Could you elaborate ?
I’m not sure I understand well all what’s said about this :
Does not optical modules use distance anymore ?
OR
Does DxO not provide way to manually enter distance when this information is not given by the couple camera-lens ?
OR
Is it only for Sony cameras ?
OR
Aren’t all combo camera-lens fully sampled by DxO ?
OR
Anything else ?
So what’s wrong ?
noname has found examples I posted of A lenses that have there distance found that the distance is looking complexly wrong and reading DistortionFocus = 128 when they were clearly different. Franky also said what appears to be the same “Ok I just checked and whatever the lens, I’m still on “infinite”.
Should this be dependent on the camera + lens couple too ?”
If PL is not reading distance right on my two lenses but the distance slider is not shown as it has a distance, even if the wrong one, users will not know its wrong. I have posted number of new images from a lens that has distance read by PL it will be interesting to see if those who are better at these thing than me find if they are getting the right distance or they are using 128 as well.
I just done 3 tests with my Tamron that PL claims to get distance from. One at .5 meter, one at 1 meter and last one at 2 meters. All are down as DistortionFocus = 128 so if that is the distance its clearly wrong. How many more camera/lenses are there if this is misreading distance are there and as the distance slider isn’t shown users thing the correct distance and lenses corrections are being used when they clearly may not be?
SAME HERE.
I fastly checked about 10 pictures took at different distances (minimum one was about 3 or 4 m, max about 20 or 30 m) with 2 differents lenses on my drive: Nikon 300 PF and nikon 500 PF with D850.
Every distortionFocus value is 128.
Photolab 6.3.1 PC here.
I have told support of my tests but had nothing back. Could be more users need to start pestering them and anywhere that might get DXOs attention and get the to sort this out. Mainly we need them to get the current camera/lenses not having there distance read done so correctly.
It even seems that this value is a 8bit data (128, mid value between 0 and 1 ( 0-256)).
@John7
All this seems to need further testing.
Maybe this 128 value is a kind of flag that runs some computation - or a parameter used in computation - when distance is in fact read from raw file ?
Or even an obsolete parameter left in place for compatibility with camera/optical samples already taken.
I really think 128 is not 128m, but 0.5 in 8bit value (even if the interface displays 128m - and this would be a display bug in the interface).
Good test would be to shot a grid or some relevant shape at different distances and see if correction is done when this 128 value appears in dop file.
Unless you already have evidences that correction is not done.
The only thing I know is the focus sliding isn’t there. I know on the lenses it is there there can br quite a lot of change
I have a hospital joy visit tomorrow and family coming in the evening for week end so it will be next week befor I could have a go. This is the problem of having DXO ignoring the fourm if there is a “simple” explanation they could tell us. But support have been very quiet since I contaxed them on this.
This could be a display and slider bug, but not a lack of reading focus distance when raw file provides focus distance.
D850 + 300mm PF and 500 mm PF should provide focus distance. And I get this 128 value too.
Slider needed when raw not provide the focus distance would so be the only bug.
We need in absence of anyone from DXO to fo some tests as you suggest. There are test charts on the Web so need hunting down next week. Sorry for the endlesd changes I do as Im dislexic and0 unless at using the android keyboard.
yes, when focusing distance read that parameter in DxO is used for focusing distance in meters and then 128 is some magical " flag that runs some computation " indeed … let us invent some nice theory to shield sloppy work in DxO …
and with long tele lenses DxO might as well just provide the same distortion correction regardless of the focusing distance as for if the lens focused @ infinity… and then they simply put 128m ( which is infinity in DxO parlance )
if you want to do test you need short focal lens where distortions are in your face very much based on focusing distance
There is also the problem of knowing if the lense normally does have distortion. Some of my old Canon lenes did I know but the rewiews of the Tamron was it was very good at not having distortion. So with out knowing what the lense dows it could be difficult, maybe some viewere like fastview might pick it up but I cant see any that jumps out at me with my A lenses. You will be safe from changes to the erres in this post, off to hospital now for a lovely day.
Sorry to my wifes anoyance, di a fast check with the two test ones oposted and coiuld see no diffrnce between eather of them between Fast View and PL not cleare if thats down to being no distorten or PL not doing any correction, I would tend to the latter.
Not inventing. Just trying to help.
When I’ve got a problem without answer I always try to find a way to solve it. This is my nature …
But I’m not doing architectural photography, and landscape or other kind of photography I do generally does no require very precise optical distortion correction. I often uncheck it. (I don’t do wide angle landscape photography - very strange, isn’t it ? - see below).
I’m more concerned by sharpness center to side and vignette corrections, even if I mostly use prime lenses that generally does not require this as much as some other lenses. (Won’t say that for 300 PF vignette full open 
which is huge !)
Indeed possible.
How do you know 128 meters is infinity (in DxO parlance) ? Just an assomption or do you know it for sure ?
Would this make sense to have same infinity distance for a 10 mm and a 800 mm ?
300 mm landscape
how about you test it yourself - take a raw file where DxO PL6 is known to read and use focusing distance and see what DxO displays you and writes to .DOP when it sees focusing distances past 128m… that simple
you don’t need to do such kind of photography to get issues - you shot can have something like a fence or a door frame behind a person and visibly distorted just because DxO did not automatically select a proper distortion correction instead leaving if for you to fix manually - which might time consuming ( and in cases with Fuji raws it can’t even be automated manually - like with the script I suggested to fix DOP files for Sony raws ) … it is not about precise corrections, it is about simply using the information available in raw file
that might be simply difficult to see vs geometric distortion that is easy to spot … however if, if DxO indeed varies application of sharpening, NR, etc, etc based on focusing distance ( for some lens ) that is just an additional icing ( just not very noticeable ) on the s$$$t cake… unless you fix manually each and every raw shot in DxO PL ( and in case of Fuji raws you can’t even script that ) you are not getting the “best” results out for lenses where DxO creates different corrections for different focusing distances - and with Fuji raws it is again the reason why raw converter shall offer an option to use manufacturer’s own optics correction and let users decide what they want to do - to use DxO optics modules or just use Fuji’s own optics corrections readily available in raw files ( unlike focusing distance which is not - at least here is not DxO’s fault ) … not to mention that users in many cases do not need to wait for DxO to create modules at all and can simply use manufacturer’s optics correction right away
Not all files carry the focus distance tag though. This also depends on makers and whether the file has been touched by some other app or not.
Canon’s “Digital Photo Professional” replaces the FocusDistance tag by other tag(s) that exist in DPP “recipes” only…and I’m not sure if PhotoLab reads Canon VRD tags…and I suppose that DPP does the same on both Win and Mac, which sould make DPL react similarly on both platforms.
This is why I told generally.
And anyway, I like to get what I buy. And I bought it. So it should work.
I don’t know if they sample this by distance.
I generally need to play with sharpness slider which is generally too high with new 1 default value.
So it anyway seems to have some subjective appreciation in how this tool operate. Don’t know if an objective “make my lens better” tool is fully realistic.
this was about Sony ARW raw files that do carry known tags and DxO PL6 not using that information … what .CR2/.CR3 files have to do with this discussion about Sony .ARW files ? if there are no known tags with focusing distance like in Fuji RAF raw files than nobody blames DxO for not getting that info ( nobody does ) … but if tag is present, known - then go and fetch it, instead sitting for years w/o any progress … and this is even more stupid given the claim about the effort allegedly invested in lens testing ( not that sane people believe that all combos are really physically tested, not to mention that proper test like LensRentals does involves many samples of lenses and not just one )
I do not know either, but what if ? chromatic aberrations certainly can be affected by changing the way optical elements of the focusing group are positioned inside the lens , again that might be less noticeable vs geometric distortions ( and of course lateral CA are just how RGGB,CMYE, whatever other sensel arrangements we have, are geometrically distorted vs each other )
