Better AF results with Vivid plus increased Sharpness and Contrast?

[kirschm]

I off and on hear comments in Olympus micro 4/3 posts that the reliability of the autofocus gets better by setting picture mode to Vivid and/or maximizing the sharpness and contrast settings (of course not recommended for JPG-shooters). Since I shoot RAW only I would not have a problem with this... Is this theory true or is it only wishfull thinking?

Preliminary results as of May 25th:
o OzRay: Picture Mode 'Vivid' seems to have most impact on better AF
o OzRay: More Sharpness or Contrast might have no impact at all (kirschm: but it might matter, if it was set to minimum before?)
o DonParrot: Vivid as a prerequisit to improve AF significantly. Increasing Sharpness/Contrast as a further finetuning tool towards better AF

 

[yakky]

Seem unlikely since the camera shoots the raw first then produces the jpeg. Besides, of the issues M43 bodies have, since when it reliable autofocus an issue?

 

[pdk42]

I think there is some logic in this train of thought. AF works by contrast detection on edges in the focus zone and if the AF processing is in the pipeline after demosaicing, levels, sharpening etc then changing any of them will have an impact on AF. Of course, it could equally be true that AF is independent of these settings. I guess it needs either some scientific testing or someone who knows the internals to determine formsure.

 

[fredlong]

Better AF results with Sharpness and Contrast at maximum?

Seem unlikely since the camera shoots the raw first then produces the jpeg. Besides, of the issues M43 bodies have, since when it reliable autofocus an issue?

Well, you're looking at an image in real time that reflects the JPEG settings before firing the shutter, so...

This has come up before and people who say say they've tried it say it works. It doesn't seem to me, to make enough of a difference to notice. But I'm using a camera from the historic collection, a Panasonic G2.

Fred

 

[SkiHound]

Gary Ayton has some pretty extensive information posted on setting up the E-M5. He says it improves AF. I've tried and subjectively I think it improves AF. I may just be drinking the Kool-Aid but my impression is it improves both AF speed and accuracy.

 

[wushumr2]

It works. It's contrast detect, so raising the contrast absolutely has an effect. Sold my E-M5 to a girl and when I was showing it to her, she remarked on the focus being a bit slow. Bumped contrast from -1 to 1 and it was miles better under low light.

 

[usayit]

Wow... totally counter of what I would expect since it takes effort on the camera to create JPG and increase the contrast.

Gotta try it myself.

I have noticed that increasing the EVF refresh rate to high does improve AF. In theory its because the processing required for AF is occurring more frequently.

 

[dhazeghi]

Wow... totally counter of what I would expect since it takes effort on the camera to create JPG and increase the contrast.

That work is on the part of a dedicated chip so it's extremely fast. Plus, as Fred said above, it's work that must be carried out regardless since the display (LCD or EVF) depends on it.

 

[kirschm]

Gary Ayton has some pretty extensive information posted on setting up the E-M5. He says it improves AF...

http://www.ayton.id.au/wiki/doku.php?id=photo:olympusomdem5

Here he says: "...it seems C-AF and S-AF is even faster and C-AF more reliable if you select Vivid picture mode instead of natural, and worse if you select the low contrast Muted mode - presumably, the AF system is using live processed images rather than raw data. Just shoot in RAW+jpeg and you can have the best of both worlds"

I hope this does not mean that I need to shoot RAW+JPG instead of RAW only?

 

[OzRay]

I just did some tests using my 35-100mm f2, which doesn't have the SWD and switching to vivid mode did improve focus speed. I switched between normal and vivid several times, using different subjects and forcing the lenses to go through the full focus movement and it really does work. The E-M1 must use both PDAF, and CDAF, with 4/3 lenses to assist in focussing for this to work.

 

[ahinesdesign]

I tried it with my G5 and 12-35mm and it seems to improve the speed of AF lock (although I have no way of gathering any actual data on time to acquire focus...). The 12-35mm is already very fast to focus, so even a "big" improvement may not necessarily be that easy to determine with any certainty. In good light and/or high-contrast scenes, it may not make much difference. I would expect cranking up sharpness and contrast to most affect AF speed/accuracy in situations with lower contrast that were more challenging for CDAF systems to begin with.

My test was against my usual custom JPG mode of "Natural" with sharpness, contrast, and noise reduction all at -2, and a new custom "Vivid" JPG mode with sharpness and contrast cranked up to +2. I'm not a fan of the JPG output, but I can use the other custom mode when I need the JPGs to look better or I'm not recording RAW+JPG.

I have a photo shoot in the morning where I'll test the boosted sharpness+contrast theory to see if it really does improve AF speed and accuracy. I've done the same shoot several times with a variety of lenses, so I've got some past images to compare against...

 

[glenchen68]

this piques my interest as a new owner of an EM10... does it work with just contrast upped instead of contrast plus saturation?

I say this because saturation in Oly cams are already blown IMO. I am usually a JPEG shooter.

 

[kirschm]

I have a photo shoot in the morning where I'll test the boosted sharpness+contrast theory to see if it really does improve AF speed and accuracy...

What's also good to test: Does 'too much tuning' (everything set to the max, plus Vivid, plus I don't know) may have negativ impacts (e.g. quick, but wrong focussing).

 

[OzRay]

this piques my interest as a new owner of an EM10... does it work with just contrast upped instead of contrast plus saturation?

I say this because saturation in Oly cams are already blown IMO. I am usually a JPEG shooter.

I tried adjusting both the 'vivid' setting as well as the sharpness/contrast. The latter made no difference, but setting to vivid mode was noticeable. It made no difference doing both.

 

[glenchen68]

I tried adjusting both the 'vivid' setting as well as the sharpness/contrast. The latter made no difference, but setting to vivid mode was noticeable. It made no difference doing both.

I prefer natural but I guess I could live with vivid if that alone really does improve AF lock speed. Not too keen on pushing saturation to max though..

 

[OzRay]

I prefer natural but I guess I could live with vivid if that alone really does improve AF lock speed. Not too keen on pushing saturation to max though..

I exclusively shoot RAW, so the settings are irrelevant for me.

 

[glenchen68]

I exclusively shoot RAW, so the settings are irrelevant for me.

i am staying put with JPEG. Will try just upping contrast alone to see if there is a difference..

 

[dejongj]

I like the idea....I do find it very hard to evidence though as my lenses snap immediately into focus as they do...Any suggestion on how to evidence this?

 

[kirschm]

I like the idea....I do find it very hard to evidence though as my lenses snap immediately into focus as they do...Any suggestion on how to evidence this?

I think there are at least 2 ways:

1) Often the AF immediately finds the focus... but the focus was wrong (even, if there was nothing with different distances within the focus field, just check the focus distance in the Exif Info)

2) At night / at low contrast... perhaps focus hunting stops hunting

 

[kirschm]

I tried adjusting both the 'vivid' setting as well as the sharpness/contrast. The latter made no difference, but setting to vivid mode was noticeable. It made no difference doing both.

See my 'Preliminary results' in the first post:
OzRay: More Sharpness or Contrast might have no impact at all (kirschm: but it might matter, if it was set to minimum before?)

Any evidence for this?