The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

IQ4 recording at wrong image format, 14 bit instead of 16 bit EX

Steve Hendrix

Well-known member
Sure, raw files size is 109MB, for all the images that I took, they show up as IIQS in C1, and on the camera playback screen. even as the main camera screen as shown in my screen captures shows 16 BitEX.

The true 16 bitEX files I have are all 186 to 200MB, close to 2x the file size.

Going through all of these settings, I remembered that the camera had the sensor plus mode. It shows only 14 bit, I only looked at it once way back, but I though it was available in 16 bit (sensor plus), but I may easily have that wrong.

Until you go back and forth in the image quality settings, all images are taken in the IIQS size.

Paul C

Paul, I don't think the Sensor Plus mode has ever shot anything but 14bit files.


Steve Hendrix/CI
 

dchew

Well-known member
Can I ask you to confirm that the file size of the captured-during-the-but raw files indicate that the bug is the actual bit depth rather than the bug being the mislabeling the metadata field?
Both of the images shown below are the same except for the issue discussed. In Capture One, __21 one is "IIQ S Ex" format, __22 is "IIQ L 16bit Ex" format.

Dave

 

onasj

Active member
I can confirm that this serious bug occurs on my IQ4 with the latest firmware as well. Interestingly, it can only be fixed by selecting one of the 14-bit quality settings and then selecting back the 16-bit quality of choice. If you select 16L or 16ex it will remain bugged until you pick a 14-bit setting.

It’s an especially bad bug because:
1) The back SAYS it’s in a 16-bit mode, but it’s actually in 14-bit S mode
2) The bugged state occurs every time you power up the back
3) There’s no indication that your photos are being captured in the wrong quality setting until you review them (this can be done in camera by turning on the “show details” button in review mode. So there is no doubt that the bug is occurring, even without transferring the files to a computer).

The combination of (1-3) means photographers using this firmware are likely capturing photos in the lowest quality 14-bit mode (without pixel binning) when they don’t intend to. Which is quite a problem!

Regarding the point about the firmware being a beta... ALL of the IQ4 firmwares have seemed like a beta, unfortunately. Between the split image review bug, non-capture bug, vertical orientation scroll bug, and now this one (arguably the most serious bug yet given its prevalence and lack of errors indicating something is wrong)... by my count ALL of the IQ4 firmware releases to date have had a serious bug.

Phase should announce to its customer base the steps it is taking to minimize such serious bugs in future firmware releases.
 

vonalpen

Active member
Same problem here, never realized it so far, so thanks to Paul!
However, as I have been doing a lot of frame averaging lately, while checking the file formats in C1, I now found that my IQ4 150 stores all files as 16L again as soon as I had taken a frame averaging image!
But by turning the back off, the 16L setting is lost again....
So the work around seems to be to change the file format to 14bit and right away back to 16bit every time you turn on the back!
I have checked this, works for me....
But sure hope Phase One reacts soon with a fix - what else are we missing???
 

vjbelle

Well-known member
My ordered back is not here yet but I'm on the fence to say no way......

This is all really uncalled for..... at the best price you can possibly get this should still never happen. Something is going on at Phase that does not sit right..... the only way to describe this is sloppy QC and that means that Phase is cutting corners.

I'm more than a little concerned.....

Victor
 

Craig Stocks

Well-known member
Sadly it’s becoming a bit embarrassing to own an IQ4150. I’m really concerned about Phase One’s ongoing incompetence.
 

Ferg

New member
I noticed the image count today but didn’t think about it until now. I’d also noticed less quality on image review the past few days, but didn’t double check. Luckily it’s just been snapshots and test shots since I updated.

Ps first post on here
 

earburner

Member
Correction, it does effect my back... glad I know about it. well spotted :D

Just found another bug... in my laziness I switched on auto iso to just snap a test shot and... it don't work... lol
 

algrove

Well-known member
Is all I know is that I did not buy an IQ4150 to shoot in 14bit.

So to make sure the workaround is:

1. Power on
2. Set to 14 bit
3. Then even without taking an mage set to 16bit extended.
4. Now ready to capture in 16 bit.

Is this correct?

Can others confirm the "Auto ISO" problem. Thanks.
 

GrahamWelland

Subscriber & Workshop Member
I agree that it’s embarrassing for Phase One but what we’re seeing here is the classic software issues that occur when you fundamentally replace the entire OS/application stack and start again. In my business this tends to be the v2 rewrite ahead of v3 return to stability.

Ultimately though the IQ4150 was released too soon and obviously needed (still needs) more testing before hitting the streets.
 

Boinger

Active member
I agree that it’s embarrassing for Phase One but what we’re seeing here is the classic software issues that occur when you fundamentally replace the entire OS/application stack and start again. In my business this tends to be the v2 rewrite ahead of v3 return to stability.

Ultimately though the IQ4150 was released too soon and obviously needed (still needs) more testing before hitting the streets.
I wouldn't mind as much if once the bug is found, and such a critical one at that. It should be hotfixed within a few days at most.

No need to wait till the next firmware iteration which is likely what will happen here.
 

GrahamWelland

Subscriber & Workshop Member
I wouldn't mind as much if once the bug is found, and such a critical one at that. It should be hotfixed within a few days at most.

No need to wait till the next firmware iteration which is likely what will happen here.

Agreed, if it’s such a simple fix. Although if they are vigilant it should invoke a full regression test to make sure that the fix didn’t break something else. We follow a full Devops process where developers are responsible for ultimate quality plus end to end standard test suites vs ad hoc user tests - that’s what’s showing up in Phase One’s release approach. You really need a standard test suite of known use cases that have to pass every time.
 

MGrayson

Subscriber and Workshop Member
As a non-PhaseOne owner, I'm happy to see a functioning and supportive community. If this had happened with other subcultures, we'd see:

Fuji: "This is a small price too pay for such rapid innovation."
Hasselblad: "DJI doesn't care about them! They're DOOMED!"
Leica: *iz ded*
Sigma: "14bits is better, anyway."
The measurement crowd: "Well, the definition of a bit depends on quantum read noise crosstalk well depth. The IQ4150 actually records 15.7*Sqrt[-1] bits"
The film crowd: "HCB didn't need bits!"
Apple: "The 14bits are not upgradeable. I get 18bits for half the money from Dell!"

:chug:
Matt
 

Paul2660

Well-known member
I was able to do a simple regression process on my back, and can pinpoint that it's Frame averaging that created the problem. I updated on 07/03/19, and used the back without frame averaging on a shoot, came back the next morning to test Frame Averaging.

The first series of images during my testing are all 16 Bit EX or non EX, but all are 16 bit. I am sure during those tests I powered the back on and off several times but all the files are the correct size and are 16 bit.

Later that day I came back to the same location and tried some different angles but had turned off the back. From that point on for the next 500 or 600 frames everything was shot in 14 bit low quality. I always check the indicator showing 16 bit EX or non EX, but never noticed the count for frames remaining had gone up by close to 2x. Once you use frame averaging, this bug is set and stays in place even after the back is powered off, so it's a different type of bug one that appears to have to be enabled by a feature, but once enabled stays in place.

This simple fact, still amazes me that none of the P1 developers, or "Professional" shooters caught this as it had to happen to them also, unless this issue was caused by some last minute tweak to the firmware, on P1 knows that.

I still say, this firmware was NOT released as public beta, after rechecking Phase One's main page and doc's, there is no mention of it being a beta with the all the normal literature about use and risk of a beta. Also I know of no other firmware for a IQ that was released as a Public Beta, if anyone knows of one, please correct me. I am fully aware of IQ firmware in beta being given to an end user (I have had it happen) with an understanding that it is a beta and thus be aware of issues. If Phase goes back and changes their release notes on the this firmware to say it's a Public Beta, then shame on them. It's not in the release notes I pulled down on 07/03/19 or yesterday.

Many may find this is not a big deal, that's their prerogative. Net to me is I would have gone on making this same mistake and continuing to shoot the lowest quality from the back as I did not realize the issue was there until someone else pointed out to me. All I know is what I can see in the files. Like another poster, I had also started to wonder about increases in noise in the reviews of the images, but due to the heat in my location I just wrote it off to the back getting warm. There is almost 100MB of file size difference between the 14bit and 16 Bit Ex, the average 14 LOW res image from the IQ4 being close to equal to a High res 14 bit file from a Nikon Z 7.

For me this is as bad as the bug that was allowed to go out day one on non auto rotation of portrait images. And that persisted for almost 6 months. Auto rotation of a portrait image is basic camera operations 101, if you can't auto rotate you should not be releasing a camera yet.

Paul C
 

dougpeterson

Workshop Member
I agree that it’s embarrassing for Phase One but what we’re seeing here is the classic software issues that occur when you fundamentally replace the entire OS/application stack and start again.
I 100% agree with this. In this case they also changed the hardware and UI at the same time as the OS/application stack. It was a gutsy move, and one that, in the long run, almost always works in the favor of both customer and company. You need only look at camera platforms that don't get fundamental re-writes and see how they end up languishing quietly into the good night; they don't die, they just... fade away. The XF was a fundamental rewrite of the DF and (after a few rocky months right at the start) has been incredible for P1 and its customers. The IQ was a fundamental rewrite of the P+ and has been incredible for P1 and its customers. The IQ4 was another fundamental rewrite, and in fact was a bigger rewrite than the other two examples.

The problem is that explanation is only good for a few months.

It's now been a bit over two quarters since the system started shipping. Phase One needs to address this; pronto.
 

dougpeterson

Workshop Member
I still say, this firmware was NOT released as public beta, after rechecking Phase One's main page and doc's, there is no mention of it being a beta with the all the normal literature about use and risk of a beta. Also I know of no other firmware for a IQ that was released as a Public Beta, if anyone knows of one, please correct me.
You are 100% correct. The firmware was not beta, and this kind of bug simply should not be in a non-beta release of firmware.
 

onasj

Active member
Interesting observations, Paul. I have not had a chance yet to take any frame-averaged images but my IQ4 nevertheless has the bug. I may have touched the "sigma" button that opens the frame averaging menu. But I definitely have not captured any frame-averaged images with it, yet my back still has the bug. If you are correct that a freshly updated IQ4 won't have the bug, then perhaps simply opening the frame averaging menu causes the problem?

Kudos to Steve/CI for posting about the bug on the CI website; other dealers and Phase One should do the same given that photographers who are using the latest firmware will likely be capturing in a lower quality than intended.

I agree with most other posters in this thread that this bug *is* a big deal, not only for the (small but very expensive!) quality difference between 14 and 16 bit, but also for what it says about Phase One's continued firmware issues and counterproductive lack of communication about fixing serious bugs.


I was able to do a simple regression process on my back, and can pinpoint that it's Frame averaging that created the problem. I updated on 07/03/19, and used the back without frame averaging on a shoot, came back the next morning to test Frame Averaging.

The first series of images during my testing are all 16 Bit EX or non EX, but all are 16 bit. I am sure during those tests I powered the back on and off several times but all the files are the correct size and are 16 bit.

Later that day I came back to the same location and tried some different angles but had turned off the back. From that point on for the next 500 or 600 frames everything was shot in 14 bit low quality. I always check the indicator showing 16 bit EX or non EX, but never noticed the count for frames remaining had gone up by close to 2x. Once you use frame averaging, this bug is set and stays in place even after the back is powered off, so it's a different type of bug one that appears to have to be enabled by a feature, but once enabled stays in place.

This simple fact, still amazes me that none of the P1 developers, or "Professional" shooters caught this as it had to happen to them also, unless this issue was caused by some last minute tweak to the firmware, on P1 knows that.

I still say, this firmware was NOT released as public beta, after rechecking Phase One's main page and doc's, there is no mention of it being a beta with the all the normal literature about use and risk of a beta. Also I know of no other firmware for a IQ that was released as a Public Beta, if anyone knows of one, please correct me. I am fully aware of IQ firmware in beta being given to an end user (I have had it happen) with an understanding that it is a beta and thus be aware of issues. If Phase goes back and changes their release notes on the this firmware to say it's a Public Beta, then shame on them. It's not in the release notes I pulled down on 07/03/19 or yesterday.

Many may find this is not a big deal, that's their prerogative. Net to me is I would have gone on making this same mistake and continuing to shoot the lowest quality from the back as I did not realize the issue was there until someone else pointed out to me. All I know is what I can see in the files. Like another poster, I had also started to wonder about increases in noise in the reviews of the images, but due to the heat in my location I just wrote it off to the back getting warm. There is almost 100MB of file size difference between the 14bit and 16 Bit Ex, the average 14 LOW res image from the IQ4 being close to equal to a High res 14 bit file from a Nikon Z 7.

For me this is as bad as the bug that was allowed to go out day one on non auto rotation of portrait images. And that persisted for almost 6 months. Auto rotation of a portrait image is basic camera operations 101, if you can't auto rotate you should not be releasing a camera yet.

Paul C
 
Top