A problem with RAW files

[Detail Man]

We did it !

I joined-up on the RAW Therapee Forum, and posted this bug-report:

http://www.rawtherapee.com/forum/viewtopic.php?f=3&t=3620#p25673

DM ...

 

[Nicholas Bell]

Win 7 Home Premium 64 bit

 

[peweuk]

Thanks for all your input guys.

peweuk, is it true that a brown colored wall appears near the far-left edge in your indoor shot posted ? It's rather hard for me to tell by looking. I had originally interpreted that to be an absence of image-information (due to it's solid "uniform-field" appearance) ...

Yes there is a wall on the right ( sorry I didn't mention it - also the wall is in reality green, not the beige or brown it appears to be in the shot, which I can see confused the issue).

A raw file actually has more detail at the edges than is processed in the main stream raw converters such as ACR or silkypix. Raw therapy is not cropping off teh very edges captured in the raw file.

I sort of knew that the raw file was larger, but didn't realize that different software displayed it differently as both SP and ACR are the only two I had experienced and as you say they 'crop'.

SO the conclusion I seem to be getting from this thread seems to be that the 'fright' I got was more due to RT than any significant failing of the camera.

But there are 2 things which still confuse me

1. Why is the clipping only at 25mm. Is that perhaps due to the position of the lens at that setting not being quite far enough forward to clear the edges of the Barrel - but thereafter clearing it - or some other reason. (Similar to the appearance with low zoom when adding a TC to the camera).

2. Why is the clipping not even on all four corners - this would imply that the sensor is not central to the lens/barrel (assuming 1. is correct) - if that is the case, is it a problem that needs addressing?

Thanks again guys.

 

[Steen Bay]

But there are 2 things which still confuse me

1. Why is the clipping only at 25mm. Is that perhaps due to the position of the lens at that setting not being quite far enough forward to clear the edges of the Barrel - but thereafter clearing it - or some other reason. (Similar to the appearance with low zoom when adding a TC to the camera).

It's quite common these days that lenses on compacts are designed like that, with lots of distortion at WA, and sometimes severe vignetting too. The images are meant to be corrected in-camera, or in the RAW converter, before we see them.

2. Why is the clipping not even on all four corners - this would imply that the sensor is not central to the lens/barrel (assuming 1. is correct) - if that is the case, is it a problem that needs addressing?

The lens could be centered a bit better, but that's probably the case on most cameras.

 

[Detail Man]

peweuk wrote:

1. Why is the clipping only at 25mm. Is that perhaps due to the position of the lens at that setting not being quite far enough forward to clear the edges of the Barrel - but thereafter clearing it - or some other reason. (Similar to the appearance with low zoom when adding a TC to the camera).

2. Why is the clipping not even on all four corners - this would imply that the sensor is not central to the lens/barrel (assuming 1. is correct) - if that is the case, is it a problem that needs addressing?

Troubling, isn't it ? If you read my (edited and addended) bug-report here:

http://www.rawtherapee.com/forum/viewtopic.php?f=3&t=3620&p=25673#p25673

... you can see what I think that the RT 4.x developers ought to do about it.

(As I already posted), I recommend that you download a RW2 or two from the Photography Blog FZ150 revue Image Samples web-page (which I am unable to access), and see if a (3rd) FZ150 has the same non-symmetrical bias towards the right edge. If so, it (may) well be a result of the FZ150 design (and/or a lack of quality control in FZ150 production-testing).

It is, however (to me) unclear whether what can be seen is the result of an actual "vignetting" effect, or (may) also result from actual software code-routines in RT 4.x itself ...

It's funny that neither the FZ50 or the FZ28 (RAW and RW2) image-files, respectively cause this problem (at all, at any Focal length, including full wide-angle). But maybe not so funny for FZ150 owners ? Panasonic's use of a larger number of extra photo-sites (4144 in the horizontal dimension) may have something to do with this phenomena appearing in the FZ150 (and FZ100?).

The RT 4.x developers may (possibly) "kick it back" to Panasonic (if Panasonic is responsible for the asymmetry). In any event, it seems (to me) that the RT 4.x developers need to work around whatever kinds of manufacturing flaws that Panasonic is "pumping out" - or else RT 4.x will remain unusable for FZ150 RW2 (unless the user specifically has to crop the previewed image-frame in order to achieve acceptable results).

Nevertheless, the 64-bit versions of RT 4.x are here said to not exhibit the same effects. For us folks with 32-bit Win OS only, that may be a disadvantage ...

You're on your own at this point. I'm sure that some luminaries will guide you farther with their deep understanding of the subjects at hand (as well as subjects that exist only within their minds).

DM ...

 

[peweuk]

Thanks again Guys.

I have put out some feelers to try and get some more FZ150 (and FZ100) 'real life' raw sample.

From what you guys have said so far I feel more comfortable that what I have is not a unique or serious problem, and I can live with that, although I will keep any eye on any developments at the RT forum.

I have got ACR with Lightroom and Photoshop, so maybe I am best to just stick with those for dealing with RAW, as they seem to display correctly (to my eye anyway).

The issue only arose because I decided to try RT after seeing another post about it.

All this fuss because I thought I'd try out a 'freebie' ;-)

Thanks again guys.

 

[Egret]

A very unlikely possibility, but do you have a filrer on the lens? I have read that at maximum wide-angle some wide angle lenses may show this type of effect with filters which have a thick (wide) rim that projects forward enough to be seen by the lens at wide angle. As I understand it, this is one of the reasons for the more expensive thin rim filters. I do not know if the wide angle on the FZ150 is wide enough to show this effect. Obviously if you have no filter it can not be that.

 

[Detail Man]

peweuk wrote:
All this fuss because I thought I'd try out a 'freebie' ;-)

Well, RAW Therapee has a lot going for it (and is in continual improvement) - all at no cost to you !

Have been able to get to the Photography Blog (PB) FZ150 revue Sample Images. There "bookshelf" test-set appears to likely be the only FZ150 RW2s that were shot at full wide-angle. Note that the FZ150 provided to PB is not asymmetrical :

http://img.photographyblog.com/reviews/panasonic_lumix_dmc_fz150/sample_images/panasonic_lumix_dmc_fz150_13.rw2

... where as the FZ150 provided to IR (like yours) is asymmetrical :

http://www.imaging-resource.com/PRODS/FZ150/FZ150hVFAWB.RW2.HTM

Nevertheless, I find that (even) when using the RT 4.0.4.2 geometric distortion-correction tools to correct the (symmetrical) PB RW2, and when cropping to a 4000x3000 image-size (only), there still is some corner-vignetting visible.

So, it seems that the RT developers may have some work to do ... (at least) there is a chance that the RT developers may respond to feedback, and (as a result) do something about the situation (for the Win 32-bit version of RT 4.x).

On the other hand, Panasonic's QC department may have unit test-standards to shore-up and improve upon. The chances that the monolithic and historically unresponsive Panasonic (once they are shipping a product, and the pipeline is already full of such units) would ever listen to (much less respond) to any user(s) complaints about manufacturing standards are about the same as a "snowball's chance in hell" ...

The asymmetry of (2 of 3 here sampled) FZ150 lens-systems (including the one provided to IR) serves to worsen the problems that RT 4.x is exhibiting . Thus, I can see that the RT developers may well face a conundrum here - in that their best-made tweaking efforts may not be able to compensate for what Panasonic chooses to be their production-test standards (or lack thereof)

 

[peweuk]

Obviously if you have no filter it can not be that.

I have a filter on the camera, but have tried it without - same results.

 

[peweuk]

peweuk wrote:
All this fuss because I thought I'd try out a 'freebie' emoticon - wink

Well, RAW Therapee has a lot going for it (and is in continual improvement) - all at no cost to you !

I personally am not sure what the advantages (if there are any technical advantages as well as cost) RT may have over other programs due to having, as yet, no experience of working with RAW files - although I am looking to change that over the foreseeable future.

In the short term it may be that other programs offer an 'easier' route (ie not so complicated or so much work) to achieve acceptable results whilst 'learning', maybe not - until I experiment I won't know that.

Have been able to get to the Photography Blog (PB) FZ150 revue Sample Images. There "bookshelf" test-set appears to likely be the only FZ150 RW2s that were shot at full wide-angle. Note that the FZ150 provided to PB is not asymmetrical:

http://img.photographyblog.com/...ple_images/panasonic_lumix_dmc_fz150_13.rw2

... where as the FZ150 provided to IR (like yours) is asymmetrical:

http://www.imaging-resource.com/PRODS/FZ150/FZ150hVFAWB.RW2.HTM

Thank you for taking the time and trouble to locate and evaluate these.

You are right that the standards set by Panasonic do seem to allow what appears to be a significant difference in 'quality'.

However from my point of view, in relation to the reasoning behind my original post, it appears that what I perceived to be a problem is not isolated, and may by many not be regarded as a problem.

I even wonder how many other cameras (different makes and models) have similar issues if viewed in RAW, but are either accepted as normal, ignored or not even known by the users.

Although I bought the FZ150 knowing it was not the perfect camera (what is?) I bought it as it offered my what I was looking for within the budget I allowed.

My concern at seeing the 'clipping' in the RAW files was only that the camera may have a fault that needed dealing with - and would be dealt with by Panasonic within the Warranty period.

It now appears that this is not the case - so I am back to being content again as the JPG images I get are acceptable for my needs, and if with processing RAW I can in due course improve quality, even better.

Thanks again for all the support and input.

 

[Detail Man]

peweuk wrote:

However from my point of view, in relation to the reasoning behind my original post, it appears that what I perceived to be a problem is not isolated, and may by many not be regarded as a problem.

Well, when you perform the geometric distortion-corrections in RT 4.0.4.2 (32-bit version, at least) for the Photography Blog test-shot that I previously linked-to, and a 4000x3000 pixel sized preview is not free from this vignetting effect (and from the symmetrical FZ150 lens-system), it's a problem

This effect is even more pronounced in the asymmetrical case (in your unit, and the Imaging Resource tested unit). The FZ50 RAWs and FZ28 RW2s do not exhibit these problems whatsoever

 

[Steen Bay]

I even wonder how many other cameras (different makes and models) have similar issues if viewed in RAW, but are either accepted as normal, ignored or not even known by the users.

Think it's likely that the "issues" (which aren't really "issues", but more a design compromise) aren't even known by most users. For example, I posted a link above, showing the rather extreme distortion in an uncorrected S100 WA image, but normally people will never see it, because it's corrected automatically by the camera or the RAW converter. What people do notice however, is if the lens is so seriously decentered that it shows in the normal/corrected images, like it's the case with DPR's S100 studio test shots (and the published images are even the best from two tested cameras).

 

[Gerry Pasternack]

This vignette just popped up on one of my RAW files. Do not think it was there when I first loaded the file. I'll continue to investigate.

--
Gerry

Take a look at my photography at:
gerryp123.zenfolio.com

 

[Detail Man]

For those reading this particular thread in reference to the RAW Therapee bug-report posted on the RAW Therapee Forum, please (also) refer to information posted on this other DPR PTF thread:

http://forums.dpreview.com/forums/readflat.asp?forum=1033&message=39907061