1:1 at f/36 Versus 1:2 at F/8 Cropped

Eyelight

Senior Member
.......If I shoot at 1:1 f/32 or at half of that magnification at f/8, crop the shot and double in PS, do you think I'll lose 1/3th of my sharpness? If your answer is yes, you should close down.

But to each their own.

Just continuing the discussion in a different place since it had more to do with macro work.

My thinking is either method has it's place in the toolbox and one may work better than the other on a given subject with a given camera/lens. Change the camera and/or lens or the subject and everything changes, so your joy may definitely vary.

ETA: @J-see , you may have processing tricks I don't use. You are welcome to add examples to the thread.

#1 - 1:1 at f/36
DSC_2127_150303_1024_002.jpg


#2 - f/8 cropped from #4
DSC_2128_150303_1024_003.jpg


#3 - f/8 cropped from similar shot to #4 with slightly different focus point.
DSC_2129_150303_1024_004.jpg


#4 - 1:2 at f/8 (1:2 was eyeballed). The shot #2 was cropped from.
DSC_2128_150303_1024_003-2.jpg


Crop of #1 - f/36
DSC_2127_150303_1024_001.jpg


Crop of #2 - f/8 that was cropped from #4
DSC_2129_150303_1024_001.jpg
 
Last edited:

J-see

Senior Member
It's not the macro period here but I'll shoot something inside later on or tomorrow. At the moment I'm still trying to get my eyes fully open. ;)

I don't know what method you use to scale but in PS, bicubic softer is usually selected, or lanczos, to scale up. You can only process after scaling since any sharpening is scale-dependent.

It's difficult taking two identical shots at different magnification since any angle you got, you have to go back and get up to get the same at a lower magnification. Easiest is shooting perfectly in level, then you only need to go back the distance.
 
Last edited:

mikew_RIP

Senior Member
A mixture of exposure,brightness and contrast,shadows and highlights,two complete high pass @1 with some of the adjustments in between and then a selective high pass @ 2.8 for just the OOF areas
 

J-see

Senior Member
A mixture of exposure,brightness and contrast,shadows and highlights,two complete high pass @1 with some of the adjustments in between and then a selective high pass @ 2.8 for just the OOF areas

It's nice work. I was experimenting with Gaussian blur - High pass sharpening the other day and even when I see potential in it for some shots, I didn't get it good enough this far. I only wonder how much of all those tricks I can do on my Adobe-low Mac. It's hard to say bye-bye to PS.
 

J-see

Senior Member
Let's do this different. We don't need to wonder if DoF is identical at 0.5X*f/8 and 1x*f/32 since that's easy to calculate. The only question is if an image scaled larger would lose more or less quality than closing down does.

There's a much easier approach to that which doesn't require secure positioning of the cam and measuring distances since the DoF is that small, anything done wrong can have massively different results.

All we need is this:

001.jpg

002.jpg

003.jpg

004.jpg

005.jpg


This is purely about the effect scaling has on the quality of our original shot. About all of these have been subjected to different scaling permanently affecting them and their quality. None has been changed at any other level, no exposure adjustments or additional sharpening. Only scaling algorithms have been at play.

Pick your best but select carefully since I am aware of every bias at play during this sort of selection and how to get around them. ;)

Anyone can still cheat but it'll require effort.

I used the prehistoric version of PS I have on this PC. It's not the best approach possible to scaling.

That'd be something like this: Blow Up Examples - Alien Skin Software
 
Last edited:

Eyelight

Senior Member
Let's do this different. We don't need to wonder if DoF is identical at 0.5X*f/8 and 1x*f/32 since that's easy to calculate. The only question is if an image scaled larger would lose more or less quality than closing down does.

There's a much easier approach to that which doesn't require secure positioning of the cam and measuring distances since the DoF is that small, anything done wrong can have massively different results.

................................

This thread is about macro work and specifically about this:

.......If I shoot at 1:1 f/32 or at half of that magnification at f/8, crop the shot and double in PS, do you think I'll lose 1/3th of my sharpness? If your answer is yes, you should close down.

But to each their own.

In another post you said:

Of course the one with more depth is better than the other, visually. But there are more roads that lead to Rome and not all of them are old and rocky. Closing down is not the best option but there's no need to believe me, just test if I'm right or not and see for yourself.

Giving plausible credence to what you said, I took the time to test and I would have to say you were not right or at least that I cannot prove you were right.

There is no reason to continue unless you can produce an 1:2 f/8 shot as stated that appears as sharp or sharper than a 1:1 f/36 shot of the same subject that has some depth.

If not, it's simply time to move on.
 

J-see

Senior Member
This thread is about macro work and specifically about this.

It's f/8 vs f/32. Small difference but anyways.

Like I said, you don't need to believe me, test it yourself. Of the five shots I posted here, I don't see any fall apart at the same rate closing down affects a shot which simply shows scaling does not affect a shot too much.

If I got time I'll set up my tripod and take some shots but it'll cost time since I can not eye-ball anything when I'm positioning a DoF of around 1mm. I can only do a comparison when measuring the exact distances.
 

Woodyg3

Senior Member
Contributor
Eyelight, your shots show clearly that 1:1 at f/36 provides a better, sharper picture than a 1:2 f/8 cropped. Again, this is the laws of physics at work.
 

J-see

Senior Member
Eyelight, your shots show clearly that 1:1 at f/36 provides a better, sharper picture than a 1:2 f/8 cropped. Again, this is the laws of physics at work.

The DOF is identical for f/8 at 0.5x and f/32 at 1x. That's the law of physics too. It does not make any difference to that. If the DoF is not identical for both, something went wrong during the shoot. F:36 is 1/3th a stop too much and the 0.5x is slightly more than that. But it doesn't matter since we can calculate it and know them to be identical.


The real question is if you lose more quality scaling than by diffraction.
 
Last edited:

J-see

Senior Member
Ok, I finally gathered enough courage to start this task. I know how difficult it is to get a similar image at two different magnifications and alas, mine isn't perfect either. You almost need a rail system for this and it's not that important I'm going to construct all that. This already got on my nerves in the end. DoF isn't placed identical but with an FX, I have even less wiggle-room.

I shot the 1x at f/32 and then measured the distance to get as close to 0.5x as possible. That was harder than I remembered. I shot the same scene at f/8 and that was it. Both have been live-view focused as best as possible and shot with MuP. All processing is identical besides scaling. The ones without EXIF are scaled in another program and then either used as, or scaled to normal size and then used as.

First the full shots:

32normal.jpg

8x1normal.jpg

8x2.jpg

I'll post details in a bit, I have one in the wrong format.
 
Last edited:

J-see

Senior Member
DoF we know to not be different, let's even simply agree that details aren't too different even when they probably are if I could do this perfectly.

Even if all would be equal, what would be the not so trivial difference which makes halving still a much, much better option than closing?
 

Woodyg3

Senior Member
Contributor
Perhaps if you labeled each shot I could understand what you are showing. Right now I am not able to follow this at all.
 

WayneF

Senior Member
DoF we know to not be different, let's even simply agree that details aren't too different even when they probably are if I could do this perfectly.

Even if all would be equal, what would be the not so trivial difference which makes halving still a much, much better option than closing?

If you also showed some background detail (like maybe the green bit), certainly all would NOT be equal. DOF is about about improving the overall picture.
 

J-see

Senior Member
Perhaps if you labeled each shot I could understand what you are showing. Right now I am not able to follow this at all.

The ones with the EXIF are f/32 and F/8 as mentioned, all the rest is f/8 scaled which is all that matters. It's f/8 vs f/32.

I saved every shot using two algorithms and picked the sharpest of both. In the details there are shots of both algorithms of some but I forgot which was which and just said; f*k it and posted them all.
 
Last edited:

J-see

Senior Member
If you also showed some background detail (like maybe the green bit), certainly all would NOT be equal. DOF is about about improving the overall picture.

Wayne, how often do I need to repeat that the DOF can be calculated and you get the exact same DoF by either closing down or shooting at a lower magnification?

I don't have the tools to do this 100% accurately and can place focus at both sides of what I see as sharp in live-view. That translates in a huge different at these magnifications. When I'm fine-tuning focus in live-view I don't see anything but that small detail I am putting in focus. That's really next to nothing and I have zero oversight on the whole shot.
 
Last edited:
Top