Pocket 4K Aperture Value Issues

[BTM_Pix]

I'm finding what might be described as "interesting" behaviour when changing aperture values.

Aside from the unusually high degree of granularity (there are 66 different positions between f1.3 and f22 for example), I'm finding that there are a number of repeats within these positions.

For example, there are two f1.3s, so when you have it on f1.3 then step up to the next value it is also f1.3.

This is repeated at several other positions and is the same, for me at least, on both native MFT and adapted lenses.

At some positions, there is no difference in exposure (so if you move from the "first" f11 to the "second" f11 it appears the same) but worryingly that is not always the case.

On a Canon nifty fifty on a speedbooster for example there is a very clear exposure difference between the two f1.3 settings.

The quirkiness of it having that many steps and some repeats is something I can live with if it was consistent between lenses and was only an indication foible in terms of the repeats but its neither of those things unfortunately.

I have an understanding of why it is happening based on the BM protocol and messages I'm reading back from the camera over bluetooth but I have a limited range of lenses so obviously, it would be appreciated if any of you could check behaviour of a couple of lenses and see what happens with yours and I'd be particularly interested if anyone can try it with lenses with a wider aperture than f1.3 to see if there is a similar issue there.

[Anaconda_]

I just tested with my nifty fifty and have the same thing.

If I'm on f1.3, and try to stop down a notch, I hear the lens tick, but the image remains the same and it still says f1.3. At f.11 though, I stop down a notch and the image gets darker but it still reads f.11.

This isn't the case with 24-105 though.

When I get home, I'll try my Sigma 30mm 1.4 (f1.0 with viltrox) and see what happens. Interesting find, but don't see it have too much impact on daily shooting.

 

[BTM_Pix]

6 minutes ago, Anaconda_ said:

I just tested with my nifty fifty and have the same thing.

If I'm on f1.3, and try to stop down a notch, I hear the lens tick, but the image remains the same and it still says f1.3. At f.11 though, I stop down a notch and the image gets darker but it still reads f.11.

This isn't the case with 24-105 though.

When I get home, I'll try my Sigma 30mm 1.4 (f1.0 with viltrox) and see what happens. 

Thanks

These are the histograms from my "two" f1.3 settings on the nifty fifty and I can definitely see from the shift that they are two distinct settings on mine at least.

 

7 minutes ago, Anaconda_ said:

Interesting find, but don't see it have too much impact on daily shooting.

I know what you mean but I'm not sure its ideal that depending on which direction you approach some particular f stops will decide whether you get the "real" setting or the one that doesn't do anything, particularly if you are exposing via an external light meter?

And that even that inconsistency is in itself inconsistent depending on which lens you are using.

In the case of my lens above, I had a little bit extra in hand when I went down to the "second" f1.3 that I wouldn't have been aware of if I'd not gone past the first one.

It also has implications for anyone trying to match more than one camera as well in that you'd have to agree that both were on the same version of the f stop.

It also causes sync issues with their own app where to add to the fun you can have one setting displayed on the app and another on the camera, which again has implications for anyone such as a gimbal manufacturer making a viable remote control for it.

The root of the problem is that the camera doesn't profile the lens so the interpretation by the lens of the value that they use for the aperture ends up being a bit of a crap shoot depending on whether the lens actually has a physical setting that matches it.

[Anaconda_]

I can see it being an issue with the app, but for multi camera shoots, the difference will be minimal. But yeah, I see what you're saying and agree its worth thinking about and knowing before setting up and pressing go.

On my 30mm f1.4(f1.0) I get a tick but no value change at f1.0/.1/.2/.3/.4 - So each value essentially has two exposures. And yes, now I look at it more closely, there is a slight difference to the picture with each tick.

And lastly it does appear on the 24-105, but only on f11.

[BTM_Pix]

Thanks for checking.

I suppose its at the wide end of the aperture that its going to have to be borne in mind as there may be a bit more there that people won't discover - as with the nifty fifty - unless they go one step further than what it is reporting as the maximum aperture.

So the rule of thumb is that when you think you've reached the maximum, turn it one more step.

It a bit of a reverse Nigel Tufnel from Spinal Tap really as these go to -1 rather than 11.

 

[zerocool22]

Well its a cinema camera after all, so its main use is with manual lenses. So I hope they rather use the time to include anamorphic desqueeze then testing and update every camera lens out there. 

[Papiskokuji]

I also noticed a weird aperture behavior with the 24-105mm and the Viltrox EF-M2. I didn't notice it when I was using the lens with the latest firmware of the adapter (although it could have already been there!). Then I downgraded the firmware to get IS to work (nice find btw) and that's when I noticed it. Especially at the widest aperture, I had two different steps, showing the same F stop but different light gathering. And if I turned the dial too fast, I couldn't fully open the aperture. I had to go backward and close the aperture a bit and the turn the dial more slowly to get the lens to go wide open... Weird... I thought it was an issue with the adapter, but apparently it is also present with native mft lenses as BTM_Pix said. It would be interesting to know if it occurs with every lens controlled with electronics.

It reminds me of a bug where switching lens IS on and off with the BMPCC did change the amount of light on the sensor. The screen showed the same f stop and same iso, etc. but the image would go darker... I think it changed the aperture. There was a work around, you had to turn IS off and then on, only when the camera is on. Don't even know if they fixed it and if it was metabones adapter's fault or the camera's.

[webrunner5]

8 minutes ago, zerocool22 said:

Well its a cinema camera after all, so its main use is with manual lenses. So I hope they rather use the time to include anamorphic desqueeze then testing and update every camera lens out there. 

Yeah but doesn't every electronic lens generate Metadata? If so why would the body not be able to interrupt it correctly? Surely every say Canon lens has the very same protocol?

[BTM_Pix]

1 hour ago, zerocool22 said:

Well its a cinema camera after all, so its main use is with manual lenses. So I hope they rather use the time to include anamorphic desqueeze then testing and update every camera lens out there. 

What do you mean 'update every camera lens out there' ?

They just have to profile them to make sure they go to where they should go to.

Tell you what, how about they compromise and just profile the ones (80% of which it will affect) they use in their advertising for the camera ?

I'm not being funny but if any other camera company had an issue where they had inaccurate aperture control like this (which isn't just a cosmetic issue either) there would be an absolute shitstorm on here but weirdly, as with a few other issues, BM somehow seem to get a pass.

 Anyway, thanks for those with the cameras that have chimed in to confirm it on their lenses and I suppose just view it as an FYI for everyone who hasn't yet got one.

 

[Anaconda_]

59 minutes ago, Papiskokuji said:

Then I downgraded the firmware to get IS to work

Are there any negative aspects to this? 

[Papiskokuji]

2 minutes ago, Anaconda_ said:

Are there any negative aspects to this? 

None so far !

Yes, it is almost accepted first BM buyers are like beta testers. But it’s actually quite puzzling. Even though I have sympathy for this brand, it’s not professional to release a camera not finished. It reminds me of my first minutes with my RX100 V. I was only browsing menus and then I had this overheating warning almost right away during my first clip... Shocking !

[zerocool22]

59 minutes ago, BTM_Pix said:

What do you mean 'update every camera lens out there' ?

They just have to profile them to make sure they go to where they should go to.

Tell you what, how about they compromise and just profile the ones (80% of which it will affect) they use in their advertising for the camera ?

I'm not being funny but if any other camera company had an issue where they had inaccurate aperture control like this (which isn't just a cosmetic issue either) there would be an absolute shitstorm on here but weirdly, as with a few other issues, BM somehow seem to get a pass.

 Anyway, thanks for those with the cameras that have chimed in to confirm it on their lenses and I suppose just view it as an FYI for everyone who hasn't yet got one.

 

Yeah I get you, it might be annoying. But the camera is dirt cheap. Even RED had this problem with lenses with cameras over 30 times the cost. Allthough I don't see any topics on that topic, probably because they are using cinema lenses. Yeah BM should not have used these lenses in the advertising, in my opinion they should have thrown Master primes or cookes S4 on it. But they obviously want to target a larger market by doing this and as a result get more complaints. 

[BTM_Pix]

2 hours ago, zerocool22 said:

Yeah I get you, it might be annoying. But the camera is dirt cheap. Even RED had this problem with lenses with cameras over 30 times the cost. Allthough I don't see any topics on that topic, probably because they are using cinema lenses. 

I'm going to go out on a limb and say that there aren't many topics on here about anything to do with RED because not many people on here own them.

Oddly enough, I do though and have had an EPIC X for seven years and I'm not sure I follow when you say its got the same problem ?

It reports those same Canon lenses just fine on its adapter and reflects the actual aperture values (rather than the 64 'send it and we'll let the lens work it out' routine) correctly for the same lenses, with no repeats.

It also reports the aperture back correctly from the suitably equipped PL lenses I've got as well so its not an issue I have come across or can reproduce.

2 hours ago, zerocool22 said:

Yeah BM should not have used these lenses in the advertising,

But they obviously want to target a larger market by doing this and as a result get more complaints. 

The design of the product has an MFT mount integral to it though so it was obvious that they would use a selection of those lenses.

What they shouldn't have done is say that those lenses are fully compatible because they patently aren't.

That's got nothing to do with it being 'dirt cheap' or who they are selling it to.

Anyhoo, for what it's worth, my Android remote app fixes the problem so that it will only send the values that each particular lens has a valid matching aperture for.

So if a low rent whining prick like me can resolve it then hopefully the $400M corporation can do the same at some point.

[zerocool22]

8 hours ago, BTM_Pix said:

I'm going to go out on a limb and say that there aren't many topics on here about anything to do with RED because not many people on here own them.

Oddly enough, I do though and have had an EPIC X for seven years and I'm not sure I follow when you say its got the same problem ?

It reports those same Canon lenses just fine on its adapter and reflects the actual aperture values (rather than the 64 'send it and we'll let the lens work it out' routine) correctly for the same lenses, with no repeats.

It also reports the aperture back correctly from the suitably equipped PL lenses I've got as well so its not an issue I have come across or can reproduce.

The design of the product has an MFT mount integral to it though so it was obvious that they would use a selection of those lenses.

What they shouldn't have done is say that those lenses are fully compatible because they patently aren't.

That's got nothing to do with it being 'dirt cheap' or who they are selling it to.

Anyhoo, for what it's worth, my Android remote app fixes the problem so that it will only send the values that each particular lens has a valid matching aperture for.

So if a low rent whining prick like me can resolve it then hopefully the $400M corporation can do the same at some point.

Yeah Im sure they will fix it, but might be a good idea to create a support ticket at bm for this. I just wish I got mine already. RED fixed it in a firmware update. 

[Anaconda_]

14 hours ago, Papiskokuji said:

None so far !

Which firmware do you downgrade to? I've tried all that are available on the Viltrox site, but the Image Stabalization option on the P4k is always greyed out. 

I also noticed they released a new firmware 3.2 a couple weeks back, but that still doesn't change the IS option.

[Papiskokuji]

6 hours ago, Anaconda_ said:

Which firmware do you downgrade to? I've tried all that are available on the Viltrox site, but the Image Stabalization option on the P4k is always greyed out. 

I also noticed they released a new firmware 3.2 a couple weeks back, but that still doesn't change the IS option.

The oldest available. 2.3 if i remember correctly. IS is still greyed out but if you turn on IS on the lens, you can hear it work and stop when turned off.

[CaptainHook]

What lenses is this happening with?

I've scanned the thread but every mention seems to indicate a focal reducer of some kind which may actually be the issue, rather than direct control of an MFT lens.

[Papiskokuji]

1 hour ago, CaptainHook said:

What lenses is this happening with?

I've scanned the thread but every mention seems to indicate a focal reducer of some kind which may actually be the issue, rather than direct control of an MFT lens.

I don’t own native lenses. It was with the canon 24-105 and the Viltrox EF-M2. So with an adapter indeed. BTM_Pix seemed to have the same issue with mft glass though ?

[BTM_Pix]

6 hours ago, CaptainHook said:

What lenses is this happening with?

I've scanned the thread but every mention seems to indicate a focal reducer of some kind which may actually be the issue, rather than direct control of an MFT lens.

As per the orginal post it is happening with native (both Panasonic and Sigma that I've tested) as well as adapted lenses.

I'll be back from this job later this evening so I can have a look at an Olympus lens for completeness but I'm not expecting it to be any different.

If you connect to the camera with something like nRF Connect and register for a callback on UUID: B864E140-76A0-416A-BF30-5876504537D9 and turn the aperture wheel (or use the app) you will see the values for parameter 2 of group 0 change on every increment and should then see the instances where the value has changed but the indicated f stop hasn't.

 

[CaptainHook]

If you can let me know exact MFT lens models and describe the issue in repo steps I can follow it up. There is one known issue at the moment that some times the lens won't set the correct aperture we tell it to, but for the aperture number to stay the same on the camera UI display but the lens actually change sounds like something different. Will look into what you mention nRF Connect too.