Wikipedia:Featured picture candidates/Purple-IrisPenu.JPG

Purple Iris

Original - Iris germanica

Edit 1 - went back to originally uploaded and QI'd version, selective NR, slight local contrast enhancement, no pseudo DOF background

Edit 2 - went back once more from scratch and reworked the entire thing. The flower retains high detail as it is not blurred at all.

Reason: A high resolution, visually appealing picture of a well classified image
Articles this image appears in: Iris (genus)
Creator: myself

Support as nominator -- penubag (talk) 01:36, 13 April 2008 (UTC)[reply]
Oppose there's some strange and obvious masking gone on in PP round the edge of the bottom petal. --Mfield (talk) 03:12, 13 April 2008 (UTC)[reply]
Oppose per above, though the masking extends around pretty much the whole bloom (jaggies in upper right). Also, I don't know if it's the masking or the light, but the bloom appears to faintly glowing. Matt Deres (talk) 03:44, 13 April 2008 (UTC)[reply]
- I've fixed some of my stupid editing errors -- penubag (talk) 04:09, 13 April 2008 (UTC)[reply]
- There's still parts where the cutout is visible. Suggest you need to go right back to a much older version and work on how you originally made/feathered the selection rather than trying to blur it back in. Mfield (talk) 04:18, 13 April 2008 (UTC)[reply]
This is just PP taken past excess. To be honest, the original looks far better than any of the edits; gaussian blur makes the backround look fake, the contrast and colors are unrealistic, and the flower is still glowing. Looks like it got the glow in the third upload, like a gaussian-blur level was lightly blended with the whole image. Noise reduction is one thing, but this is something else. When I need to reduce noise, I usually start by applying creating a new layer with a low-radius, medium threshold dust and scratch filter, then, if necessary applying a small gaussian blur. I then blend it back into the original at 70%—I find the lack of noise more disturbing than the presence of it. This takes it just a bit overboard. I'd recommend taking the original and starting over with a more conservative approach. Thegreenj 04:22, 13 April 2008 (UTC)[reply]
- If you're interested, here is about where I'd put it. Thegreenj 05:21, 13 April 2008 (UTC)[reply]
  - Thank you very much for the tips, I will definitely take note of this -- penubag (talk) 22:05, 13 April 2008 (UTC)[reply]
Oppose - badly masked and blurred (and there's still bits missed off the mask, if you have Photoshop CS3 try using the Refine Edge tool on the selection). There's still bits missed from the mask too. Is it really impossible to just go back and take the photo on a lower ISO speed with a wider aperture to get the DOF? —Vanderdecken∴ ∫ξ φ
- The flower is seasonal and won't bloom for another year, I will do as you suggested though. -- penubag (talk) 22:05, 13 April 2008 (UTC)[reply]
Comment, can you upload the original non processed version and see if maybe a lighter hand can bring that up to featured? Because generally this is a well done picture. And can you more explicitly state on the photo page how you've manipulated this image so we have a clear record using Template:RetouchedPicture. gren グレン 21:20, 13 April 2008 (UTC)[reply]
- Here is the original with slight mods by Thegreenj. I've put the template on the nominated image page. -- penubag (talk) 22:05, 13 April 2008 (UTC)[reply]
  - I definitely think that is better... but, I'll put Neutral for alternative. But, I definitely like the photo and think it should be used in the articles. gren グレン 15:40, 14 April 2008 (UTC)[reply]
Oppose Very blurred. —αἰτίας •discussion• 21:26, 13 April 2008 (UTC)[reply]
Comment Added an edit from the version originally uploaded (and QI). Performed different levels of NR on the flower and the background to try and retain as much detail in the petals as possible. Slight local contrast enhancement but no saturation boost as in 'Original' and left the background unblurred. Mfield (talk) 23:02, 13 April 2008 (UTC)[reply]
Oppose both for the blur. The freddinator (talk) 01:08, 14 April 2008 (UTC)[reply]
Oppose per above. crassic!^[talk] 23:56, 14 April 2008 (UTC)[reply]

Complete rework of editing

Comment- I reedited the whole flower using your suggestions, please give feedback, thanks -- penubag (talk) 16:33, 15 April 2008 (UTC)[reply]

This should have been uploaded as a new image rather than over the top of the original as now the existing votes do not refer to the image they were judging. Mfield (talk) 16:36, 15 April 2008 (UTC)[reply]

But while we're at it... I think that you need to re-evaluate what you are trying to accomplish by editing this image. Going from the original to the newest edit, I see a huge loss in texture (bad for encyclopedic value), a gaussian blurred background (looks unnatural), masking edges and errors still (soft brushes are your friend), and a general sense of over-editing. I find it impropable that the image would pass QI in its current state. IMO, you are best sticking with the original until you have an end goal in mind, whether for improving the enclopedic or aesthetic value, or both. Remember that users are not expecting an ideal or noise free image, and I think you've still gone a bit too far sacrificing things we are looking for (detail, natural foreground-background transitions) for things we're not. I'd take a closer look at Mfield or my edit, both of which show a more reasonable amount of editing. The best thing, however, would be to get it right at the start; when the flower blooms again, set your camera at a reasonable ISO and f-stop to get a more realistic-looking effect. Thegreenj 21:39, 15 April 2008 (UTC)[reply]

Complete rework of editing (again)

While I have the flower for this year, I might as well try my best with the editing. I went back (again) from scratch, and produced Edit 2, using a 1px feathering tool to select the image. The flower is not blurred at all, thus retaining detail, however, I might add, de-noised. This time around, I hardly tampered with the original colors of the flower and left the saturation relatively untouched excluding a few minor spots. Please reassess, thank you. -- penubag (talk) 02:30, 16 April 2008 (UTC)[reply]

Give this a week or two, then renominate the last edit - with all the edits and such, it's probably gotten too complex for a simple picture to get voted on fairly. Shoemaker's Holiday (talk) 19:45, 20 April 2008 (UTC)[reply]

Not promoted . --John254 02:30, 21 April 2008 (UTC)[reply]