RAW Truth

[kevinparis]

Just as a point of illustration.. no wars started here

here is a shot taken with RAW on a E-p1 with 17mm lens

first taken through the hell known as Olympus Master.. the software that comes with the camera

second is taken through Aperture using a wee applescript i wrote to spoof the file to look like it came from a e-30

ignore the tonal stuff.. but just look at the distortion in the vertical pillars... the camera manufacturers are doing a lot in RAW that programs like Aperture don't recognise... or even have the tools to deal with... this may explain why aperture doesn't support the latest cameras like the e-p1 and the LX 3

the futures bright .. but the futures wonky

K

 

[Ranger 9]

Welcome to the world of "opcodes" -- special flags that can be written into raw files to trigger software features such as distortion correction.

Obviously Aperture doesn't read E-P 1 opcodes, at least not yet.

I wonder if you can try putting the same file through an opcode-enabled application such as Lightroom (maybe the downloadable demo version?) or the latest Camera Raw...? Do you still see the same types of bending you get in Aperture?

 

[kevinparis]

ranger...

maybe if i have time i will try other routes. I am not excessively bothered by the distortion.. but was surprised at how pronounced it was... which was my reason for highlighting it.

My worry is that even if aperture could read the information i don't think it has the tools to deal with it

ps thanks for the B/W printer feedback to Cam... I appreciate your comments


cheers
K

 

[photoworkplace]

I have no 4/3rds camera (yet)
But I do have a question
I would have expected in the optically uncorrected raw image their would be more of the image, so that correction can take place while maintaining the same field of view, but that does not seem to be the case.
Is there any cropping taking place?
There seems to be more room on top on the second one than the first (above the head of the guy with the newspaper), compared to the bottom of the both images (looking at there shoes)

 

[Terry]

Adobe products handle this just fine. If you were to bring the raw either into Camera Raw or Lightroom, it will automatically be corrected.
For me, Aperture is a victim of bad timing. When I needed to make the LR/Aperture decision they didn't support two of my three cameras. A year later I went out and bought Aperture thinking I was going to give a go to switch over. So, glad I didn't because they haven't dealt with micro 4/3 and it has been 1 year since the launch of the G1.

 

[nostatic]

I am thinking of ditching Aperture for LR simply because Apple seemingly has no interest in timely support of cameras. They still do not support the DLux4/LX3, or E-P1, or G1. I am happy with the application otherwise, and prefer the UI to LR, but I pretty much gave up shooting raw with the DLux4 because it was too much of a pita to use another program. Same when I had the E-P1.

Aperture seems to be a red-headed stepchild at Apple. FinalCut has had multiple updates while Aperture sits in a state of suspended animation (pun intended).

 

[Terry]

Probably because the LX3/D-lux4 have lens corrections.

 

[nostatic]

As does the E-P1. And the G1 I believe. What I fail to understand is how LR can support them but Aperture doesn't. I could see it if every main software package said, "nope, we can't handle the in-camera lens correction, but it seems that only Apple throws up their hands. Frustrating and I can swap my Nik plugins from Aperture to LR if I want to switch. I do prefer the Aperture interface, but hate having a more convoluted workflow. I'm about to consolidate my various libraries from the past 4 years so now is the time to switch if I'm going to...

 

[barjohn]

The thing many people fail to catch with many of the P&S and even some high priced cameras like an M8 or M9 and probably the X1 is that some vendors process lens corrections in their processing even for RAW files while other vendors provide information in the RAW files that can be used in post processing to make the corrections but if not used you see the lens' flaws (distortion or CA). In the M8 and M9 coded lenses give you better results because Leica can process in corrections based on the known characteristics of the lens. No one will ever know how really good or bad the X1 lens is since you can't remove it from the processed image. If Oly or Panny had processed in the corrections into the RAW file you would have no idea how good or bad the distortion or CA was unless you tested the lens by itself. I think that with known lenses they should process in the correction so that no matter what post processing software you use the images will appear to be optically the same.

 

[pellicle]

of course you could always use dcraw to convert your file to a tiff and see what it is like natively...

also, this is lens dependent and is not in my opinion a camera issue. I've used stuff like PTLens for some time which already has a database of lenses that it can work with (reading the exif of JPGS too)

What I think it means is that you need to consider carefully lens testing results in future ... was it 'inadvertently' post processed?

is that a problem?

If the designers of lenses can make better / cheaper lenses with some compromises which need to be fixed in software (or can be fixed in software for less than re-balancing other design equation parameters) then isn't that just a benefit of digital?

I get better pictures out of my 1940's Vaskar lensed Bessa 6x9 film camera by using PTLens ... sharper edges. Seems like only a good thing to me.

 

[Brian Mosley]

Yes, this has been a pain for me too... the 20mm f1.7 is much better with the E-P1, and also check out PhotoAcute if you want a distortion/CA corrected DNG file for your favoured raw converter.

Unfortunately, I think PhotoAcute only runs on Windoze.

Cheers

Brian

 

[kevinparis]

morning....

in response to photoworkspace - yes the second image was cropped - I went back and did the Olympus Master version purely to show the lack of distortion.

to the rest of you

I am an Aperture user by default.... I worked at Apple over the period when Aperture was first launched and have had to present it at trade fairs and other events as well as write demos.

I still like its workflow and am annoyed like you that it doesn't support the latest cameras - but I am guessing that this is a result of the the way that Apple handle the RAW processing at the core of the OS and that they didn't build in the algorithms to correct distortion. I reckon that any camera out there that uses this sort of correction will have problems with Aperture, and indeed iPhoto

Adobe on the other hand because they have to develop for both windows and Mac had to develop ACR as their RAW engine and had the foresight to build these sort of control s in there

cheers
K

 

[ChrisJ]

It raises the question as to why Olympus haven't reduced this distortion optically which in a single focal length lens they could and IMO should have done. These lenses aren't cheap in all other 'normal sized' optics this problem is addressed in lens which makes the lens design more complicated (and expensive).

As some of you may know I've been working with the old Pentax Auto 110 lenses none of which including the 18mm f2.8 have this problem so it can't be a small size thing.

May be our wrath should be directed more at Olympus for not giving us a higher quality lens the price would suggest were getting, rather than Adobe.

Chris

 

[Brian Mosley]

Chris, I generally agree with you - I know that other people say 'get over it' and accept that these lenses have been 'designed' to be easy to correct... but that sounds like apologetic / marketing BS to me - I'd like to see the 20mm f1.7 become the minimum standard of IQ for m4/3rds from any manufacturer!

Cheers

Brian

 

[kevinparis]

This throws up another issue that may or may not be a problem.

If deficiencies in the optics of a lens are corrected in the camera by means of software, then presumably all the manufacturers of a system - in this case Olympus and Panasonic at the moment, will have to agree on a common way to deal with these corrections.

They will have to recognise which lens they are dealing with and apply appropriate corrections, despite each manufacturer having their own processing engines.

Maybe its the lens that tells the camera body what corrections to apply, otherwise you would have to update the camera firmware everytime a new lens came out.

Think what I am trying to get at is that there is a danger that unless the whole thing is tightly controlled you could end up with a situation where a lens from one manufacturer worked better on their camera than it would on another from supposedly the same system.

Olympus and Panasonic can manage this at the moment - but if more manufactures join micro 4/3 then it will probably get hard to maintain compatibility. It also may be a barrier to third parties producing lenses for the systems as they may not wish to share information at this level.

Anyway... nothing to lose sleep over.. just some thoughts

K

 

[Brian Mosley]

I think the design does allow for both distortion and CA 'parameters' to be supplied by the lens and stored in the raw file - but even Olympus and Panasonic haven't seemlessly coordinated their approach...

As you say, other manufacturers may face an artificial barrier... or, maybe they will just produce good optics :deadhorse:

Cheers

Brian

 

[Terry]

This was supposedly done with the new DNG standard and the opcode specification.

This throws up another issue that may or may not be a problem.

If deficiencies in the optics of a lens are corrected in the camera by means of software, then presumably all the manufacturers of a system - in this case Olympus and Panasonic at the moment, will have to agree on a common way to deal with these corrections.

They will have to recognise which lens they are dealing with and apply appropriate corrections, despite each manufacturer having their own processing engines.

Maybe its the lens that tells the camera body what corrections to apply, otherwise you would have to update the camera firmware everytime a new lens came out.

Think what I am trying to get at is that there is a danger that unless the whole thing is tightly controlled you could end up with a situation where a lens from one manufacturer worked better on their camera than it would on another from supposedly the same system.

Olympus and Panasonic can manage this at the moment - but if more manufactures join micro 4/3 then it will probably get hard to maintain compatibility. It also may be a barrier to third parties producing lenses for the systems as they may not wish to share information at this level.

Anyway... nothing to lose sleep over.. just some thoughts

K

 

[jonoslack]

HI Kevin - Brian - Terry - Nostatic and others.
I'm also an Aperture user, like Nostatic I've been considering changing to lightroom because of this.
But I still like Aperture better - both the interface and the raw processing. I was hoping that this issue would be fixed with Snow Leopard, but apparently not. Changing to Lightroom is a huge task, if I thought it was better I would embark on it . . . but I don't.

Chris - I understand that it's easier, cheaper and SMALLER to design lenses without having to deal with the correction, but to put it in the software instead. However, IMHO it would be better if this correction was done when the RAW file was produced rather than as instructions for the processor (even better if there was an option to switch it off).

Leica have added vignetting and IR correction to the DNG files, which can then be opened by any DNG aware program, which, to be honest, is one of the biggest reasons my Olympus E-P1 is in Cornwall with Silas and not sitting in my camera bag.

Terry - you're right, it's in the DNG standard, but converting E-p1 files to DNG using Adobe converter doesn't embed these lens corrections. I'd be quite happy if Olympus and Panasonic output DNG files with compliant lens corrections which compliant software would deal with (in which case Apple would really have to get up off their backside and provide proper core support). At the moment it seems to be entirely proprietary and unregulated, and I don't think one can blame Apple for not immediately coming up with support for something that looks like a moving target.

I think this is a serious issue, It's really a deal killer with me and m4/3 (incidentally, just in case someone doesn't realise, this is NOT an issue with 4/3 cameras and lenses, only m4/3). Not just because I can't use Aperture, but because it really does seal you in with converters which will deal with the problem.

What happens in 5 years? I want to keep my library as RAW, I DON'T want to have to convert everything to tiff (which also rules out C1). To me that means subscribing to DNG - with the E3 and the Sony I can convert the files to DNG and still get the benefit of the camera specific profiles in Aperture (or lightroom) - with m4/3 I can't do that and keep the lens corrections.

For me that means no m4/3, really it's as simple as that.

 

[kevinparis]

Just as an aside - we have to remember that in relative terms the modern lenses we are talking about are really inexpensive compared to their forefathers.

Just for a giggle i googled some old price lists. I found a Minolta one from 1970 - when a 35mm 2.8 lens would cost 125 dollars. (I a sure i could find other price lists - but thought Minolta was a good middle of the road reference point)

using this page http://www.dollartimes.com/calculators/inflation.htm

I discovered that in todays money that would be nearly 700 dollars

The inverse of that is that the 300 dollars you pay today for the 17mm is only 50 dollars in 1970 money.

I know there are other variables to take into account - but it does put things into perspective

Anyway - the conversion website is fun...

K

 

[Terry]

Chris, I generally agree with you - I know that other people say 'get over it' and accept that these lenses have been 'designed' to be easy to correct... but that sounds like apologetic / marketing BS to me - I'd like to see the 20mm f1.7 become the minimum standard of IQ for m4/3rds from any manufacturer!

Cheers

Brian

Then what good does it do to start threads calling the lens the most underrated m4/3 lens. Not sure how that gets the right message to the manufacturers.

http://forums.dpreview.com/forums/read.asp?forum=1041&message=33244076

I look at the situation 2 ways. I'm pretty convinced the reason that we see the soft corner problems on the legacy wide angle lenses is related to the complications that the M8/M9 has with the short register distance and the need for micro lenses. If the only way to go wide in this format is to make lenses with corrections then I'm OK (the tradeoff being avoiding a hugely expensive sensor). Technically, I could be way off base and I'm not a physicist but I do think physics is involved. I'm specifically talking about the 7-14 here. I'm not sure where the crossover point (what focal length) lens design gets easier.