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 frame averaging vs ND filter

Maxx9photo

Active member
I've found this page and found very useful,


I did quick short trip yesterday to take some shots on using filter vs frame averaging.

My take on this, using ND is faster in process but of course added weight on travel, What do you guys think?
 

Ray Harrison

Well-known member
For me, I find frame averaging to be quite a bit faster in process. At the very least, I'm not waiting on a black frame. And as you say, it's much lighter.
 

SrMphoto

Well-known member
I've found this page and found very useful,


I did quick short trip yesterday to take some shots on using filter vs frame averaging.

My take on this, using ND is faster in process but of course added weight on travel, What do you guys think?
Why is using ND faster in process than using in-camera frame averaging? What added weight? ND filters are not that heavy, aren’t they?
 

Paul Spinnler

Well-known member
In my experience frame averaging doesn’t work well on things like cars and headlights so I generally keep a ND filter on me as well.
It may be a use case thing - if you use it to shoot landscapes or seascapes etc then FA is a godsend. A clear differentiator in the marketplace.

I haven’t tried it in cityscape scenarios though which is what I think you describe, effectively.
 

Maxx9photo

Active member
Why is using ND faster in process than using in-camera frame averaging? What added weight? ND filters are not that heavy, aren’t they?
Heavy is not definitely the case, taking space in the bag yes. I’ve been using Lee SW 150 filter system so it does take space and added weight.
 

tcdeveau

Well-known member
It may be a use case thing - if you use it to shoot landscapes or seascapes etc then FA is a godsend. A clear differentiator in the marketplace.

I haven’t tried it in cityscape scenarios though which is what I think you describe, effectively.
for sure! Cityscape is a specific use case but 100% agree that for landscapes and seascapes, it’s a god send. The files are SO clean too, definitely less noisy than typical long exposure with ND (which is a perfectly acceptable level of noise to me - not complaining just saying the FA ones are so much cleaner)
 

Mexecutioner

Well-known member
I’ve experienced the opposite with FA. Faster, easier, less weight and space than the clunky filter systems from the past.
 

FloatingLens

Well-known member
It may be a use case thing - if you use it to shoot landscapes or seascapes etc then FA is a godsend. A clear differentiator in the marketplace.
I‘d really appreciate Hasselblad to come with their own take on Frame Averaging. Some kind of patent agreement between the two companies would be awesome IMO. 🤩
 

SrMphoto

Well-known member
I‘d really appreciate Hasselblad to come with their own take on Frame Averaging. Some kind of patent agreement between the two companies would be awesome IMO. 🤩
I would love to have in-camera frame averaging (like Phase One and Olympus m43 cameras), but my first choice of Phase One features would be a live raw histogram.
 

tcdeveau

Well-known member
I‘d really appreciate Hasselblad to come with their own take on Frame Averaging. Some kind of patent agreement between the two companies would be awesome IMO. 🤩
let’s keep in mind that just because P1 has it, and hasselblad doesnt, that doesn’t mean there’s a patent issue between P1 and HB that is keeping FA off of HB cameras. I could do FA with the original A7R I had almost 10 years ago, and other systems have this functionality too, so it’s extremely unlikely P1 has a patent that is keeping HB from putting FA on their cameras.
 

Ray Harrison

Well-known member
Frame averaging is best done at base ISO, AFAIK. Does anyone run frame averaging at high ISOs, and why?
With Phase's approach to frame averaging, you decide on what exposure you want for a single frame given conditions of the scene. You have more freedom over the aperture, ISO and single-frame shutter speed. You then decide on how long you want to average things together. ISO can be higher because the FA itself averages out the (random) noise. Additionally, when you have a need to be "gapless" in your sequence, you can tweak the ISO to get the shutter speed you want. Phase actually recommends ISO 100 as a starting point, not base ISO, because they've found there's more of apossibility to get overblown highlights.

Olympus' implementation is more limited, I find, because they're actually couching it in terms of ND filter factors and you've got specific shutter speed restrictions.
 

SrMphoto

Well-known member
With Phase's approach to frame averaging, you decide on what exposure you want for a single frame given conditions of the scene. You have more freedom over the aperture, ISO and single-frame shutter speed. You then decide on how long you want to average things together. ISO can be higher because the FA itself averages out the (random) noise. Additionally, when you have a need to be "gapless" in your sequence, you can tweak the ISO to get the shutter speed you want. Phase actually recommends ISO 100 as a starting point, not base ISO, because they've found there's more of apossibility to get overblown highlights.

Olympus' implementation is more limited, I find, because they're actually couching it in terms of ND filter factors and you've got specific shutter speed restrictions.
Thank you for the reply.
IQ3 100MP and IQ4 150MP have a base ISO of 100 (see P2P measurements). So it makes sense to recommend ISO 100. ISO 50 is an extended ISO and, therefore, to be avoided.
Is there a case where averaging 800 frames at ISO 800 is better than 100 frames at ISO 100? Isn't it preferable to maximize the IQ of a single frame even when frame-averaging?
To use gapless averaging, the exposure must use shutter speeds equal to or slower than the sensor's readout time. A high ISO could make the shutter speed too fast and therefore introduce gaps in averaging.
 

Ray Harrison

Well-known member
Thank you for the reply.
IQ3 100MP and IQ4 150MP have a base ISO of 100 (see P2P measurements). So it makes sense to recommend ISO 100. ISO 50 is an extended ISO and, therefore, to be avoided.
Is there a case where averaging 800 frames at ISO 800 is better than 100 frames at ISO 100? Isn't it preferable to maximize the IQ of a single frame even when frame-averaging?
To use gapless averaging, the exposure must use shutter speeds equal to or slower than the sensor's readout time. A high ISO could make the shutter speed too fast and therefore introduce gaps in averaging.
The key is that there are a lot of ways you can approach it. It really comes down to what you want to do creatively. A lot of it depends on how long you want to average a scene. In general I do tend to use lower ISOs. That said, if I look at a scene and and judge cloud or water movement to be something I want to average for a particular period of time at a given aperture and a higher ISO gives me the ability to do that, I'm not afraid to use it. To me, I think the question is less "if 800 at 800 is better than 100 at 100", it's that I can use either depending on my needs without appreciable loss of IQ. The IQ4-150 also exhibits a fair amount of ISO invariance at full-stop increments from ISO 400 so there's some flexibility there too.
 

ThdeDude

Active member
In my experience frame averaging doesn’t work well on things like cars and headlights so I generally keep a ND filter on me as well.
Since the longest shutter speed on the iPhone is currently one second, I use a frame average app, Show Shutter, and I also have noticed that.
Wonder whether the algorithm in frame averaging uses the median ("middle" value) and not the mean ("average" value). Median would discount bright, but moving sources like the lights of a car. In contrast, with film its more like the "average" value just only divided through one, or in other words, all protons with no temporal weighting.
 
Last edited:
Top