Phase IQ 180 Focus checking tool

[scott morrish]

I have had a quick look at the IQ 180 today, and am particularly interested in the real-world usefulness of the new focus checking tool. Does it work?

I only had a little time, but my first few frames were on a DF through a Schneider LS 110mm (on a tripod, mirror up, 3 sec delay etc, F11), and i was surprised how little it detected as being in focus. (I guess surprise is not really the right word - the reason i am interested in the feature is the difficulty i have had in achieving consistent and accurate focus with a P65+ since switching to the DF). I might be doing something obviously wrong - so i am not jumping to any conclusions - i am certainly hoping this feature works!

From my brief look - low 'level' settings seem to inaccurately indicate a wider area to be in focus than actually is (when later viewed in C1), whilst high 'level' settings don't show anything to be in focus.

Can anybody offer advice on the 'level' settings they trust? ... or is it a setting that gets changed depending upon the subject (texture / contrast etc)?

Thanks for any help,
Scott

[Guy Mancuso]

I found 31 to be pretty good with the 110. But you could go a few points lower

I think the default is like 50 but not sure

Its certainly worth testing and getting a feel for the settings and your system.

[michael]

Remember, it's not really detecting focus, but rather contrast. These correlate well most of the time, but not always. For example, with the lens focused on infinity, a cloud on the horizon might not show as being in focus, because it's low contrast, while a tree a bit closer ( though still in focus ) will.

Michael

[Conner999]

Using it tethered with an SLR. It works ok, but as Michael indicates, it detects contrast and (in it's current form) is very, very easily thrown off. In my case it's for portrait work. It's usefulness in it's current form will depend on a lot of user and scene-specific variables. I trust the Mk1 eyeball and if C1 agrees with me great, but I currently don't trust it on it's own.

Maybe in future refinements it would be a trusted indicator of focus vs. 'it might be focus, it might be contrast, it might be both', but for me right now it's something neat to watch. Maybe if they could tie-in the focus point to increase the accuracy probability  (e.g. if it detects an 'in focus' condition at the point of focus, odds are it's in focus).

[scott morrish]

Thanks Michael and Guy.

The odd thing is - based on the few exposures i have made so far - that contrast detection can be very misleading as an indicator of focus. I looked at an image where almost nothing that was actually in focus, registered as such via the focus mask. At the same time lots of areas that were definitely out of focus, but which happened to be high contrast, were highlighted as being in focus. I shall experiment further...

And having just seen Conner999's post whilst i was typing - that makes sense from what i have seen so far.

Thanks to you all,
Scott

[Jack Flesher]

I assume you are meaning the focus mask on the IQ180 review image? First off, you need to know that you can set the tolerance for the mask itself to suit your needs.  I find it works very well, but needs a tighter setting to be accurate.  For example, I find 35 a little generous, and 50 pretty tight, almost accurate to the actual pixel level.  I still need to experiment more, but will probably settle on somewhere around 40 for my landscape work.

[Guy Mancuso]

Yes i was shooting a model outside and should have said that as well. So 31 seemed good there but like everyone has said you need to experiment to find your happy spot and contrast will have a lot to do with it.

[gazwas]

Never got on with the focus mask tool in C1 so don't have high hopes for the one built into the IQ backs. I always thought it was very misleading calling it a focus mask as it doesn't actually establish correct focus but contrast. As to tweaking the IQ backs's settings depending on the subject seems a like a very hit and miss approach and without the IQ's amazing screen I'd never trust this feature. As suggested, linking this information to an AF focus point or user set (tech camera) focus area would offer better information but still very dependant on lighting contrast, subject matter and your best last used mask settings.

[Jack Flesher]

Not sure what your point is --  it's only available on the IQ backs "with their incredible screen" to begin with.  As for trust, any craftsman needs to learn how to get the most out of their tools.  Personally, I am finding the focus mask on the IQ back VERY useful, YMMV... 

[bcooter]

As for trust, any craftsman needs to learn how to get the most out of their tools. 

True.  But a hammer is not is not a laser.

IMO

BC

[gdh]

Interesting--I came up with 39--distinction without a difference.  You all keep saying contrast is the thing being measured. There has to be more than that.  I've intentionally included nearly black/white contrasty items slightly out of focus with less contrasty area in focus and the mask seems to work, only picking the in-focus areas to highlight. I don't profess to be the most technically astute, so maybe some can explain this a little more technically, as I'm interested and to best work with your tools, it's nice to know the why's.

I assume you are meaning the focus mask on the IQ180 review image? First off, you need to know that you can set the tolerance for the mask itself to suit your needs.  I find it works very well, but needs a tighter setting to be accurate.  For example, I find 35 a little generous, and 50 pretty tight, almost accurate to the actual pixel level.  I still need to experiment more, but will probably settle on somewhere around 40 for my landscape work.

[gazwas]

Not sure what your point is --  it's only available on the IQ backs "with their incredible screen" to begin with.  As for trust, any craftsman needs to learn how to get the most out of their tools.  Personally, I am finding the focus mask on the IQ back VERY useful, YMMV... 

In other words, would you 100% take what the focus mask was telling you was in focus it it didn't have the new screen to to back it up?

I hate pompous comments like that....."craftsman". I'm not quite sure how this makes any sense when all you're doing is altering a number in the preferences to a setting which looks good without really knowing how the the mask actually measures focus. Yep, 37 units of mask focus thingy looks good. Seems more best guess to me. 

[Jack Flesher]

True.  But a hammer is not is not a laser.

IMO

BC

I think a finishing plane is a better comparative example James.

[Jack Flesher]

In other words, would you 100% take what the focus mask was telling you was in focus it it didn't have the new screen to to back it up?

At this point, and at a setting of 40 I do.

I hate pompous comments like that....."craftsman".

You may hate it, but the fact has been proven over and over again for the past couple thousand years

[scott morrish]

Okay - after a bit more playing - i think i get it now.

38 - 40 looks about right but i'd need to use it more on a wider range of subjects to be sure.

FWIW i wouldn't rely on it as a guarantee of focus at all. I'd use it more as a rough guide 'if and when' the subject (usually landscape for me) happens to coincide with the focus masks tolerances.

Focus masks aside however...the quality and resolution of the screen is a vast improvement?