Quote from: leszek on February 07, 2008, 16:20

2)   I do not think that the image has too many artifacts � if any, but most definitely it has issues, and in the end I think that I have botched this one up rather badly. The final result is not what it could be � and �artifacts� is a convenient scapegoat here. I put this vase on the table, plyed with lighting � and most of people refer to as �artifacts� are actually there � mostly internal reflections and distortions due to rapidly changing thickness.

I fear with this you have not seen what some of us are, glass is smooth and there are jagged artifacts which just should not be there at 100% percent, period.

I'm afraid thesentinel is right. Please don't take this as harsh criticism, it's not my intention at all, but really that glass has jagged edges in your photo and I highly doubt is like that in reality. If I wouldn't know anything about the image I'd say it's a compressed jpeg scaled up to greater size, it really looks that way. Since your camera is innocent and beside the focus problems you're evidently able to shoot (I wear strong corrective glasses too and focusing is a pain sometimes, sigh, I know the feeling!) the problem is in post-processing.
In CameraRaw you have to check you didn't inadvertely a different image size than the original one (the pull-down setting is in the bottom of the ACR window), then put to zero both sharpening and noise reduction.

P.S. the points where there are jagged edges are the ones already pointed out by araminta. These ones don't look like "natural" glass imperfections, neither a focus problem. Those ones come from a poor image interpolation.

Example of jagged edges magnified at 600% (but still very visible at 100%)



You see those hugely squared borders? Is the glass made that way? I doubt so.

will redo
 

araminta and ale1969: this is exactly the kind of feedback I am looking for. Thanks a lot - one picture is worth a thousand words

There is no question  about the stem - something went wrong there - and I will track it down (I have to say here that I usually do not produce this kind of thingies...)

Regarding the 600% crop posted by ale1969 - I will have a look at the RAW file - but I am afraid it looks the way it looks: even at 100% one has to see the jaggies: there is a lot of curvature and color change in that area - and the resolution of human eye is about 300 pixels/inch at a normal viewing distance of about 10" (or 6 to 8 lines per milimeter if you prefer). At 100% on the screen (which has 80 to 90 pixels per inch) you see it at least 3x magnified (depending on the resolution of the monitor) - which means that you have to see individual pixels. And with a sudden change in color - you will see jaggies.

So - yes, it is quite possibly that these are the issues flagged by the reviewer, and if this is so - I think I know how to deal with it. Whether or not they should be described as "artifacts" is another story - but it is quite acceptable to me that they should not be there - and I think this can be achieved by different lighting or slight change of angle.

The stem - it is a cockup though, no questions asked.
The only saving grace is that I printed it 10"x8" and it does not look too bad, really

The whole point of this thread was to find out exactly what and where the "artifacts" are, so I can deal with the issue - so thanks again for your help.

"Overfiltering" is another favourite of the reviewers - and I have an example where overfiltering was quoted as a reason for rejection - while the image was a straight conversion from RAW (all settings in neutral), plus 250% unsharp mask with 0.4 pixel radius - so sometimes I have difficulties with accepting things at a face value.

So - I will check the RAW and will re-shoot the glass to see what can be improved and to what degree the arifacts are really artifacts (not questioning the fact that whatever they are, they should not be there).

As I said before - thanks for posting the 600% crop. I see that I need to clarify a bit what I said in the previous post: I was not questioning the usefulness of this - all I meant was that you will always see jagged edge on a 100% crop if the contrast is high - as in the posted crop. No one can help it because human eye can resolve pixels on the screen (that's why I put the numbers in).

Yes - I can see the jagged edge there at 100% - and it is in RAW file too. It shouldn't be there -  but this is no JPEG artifact. I will post 100% crop from RAW with all settings in neutral. The pic is quite possibly botched up - but I really want to get to the bottom of the real issue (focusing and possibly RAW conversion aside), not to start chasing non-existent issues. That's why I apprecciate a lot when people include a crop to illustrate, as only then we all can be sure what it is that we are talking about.

So - I am not arguing, really

Point taken. It is evident that I did something wrong - as witnessed by a great number of excellent shots of glass.  Believe it or not - I am more interested in getting to the bottom of it than defending myself. Just give me a couple of hours to get back home and have a look at RAW. I indeed played with curves - but I wouldn't think it was excessive. Maybe I was wrong.

The camera: 1DMkII + 24-70/2.8L. RAW with minimum of adjustments � barely touched in PS. ... Saved as JPEG in CS3 with quality=12 (max).

The 1DMkII has a resolution of 3504 x 2336 (8.2 MP).  The image as posted has a resolution of 3000 x 2400 (7.2 MP).  Why the vast difference in resolution?  How did the height get to 2400 when the camera can only do 2336?

"So the reviewer was spot on"

Well - there is a comparison of RAW with no adjustments and the finished image - just above. You need to tell me if these are artifacts or not. I still would not call it artifacts. The lighting was not the best, the stem out of focus, but jaggies are there and will be there in anybody's picture when photographing similiar subject.

I just have spent about 3 hours experimenting with light and different lenses. And different subjects. I am getting perfectly sharp and well exposed images - of different subjects. This glass vase, while beautiful to look at, is extremely difficult to photograph.  It needs different lighting - and trying to produce white (or close) background simply does not work. I put separate light on the background trying to blow it - and it worked, except that the reflected light went straight through the clean glass into the camera - swallowing most of  detail in the glass. The contrast between white background and the glass is very low - that's why perceived sharpness is low. The light needs to be set up to produce reflections on the edges (except that the corners are rounded, so no cigar). This leaves me with underexposing the picture , playing with curves - and using clipping path to produce 100% white background.

I guess if you tried to do that (or if you mastered photographing this kind of subject) - then you know what I am talking about. If not - then just for the kicks - get a similiar piece of glass and let me know how it went.

I made several errors - but the largest one was the selection of the subject. The total DOF is under 3" - which is less than the diameter of the vase (and less than the length of the stem). I could go on and on. So - I think I know what went wrong, and I mostly know now what I would do better next time.

Why I have been rejected so many times ? I guess the type of photos I take and the amount of processing I put in the images is a big no-no at iStock. I am not arguing with that - but I am not entirely sure that I want to conform. Some may say that (255,255,255) background equals quality. For me - it is just painting pixels for the sake of it (if a clipping path is provided - then the designer can extract the subject in 3 seconds anyway). I guess my perception of "quality" ends somewhere at 50% magnification - which on the screen still produces an image about 20" high (or wide).

You see - I am not really arguing with the reviewers. They are 100% right - according to their standards. I just need to decide whether it is worth my time to meet those standards, and essentially converting to entirely different type of photography to meet their requirements.

I am not doing it for money - I would have to have many hundreds of downloads a day to match my daily job (aircraft design). I just like taking pictures that I like, not pictures which have debatable jaggies at 100% or at 200%. How long is a piece of string, after all ?

I am still very grateful for comments and help I have received here. It opened my eyes to couple of issues and helped me to find out that the bloody comp was misbehaving. It helped me to look with a different perspective at what is important at a microstock site. Perhaps it even helped me to take better pictures - although I will let others to decide.

Cheers 

Thanks - GeoPappas and yingyang0. This is a useful bit of information
I usually do not upsize as such - just crop 8x10 at 300 ppi (most frequently using the whole 2336 pixels. So - strictly speaking there is no much upsizing there  some 3% in one direction only. But - thanks again, I will keep the image size according to your advice.

Thanks CCK. This is again a useful bit of info. I guess I will have to change the workflow: first RAW conversion and REALLY minor adjustments, then post-processing for the other sites - hey, no post-processing means taking away half of the fun

As much as possible ? Well, it probably is OK, but I can assure you - not for initial application 

i think the no editing rule mostly only applies to istock.  Other sites seem to have good sales with 'photoshoped' images.