you that at some point you would have wanted more memory so the acquisition is useless) instead i care more about automatic history mode with LOTS of acquisitions

[JPortici]

A question for anyone with a scope that has the 'history' mode:

Does changing the settings between triggers, e.g. changing vertical or horizontal scale, or enabling an additional channel, clear the history buffer? If not, do such changes get reflected in the on-screen display of the settings, when reviewing history?

Usually, yes. (Picoscope, Lecroy and siglent that i've used) It can be a pain but you get used to it. I think it depends on how the scope performs operations on the data (assumes current v/div for calculation, memory per channel and so on)

[2N3055]

To me convoluted process is to set memory depth, mentally calculating what timescale it will be at target sample rate, mentally calculate how much of that I have to set as pretrigger to make sure I have enough pre trigger capture once I'm at the target timebase. Then I set target timebase, tightly set around trigger and then keep capturing until I find something interesting, and then, without any sense where exactly in the whole buffer I am i keep twiddling around with timebase looking at data searching for something interesting.

Problem is that Dave misunderstood (or ignored) original reason why Nico is such a proponent of this.
Nico uses it for slow, long samples, where he looks at long bursts of data on ,say, CAN bus.  He keeps looking at one packet or event, and when that even happens he wants to know what happened around it. He doesn't stop captures, but works in RUN/Normal triggering mode (not Auto), from trigger to trigger event, and he initiates data transfers from equipment manually, or has them set up to occur slow enough that he has time to manually stop scope if needed.
It is very deliberate, he doesn't mind that scope needs 1 second per capture to process, and it IS NOT interactive scope work. You cannot have interactive scope work with 2 triggers per second.

As I already documented, Keysight scope DOESN'T work like that (Nico uses his R&S RTM3000 for this) in RUN mode.  In RUN mode in between triggers, it has only screenfull of data available. Only after you STOP, it will gather all buffers and reassemble them. And if you are lucky that you are working in less 5 us/div, you will have data up to 20 us/div. If you are at 20 us/div and up (50us/div in single mode), there is NO data outside...

I had DS1074Z. I pretty much never used fixed memory length. I used AUTO mode pretty much exclusively. Setting scope to capture too long capture on every trigger makes scope slow. Every one. One of the reasons Tektronix get their bad reputation is manual memory settings. People put them to 20MS and then complaint how it non responsive. 
Suddenly that is a good thing. It is not. That being said I DID USE IT from time to time. For instance to look at gated PSU startup. Set it for MAX mem, set timebase for switch on event, look at it and then if there was something wrong, pull out to see what was happening around it. But with Picoscope's excellent zoom implementation, it behaves exactly the same the other way around.

Siglent 2000X+ seems to work exactly as Picoscope.  And it is not "badly implemented manual memory length", but a "auto memory management with maximum length" setting. It is there to control strategy to balance huge memory with how will sampling rate drop off as you go to slower timebases. Siglent should have better explained and/or named this option. Siglent doesn't have "fixed manual memory size"
Also, if it is possible, they should think of implementing Keysight like feature in user configurable option, that they give up history mode and do single shot captures in full length as set. That would make them almost exactly same as Keysight (same at slow timebases). They might even do "full Keysight" and set minimum sampled length.  Whatever.

That being said, if customers think this is necessary, manufacturers should try to address it.

[nctnico]

To me convoluted process is to set memory depth, mentally calculating what timescale it will be at target sample rate, mentally calculate how much of that I have to set as pretrigger to make sure I have enough pre trigger capture once I'm at the target timebase. Then I set target timebase, tightly set around trigger and then keep capturing until I find something interesting, and then, without any sense where exactly in the whole buffer I am i keep twiddling around with timebase looking at data searching for something interesting.

That is not how it works! If you don't get it then you don't get it but by riduculing other people's workflow you only make a fool out of yourself.

[2N3055]

To me convoluted process is to set memory depth, mentally calculating what timescale it will be at target sample rate, mentally calculate how much of that I have to set as pretrigger to make sure I have enough pre trigger capture once I'm at the target timebase. Then I set target timebase, tightly set around trigger and then keep capturing until I find something interesting, and then, without any sense where exactly in the whole buffer I am i keep twiddling around with timebase looking at data searching for something interesting.

That is not how it works! If you don't get it then you don't get it but by riduculing other people's workflow you only make a fool out of yourself.

It is exactly how it works. Unlike you, I like to have a guarantee I will capture something before I start doing it. You are doing this not because it is smart thing to do it, but because it is the lazy way, and you hate zoom mode. So you do it this way. If it works for you , that's nice.
I do it in opposite direction and get same results and find that much easier.
But both you, and now Dave apparently, try to sell this as some panacea, and will do that without any negative impact. Which is not true. It has consequences.
In my mind it is not that Siglent manages memory wrong way, but that zoom mode should be implemented to be more user friendly.  I LIKE how zoom mode has DETERMINISTIC dual time bases. They should shuffle windows differently and make them more configurable, but it is a display and user interface  problem, not capture problem.  This is exactly what R&S and Keysight does: it sets timebase (internaly) for long capture, and then shows you only part of captured waveform. It is just U/I that shows you part of waveform, using basically "virtual timebase".

And I get it. I read what you wrote on this topic before. 

By the by, it is you who is proponent of idea that whoever is not agreeing with you is stupid. And this is not first topic you had that condescending tone..
As I said before, you are wonderful engineer with waste experience. I respect that. And I agreed and disagreed with you on many topics. And I will keep doing that at my own conscience.

[nctnico]

To me convoluted process is to set memory depth, mentally calculating what timescale it will be at target sample rate, mentally calculate how much of that I have to set as pretrigger to make sure I have enough pre trigger capture once I'm at the target timebase. Then I set target timebase, tightly set around trigger and then keep capturing until I find something interesting, and then, without any sense where exactly in the whole buffer I am i keep twiddling around with timebase looking at data searching for something interesting.

That is not how it works! If you don't get it then you don't get it but by riduculing other people's workflow you only make a fool out of yourself.

It is exactly how it works. Unlike you, I like to have a guarantee I will capture something before I start doing it. You are doing this not because it is smart thing to do it, but because it is the lazy way, and you hate zoom mode. So you do it this way. If it works for you , that's nice.
I do it in opposite direction and get same results and find that much easier.

You are free to find a workflow easier. But it still means you go through way more steps to achieve that result than I do (which is also fine by me). The downsides you see is that you can not be 100% sure you capture an event due the limit of the captured time interval. I don't care about that (which some people may find weird or even offensive; I'm a 'the glass is half full' person). In some cases I have no idea of the time scale anyway. I could spend half an hour to try and look it up or spend 5 seconds to poke with a probe. If the event isn't there then the measurement at least showed the time scale of the signal and there are alternative ways available to capture a specific event so no worries and no time lost. But when the event is in the memory then it is a massive saving in time & effort because I did two steps in one: analyse the problem and found the cause. Or fixed the last bug and verified I introduced no new bugs in one go. For most of the stuff I work on 10Mpts of memory is deep enough to cover 9 out of 10 cases but more memory is always better. In the end the result counts; the journey not so much.

[Elasia]

To me convoluted process is to set memory depth, mentally calculating what timescale it will be at target sample rate, mentally calculate how much of that I have to set as pretrigger to make sure I have enough pre trigger capture once I'm at the target timebase. Then I set target timebase, tightly set around trigger and then keep capturing until I find something interesting, and then, without any sense where exactly in the whole buffer I am i keep twiddling around with timebase looking at data searching for something interesting.

That is not how it works! If you don't get it then you don't get it but by riduculing other people's workflow you only make a fool out of yourself.

It is exactly how it works. Unlike you, I like to have a guarantee I will capture something before I start doing it. You are doing this not because it is smart thing to do it, but because it is the lazy way, and you hate zoom mode. So you do it this way. If it works for you , that's nice.
I do it in opposite direction and get same results and find that much easier.
But both you, and now Dave apparently, try to sell this as some panacea, and will do that without any negative impact. Which is not true. It has consequences.
In my mind it is not that Siglent manages memory wrong way, but that zoom mode should be implemented to be more user friendly.  I LIKE how zoom mode has DETERMINISTIC dual time bases. They should shuffle windows differently and make them more configurable, but it is a display and user interface  problem, not capture problem.  This is exactly what R&S and Keysight does: it sets timebase (internaly) for long capture, and then shows you only part of captured waveform. It is just U/I that shows you part of waveform, using basically "virtual timebase".

And I get it. I read what you wrote on this topic before. 

By the by, it is you who is proponent of idea that whoever is not agreeing with you is stupid. And this is not first topic you had that condescending tone..
As I said before, you are wonderful engineer with waste experience. I respect that. And I agreed and disagreed with you on many topics. And I will keep doing that at my own conscience.

I agree its just a UI problem for the siglent anyway now that i figured out how to use the thing lol

I do like the configurables but they need a couple more and a way to make it work in a more natural mode by just turning the knob out.. if you could hide the preview bar and they put the zoom timebase below with the other timebase data, who could tell the difference?

[2N3055]

You are free to find a workflow easier. But it still means you go through way more steps to achieve that result than I do (which is also fine by me). ...

It's same amount of steps. You Zoom Out,  I Zoom In.

And on Picoscope with big screen and mouse scroll wheel it is so intuitive I don't even think about it.
That is for debug work similar to what you explained in your original post.

For interactive work ( what I call "scoping around") I use Keysight and twiddle knobs all the time. I will twiddle timebase in and out, and will be in Auto mode. Working similar to how you would with analog scope. Then I will stumble upon a signal that has frequency of occurring slower than what auto mode is auto triggering and  I want it to stay on screen between triggers so triggering is stable. It might be something interesting so I press Single for it to capture it to Stop so I can take a look at what it is. This is the point where I can :

1 Change timebase to be longer to make sure to capture all of it (twiddle knob)
2. Arm for single and wait for trigger (press button)
3. Change timebase and horizontal position to look at signal in detail (like a full screen zoom in/out)

Or I can do this:

1. Go into scope setting and set acquisition memory for this (enter number of keystrokes and twiddles for your scope)
2. Arm for single and wait for trigger (press button)
3. Change timebase and horizontal position to look at signal in detail (like a full screen zoom in/out)
4. Once you move form this point go into scope setting and set acquisition memory for general work.

There is no even need for zoom mode. Zoom mode will add one more step to first procedure, and it will use some screen. OTOH it will allow you to have overview over where in the buffer are you, which can be confusing when you have zoom factors of 10000x. So with zoom mode it is same number of steps..

[Martin72]

Yes, on the Siglent you literally change the memory depth on the menu and the memory depth shown in the bottom sample window doesn't change. It's insane.

That´s not completely correct…

In the menu, you could change the memdepth to 10k, 100k, 100M and 200M, so far so good.
But it will ony change in real (bottom sample window) when the timebase is in "the range".
I had created a little table:



So, when you are in a timebase that "allows" 200Mpts memory, you could choose want yout want and it will be used (and displayed).
For example, when you are in a timebase where only 40Mpts are "possible", you could choose 100K or 10K, but you got no effect when choosing 100Mpts or 200Mpts.
So in my opinion, the selectable memory size is "semi-manual"....
When it´s not a big deal to change it in the firmware, siglent could let the user decide to have always the full memory control or change to auto mode.
And all will be happy...

[maginnovision]

I think we already established it's a very limited(4 or 5 options?) memory LIMIT. Not a depth selection.

[2N3055]

Yes, on the Siglent you literally change the memory depth on the menu and the memory depth shown in the bottom sample window doesn't change. It's insane.

That´s not completely correct…

In the menu, you could change the memdepth to 10k, 100k, 100M and 200M, so far so good.
But it will ony change in real (bottom sample window) when the timebase is in "the range".
I had created a little table:

(Attachment Link)

So, when you are in a timebase that "allows" 200Mpts memory, you could choose want yout want and it will be used (and displayed).
For example, when you are in a timebase where only 40Mpts are "possible", you could choose 100K or 10K, but you got no effect when choosing 100Mpts or 200Mpts.
So in my opinion, the selectable memory size is "semi-manual"....
When it´s not a big deal to change it in the firmware, siglent could let the user decide to have always the full memory control or change to auto mode.
And all will be happy...

Thanks for this Martin.
That confirms my statement the setting is "auto mode max memory setting", meaning it will autorange memory, trying to keep sample rate highest possible until timebase becomes slow enough that it will limit memory and start slowing down sampling. That is exactly how Pico handles memory.
I also agree that if possible Siglent should enable user sacrifice history mode, and use full length sampling if user for some reason wants that . Making existing History mode they did was more complicated than just making it simply stupidly sample fixed amount.

[Martin72]

I think we already established it's a very limited(4 or 5 options?) memory LIMIT. Not a depth selection.

Sounds logically to me...

[nctnico]

You are free to find a workflow easier. But it still means you go through way more steps to achieve that result than I do (which is also fine by me). ...

It's same amount of steps. You Zoom Out,  I Zoom In.

And on Picoscope with big screen and mouse scroll wheel it is so intuitive I don't even think about it.
That is for debug work similar to what you explained in your original post.

For interactive work ( what I call "scoping around") I use Keysight and twiddle knobs all the time. I will twiddle timebase in and out, and will be in Auto mode. Working similar to how you would with analog scope. Then I will stumble upon a signal that has frequency of occurring slower than what auto mode is auto triggering and  I want it to stay on screen between triggers so triggering is stable. It might be something interesting so I press Single for it to capture it to Stop so I can take a look at what it is. This is the point where I can :

1 Change timebase to be longer to make sure to capture all of it (twiddle knob)
2. Arm for single and wait for trigger (press button)
3. Change timebase and horizontal position to look at signal in detail (like a full screen zoom in/out)

Or I can do this:

1. Go into scope setting and set acquisition memory for this (enter number of keystrokes and twiddles for your scope)
2. Arm for single and wait for trigger (press button)
3. Change timebase and horizontal position to look at signal in detail (like a full screen zoom in/out)
4. Once you move form this point go into scope setting and set acquisition memory for general work.

There is no even need for zoom mode. Zoom mode will add one more step to first procedure, and it will use some screen. OTOH it will allow you to have overview over where in the buffer are you, which can be confusing when you have zoom factors of 10000x. So with zoom mode it is same number of steps..

No. A common procedure for me is to check a detail of a bus decoded message or trigger on a very specific part of a signal. When zoomed out too far, I won't be able to inspect the detail. Depending on the results I zoom out if necessary. That use case works much faster when the scope just uses the amount of memory I selected instead of needing zoom. Your method would require to reset the time base to a low time/div, do the acquisition, then zoom in to see the result and repeat again. Doing this 10 times in a row gets frustrating quickly and the scope gets in the way of being productive. With my method you can just leave the scope at whatever setting and zoom in/out without needing to reset. It really is easier because I only need to use two knobs on the scope: horizontal position and time base and their exact setting is less critical.

[2N3055]

You are free to find a workflow easier. But it still means you go through way more steps to achieve that result than I do (which is also fine by me). ...

It's same amount of steps. You Zoom Out,  I Zoom In.

And on Picoscope with big screen and mouse scroll wheel it is so intuitive I don't even think about it.
That is for debug work similar to what you explained in your original post.

For interactive work ( what I call "scoping around") I use Keysight and twiddle knobs all the time. I will twiddle timebase in and out, and will be in Auto mode. Working similar to how you would with analog scope. Then I will stumble upon a signal that has frequency of occurring slower than what auto mode is auto triggering and  I want it to stay on screen between triggers so triggering is stable. It might be something interesting so I press Single for it to capture it to Stop so I can take a look at what it is. This is the point where I can :

1 Change timebase to be longer to make sure to capture all of it (twiddle knob)
2. Arm for single and wait for trigger (press button)
3. Change timebase and horizontal position to look at signal in detail (like a full screen zoom in/out)

Or I can do this:

1. Go into scope setting and set acquisition memory for this (enter number of keystrokes and twiddles for your scope)
2. Arm for single and wait for trigger (press button)
3. Change timebase and horizontal position to look at signal in detail (like a full screen zoom in/out)
4. Once you move form this point go into scope setting and set acquisition memory for general work.

There is no even need for zoom mode. Zoom mode will add one more step to first procedure, and it will use some screen. OTOH it will allow you to have overview over where in the buffer are you, which can be confusing when you have zoom factors of 10000x. So with zoom mode it is same number of steps..

No. A common procedure for me is to check a detail of a bus decoded message or trigger on a very specific part of a signal. When zoomed out too far, I won't be able to inspect the detail. Depending on the results I zoom out if necessary. That use case works much faster when the scope just uses the amount of memory I selected instead of needing zoom. Your method would require to reset the time base to a low time/div, do the acquisition, then zoom in to see the result and repeat again. Doing this 10 times in a row gets frustrating quickly and the scope gets in the way of being productive. With my method you can just leave the scope at whatever setting and zoom in/out without needing to reset. It really is easier because I only need to use two knobs on the scope: horizontal position and time base and their exact setting is less critical.

For that case I do exactly same: I set timebase to long, zoom in to detail and keep retigering. On each retrigger i see my detail and get full buffer.... List shows all decode from full buffer, detail shows detail decode underneath.

[Elasia]

No. A common procedure for me is to check a detail of a bus decoded message or trigger on a very specific part of a signal. When zoomed out too far, I won't be able to inspect the detail. Depending on the results I zoom out if necessary. That use case works much faster when the scope just uses the amount of memory I selected instead of needing zoom. Your method would require to reset the time base to a low time/div, do the acquisition, then zoom in to see the result and repeat again. Doing this 10 times in a row gets frustrating quickly and the scope gets in the way of being productive. With my method you can just leave the scope at whatever setting and zoom in/out without needing to reset. It really is easier because I only need to use two knobs on the scope: horizontal position and time base and their exact setting is less critical.

This is exactly how the siglent works now when setup properly.. its pretty nice, i just let it sit there and auto trigger and it marches along

[EEVblog]

That confirms my statement the setting is "auto mode max memory setting", meaning it will autorange memory, trying to keep sample rate highest possible until timebase becomes slow enough that it will limit memory and start slowing down sampling. That is exactly how Pico handles memory.

Then damn well call it Auto then.
Most other scopes have Auto and manual, but it appears the Siglent only has Auto, and that "Auto" mode will never use maximum memory below a certain time base setting.
It's completely inflexible for no reason.

I also agree that if possible Siglent should enable user sacrifice history mode, and use full length sampling if user for some reason wants that . Making existing History mode they did was more complicated than just making it simply stupidly sample fixed amount.

Yes, they went to the effort to implement it this way, very deliberately sacrificing manual control over memory at a huge number of timebase settings. No excuse for this. There is zero downside to not allowing users to select the memory depth they want at all timebase settings.

[tautech]

We look forward to your followup video Dave when you discover the capabilities of 2 obscure buttons on the Siglent front panel.   

Tautech, why not show the steps to achieve this without cheekiness? Remember that people that come across this thread may be trying to make a purchase decision - since they will not have a Siglent in front of them to explore the mysterious buttons, they will go elsewhere.

I urge you refresh your memory with study of the first 3 pages in the SDS2kX Plus thread.
Maybe the reasoning behind Siglents acquisition strategy will become clear. 

[edigi]

I urge you refresh your memory with study of the first 3 pages in the SDS2kX Plus thread.
Maybe the reasoning behind Siglents acquisition strategy will become clear.

Care to make a quick summary?

DSO is a toolbox, and having one more tool in the box (be it zooming out or something else) can become handy sometimes and can shorten the workflow even if you typically don't need it. Mostly in single mode, where hard to repeat signals are checked and where DSOs excel.
Although Dave captured one possible design point that influenced the current behaviour with the always on history mode but my feeling is that we're still probably far from understanding the depth of this choice (as I guess few people were involved  in DSO design from here).
For example if you can zoom out what else you can do with the more data then just looking on it? Can you re-apply FFT (or generally math) or will there be a gated FFT (I don't know if it's even possible today apart from the implicit power of 2 gating)?
How does this work with DSOs that can zoom out (I guess there can be vastly different design choices and end results...) Is it only about that you can look the data and take measurements with it (which become gated if zoomed out or re-applied?) or do more as well?
OK, last few questions are mainly for those who can check how the zoom out mode works in those scopes that support it but still a lot of unexplored questions...

[Elasia]

We look forward to your followup video Dave when you discover the capabilities of 2 obscure buttons on the Siglent front panel.   

Tautech, why not show the steps to achieve this without cheekiness? Remember that people that come across this thread may be trying to make a purchase decision - since they will not have a Siglent in front of them to explore the mysterious buttons, they will go elsewhere.

I urge you refresh your memory with study of the first 3 pages in the SDS2kX Plus thread.
Maybe the reasoning behind Siglents acquisition strategy will become clear.

He does have a point... I had no clue zoom worked like that even after reading that originally before i picked up the scope and then forgot about it. It's not automatically natural/instinctual to setup and thats why it flummoxes a number of people and to a large degree they have a valid point.  The scope does indeed allow timebase expansion already even exactly how nctnico works and is fanatic about / how Dave expects it to work as well so there is no reason they cant also make it an option to default work in that manor if someone wants it to WITHOUT selecting zoom.. in fact i'd encourage you to give them that direct feedback as it is something i'd like to see as well even after mastering zooming.

The ability to do that is just a matter of hiding the zoom ui components and reduce all of it into a timebase/memory selector based on the same table Martin72 posted.. they dont need to get fancy about it and shouldnt take one of their programmers no  more than a couple of hours to tinker around with the UI

This is only a UI interface problem and NOT with the base functions of the scope itself

[nctnico]

I urge you refresh your memory with study of the first 3 pages in the SDS2kX Plus thread.
Maybe the reasoning behind Siglents acquisition strategy will become clear.

Care to make a quick summary?

DSO is a toolbox, and having one more tool in the box (be it zooming out or something else) can become handy sometimes and can shorten the workflow even if you typically don't need it. Mostly in single mode, where hard to repeat signals are checked and where DSOs excel.
Although Dave captured one possible design point that influenced the current behaviour with the always on history mode but my feeling is that we're still probably far from understanding the depth of this choice (as I guess few people were involved  in DSO design from here).
For example if you can zoom out what else you can do with the more data then just looking on it? Can you re-apply FFT (or generally math) or will there be a gated FFT (I don't know if it's even possible today apart from the implicit power of 2 gating)?
How does this work with DSOs that can zoom out (I guess there can be vastly different design choices and end results...) Is it only about that you can look the data and take measurements with it (which become gated if zoomed out or re-applied?) or do more as well?
OK, last few questions are mainly for those who can check how the zoom out mode works in those scopes that support it but still a lot of unexplored questions...

In general FFT follows what is on screen or can be gated (R&S for example). Math and measurements usually follows the memory depth but it can be gated on some scopes where math can be based on actually acquired data (Tektronix for example) or decimated data (Keysigh and R&S for example).

See this video I recorded on a MicSig TO1000 for example where I scroll through a 14Mpt record with FFT and some measurements enabled (you have to click on it in order to view it on Flickr):

[tautech]

Announcement from another thread:

Let's see how Siglent reacts to this video and the one showing the comparison to their competitors. Perhaps there has been a conversation between the CEO and the R&D department already:

CEO: Why are our oscilloscopes different compared to all of the competitors?
R&D: The people from Lecroy said...
CEO: Don't care about what Lecroy said. Dave made us look stupid! And dumb! And more stupid. How much time to fix?
R&D: A couple of hours tops I guess.
CEO: Make it so! New firmware by the end of the week!
R&D: OK boss.

Alternative:
~1 month ago resulting from beta tester emails:

CEO: We need allow for capture zoom out.
R&D: Why ? Pico and LeCroy don't support it.
CEO: Apparently some can't embrace the capture analysis tools we offer so we need to make it much easier for them.
R&D: OK we will look into it but we really should be fixing the few bugs already reported.
CEO: Yes, OK then get onto it when you can.

Yes you heard it here first !
ETA, unknown.

[Martin72]

The beauty of dso is, you can change nearly everything in firmware.
So, in case of siglent, why don´t change the memory management, letting the user allow to choose for himself always the maximum memory - or choosing it to auto.