How fast can you shoot your normal (50mm equivalent) lens handheld, and without optical image stabilization?

Question

How fast can you shoot your normal (50mm equivalent) lens handheld, and without optical image stabilization?

I have had the FUJINON XF18-55mmF2.8-4 R LM OIS for about two years now. I am very happy with it, but I bought the FUJINON XF35mmF2 R WR because I wanted a smaller lens.

The zoom is f/3.5 at the 35mm setting, so I gain about 1.5 stops of light with the f/2 prime. Apart from that I'm okay with bumping the ISO from the auto 3200 I normally use to auto 6400, so I gain about 2.5 stops of effective exposure which I hoped would be enough to offset the loss of optical image stabilization, but it's not even close.

I did some tests and I can get a sharp image with the 35mm prime at 1/128s only about 80% of the time and, even then, it's not a perfectly steady shot. It's much sharper from a tripod. At 1/64 I can only get a sharp image about 30% of the time, and at 1/32s I can't get a sharp image at all.

With the zoom lens at 35mm I can get a sharp image at 1/16s 100% of the time, at 1/8s 95% of the time, and at 1/4s 60% of the time. And if I get a sharp shot, it's as sharp as I can get it from a tripod.

I actually mostly shoot film on Nikon SLRs with old AI/AI-S manual focus lenses, so I believe I have enough experience "bracing myself" for getting steady shots without OIS. With my film cameras I get 100% sharp shots at 1/30s with a 50mm or 85mm lens. I shoot my 135mm lens at 1/60 with no problem.

In other words, with my film camera I can easily beat the 1/f rule of thumb for maximum shutter speed, but I'm nowhere near as good with my digital Fuji camera. And the SLR even has slapping mirrors which would in theory make this harder than on my digital camera.

The problem seems to be that on my Nikon FM3a SLR the shutter release is very smooth, there are no kinks and detents. I can smoothly squeeze the shutter release button, I don't even have to hit the bottom.

On the Fuji there are annoying middle detents, and I have to fully depress the shutter release button to make an exposure. I can set a two second delay, but it's very annoying and still nowhere near as good as what I can get on film.

So what's the problem here? If you're hand-holding your digital camera, what shutter speed can you use and still get sharp shots 99% of the time with a 50mm equivalent lens?

I was thinking that the weight of the film and digital setups are different, so I made a comparison:

Fuji X-T10: 372g
Fuji X-E2: 344g

Nikon FM3a: 595g
Nikon FA: 625g

Fuji XF35F2: 165g
NIKKOR 50mm f/1.4 AI-s: 250g
Nikon Nikkor 85mm f/2 AI-s: 313g
Nikon 135mm f/2.8 AI-s: 435g

Fuji X-T10+XF35F2: 537g
Nikon FM3a+50mm: 845g

So there is 57% more weight with the Nikon setup.

I have also did some more tests between my X-T10 and my X-E2 (same sensor), and strangely enough I am much sharper on my X-E2 at 1/f shutter speeds.

I am comparing 50+ MP film scans downsampled to 16MP (same resolution as X-T10) viewed on the same 5k display at 1:1. I shoot mostly Velvia 50 and sometimes Ektar 100. I have done some more tests, it appears that if I shoot in continuous-low mode (I think around 3fps), the first and last frames are pretty shaky, but the middle frames are sharp even at 1/15 (again I'm talking about the same 50mm-equivalent lens).

Later edit: This is NOT a duplicate of How can I determine the minimum shutter speed to avoid blur from camera shake? because this question is about the differences in camera shake behavior for specific different systems, not about camera shake in general.

Answer 1

A very rough rule of thumb is the maximum shutter time in seconds is the reciprocal of the 35mm-equivalent focal length in mm. That means, for example, that you should keep the shutter speed at 1/50 s or faster for a 50 mm lens (or 1/80 s if you are using a crop sensor so that the 35mm-equivalent focal length is 80 mm).

Of course many things can alter that tradeoff. If you can lean your body against something solid, you can use a lower shutter speed. If you can lean the camera against something solid, you can use a even lower shutter speed.

Your own skill matters too. Yes, taking slow pictures without much blur is something you can learn and get better at. You hold your body and arms just right, breath just right, slowly squeeze as apposed to push the shutter button, etc.

There is always some random variation and chance. At slow shutter speeds, take several pictures of the same thing. Some will have more wobble than the others, even though you think you did everything the same.

Answer 2

I am comparing 50+ Mpix film scans downsampled to 16Mpix (same resolution as X-T10) viewed on the same 5k display at 1:1. I shoot mostly Velvia 50 and sometimes Ektar 100. I have made some more tests, it appears that if I shoot in continuous-low mode (I think around 3fps), the first and last frames are pretty shaky, but the middle frames are sharp even at 1/15 (again I'm talking about the same 50mm-equivalent lens).

It seems you've isolated the issue to somehow being related to when you move your index finger to press or release the shutter button with the X-T10. Since you're not experiencing the same issue on the X-TE2 or the Nikon film camera, your shooting technique in general does not seem to be the culprit. The way you are required to hold the X-T10 to operate it could be a contributing factor, or vibration introduced by the actual movement of the shutter button (as distinct from your movement actuating it) could be part of the problem.

There are still some differences in magnification that must be acknowledged, though.

Your film scans downsized to 16MP and viewed at 100% on the same monitor as your files from the X-T10 are not being enlarged at the same degree of magnification as the files from the X-T10. This is because the film negatives are 36x24mm whereas the sensor of the X-T10 is only a bit less than 24x16mm. So even though they are the same resolution and display size, you are still magnifying the X-T10 images linearly by 1.5X more.

To get equal enlargement from the film scans you'd need to downsize them to about 36MP rather than 16MP. Then the center 16x24mm/16MP of the 36MP FF image would occupy the same angular size as the 16MP files from your X-T10 sensor. It's no different than comparing the 36MP FF D800 with the 16MP APS-C D5100. They both have the same pixel density and pixel size, but the large sensor has 2.25X more of the same sized pixels. If you enlarge both images to the same number of pixels per inch, the image from the FF sensor will be 1.5X longer and 1.5X higher, for a 2.25X greater area, than the image from the APS-C sensor.

This difference in magnification could maybe explain some of the difference between the amount of relative blur you are getting with each camera. But if that were the case then the X-TE2 images should be just as blurry as the X-T10 images, since both cameras share the same sensor and would be enlarged by the same amount. Since you report that your results are much better with the X-E2 than with the X-T10, magnification/enlargement ratios can be eliminated as an explanation for the differences you are seeing. It almost certainly has to do with the movement required to actuate the shutter button or the movement of the shutter button itself.