Why don't more brands have something like IBIS? After more than 2 years I'm still awed every time I take an indoor shot with my E-M5, even of people, at 1/15 of a second ISO 200, and get great results. Sony's is moving away from the Alpha serious which does have it and toward the e-mount which doesn't. Pentax still has it and I respect them for it. But Fuji, Nikon, Canon, Panasonic, not on the horizon.
Even if the Fuji X-T1 is an amazing camera, but who cares if one is forced to take photos at 2 or 3 f stops higher ISO than I'd have to with the OMD series. I'd rather have my E-M5 or if I was going to earn more money from photography, buy an E-M1. I'm aware that IBIS doesn't matter with say photography of football at night or race cars, requiring high shutter speeds, but it helps in more situations than I expected.
The reason Panny won't abandon OIS is because they're a much more video oriented company than Oly. And IBIS sucks for video.
That's a typo, right? I am sure you meant to say that OIS is worse for video (as it is by comparison, at least as currently implemented).
Wrong. If we were to compare just OIS and IBIS, OIS would be better.
Please let us all know how you did the comparison. And if you didn't, how would you know which is better?
Oly does a combo deal, adding DIS to their IBIS. As far as I know, there's never been any Oly implementation of video IBIS that hasn't also included DIS.
If Oly uses a combination of DIS and IBIS for video (as seems likely, see below), what's wrong with that as long as it produces good results.
Nothing wrong with it, but they never put it in the specs for the E-M5 and E-M1, as far as I know. And with no option to only use IBIS, we can't accurately compare the two.
We can accurately compare two technical solutions, IBIS+DIS versus OIS, and the former wins.
Furthermore, we don't know what relative contribution the IBIS gives and the DIS gives.
It would be nice to have the option to just use IBIS.
Why is that if IBIS+DIS gives better results?
Also, it's important to note that DIS can be done in post with a slight reduction in sharpness and possibly with better results, given the difference in computing power you have in real time vs with your computer. It's also something that can be added to OIS if the camera body supports it.
Well, at present it doesn't.
There's nothing inherent in IBIS that makes DIS easier to implement.
Neither of us knows. But it remains a fact that it is implemented with IBIS but not with OIS.
I realize that some people may believe that Oly IBIS is buttery smooth for video,
Well, regardless of how it is achieved, Oly stabilization actually is "buttery smooth" for video.
Using methods that introduce artifacts into the video.
What artifacts specifically due to DIS are you talking about?
Digital stabilization is known to introduce artifacts, whether in real time or in post, because it's simply a software analysis of the motion. Not to mention the fact that you're doing the crop in real time.
The fact that it is software- rather than hardware-based doesn't necessarily imply that it introduces artifacts. I asked you specifically what artifacts you are talking about. You dodged the question.
but what they don't realize is that Oly has implemented a form of DIS on every MFT camera where they've done IBIS for video. So, when you see Oly's IBIS in action on YouTube, you're seeing a combination of IBIS and DIS. Note that I don't believe this is done for stills, as far as I know.
What makes you think that it is a combination of DIS and IBIS?
Even if Oly is using DIS with IBIS, why do I say it sucks, you might ask. Because IBIS is a much greater energy hog than OIS and will suck much more power when used continuously.
Do you have any evidence of that or is it just speculation on your part?
Yes. People that I've talked to that have used Oly cameras for video with IBIS have always told me that the video is cropped when IBIS is activated (as compared to no IBIS). If I'm wrong then let me know.
Cropped video to achieve stabilization means that DIS is being used. Not maybe but definitely.
Yes, video with stabilization enabled is cropped slightly, which suggests that sensor-shift is supplemented by DIS. This is also what Imaging-Resource reports in their review of the E-M10:
"For video recording, Olympus has also included a hybrid IS system utilizing both sensor-shift and digital image stabilization"
http://www.imaging-resource.com/PRODS/olympus-e-m10/olympus-e-m10A.HTM
This is precisely the reason IBIS doesn't run all the time (only when the shutter is pressed) while OIS does.
It's a good idea to turn things off when they are not required. Why doesn't Pany do the same?
Why no option to keep it running all the time? Very strange if you ask me. I've theorized that it's either due to excessive power consumption or possibly heat generation.
Why would you want to run it all the time even if not required? Recent implementations (E-M1, E-M10) do in fact run it all the time when in magnified view since in that case it is indeed useful.
Because, who wants to look at a shaky image when composing a shot. This is especially true of telephoto lenses. It's distracting and annoying.
That's not an answer to the question I asked. The question was: Why would you want to run it all the time even if not required? When required, it either is on (because of magnified view or video mode) or can easily be turned on (by half-pressing the shutter button). Consequently, I never have to look at a shaky image when composing a shot with IBIS.
And if Panasonic feels that turning on/off OIS when you press the shutter doesn't add anything, who am I to argue?
And if Olympus feels that keeping IBIS on even when not required doesn't add anything, who am I to argue?
Perhaps in the final analysis, they found no benefit. From a consumer perspective, you certainly can't argue that this gives you any added benefit when shooting.
According to Panasonic themselves, keeping it off until the time of exposure (when they allowed that, as they did prior to the G3) increases the effectiveness of the system during exposure by 20 percent (see point 12 in the FAQ list)
http://shop.panasonic.com/shop/support-digital-cameras?support=y
the reason being that the OIS lens group is centered at the beginning of the exposure and therefore has more leeway to move in all directions.
The only benefit would be if the MFR felt it was more efficient or saved power.
With respect to efficiency, see above.
And it seems clear Panasonic felt it didn't.
So are you saying that Panasonic never makes any mistake?
I certainly see very little power savings with OIS off vs on.
And you still have not given an explanation of why Oly doesn't give the option to keep it running all the time under any circumstance you want. You certainly can't argue that this would not be beneficial.
I am certainly arguing that it wouldn't be beneficial. What would be the benefit of keeping it running even when not required?
If you do, it's a clear demonstration of bias on your part.
Not unless you can show that it is beneficial to keep it running when not required, which you evidently can't.
The fact that you would fault Panny for not giving the option to turn on OIS only when the shutter is pressed yet you don't fault Oly for not giving the option to turn on IBIS all the time is a clear indication of bias, and you'll have a hard time convincing me or any objective reader otherwise.
See above.
And I stand by my assertion that there is probably some technological reason behind it.
For example the desire not to waste power, which is a good thing, isn't it?
There's simply no plausible reason not to give the user the option of leaving it on continuously unless there were a technological limitation.
It does run continuously when that's beneficial. See above.
Uh-huh.
So this is all you can muster in the way of argument?
And this jives with my understanding of IBIS and the energy requirements of moving the sensor (vs. moving a lens element).
And why would Panny give the option of only activating OIS when you press the shutter. Sounds ridiculous to request it to me, as I see no benefit to it (given the minimal power requirements of OIS). Panny gives you the option of turning on/off the OIS.
What are the OIS and IBIS power requirements? Got any facts and figures?
No. They don't release those and it's very difficult to measure.
Fine that you recognize that we have no information on that.
And it's very probable that Oly uses a less energy hogging form of IBIS when video is engaged (hence the reports of less noise from IBIS in video than stills mode).
Yes, they do. Less energy is required for video stabilization than for stills since the objective is different. With stills, you want each frame as sharp as possibles. Hence IBIS has to move fast. With video, the primary objective is to eliminate the jerkiness between frames. Hence IBIS can move at a more leisurely pace.
Sounds dubious to me. Some would say grasping at straws. You want every frame as sharp as possible with video as well.
Actually you don't inasmuch as that sometimes conflicts with the desire to eliminate jerkiness across frames, which is of greater importance perceptually. But possibly, Oly's combination of sensor-shift and DIS helps them accomplish both.
Of course it helps with the shake, but DIS is known to negatively impact image quality in other ways.
In specifically which ways is DIS known to negatively impact image quality and where's your evidence?
To see how effective it can be, again, take a look at the Sony action cam, which only uses DIS. You can walk with that and have a buttery smooth image. When something is that smooth, it's often clear that DIS is being used. But, again, it comes at a cost. And it's always nice when you have the opportunity to shut it off and just go with the mechanical stabilization.
And, again, no opportunity to really compare IBIS to OIS at wider angles. However, based on what I've seen, OIS performs better than IBIS at telephoto FL's, even with DIS on Oly cams. DIS works better at wider angles, and it shows in Oly's implementation.
What specifically is it that you have seen that makes you think OIS performs better than IBIS+DIS at telephoto FLs?
Why does IBIS suck so much more power, you might ask. Because the sensor element is often much larger and heavier than the lens element that has to be controlled, hence more power consumed to move it (and also slower and less responsive movements of the element).
Do you have any evidence of that or is it just speculation on your part?
Yes, I have evidence of that. Aside from my sources,
Which are these sources? Links please.
I've actually taken apart some old gear to have a look at it.
Exactly what old gear did you take apart, exactly what did you see and measure, and exactly what conclusions about power requirements can you draw from these observations?
That's brings me to another point.
IBIS is less effective than Power OIS (by about 0.5 to 0.8 stops).
At least some implementations of Power OIS (as opposed to Mega OIS) were, when tested while shutter shock was still a problem. Below you find what I think is the latest list of the results from Lenstip/optyczne.pl that I have compiled:
http://www.dpreview.com/forums/post/52944375
And that difference will likely increase as the focal length is increased. The reasons are two-fold. First, the lens element being controlled will likely be much smaller than the sensor, so it can be moved quicker. This becomes more important as the focal length increases, because hand shake is amplified at greater focal lengths. Also, the compensatory movement of the lens element (OIS) can be magnified on the sensor, thus providing greater shake compensation than the actual motion of the lens element. This obviously can't be done with IBIS.
Here I have some new and interesting facts to offer as a counterpoint to your speculations. As we both know, the real effectiveness of the different stabilization systems couldn't really be tested with longer FLs in the past due to shutter-shock problems. Below 1/250 or 1/320, shutter shock spoiled the fun regardless.
With the electronic first curtain shutter (0-second anti-shock) now available on the E-M1, E-M10, and E-P5, we are finally in a position to do the test and of course I did it. I tried the 100-300 at 300 on my E-M1 and ... my preliminary results (I have still to do more extensive testing) indicate quite clearly that the E-M1 IBIS is significantly better than the 100-300 OIS once the shutter-shock problem is out of the way. At the same shutter speed, the number of keepers is clearly higher with IBIS than with OIS. And the lowest speed worth trying is lower with IBIS as well.
Here's a modest example of what I am talking about. This is a 100-percent crop shot with the Oly 75-300 I am now trying out as a replacement for the Pany 100-300. Hand-held with IBIS at 300 mm and 1/10.
Aside from the fact that I see no actual comparisons here,
None was intended. I just illustrated that IBIS can help me manage some six EV below the 1/EFL rule at 300 mm. I haven't yet managed to accomplish that with OIS.
A single shot proves absolutely nothing.
The crop wasn't meant to prove anything. It was merely meant as an illustration.
And I've come to the conclusion that we need unbiased laboratory tests when it comes to these issues. Otherwise, we get unscientific statements such as I "managed six stops", which is meaningless.
In exactly what way is the statement meaningless? I have yet to manage the same with OIS.
you'll forgive me if I actually wait for an objective comparison.
Unlike you, I have no problem with impartiality. If OIS were better, I'd say so and use it whenever possible. If you think I am trying to hide anything, why would I have compiled the list I linked to above, showing that Power OIS is better than IBIS as long as shutter shock is still in the reckoning? And why would I state, as I have many times over, that, with my E-M5, I prefer the 14-45 OIS over IBIS since the OIS of that particular lens (not others) actually helps counteract shutter shock?
I think you are biased, as I mentioned earlier. The fact that you see nothing wrong with Oly not giving continuous stabilization yet you fault Panny for not giving the option of just turning on OIS when the shutter is pressed is a strong indication of bias in my book.
It's a nonsensical stand as there are clear and unequivocal benefits to continuous stabilization to anyone who would spend a second thinking about it (lack of shake when composing for one). While the thing you fault Panny for, I have yet to see you provide one lick of evidence that it provides any benefit in terms of power or effectiveness. And absent any laboratory tests demonstrating a benefit, there is simply no benefit to a shooter in a general sense from having this. In other words, the thing that you fault Panny for can only demonstrate benefit under laboratory conditions. It's not something that someone would ever put on their wish list, as there's no clear benefit to it (under any circumstance, as far as I can tell).
All of that has already been dealt with above.
Also, it's important to note that the 100-300 is an old lens with mega OIS.
Is there any other Pany lens than this that goes to 300 mm?
I don't know.
Well, the rest of us know that there isn't.
So, why doesn't Panny do IBIS along with OIS. I believe this has everything to do with competition. If Panny were the sole player in MFT, I have a feeling they would. But, for now, I believe they want to protect their lens OIS. And I don't blame them
What exactly do you mean when you say they want to "protect their lens OIS"? What specifically is their objective? And how does abstaining from IBIS help them accomplish that?
If someone is a Panasonic body loyalist and Panny provides in body stabilization for all their cameras, there's less incentive to buy Panasonic lenses.
So you mean that Pany sticks to OIS in order to limit competition on the lens side? That may well be, but it is hardly a convincing argument to those of us who are on the consumer rather than manufacturer side, is it?
No, they stick to OIS because it's a technically better system. But decisions are often financially motivated and, in order to protect their technically better system, they have to make compromises, IMO. Note that this is all conjecture on my part as I'm not sitting in on their boardrooms.
If it were indeed technically better, no financial motivations would be needed.
IMO, OIS and IBIS have their benefits and disadvantages. The only disadvantage of OIS however is the fact that it doesn't work on all lenses. But OIS is a technically better system than IBIS when a lens has it.
Certainly not as judged by my tests with the 100-300 and OIS versus E-M1 IBIS.
For me, being a purist and understanding the benefits of in lens OIS, I would still buy the Panny lenses.
Noone denies you the privilege of buying whatever you want. As a non-purist, I simply prefer to use what is best rather than most pure.
I'm exactly the same. I meant, as a purist, I stick to what I know is best, rather than go with unscientific marketing hype. That's why I stick to OIS when available.
So we simply disagree on what is best. And what is best is a matter of evidence, right?
And as a video shooter and one who knows the disadvantages of IBIS at long focal lengths,
On what basis do you know the disadvantages of IBIS at long focal lengths? Where's your evidence?
I don't want to see MFT go down the wrong path with IBIS only. Without OIS, MFT will be severely handicapped in some areas.
Which areas?
For most of the shooting done on this forums, I'm sure IBIS is a godsend, but the typical user should understand the very limited scope of interest of people on this forum and understand that this is the reason for the bias toward IBIS.
See my question immediately above.
But others would believe that stabilization is stabilization and just go with whatever.
A third option is to believe, as I do, that different stabilization systems are different means to the same end and then choose the best means to that end.
In summary, it's the limited interests (no video, etc.) on this forum that are the reason so many people don't understand why things are done the way they are. No offense.
Although you are right that most people on the forum are more focused on stills than on video, that certainly doesn't prevent them from understanding what's going on with regard to stabilization systems.
You're right.
Thanks!
It's ignorance of other aspects as well.
As well? I didn't say anything about ignorance.
I look forward to your response Anders.
Glad to hear that.
