Should af sharpness equal live view sharpness?

[Murren]

I've been testing out a lens to compare the sharpness of the image taken with af to the one taken with liveview. No matter how much i microadjust it, I cannot get it as sharp. Is this normal?

 

[AlanF]

I've been testing out a lens to compare the sharpness of the image taken with af to the one taken with liveview. No matter how much i microadjust it, I cannot get it as sharp. Is this normal?

No, it is not normal for me. They are usually the same. Perhaps you have a problem with shutter/mirror shake affecting the AF images at lower shutter speeds? You could detect this by comnparing the shots at high shutter speeds of 1/1000s or faster.

 

[Murren]

I used a tripod, timer and mirror lockup. There wasn't enough light for a very fast shutter speed, but the same settings were used for the liveview and af shots.

 

[Joules]

I used a tripod, timer and mirror lockup. There wasn't enough light for a very fast shutter speed, but the same settings were used for the liveview and af shots.

A tripot will not cancel out the shock of the mirror moving. In LiveView, there's even less of that if you're also using an electronic first curtain shutter. So depending on your actual shutter speed, that may explain a difference.

Maybe share some more details about what you're using? Camera, settings, lens?

 

[Murren]

I was under the belief that mirror lockup negates that? I'm using a 50d, efs 55-250mm at 250mm (IS off). 1/250s, f5. 6.

Manual focusing with liveview x10, and taking the same shot with viewfinder/af (at all microadjustments). Comparing the photos, none are as sharp as the liveview one.

 

[Joules]

Oh sorry, I somehow mentally skipped the mirror lock up part.

So when you focus in LiveView, then go into OVF mode but only use that to take the picture without AF, are you also seeing the difference? Or is it only if you focus AF through the viewfinder?

 

[Viggo]

They should be pretty much the same, but only if you use one shot and not servo. But it is tricky. When calibrating you need at least 10 EV light and test at 50x focal length and take at least 5 shots at each AFMA value, with pull it completely oof between shots. Then compare them in Lr at full 1:1.

one of the many reasons I don’t miss dslr’s