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!

Phase One IQ4 - Feature Update 1

dougpeterson

Workshop Member
Of course, having said that, the outcome of the ISO 1600 image was still better in the deep shadows than a single ISO 50 long exposure in "normal" mode
That's really quite profound.

I think if you'll be up above ISO 1600 it will behove you to average a very large number of frames.

I share your hope that the 2 sec max will be raised somewhat. But I agree that the reason the limit is there is probably well founded.

Re single pixel noise, definitely make use of the single pixel noise reduction slider. In most cases it has very little impact on the overall image (e.g. very minimal, or no, degradation of actual detail) but does a great job cleaning up those single pixels.

Now... post some pics! :)

Doug
 

Audii-Dudii

Active member
So I'm faced with a dilemma here when shooting in low light: since the maximum length of every single exposure is only 2 seconds, I'm being forced to use a higher ISO when doing frame averaging, which besides noise also causes quite some hot pixels in the shadow areas.
Alternatively, you can always use whatever exposure length you need / want and average (or median-blend, if you're trying to eliminate people from the photo) however many frames you need / want the old-fashioned way, which is to do it yourself during post-processing.

While having an averaging function available in-camera is certainly convenient, it's not essential nor is it always the optimal way to blend multiple files, which is why Photoshop, for example, offers many different blending modes.

And this technique is also useful with other backs besides the IQ4, as I was successfully using it with my P30+ for long-exposure, nighttime photography way back in 2010-2013, as well as again just recently to take this test photo down the street from my house (consisting of seven, 30-second exposures, median-blended in Photshop) before it stopped working for whatever reason. :(



P.S.: Using ISO 1600 instead of ISO 50 as a workaround for the exposure length limitation of this function will also cost you several stops of dynamic range -- 13.11 v. 9.85 -- as Bill Claff's testing at Photons to Photos clearly shows: Photographic Dynamic Range versus ISO Setting
 
Last edited:

beano_z

Active member
Alternatively, you can always use whatever exposure length you need / want and average (or median-blend, if you're trying to eliminate people from the photo) however many frames you need / want the old-fashioned way, which is to do it yourself during post-processing.

While having an averaging function available in-camera is certainly convenient, it's not essential nor is it always the optimal way to blend multiple files, which is why Photoshop, for example, offers many different blending modes.

And this technique is also useful with other backs besides the IQ4, as I was successfully using it with my P30+ for long-exposure, nighttime photography way back in 2010-2013, as well as again just recently to take this test photo down the street from my house (consisting of seven, 30-second exposures, median-blended in Photshop) before it stopped working for whatever reason. :(



P.S.: Using ISO 1600 instead of ISO 50 as a workaround for the exposure length limitation of this function will also cost you several stops of dynamic range -- 13.11 v. 9.85 -- as Bill Claff's testing at Photons to Photos clearly shows: Photographic Dynamic Range versus ISO Setting
Agreed, but that's really besides the point for me. I'm just trying to raise some awareness from my observations as to the newly added function by Phase One on the IQ4 back, and I personally know that there are people who are very much looking forward to this feature and it was also one of the reasons for their upgrade.

Hence, I think it's good to post my findings here so that potentially phase one or their partners will see this and perhaps advice on either how to adjust our shooting styles to how to improve future iterations of the firmware.

Anyway, like I said above, it's good to know my limitations at the moment. In my next round of testing I'll see if I can find a workaround that still allows me optimal image quality and not having ta carry around stacks of filters :grin:
 

Wayne Fox

Workshop Member
Had a few minutes to test drive the new frame averaging feature. to me the idea of no ND filters even in bright sunlight had a lot of appeal, but after working with it and now understanding it better, I can see the ND filters will remain in the bag.

I found that if I did a somewhat long capture, simulating a minute, even if the shutter speed didn’t provide continuous exposure but instead was a “gap” shutter speed, I might be able to get away with it. I think the longer the time (in minutes), the more likely this will work. Below is a 1/15th of a second shutter speed simulating a one minute exposure.
190705_Balboa_P0000376_temp.jpg

Using 1/6th of a second simulating a 2 second exposure left me with magenta discolored highlights. So here ND’s would have been much preferred.

190705_Balboa_P0000380_temp.jpg
Scenic_Session copy.jpg

Similarly I tried a 30 second simulation at 1/30th of second shutter speed toward the sun, and had major magenta artifacts.
190705_Balboa_P0000393_temp.jpg
Scenic_Session_and_Phase_One_IQ4_-_Feature_Update_1_-_Reply_to_Topic.jpg

I must have misunderstood exactly what was going on (or I just didn’t think it through), but there are some major limitations based on the electronic shutter “scan” time or whatever it’s called.

You have to get to 0.5 seconds to have what is called “continuous” exposure. This means that if you simulate 30 seconds with a .5 second exposure time, you will get 60 frames averaged together with no gaps, and the results will most likely be nearly identical to using an ND.

but once you go faster than .5 seconds, you will get 68 frames averaged, no matter what the shutter speed. So 30 seconds at 1/30th of a second doesn’t average 900 frames. whether its 1/10th or 1/500th, you will get 68 frames averaged.

My hope was the ability to simulate 2 second by averaging 30 1/15th of a second exposures, things like that. I really didn’t think it through because I just sort of focused on the idea of the original discussions to eliminate ND filters but now I can see this is probably only in extreme long exposure situations (minutes).With moving water, unless you can get to ½ second or want “smoke” water with really long exposure times, you’ll need to stick with ND filters. I question whether it will be much use on things like waterfalls. I really don’t want to take a chance on colored artifacts because of all the water moving, and most waterfalls I don’t expose for minutes, more like 5 - 20 seconds.

Certainly a lot more testing to see when it might be useful. In typical lower light situations when I shoot this type of work, I might find it more useful. I didn’t stick around for sunset because I discovered Balboa Beach faces southwest, and in the summer the sun doesn’t set over the water. So original plan of silhouetting the pier into the sunset was a bust.
 

Boinger

Active member
Had a few minutes to test drive the new frame averaging feature. to me the idea of no ND filters even in bright sunlight had a lot of appeal, but after working with it and now understanding it better, I can see the ND filters will remain in the bag.

I found that if I did a somewhat long capture, simulating a minute, even if the shutter speed didn’t provide continuous exposure but instead was a “gap” shutter speed, I might be able to get away with it. I think the longer the time (in minutes), the more likely this will work. Below is a 1/15th of a second shutter speed simulating a one minute exposure.


Using 1/6th of a second simulating a 2 second exposure left me with magenta discolored highlights. So here ND’s would have been much preferred.



Similarly I tried a 30 second simulation at 1/30th of second shutter speed toward the sun, and had major magenta artifacts.

I must have misunderstood exactly what was going on (or I just didn’t think it through), but there are some major limitations based on the electronic shutter “scan” time or whatever it’s called.

You have to get to 0.5 seconds to have what is called “continuous” exposure. This means that if you simulate 30 seconds with a .5 second exposure time, you will get 60 frames averaged together with no gaps, and the results will most likely be nearly identical to using an ND.

but once you go faster than .5 seconds, you will get 68 frames averaged, no matter what the shutter speed. So 30 seconds at 1/30th of a second doesn’t average 900 frames. whether its 1/10th or 1/500th, you will get 68 frames averaged.

My hope was the ability to simulate 2 second by averaging 30 1/15th of a second exposures, things like that. I really didn’t think it through because I just sort of focused on the idea of the original discussions to eliminate ND filters but now I can see this is probably only in extreme long exposure situations (minutes).With moving water, unless you can get to ½ second or want “smoke” water with really long exposure times, you’ll need to stick with ND filters. I question whether it will be much use on things like waterfalls. I really don’t want to take a chance on colored artifacts because of all the water moving, and most waterfalls I don’t expose for minutes, more like 5 - 20 seconds.

Certainly a lot more testing to see when it might be useful. In typical lower light situations when I shoot this type of work, I might find it more useful. I didn’t stick around for sunset because I discovered Balboa Beach faces southwest, and in the summer the sun doesn’t set over the water. So original plan of silhouetting the pier into the sunset was a bust.
Those aren't really artifacts, What you are seeing is individual wave structure.

If you took 60 frames and averaged in photoshop it produce the same result.

Simple reason is because its not recording the movement its recording multiple exposures.

You would just have to take many more frames to get the same effect.

So if you want to simulate a 2 min exposure at say 1/15th of a second. Then do the math out so it would give you 1800 frames needed to get the same effect of a continuous shot.

Even then there might be variables that don't truly replace an nd filter. Like if the tide changed etc.
 

etrump

Well-known member
I was not so fortunate to have fog on my last visit and doubt it will happen on this visit either.

Your image is excellent and the fog adds so much emotion to this incredible scene.

Indeed, early worked for me for this shot. After 5AM there's just too many people.

Actually the morning I took this image there was just one other person taking almost the same framing with his Alpa STC and an IQ back. Last year when I got my Phase One system my wife told me just the day I got it: "I knew you wanted one of those since we went to Prague a year ago". The girl was paying attention. :)
 

etrump

Well-known member
I discovered the same thing Wayne. I'm thinking a 2 or 3 stop ND with a polarizer should get me to 1/2 second exposure in almost any lighting which I'm hoping will be long enough to similate a true ND long exposure without the additional noise. So I'm trading 6/10 & 15 stop NDs for one - hopefully.

Had a few minutes to test drive the new frame averaging feature. to me the idea of no ND filters even in bright sunlight had a lot of appeal, but after working with it and now understanding it better, I can see the ND filters will remain in the bag.

I found that if I did a somewhat long capture, simulating a minute, even if the shutter speed didn’t provide continuous exposure but instead was a “gap” shutter speed, I might be able to get away with it. I think the longer the time (in minutes), the more likely this will work. Below is a 1/15th of a second shutter speed simulating a one minute exposure.
View attachment 142881

Using 1/6th of a second simulating a 2 second exposure left me with magenta discolored highlights. So here ND’s would have been much preferred.

View attachment 142880
View attachment 142882

Similarly I tried a 30 second simulation at 1/30th of second shutter speed toward the sun, and had major magenta artifacts.
View attachment 142883
View attachment 142884

I must have misunderstood exactly what was going on (or I just didn’t think it through), but there are some major limitations based on the electronic shutter “scan” time or whatever it’s called.

You have to get to 0.5 seconds to have what is called “continuous” exposure. This means that if you simulate 30 seconds with a .5 second exposure time, you will get 60 frames averaged together with no gaps, and the results will most likely be nearly identical to using an ND.

but once you go faster than .5 seconds, you will get 68 frames averaged, no matter what the shutter speed. So 30 seconds at 1/30th of a second doesn’t average 900 frames. whether its 1/10th or 1/500th, you will get 68 frames averaged.

My hope was the ability to simulate 2 second by averaging 30 1/15th of a second exposures, things like that. I really didn’t think it through because I just sort of focused on the idea of the original discussions to eliminate ND filters but now I can see this is probably only in extreme long exposure situations (minutes).With moving water, unless you can get to ½ second or want “smoke” water with really long exposure times, you’ll need to stick with ND filters. I question whether it will be much use on things like waterfalls. I really don’t want to take a chance on colored artifacts because of all the water moving, and most waterfalls I don’t expose for minutes, more like 5 - 20 seconds.

Certainly a lot more testing to see when it might be useful. In typical lower light situations when I shoot this type of work, I might find it more useful. I didn’t stick around for sunset because I discovered Balboa Beach faces southwest, and in the summer the sun doesn’t set over the water. So original plan of silhouetting the pier into the sunset was a bust.
 

etrump

Well-known member
I'm reposting these images here (which probably should have landed here in the first place).

I've found the frame averaged files to be a welcome addition to my usual bracketed exposures.

Just for fun I exposed for the street lighting and came up with a 5 minute frame averaged exposure of probably 1/4 second at ISO 50:

01DF206E-DF41-4A31-AF72-179B3EEA9C7D.jpeg

In capture one I was able to raise the shadows enough to do what I would normally do with 5 exposures. The shadow detail was so clean, even after this extreme an edit no noise was visible at full resolution. Normally this would take a day or two of blending those 5 exposures. I am keen to print it and see how it compares to my normal workflow.

E6BE00E1-227E-45DF-A284-688F1D4225F1.jpg
 

tashley

Subscriber Member
I think I'm being thick but I cannot get this to work: I have installed FW 5.00.20 and then put the back onto my Alpa TC, set it to use ES, gone to live view and set the shot parameters and can't see the options for frame averaging when I swipe in from the right. There must be a good reason for this but I cannot for the life of me work out what it is!

Any thoughts?
 

dougpeterson

Workshop Member
I think I'm being thick but I cannot get this to work: I have installed FW 5.00.20 and then put the back onto my Alpa TC, set it to use ES, gone to live view and set the shot parameters and can't see the options for frame averaging when I swipe in from the right. There must be a good reason for this but I cannot for the life of me work out what it is!

Any thoughts?
From the ES menu (the one with shutter/aperture/iso), drag in from the right.
 

tashley

Subscriber Member
From the ES menu (the one with shutter/aperture/iso), drag in from the right.
That's what I was doing- it showed the level only, not the frame averaging symbol.

It's all academic now because I thought I'd do a firmware factory reset (with a fresh battery....) as a precursor to reinstalling feature set 4 and then the 5.00.20 but the back is now totally bricked.

I wanted silky smooth long exposures, now I have a paperweight. Grrrr..... flakey.
 

dougpeterson

Workshop Member
Using 1/6th of a second simulating a 2 second exposure left me with magenta discolored highlights. So here ND’s would have been much preferred.
This is because your individual exposures were too bright, causing purple fringing. Next time aim to under expose a bit to make sure your highlights are not blown in any individual frame.

You have to get to 0.5 seconds to have what is called “continuous” exposure. This means that if you simulate 30 seconds with a .5 second exposure time, you will get 60 frames averaged together with no gaps, and the results will most likely be nearly identical to using an ND.
1/2 second for gapless exposure in 16-bit IIQ
1/4 second for gapless exposure in 14-bit IIQ

Do note that if you're close to gapless the effect will be very very similar to being exactly at gapless. Think about the "temporal fill factor". If you're at 1/60th of a second in 16 bit then for 1/60th of a second the sensor is exposing and for 29/60th of a second the sensor is not exposing, which means you're only exposing 3% of the time. At 1/8th of a second in 14 bit the fill factor would be 50% which would lead to visual movement pretty similar to a single long-exposure with a strong ND, especially if the movement is repeating (e.g. many crashing waves, not one set of headlights driving off into the distance).

Most waterfalls I don’t expose for minutes, more like 5 - 20 seconds.
Which I suspect you can get to with a 2-stop ND Filter or a polarizer, rather than needing a, say, 10 stop filter.

With a 2-stop filter you can still use live view focusing, look through the lens (if on an SLR rather than tech camera), still test the exposure very quickly.

And if you have no ND whatsoever (e.g. you forgot one or are packing very light) you can still do it, but, as you say, you'd need a lot of averaging (e.g. as you say, several minutes of exposures averaged) to adequately smooth out the motion.
 

etrump

Well-known member
There is a reset procedure, have you tried that?

That's what I was doing- it showed the level only, not the frame averaging symbol.

It's all academic now because I thought I'd do a firmware factory reset (with a fresh battery....) as a precursor to reinstalling feature set 4 and then the 5.00.20 but the back is now totally bricked.

I wanted silky smooth long exposures, now I have a paperweight. Grrrr..... flakey.
 

kdphotography

Well-known member
That's what I was doing- it showed the level only, not the frame averaging symbol.

It's all academic now because I thought I'd do a firmware factory reset (with a fresh battery....) as a precursor to reinstalling feature set 4 and then the 5.00.20 but the back is now totally bricked.

I wanted silky smooth long exposures, now I have a paperweight. Grrrr..... flakey.
The latest (beta) fw 5.00.20 requires the previous fw to be installed. The fw numbering convention can be confusing because certain firmware are contained within another "wrapper" number, which probably is because it updates both the IQ and the XF body together. It probably is slightly better then previous methods, and actually, I'd rather be able to just concern myself with one fw and be done with the update (which is supposed to be how it works). My update did not go as smoothly but with dealer support and guidance from the Mothership, you should be able to get your IQ4 updated. Mine also "pretended" to be bricked, but it just took a bit to get booted back up with its original firmware. It took some work, but thanks to Brad "Bad Brad" Kaye, all is working great!

Now I just need to get caught up and get my head wrapped around this Frame Averaging feature!

It would be great to see a write-up in one place on how to use these added features rather than perusing several pages of forum threads...

ken
 

etrump

Well-known member
Found the reset in an old email from brad at ci. Can’t hurt since you are bricked:

2 button reset is performed with the digital back disconnected from camera, protective plate on, sitting on table surface.
Hold down the top two hard silver buttons, power the unit on and while still holding the buttons, count to five and then release.


:
That's what I was doing- it showed the level only, not the frame averaging symbol.

It's all academic now because I thought I'd do a firmware factory reset (with a fresh battery....) as a precursor to reinstalling feature set 4 and then the 5.00.20 but the back is now totally bricked.

I wanted silky smooth long exposures, now I have a paperweight. Grrrr..... flakey.
 

dougpeterson

Workshop Member
2 button reset is performed with the digital back disconnected from camera, protective plate on, sitting on table surface.
Hold down the top two hard silver buttons, power the unit on and while still holding the buttons, count to five and then release.
Make sure to do this (and any firmware operation) with a completely full battery.

After your back has successfully restored its factory firmware, follow the step by step instructions for the firmware update here.

If you're still having problems (or even if you just want someone to be with you on the phone as you do these steps, or would prefer someone else do them for you), make sure to reach out to your dealer. It's their job to help! :)
 

tashley

Subscriber Member
Thanks all. I eventually managed a de-brick and update and all now works. Thanks for all the suggestions. Now time to go play!
 

Wayne Fox

Workshop Member
Ok I’ll bite. 5 minute frame average exposed for the highlights (ISO 50, f/16, 2 seconds) then shadows raised in post.

View attachment 142904
Cool. wondering if you could add 5 stops of ISO (so 1600), cut the exposure to only 1 minute if the results would be nearly identical. should be enough frames for averaging to eliminate the noise as well.
 
Top