gypsythief

Not too long ago, in a land not very far away, I came across the concept of focus merges to dramatically increase depth-of-field of photos, especially macro ones. This was one of the reasons I bought Affinity Photo, as having learned the technique I had a requirement for affordable software the could do a good job of a focus merge. The focus merges I have done thus far have typically been hand-held and of fairly large 'macro' subjects such as mushrooms and flowers. Generally, these consist of 10-30 shots that Affinity Photo churns through and merges in very short order. Due to the hand-holding, there are often a few ghosting artefacts around some edges, but nothing that a bit of AffinityPhotoShopping can't sort out.
Yesterday, I came across an unfortunately-demised beetle which seemed to be a perfect opportunity to try a more technical, tripod based, photo merge. One of the reasons for this post is to showcase the ability of Affinity Photo which coped with a quite staggering amount of data (the other reason is a desire to share what I think is a quite stonking photo!).
To achieve the focus merge, I took 270 20-megapixel raw photos, totalling 7.8 GB of data. These I then processed into 16 bit/channel PNGs using RawTherapee which came to 24.7 GB. I then fed them into Affinity Photo's Focus Merge and let it churn away. It took just over 1.5 hours to complete the merge, and I'm very pleased with the results:

In case anyone is interested in the hardware involved, here's the list:
Camera: Panasonic DC-G9
Lens: Olympus 60mm f/2.8 Macro
Images shot at: 0.5 sec, f/3.5, ISO 200
CPU: AMD Ryzen 1500X
RAM: 16 GB DDR4 2666
GPU: nVidia Geforce GTX 970 
I largely seemed to be CPU-bound during the various processing stages. Memory usage from RawTherapee and Affinity Photo pottered around the 2-3 GB mark; RawTherapee thrashed all 4 cores / 8 threads in the Raw -> PNG conversion (but still took ages; I left it going over night, so I don't know quite how long...) , whilst Affinity Photo used 1 core / 2 threads for the stack. A lower core count / faster GHz processor would therefore seem to be an advantage during the stacking process in Affinity Photo, whilst RawTherapee clearly liked all the cores it could get it's sticky little paws on. I included my GPU in the list as Affinity Photo does seem to use it a bit: 5-8% GPU activity shows up in Task Manager during Focus Merges; I've no idea if it makes a difference though.
In conclusion, I'm very impressed with the ability of Affinity Photo to deal with huge amounts of data without choking. A few more threads might be nice though

Busy week, so a bit late posting back! The lady bird was unfortunately a bit too dead, so I won't post the photo I took as it wasn't great. I did find GBracket to be a really useful app, though. 
First off, it showed me that the stepping is not arbitrary, as I stated in my fourth post:
"You specify the number of shots and the "stepping". This is an arbitrary number from 1-10 and the amount it moves the lens' focus varies wildly with lens, focal length and focus distance, so it's guess work as to the right number."
Going by how GBracket works, the focus on the lens is divided into 1024 steps from minimum focus to infinity. The stepping of the focus stack directly corresponds to these 1024 steps, so a stepping of 1 gives 1024 shots from minimum focus to infinity, whilst a stepping of 10 gives 102 shots for the same range.
GBracket allows you to select the near and far focus points of your stack and shows the focal distance in terms of the 0 to 1024 steps of each focus point. It is then some simple maths to calculate the number of steps your focus merge covers and select a stepping that gives a reasonable number of shots to cover the required range.
Whilst the lady bird was a bit too decayed, I found another beetle like the one in my original post. Using GBracket I was able to photograph it with an optimised number of photos rather than the somewhat "spray and pray" method of the photo I first posted. I took this earlier this week; I haven't kept the original shots so I don't know exactly how many there were, but it was around 30:

Unsurprisingly, a tenth of the shots from my first post stacked 10 times faster, so the optimisation provided by GBracket* was very useful.
*If this is starting to sound like an advert for GBracket, I apologise, it isn't meant to be. I did find it very useful though!

Busy week, so a bit late posting back! The lady bird was unfortunately a bit too dead, so I won't post the photo I took as it wasn't great. I did find GBracket to be a really useful app, though. 
First off, it showed me that the stepping is not arbitrary, as I stated in my fourth post:
"You specify the number of shots and the "stepping". This is an arbitrary number from 1-10 and the amount it moves the lens' focus varies wildly with lens, focal length and focus distance, so it's guess work as to the right number."
Going by how GBracket works, the focus on the lens is divided into 1024 steps from minimum focus to infinity. The stepping of the focus stack directly corresponds to these 1024 steps, so a stepping of 1 gives 1024 shots from minimum focus to infinity, whilst a stepping of 10 gives 102 shots for the same range.
GBracket allows you to select the near and far focus points of your stack and shows the focal distance in terms of the 0 to 1024 steps of each focus point. It is then some simple maths to calculate the number of steps your focus merge covers and select a stepping that gives a reasonable number of shots to cover the required range.
Whilst the lady bird was a bit too decayed, I found another beetle like the one in my original post. Using GBracket I was able to photograph it with an optimised number of photos rather than the somewhat "spray and pray" method of the photo I first posted. I took this earlier this week; I haven't kept the original shots so I don't know exactly how many there were, but it was around 30:

Unsurprisingly, a tenth of the shots from my first post stacked 10 times faster, so the optimisation provided by GBracket* was very useful.
*If this is starting to sound like an advert for GBracket, I apologise, it isn't meant to be. I did find it very useful though!

@ gypsythief: You are using a Panasonic G9? I'm stacking with the FZ1000. But, if you use the ImageApp and GBracket too, stacking is nearly full automatic. You then set the middle sharpness on your cam, the nearest and the farest point on GBracket then, select the stepping (1-10), some parameters you can also change as ISO, shutter speed, aperture, and then GB does the rest. GB has been programmed by a German doctor (Holger Kremmin), and runs only in combination with the ImageApp - on IOS or Android. What you also can do, are HDR-Stacks (up to 7 series) and - HiRes-Stacks with the G9. The largest number of pictures for a stack is 1025 - and, if you do extremly close up, you get near to it.
This is an example done with the Raynox 250 close-up lens mounted on my fz1000 - a butterfly-detail, consisting of 233 raw's, and a prepared beetle too, with raynox 150..
 

Not too long ago, in a land not very far away, I came across the concept of focus merges to dramatically increase depth-of-field of photos, especially macro ones. This was one of the reasons I bought Affinity Photo, as having learned the technique I had a requirement for affordable software the could do a good job of a focus merge. The focus merges I have done thus far have typically been hand-held and of fairly large 'macro' subjects such as mushrooms and flowers. Generally, these consist of 10-30 shots that Affinity Photo churns through and merges in very short order. Due to the hand-holding, there are often a few ghosting artefacts around some edges, but nothing that a bit of AffinityPhotoShopping can't sort out.
Yesterday, I came across an unfortunately-demised beetle which seemed to be a perfect opportunity to try a more technical, tripod based, photo merge. One of the reasons for this post is to showcase the ability of Affinity Photo which coped with a quite staggering amount of data (the other reason is a desire to share what I think is a quite stonking photo!).
To achieve the focus merge, I took 270 20-megapixel raw photos, totalling 7.8 GB of data. These I then processed into 16 bit/channel PNGs using RawTherapee which came to 24.7 GB. I then fed them into Affinity Photo's Focus Merge and let it churn away. It took just over 1.5 hours to complete the merge, and I'm very pleased with the results:

In case anyone is interested in the hardware involved, here's the list:
Camera: Panasonic DC-G9
Lens: Olympus 60mm f/2.8 Macro
Images shot at: 0.5 sec, f/3.5, ISO 200
CPU: AMD Ryzen 1500X
RAM: 16 GB DDR4 2666
GPU: nVidia Geforce GTX 970 
I largely seemed to be CPU-bound during the various processing stages. Memory usage from RawTherapee and Affinity Photo pottered around the 2-3 GB mark; RawTherapee thrashed all 4 cores / 8 threads in the Raw -> PNG conversion (but still took ages; I left it going over night, so I don't know quite how long...) , whilst Affinity Photo used 1 core / 2 threads for the stack. A lower core count / faster GHz processor would therefore seem to be an advantage during the stacking process in Affinity Photo, whilst RawTherapee clearly liked all the cores it could get it's sticky little paws on. I included my GPU in the list as Affinity Photo does seem to use it a bit: 5-8% GPU activity shows up in Task Manager during Focus Merges; I've no idea if it makes a difference though.
In conclusion, I'm very impressed with the ability of Affinity Photo to deal with huge amounts of data without choking. A few more threads might be nice though

Not too long ago, in a land not very far away, I came across the concept of focus merges to dramatically increase depth-of-field of photos, especially macro ones. This was one of the reasons I bought Affinity Photo, as having learned the technique I had a requirement for affordable software the could do a good job of a focus merge. The focus merges I have done thus far have typically been hand-held and of fairly large 'macro' subjects such as mushrooms and flowers. Generally, these consist of 10-30 shots that Affinity Photo churns through and merges in very short order. Due to the hand-holding, there are often a few ghosting artefacts around some edges, but nothing that a bit of AffinityPhotoShopping can't sort out.
Yesterday, I came across an unfortunately-demised beetle which seemed to be a perfect opportunity to try a more technical, tripod based, photo merge. One of the reasons for this post is to showcase the ability of Affinity Photo which coped with a quite staggering amount of data (the other reason is a desire to share what I think is a quite stonking photo!).
To achieve the focus merge, I took 270 20-megapixel raw photos, totalling 7.8 GB of data. These I then processed into 16 bit/channel PNGs using RawTherapee which came to 24.7 GB. I then fed them into Affinity Photo's Focus Merge and let it churn away. It took just over 1.5 hours to complete the merge, and I'm very pleased with the results:

In case anyone is interested in the hardware involved, here's the list:
Camera: Panasonic DC-G9
Lens: Olympus 60mm f/2.8 Macro
Images shot at: 0.5 sec, f/3.5, ISO 200
CPU: AMD Ryzen 1500X
RAM: 16 GB DDR4 2666
GPU: nVidia Geforce GTX 970 
I largely seemed to be CPU-bound during the various processing stages. Memory usage from RawTherapee and Affinity Photo pottered around the 2-3 GB mark; RawTherapee thrashed all 4 cores / 8 threads in the Raw -> PNG conversion (but still took ages; I left it going over night, so I don't know quite how long...) , whilst Affinity Photo used 1 core / 2 threads for the stack. A lower core count / faster GHz processor would therefore seem to be an advantage during the stacking process in Affinity Photo, whilst RawTherapee clearly liked all the cores it could get it's sticky little paws on. I included my GPU in the list as Affinity Photo does seem to use it a bit: 5-8% GPU activity shows up in Task Manager during Focus Merges; I've no idea if it makes a difference though.
In conclusion, I'm very impressed with the ability of Affinity Photo to deal with huge amounts of data without choking. A few more threads might be nice though

@John Rostron Some good questions and points here which I'll address a bit out of order, as that should make more sense.
Q: Was it by focussing or with a rack? Was it automated?
A: It was automated: the camera has a focus bracketing function. You specify the number of shots and the "stepping". This is an arbitrary number from 1-10 and the amount it moves the lens' focus varies wildly with lens, focal length and focus distance, so it's guess work as to the right number. After that, you press the shutter button and leave it to it until it's done. In terms of further automation, in both RawTherapee and Affinity it was simply a case of loading in all the photos and batching them whilst I did other things, the former whilst sleeping, the later whilst walking the dog!
Q: Why 270 images?
A: I wanted as sharp an image as possible. As I mentioned above the stepping is largely guess work so I simply set it to 2. I tried 30 shots to start with and the focus only crept forward a bit, so I simply cranked the number of shots up to 300. This nicely covered the scene, so I stuck there. The first 30 were totally blurry, and were discarded. I could have tweaked the stepping to reduce the number of images, but I started to get curious about whether Affinity Photo would handle so much data. Which it did, quite admirably.

The depth of the image was much more than 27mm; the beetle alone was about that (it was a very big beetle!). The vetch in the background is running off for some distance at about 45°. I don't have the scene set up anymore so I can't give you an exact figure, but I do have some maths: I've just tested my camera and lens photographing along a ruler using the same settings I used for that shot, and 25 shots moved the focus about 1mm, so 0.04mm per shot. At 270 shots, that would give a depth of the image of 108mm. This agrees with what I would estimate from memory.
In terms of how many shots I could have taken, the quick tests shots I mentioned above showed a depth of field of about 2mm, which ties with what Olympus claim in their depth of field chart for this lens at f/3.5 between 0.19 and 0.3 metres. With a 2mm depth of field, I'd probably go for 1 photo/mm for maximum sharpness and to give a good overlap of focus for the stacking. So I would still have needed 100 photos for the 100mm image depth, but that would have been nearly I third of the processing time.
Edit: Thinking about it, I was probably a bit further back; the depth of field chart shows 2-4mm depth of field between 0.3 and 0.4 metres; averaging that to 3mm, and I could have done the stack in 66 shots. That would have definitely sped things up!
(I should also mention that this was all done at around 10 o'clock on Friday night after a few cold ones, so there was much more enthusiasm than technical analysis!)

Not too long ago, in a land not very far away, I came across the concept of focus merges to dramatically increase depth-of-field of photos, especially macro ones. This was one of the reasons I bought Affinity Photo, as having learned the technique I had a requirement for affordable software the could do a good job of a focus merge. The focus merges I have done thus far have typically been hand-held and of fairly large 'macro' subjects such as mushrooms and flowers. Generally, these consist of 10-30 shots that Affinity Photo churns through and merges in very short order. Due to the hand-holding, there are often a few ghosting artefacts around some edges, but nothing that a bit of AffinityPhotoShopping can't sort out.
Yesterday, I came across an unfortunately-demised beetle which seemed to be a perfect opportunity to try a more technical, tripod based, photo merge. One of the reasons for this post is to showcase the ability of Affinity Photo which coped with a quite staggering amount of data (the other reason is a desire to share what I think is a quite stonking photo!).
To achieve the focus merge, I took 270 20-megapixel raw photos, totalling 7.8 GB of data. These I then processed into 16 bit/channel PNGs using RawTherapee which came to 24.7 GB. I then fed them into Affinity Photo's Focus Merge and let it churn away. It took just over 1.5 hours to complete the merge, and I'm very pleased with the results:

In case anyone is interested in the hardware involved, here's the list:
Camera: Panasonic DC-G9
Lens: Olympus 60mm f/2.8 Macro
Images shot at: 0.5 sec, f/3.5, ISO 200
CPU: AMD Ryzen 1500X
RAM: 16 GB DDR4 2666
GPU: nVidia Geforce GTX 970 
I largely seemed to be CPU-bound during the various processing stages. Memory usage from RawTherapee and Affinity Photo pottered around the 2-3 GB mark; RawTherapee thrashed all 4 cores / 8 threads in the Raw -> PNG conversion (but still took ages; I left it going over night, so I don't know quite how long...) , whilst Affinity Photo used 1 core / 2 threads for the stack. A lower core count / faster GHz processor would therefore seem to be an advantage during the stacking process in Affinity Photo, whilst RawTherapee clearly liked all the cores it could get it's sticky little paws on. I included my GPU in the list as Affinity Photo does seem to use it a bit: 5-8% GPU activity shows up in Task Manager during Focus Merges; I've no idea if it makes a difference though.
In conclusion, I'm very impressed with the ability of Affinity Photo to deal with huge amounts of data without choking. A few more threads might be nice though

Not too long ago, in a land not very far away, I came across the concept of focus merges to dramatically increase depth-of-field of photos, especially macro ones. This was one of the reasons I bought Affinity Photo, as having learned the technique I had a requirement for affordable software the could do a good job of a focus merge. The focus merges I have done thus far have typically been hand-held and of fairly large 'macro' subjects such as mushrooms and flowers. Generally, these consist of 10-30 shots that Affinity Photo churns through and merges in very short order. Due to the hand-holding, there are often a few ghosting artefacts around some edges, but nothing that a bit of AffinityPhotoShopping can't sort out.
Yesterday, I came across an unfortunately-demised beetle which seemed to be a perfect opportunity to try a more technical, tripod based, photo merge. One of the reasons for this post is to showcase the ability of Affinity Photo which coped with a quite staggering amount of data (the other reason is a desire to share what I think is a quite stonking photo!).
To achieve the focus merge, I took 270 20-megapixel raw photos, totalling 7.8 GB of data. These I then processed into 16 bit/channel PNGs using RawTherapee which came to 24.7 GB. I then fed them into Affinity Photo's Focus Merge and let it churn away. It took just over 1.5 hours to complete the merge, and I'm very pleased with the results:

In case anyone is interested in the hardware involved, here's the list:
Camera: Panasonic DC-G9
Lens: Olympus 60mm f/2.8 Macro
Images shot at: 0.5 sec, f/3.5, ISO 200
CPU: AMD Ryzen 1500X
RAM: 16 GB DDR4 2666
GPU: nVidia Geforce GTX 970 
I largely seemed to be CPU-bound during the various processing stages. Memory usage from RawTherapee and Affinity Photo pottered around the 2-3 GB mark; RawTherapee thrashed all 4 cores / 8 threads in the Raw -> PNG conversion (but still took ages; I left it going over night, so I don't know quite how long...) , whilst Affinity Photo used 1 core / 2 threads for the stack. A lower core count / faster GHz processor would therefore seem to be an advantage during the stacking process in Affinity Photo, whilst RawTherapee clearly liked all the cores it could get it's sticky little paws on. I included my GPU in the list as Affinity Photo does seem to use it a bit: 5-8% GPU activity shows up in Task Manager during Focus Merges; I've no idea if it makes a difference though.
In conclusion, I'm very impressed with the ability of Affinity Photo to deal with huge amounts of data without choking. A few more threads might be nice though

Not too long ago, in a land not very far away, I came across the concept of focus merges to dramatically increase depth-of-field of photos, especially macro ones. This was one of the reasons I bought Affinity Photo, as having learned the technique I had a requirement for affordable software the could do a good job of a focus merge. The focus merges I have done thus far have typically been hand-held and of fairly large 'macro' subjects such as mushrooms and flowers. Generally, these consist of 10-30 shots that Affinity Photo churns through and merges in very short order. Due to the hand-holding, there are often a few ghosting artefacts around some edges, but nothing that a bit of AffinityPhotoShopping can't sort out.
Yesterday, I came across an unfortunately-demised beetle which seemed to be a perfect opportunity to try a more technical, tripod based, photo merge. One of the reasons for this post is to showcase the ability of Affinity Photo which coped with a quite staggering amount of data (the other reason is a desire to share what I think is a quite stonking photo!).
To achieve the focus merge, I took 270 20-megapixel raw photos, totalling 7.8 GB of data. These I then processed into 16 bit/channel PNGs using RawTherapee which came to 24.7 GB. I then fed them into Affinity Photo's Focus Merge and let it churn away. It took just over 1.5 hours to complete the merge, and I'm very pleased with the results:

In case anyone is interested in the hardware involved, here's the list:
Camera: Panasonic DC-G9
Lens: Olympus 60mm f/2.8 Macro
Images shot at: 0.5 sec, f/3.5, ISO 200
CPU: AMD Ryzen 1500X
RAM: 16 GB DDR4 2666
GPU: nVidia Geforce GTX 970 
I largely seemed to be CPU-bound during the various processing stages. Memory usage from RawTherapee and Affinity Photo pottered around the 2-3 GB mark; RawTherapee thrashed all 4 cores / 8 threads in the Raw -> PNG conversion (but still took ages; I left it going over night, so I don't know quite how long...) , whilst Affinity Photo used 1 core / 2 threads for the stack. A lower core count / faster GHz processor would therefore seem to be an advantage during the stacking process in Affinity Photo, whilst RawTherapee clearly liked all the cores it could get it's sticky little paws on. I included my GPU in the list as Affinity Photo does seem to use it a bit: 5-8% GPU activity shows up in Task Manager during Focus Merges; I've no idea if it makes a difference though.
In conclusion, I'm very impressed with the ability of Affinity Photo to deal with huge amounts of data without choking. A few more threads might be nice though

Not too long ago, in a land not very far away, I came across the concept of focus merges to dramatically increase depth-of-field of photos, especially macro ones. This was one of the reasons I bought Affinity Photo, as having learned the technique I had a requirement for affordable software the could do a good job of a focus merge. The focus merges I have done thus far have typically been hand-held and of fairly large 'macro' subjects such as mushrooms and flowers. Generally, these consist of 10-30 shots that Affinity Photo churns through and merges in very short order. Due to the hand-holding, there are often a few ghosting artefacts around some edges, but nothing that a bit of AffinityPhotoShopping can't sort out.
Yesterday, I came across an unfortunately-demised beetle which seemed to be a perfect opportunity to try a more technical, tripod based, photo merge. One of the reasons for this post is to showcase the ability of Affinity Photo which coped with a quite staggering amount of data (the other reason is a desire to share what I think is a quite stonking photo!).
To achieve the focus merge, I took 270 20-megapixel raw photos, totalling 7.8 GB of data. These I then processed into 16 bit/channel PNGs using RawTherapee which came to 24.7 GB. I then fed them into Affinity Photo's Focus Merge and let it churn away. It took just over 1.5 hours to complete the merge, and I'm very pleased with the results:

In case anyone is interested in the hardware involved, here's the list:
Camera: Panasonic DC-G9
Lens: Olympus 60mm f/2.8 Macro
Images shot at: 0.5 sec, f/3.5, ISO 200
CPU: AMD Ryzen 1500X
RAM: 16 GB DDR4 2666
GPU: nVidia Geforce GTX 970 
I largely seemed to be CPU-bound during the various processing stages. Memory usage from RawTherapee and Affinity Photo pottered around the 2-3 GB mark; RawTherapee thrashed all 4 cores / 8 threads in the Raw -> PNG conversion (but still took ages; I left it going over night, so I don't know quite how long...) , whilst Affinity Photo used 1 core / 2 threads for the stack. A lower core count / faster GHz processor would therefore seem to be an advantage during the stacking process in Affinity Photo, whilst RawTherapee clearly liked all the cores it could get it's sticky little paws on. I included my GPU in the list as Affinity Photo does seem to use it a bit: 5-8% GPU activity shows up in Task Manager during Focus Merges; I've no idea if it makes a difference though.
In conclusion, I'm very impressed with the ability of Affinity Photo to deal with huge amounts of data without choking. A few more threads might be nice though

Not too long ago, in a land not very far away, I came across the concept of focus merges to dramatically increase depth-of-field of photos, especially macro ones. This was one of the reasons I bought Affinity Photo, as having learned the technique I had a requirement for affordable software the could do a good job of a focus merge. The focus merges I have done thus far have typically been hand-held and of fairly large 'macro' subjects such as mushrooms and flowers. Generally, these consist of 10-30 shots that Affinity Photo churns through and merges in very short order. Due to the hand-holding, there are often a few ghosting artefacts around some edges, but nothing that a bit of AffinityPhotoShopping can't sort out.
Yesterday, I came across an unfortunately-demised beetle which seemed to be a perfect opportunity to try a more technical, tripod based, photo merge. One of the reasons for this post is to showcase the ability of Affinity Photo which coped with a quite staggering amount of data (the other reason is a desire to share what I think is a quite stonking photo!).
To achieve the focus merge, I took 270 20-megapixel raw photos, totalling 7.8 GB of data. These I then processed into 16 bit/channel PNGs using RawTherapee which came to 24.7 GB. I then fed them into Affinity Photo's Focus Merge and let it churn away. It took just over 1.5 hours to complete the merge, and I'm very pleased with the results:

In case anyone is interested in the hardware involved, here's the list:
Camera: Panasonic DC-G9
Lens: Olympus 60mm f/2.8 Macro
Images shot at: 0.5 sec, f/3.5, ISO 200
CPU: AMD Ryzen 1500X
RAM: 16 GB DDR4 2666
GPU: nVidia Geforce GTX 970 
I largely seemed to be CPU-bound during the various processing stages. Memory usage from RawTherapee and Affinity Photo pottered around the 2-3 GB mark; RawTherapee thrashed all 4 cores / 8 threads in the Raw -> PNG conversion (but still took ages; I left it going over night, so I don't know quite how long...) , whilst Affinity Photo used 1 core / 2 threads for the stack. A lower core count / faster GHz processor would therefore seem to be an advantage during the stacking process in Affinity Photo, whilst RawTherapee clearly liked all the cores it could get it's sticky little paws on. I included my GPU in the list as Affinity Photo does seem to use it a bit: 5-8% GPU activity shows up in Task Manager during Focus Merges; I've no idea if it makes a difference though.
In conclusion, I'm very impressed with the ability of Affinity Photo to deal with huge amounts of data without choking. A few more threads might be nice though

This is a known issue and is logged with our developers, I will update them regarding this issue, thank you for reporting it