Advantage of 42 megapixels?

Hi all, 

I’m considering options for focus stacking in macro work. I currently have a Sony A7rII, which is 42 megapixel. I’m achieving good results (see link) but I want to improve it.

https://vimeo.com/266781428

Option 1: the Cognisys Stackshot, this allows me to keep using the A7rII, but possible drawbacks are the additional time that it takes for this system to stack 10 - 20 photos per increment on the turntable.

Option 2: an Olympus OMD-E-M1 MkII with in camera stacking / bracketing. This camera captures bursts very quickly, however, this is a 20mp camera, half the res of the A7rII

In the case of camera vs camera, how much of a benefit is the extra resolution of the Sony? assuming the same pixel quality, deep focus etc.

Cheers for any help.

I have only manually focus stacked, but i can say the sony a7riii beats the pants of my sony a77ii 24mp crop in detail, noise ( even at 100 iso and good lighting), and I feel like I have to take less pictures with more megapixels to get the same quality results. I also believe moving the camera body to move the focus plain, instead of changing the focus of the lens.  Leads to better less distorted images because of focus breathing.

I am by no means an expert in macro photography and only mess around for fun. I encourage you to fact check the part about moving the camera vs focusing the lens. I want to say I saw it on a B and H class on youtube.

Thanks Steven, logically, I would think more MP the better. It also means you can scan smaller bugs and crop in if you have to if 1:1 mag isn’t enough to fill the sensor.

Someone suggested this, it’s far less expensive than the Stackshot. The Stackshot feels overpriced to me, and the control unit feels like 90’s era tech, however, it seems to be the benchmark.

www.wemacro.com

Hi Paul,

resolution isn’t everything:

https://www.dpreview.com/reviews/image-comparison?attr18=daylight&attr13_0=sony_a7rii&attr13_1=olympus_em1ii&attr13_2=fujifilm_xt20&attr13_3=sony_a7iii&attr15_0=raw&attr15_1=raw&attr15_2=raw&attr15_3=raw&attr16_0=3200&attr16_1=3200&attr16_2=3200&attr16_3=3200&attr126_1=1&attr171_0=1&normalization=full&widget=1&x=0.7788089480048368&y=0.6771323549151153

In my opinion the olympus is at least 1-2 steps down in terms of detail and noise in relation to the a7rII. My X-T20 is in between the two I would say. But then take the a7III, I would say it has the same detail as the a7rII but with the same resolution as the olympus…

Thanks Götz,

I’ve done a simple test by taking the photos from the scan I posted above, and downsizing them to 20mp. It’s not a particularly scientific test, but I just wanted to see if dropping the pixel count makes a noticeable difference. I’ll attach an image.

The full res version is a bit sharper but all of the features are still present in the lower res version. These images aren’t stacked at all, and shot at f22, so they’re far from “sharp” to begin with.

The advantage of the Olympus is speed. I’ve bought the wemacro rail, and will see how it goes, but if each stack takes one minute to shoot (no clue yet how long it will actually take) and I shoot say, 200 photos per scan (often more) then I’m looking at 3hrs 20 mins, just to shoot the photos. The Olympus, with it’s burst stacking will do a whole stack in 1 - 2 seconds.

I just read that the X-T20 also does stacking / bracketing? Have you tried it with Reality Capture?

I have feelings about in camera processing, and not the good kind. Does anyone have a sample of in body stacking vs stacking on pc. Is it just good enough, or equal in quality?

Hi Paul,

sorry, I never tried it but I might give it a shot soon (pun intended).

Hmm, are you sure it’s the T20? I didn’t find anything in the manual and a quick search only said something like that about the T2…

Because to be honest, Steven, I have made the opposite experience. I am now going back to in-camera JPGs where the object doesn’t merrit the last 2% of gain, which is mainly in the texture in my opinion. Of course, this all depends on the camera etc.

Unfortunately, I cannot switch off rectification in my X-T20 but the correction seems to be very good (RC shows almost no difference in the grid) and as I said, in therms of geometry (mesh) I couldn’t see much difference to processed RAWs. Certainly not enough to make the extra work and disk space worth my while.

Götz Echtenacher after testing I’m slowly coming to your side of thinking. Like many things in life it is a trade-off between quality and time. Once you hit a certain point diminishing returns kicks in and it’s just a question of how much you’re willing to put up with. For me tiffs have fallen outside of reasonable time and drive space vs the maybe slight quality difference, but haven’t quite yet reached a place of giving up total control of my jpgs.

Emphasis on the word “yet” lol

Sorry Götz, you’re right, it’s the T2

I never use RAW for photogrammetry. I use an Orangemonkie Foldio360 Turntable, which triggers the shutter every few seconds, for 48 photos per 360, so the buffer fills up and the camera can’t keep up if I shoot raw, and I don’t see an advantage in my results anyway.

What I’ve been told so far is, that for magnification upto 1:1, in camera bracketing / stacking is fine, but beyond that, say 2:1, 5:1 etc, you may be better off with a rail.

The Wemacro rail has shipped, I’ll post results as soon as I can, but I’m leaning towards also getting an OM-D or some other camera that does bracketing… annoys me that Sony doesn’t implement it.

Hey Paul,

good to know you get these awesome meshes without RAW!

I never understood the magnification ratio - could you explain that ot me in 3 words?

The rail has the advantage that you wouldn’t neet to post-process the images since the position of the cameras alters slightly - might just be enough for RC to catch the difference…

Hi Götz

1:1 macro means the subject will fill the same amount of the frame that it would if you were to physically place it directly on the sensor, at least, that’s how I understand it.

2:1 would be double that, and so on. If a lens can’t achieve at least 1:1, it’s not a “true” macro lens.

I’ll still be post processing, I’ll just be feeding RC the merged final, sharp images. Helicon Focus does batch processing, so it shouldn’t be too painful.

Here’s my latest, non stacked scan.

https://www.instagram.com/p/Bk3LkZOjQui/