Canon 1d Mark IV, 500mm F4,
ISO 1600, F5.6 1/1250
For Rachel,
Sorry I missed 6th image in the series. Here is a link to the image on my 500ox
http://500px.com/photo/18577979
Canon 1d Mark IV, 500mm F4,
ISO 1600, F5.6 1/1250
For Rachel,
Sorry I missed 6th image in the series. Here is a link to the image on my 500ox
http://500px.com/photo/18577979
Last edited by Peter Kes; 12-09-2012 at 06:02 AM.
Hi Sid - Another great shot and a slight departure from the rest of the series with the bear in the water and it looks like the light was a little flatter. I really like the low pov, the visible claw and of course the IQ. I would call it "Peek-A-Boo." There's a spot just left of middle and slightly down from the top edge that is probably just darker ice but looks like a dust spot, I would either clone it out or blend it in a little more.
Thanks for the link to #6, that one is also great with excellent interaction, yet a nice simplicity to the image.
TFS,
Rachel
Hi Sid, yep the 'Peek-a-boo' look works, but do you have any more of the reflection?
Shooting I assume from a zodiac works a treat, and being at eye level links nicely with the viewer, which I feel cannot be beaten. If you haven't yet, look at the MKIV ISO settings and what you can adjust when shooting high.
Nice work Sid.
TFS
Steve
Post Production: It’s ALL about what you do with the tools and not, which brand of tool you use.
Splendid image Sid. I like it a lot.
Andrew
Like all the others, it's great. Love the playful peek-a-boo pose and great exposure here.
![]()
Hi Sid,
I like this one the best from all your series. I agree with the above. It would be nice to see the ears in the reflection.
Is this a large crop? It seems the focus might be slightly in front of the face (and on the ice in front of the bear) and the fur seems a little soft. If so, then this image should sharpen up very nicely with Richardson-Lucy deconvolution. I could do it for you if you're interested in seeing the result.
Roger
Roger, it would be good if Sid agrees, (sorry Sid, no pressure) as a lot of our members have issues with sharpening and any additional tuition can only bolster their knowledge and understanding. A concise workflow would really bring things together nicely. For me this is the biggest hurdle with images posted. Personally, if the understanding using LR or ACR sharpening can be done, it really is, IMHO the way to go. Then it leave minimal work to be done for web and avoids artifacts.
Post Production: It’s ALL about what you do with the tools and not, which brand of tool you use.
Like the composition of this one Sid. Also like the layers of color and texture of the water and ice receding into the BG and the reflection of the eyes.
Hi Sid, the appearance of high ISO speeds on the EOS-1D Mark IV are the result of an advanced image sensor and image processing technology.
The Custom Function C.Fn II-2 High ISO speed noise reduction setting is set to a default setting (option: 0) of Standard. In order to get an idea of the noise reduction function effects, you should compare photos taken with ISO speeds of ISO1600 or higher by changing Noise Reduction from Disable to Strong. You will be able to see that approaching high ISO speeds of ISO12800, when set to Disable, noise will become more visible. Even at extremely high ISO speeds, if Standard and Low noise reduction are working, noise will be considerably reduced.
Post Production: It’s ALL about what you do with the tools and not, which brand of tool you use.
Hi Sid,
like the half hidden bear, nicely seen.
Subtle colors are working good in this one , like wise the composition in Pano format.
Agree about the visible ears , think this would bring the image up a notch.
TFS Andreas
Sid, did Roger try his magic wand on the image, curious in case I missed it, as it would be great to see the enhancement and what it can do?
cheers
Steve
Post Production: It’s ALL about what you do with the tools and not, which brand of tool you use.
Steve,
Received files from Roger this morning. Thanks a lot Roger. Here is some info he provided in the email
Attached is a half size comparison of the original and final. Final is a combination of 3x3 Gaussian with 30 iterations followed
by a 5x5 Gaussian with 4 iterations. I put that image in photoshop with the original on top, then erased the original over the bear and
select portions on the ice so the sharpened image shows through where needed. The sharpening enhances noise so this combination
shows less noise overall in the image.
Another variation that might show less color noise is to convert to LAB and do the sharpening only on the luminance channel.
While the bear does show more noise, the detail in the fur is a lot more and I prefer the sharpened image. Let me know what you
think.
Roger,
Looking at the final scale down image I am with Roger on his sharpening technique. Details are much better and noise print is very very low compared to tonal contrast sharpening. I tried smart sharpening with tonal contrast(Nikfilters) selectively on the bear and could not produce results closer to Roger's final version. Thanks again Roger.
The difference with Richsrdson-Lucy (RL) deconvolution and unsharp mask or smart sharpen is as follows. In the unsharp mask/smart sharpen methods, a bright subject tends to become larger, for example a hair. With RL, the method works to move energy using the defined point spread function into the center assuming the spread in the point spread function is due to image blur. Thus something like a hair becomes thinner and brighter with RL but becomes brighter and fatter with unsharp masking methods. The downside is RL is compute intensive. No instant feedback. On a 3 GHz I7-950 class machine, it took a couple of seconds per iteration, so the 30 iteration run took about a minute on Sid's full resolution image. Often, some unsharp mask after RL produces a very nice result. I did not do any unsharp mask on Sid's image.
Roger
Hi Sid & Roger thanks for this. I guess it's a trade off between clarity/sharpness for additional noise. What would be interesting Sid is to shoot at high ISO make the required adjustments in camera within the CF and then execute the same procedure. I think for me, the litmus would be seeing a 20x16 print. BTW any additions via Luminance channel, do you not lose detail? Certainly Sid the difference between the OP and on 500px is a noticeable difference.
Roger when you say 'No Instant feedback' is this a case of just waiting for it to do it's thing, let it run it's course, and if it's not right, revert back and then adjusting more or less iteration? How can we get to try this?
Steve
Post Production: It’s ALL about what you do with the tools and not, which brand of tool you use.
By no instant feedback, I mean you have to wait the second or two per iteration. In ImagesPlus (the software I used) you get the revised image with each iteration, but the best results are with a small blur and lots of iterations, so the change from iteration to iteration is small. I usually just make several runs then bring each image up on screen and compare them, then either choose one, or from those, choose what to try next. Having been doing this for years, I usually have an idea on what to use. On Sid's image, I made 4 runs as tests then a combination of two runs for the final. RL can also be use to repair some motion blur, but that usually takes me more tries to get it well. And on larger images like mosaics, the iteration time is longer, so a run can take many minutes (I do a lot of mosaics with landscapes and occasionally on animals).
Regarding sharpening on the luminance channel, no you don't lose detail. Unsharp mask works very well on the limunance channel too. The advantage is no changes in color around edges.
Roger
Thanks Roger, it has a nick name over here 'CSI filter'no idea why.
In LR3 the Detail slider was a kind of halo suppression, now the slider has more subtle control/effect and I tend to think of it as a 'mixer' than a 'volume' control, but I guess I will still to what i know as it suits my workflow, however always good to learn new stuff. Just for the record, I try to ensure I sharpen only the high frequency detail and use the luma noise reduction to counterbalance any excesses in the low frequency detail.
steve
Post Production: It’s ALL about what you do with the tools and not, which brand of tool you use.