leonard_richmond7
-
Posts
131 -
Joined
-
Last visited
Content Type
Profiles
Forums
Blogs
Events
Downloads
Gallery
Store
Posts posted by leonard_richmond7
-
-
Sampling Caucasian skin tones, you should get high red, green 20-25% lower, and blue lower still. If white balance is too blue, it balances out the blue and green, ending up with only red higher. That's pink! So when you see pink skin, your picture is really just too blue. If it's a little less saturated pink means more white added, so the green and the blue are even closer to the red and you need to reduce the green a little and the blue even more.
So I used Photoshop Curves to warm it up. Moved the top right Blue point from 255 down to 203 (that's 255 in, 203 out, decreasing the blue) and the Green 255 down to 230. That darkens the picture, so in the RGB channel I moved the 255 point to the left to 243 (243 in, 255 out) to brighten it back up.
Everything looks believably accurate now.<div>
</div> -
Bob Atkins compares a similar lens, the Tamron 28-300, to the Canon 75-300IS Canon 300/4L and here at photo.net:<p>
Bob says, "if you're using a monitor which gives you a 12" wide image (typical 17" monitor) and you're viewing at 1024 x 768 pixels, a 200% crop represents a section of what would be an 82" x 54" image! So if the images look a little "fuzzy", that's the reason why! At normal print sizes (say 8x10), many of these crops would look pretty sharp."
To see what an 8x10 would look like from 12 inches away, stand 8 feet back from your monitor, look at the crops and compare to the others he shows.
Bob concludes "In answer to the original question 'Does the Tamron 28-300 deserve respect', I'd have to answer that it does. While I think you could probably do better with two zooms, one covering something like 28-100mm and one covering 100-300mm, not everyone wants to do that. For those who don't, the Tamron 28-300 may be a good choice.
<p>
"The bottom line is that if you hate swapping lenses, need one small, light, lens to do everything, and you're prepared to make some compromises, the Tamron SP AF28-300/3.5-6.3 XR Di LD Aspherical (IF) Macro isn't a bad choice."<p>
Thank Bob for giving you comparison shots so you make up your own mind as to the value of the Tamron.
-
<i>"where they (optical filters) filter out a lot of the light in one particular RGB channel, the lack of detail recorded in that channel by the sensor will tend to reduce the sharpness of the image - the demosaicing software has less information to reconstruct image detail from."</i>
<p>
If you're using the filter to restore white balance to daylight, exactly the OPPOSITE is true. The filter brought the R,G and B channels back into proper proportion, and exposure is typically increased to compensate for the filter's light loss, giving the demosaicing software MORE information to reconstruct image detail.
<p>Or are you suggesting that dimmer daylight, compensated for by increased exposure, will result in less sharpness than an out of balance white balance?
<p>
Having said that, I still do not use color balancing filters for 2 main reasons:
<p>
1) Convenience.
<p>2) Accuracy. Does an 81B (or whatever) correct as accurately as my RAW converter? Probably not. Does this overcast sky need 81B or 81C? Does this amount of shade, with the surrounding reflected sunlight, require an 81EF? Are those 60 watt or 100 watt incandescent bulbs that I'm trying to balance? How much warmth do I want to leave in the incandescent scene? I find it much easier to get exactly what I want in the RAW conversion than in selecting filters.
-
Look at the bottom of <a href="http://www.dpreview.com/reviews/fujifilmf700/page10.asp">http://www.dpreview.com/reviews/fujifilmf700/page10.asp</a> to see the 9 to 9.5 stop dynamic range of the Canon S50, a small sensor P&S camera. DSLR's have slightly more dynamic range.
-
"his choice of scene/lighting wasn't very contrasty" (about the Clarkvision link posted above)
Huh? Paper white in the shade (point C in Figure 2) is 4 stops darker than paper white in the sun (point A).
On the gray scale on shaded chart, I can distinguish steps 18 & 19. Step 19 is 10 stops below paper white in the sun (point C).
The round hole at the top (point E) is 10 2/3 stops darker than sunny paper white (point C).
It is contrasty enough to show the superiority of the digital camera over the scans of either film.
Yes, there are other films with greater dynamic range - I've shot them, processed them, pushed them, pulled them, and dodged and burned prints made from them in the past. But the first time I had to recover people's faces from shadows in a digital camera exposure was the last time I ever thought of using film.
As to B&W - yes you can get more. If that's what you want, shoot it. But it's irrelevant to the comparison of color film and digital.
-
From the Clarkvision link above:
<p>
"The shadow detail in the film was confirmed visually by inspecting the film under magnification: the scans show what can be detected by visual inspection <b>plus a little more</b>." (emphasis mine)
-
See <a href="http://clarkvision.com/imagedetail/dynamicrange2/">http://clarkvision.com/imagedetail/dynamicrange2/</a> for a detailed analysis of the dynamic range of Canon 1D Mark II compared to Velvia and Kodak Gold 200. Be prepared to be surprised.
-
Vasilis,
<p>
Here's a quote from the article you provided the link to:
<p>
"... this test is only going to produce a theoretical result ... the dynamic range that I will be calculating will be much higher than any figure normally quoted for film because very small changes will be detected. At some point these changes will become irrelevant to the appearance of the image"
<p>
and later<br>
"Criterion: 1EV change in exposure results in an increase in pixel value of at least 1.
Dynamic Range = 14 stops"
<p>
A whole stop represented by a pixel value change of 1 (on the 8 bit scale). Anyone claiming that as "useful" dynamic range for a digital camera gets laughed at!
<p>
Compare the graph for small p&s digital cameras at the bottom of <a href="http://www.dpreview.com/reviews/fujifilmf700/page10.asp">http://www.dpreview.com/reviews/fujifilmf700/page10.asp</a> to the gray patch represented on the graph in the article you mentioned at <a href="http://www.path.unimelb.edu.au/~bernardk/tutorials/360/technical/hdri/results.html">http://www.path.unimelb.edu.au/~bernardk/tutorials/360/technical/hdri/results.html</a>.
<p>
Of course, larger imaging chips, having less noise, have more dynamic range than shown for those small p&s digital cameras.
-
This link has the comparison you asked for using Nikon's D100 6 megapixel camera, showing 100% crops and scans from 1.5 x 1 meter prints (60" x 40"):<br>
<a href = "http://www.borutfurlan.com/test_results.html">http://www.borutfurlan.com/test_results.html</a>
<p>
The Digital Rebel XT is 8 megapixels, so it will be a bit sharper. A 16"x20" print is MUCH smaller than the 40"x60" prints he scanned, so the 16x20 will look much sharper. End result - I think you'll be happy. I'm very happy with 16x20's I've done from my 6 megapixel Rebel, and even 20x30 (with careful uprezing).
-
I was referring to the original challenge by Dan Margulis. You can see it explained and commented on at http://www.brucelindbloom.com/DanMargulis.html.
<p>
Finally found Mike Russell's reward offer at http://www.ledet.com/margulis/ACT_postings/ColorCorrection/ACT-16-bit-2002.htm, but it was for $100 (never could find any reference to a $200 reward).
<p>
From that last url: "I applied the same extreme corrections to the image, but this time not to Ric's 8-bit image but rather a direct Photoshop conversion of Ric's 16-bit image to 8-bit. Shockingly, this completely eliminated the problem. There was no reason to prefer the version corrected entirely in 16-bit. <b>When Photoshop converts from 16-bit to 8-bit it applies very fine noise to try to control subsequent problems.</b>"
<p>
Of course, if you start with an 8 bit image, you don't get the advantage of that noise dithering to prevent banding during post processing.
-
Bill, if that's the $200 reward I think it is, it is bogus. The offerer (correctly) insists that the viewing will be done in 8 bit mode, so a conversion from 16 bit to 8 bit mode must occur. The catch is that he insists the conversion must happen BEFORE any post processing of the image rather than AFTER such post processing.
DUH! Of course you won't get any better results that way!
The fair way is RAW -> 16 bit -> post processing -> 8 bit compared to RAW -> 8 bit -> post processing.
Just applying a steep ramp up in the shadows using curves will show posterization (sometimes horrendous) in the shadows with an 8 bit image. Doing the same with a 16 bit image (even after later conversion back to 8 bit) will retain smooth transitions in the shadows.
-
The noise in the shadows is MUCH greater using Photoshop's ACR. I've read that Canon does noise reduction in its conversion, and they seem to do an excellent job.
On the other hand, I really like ACR's CA correction. Is there any way to it manually? You'd need to be able to select a single channel (R, G, or B) and either shrink or expand it, leaving the other 2 layers alone. Any one know?
-
And you are how old?
Presbyopia - stiffening of the lens in your eye - starts to be a problem somewhere around age 40. Your nearest focus distance gets farther and farther out. Eventually you'll need reading glasses (or bifocals if you already wear glasses, or bifocal contacts, or a near distance contact for one eye and far distance for the other, etc.).
-
Here's the trick for getting maximum stability with the Velbon 343E - forcefully spread the legs futher apart, so that they all bow inward a bit. You can do that by putting one leg on the ground, then spreading a second leg further apart by pulling on it before placing it on the ground, then doing the same with the third leg.
This stabilizes the tripod quite a bit more stable than simply placing the legs where they naturally go. The tripod "grips" the ground like a claw.
And of course, always use a remote shutter release.
If the wind is blowing, carefully press down on the tripod collar. Or wrap the strap of your camera bag around the collar and let the bag hang - the extra weight helps stabilize the tripod.
I have no trouble getting sharp long (multi-second) exposures at 200 mm using these tricks.
-
Forget to mention that I also lowered the green a little while adjusting curves for his face. One side of his face picked up a green tint from the light reflecting of the greenery on the cake and it looked pretty ugly. Correcting for it made the other side of his face a little magenta, but that's not nearly as objectionable. Of course, you could always treat the other side of his face differently and make it even better, but I was keeping this short and sweet.
-
Look at his white jacket - it's bluer in the shadow (front) and yellower in the brighter areas. I'd say that the tungsten light overpowered the flash. Since his face is turned down, it's getting the direct flash but only reflected tungsten light, so it's much darker and bluerthan her face.
So I corrected the general white balance using curves, while also raising the midpoint in curves to lighten the midtones. Then I selected an area a bit larger than his face, feathered it a bit, desaturated it a bit, then used curves to lighten it and lower the blue channel (to correct from daylight flash only to combination flash plus tungsten light of the reast of the scene)until it looked better. I went by your statement that his face was actually redder than hers, so I didn't go too far.
I took a whole minute playing around with it, but I think it turned out OK.<div>
</div> -
Sample 1dsMkII pictures at ISO 1600 (everything from 100 to 3200) at <a href="http://www.steves-digicams.com/2005_reviews/1dsm2_samples.html">http://www.steves-digicams.com/2005_reviews/1dsm2_samples.html</a>.
I downloaded VJ9J6744_iso1600.JPG and ran it through Noise Ninja. Wonderful. I wouldn't hesitate to print it full size - say 16x24. Probably as good as Reala 100.
-
Using Noise Ninja, my 3200 shots (1600 intentionally underexposed 1 stop) from a 300D are certainly more than acceptable for an 8x10. Of course, it is a "last resort when you have to nail a shot or lose it" as stated in an earlier post, but I wouldn't hesitate to use it rather than lose it - the results can be very good after noise removal.
Just be careful when sharpening. I select the shadows (in Photoshop, that's Select > Color Range > Shadows (in the Select box at the top), then invert the selection so that everything BUT the shadows are selected. Then use Unsharp Mask on the selection. That way, you won't exaggerate the noise in the shadows.
The 20D shots should be even better.
-
See <a href="http://www.luminous-landscape.com/tutorials/s30-as-meter.shtml">Digital Point & Shoot As Lightmeter, The Canon S30 as Pocket Histogram, For Large Format Photographers</a> at <a href="http://www.luminous-landscape.com">www.luminous-landscape.com</a>.
-
<i>"The 20D will have more dynamic range than any P&S (meaning you'll have a better range of tones to work with)" </i>- Carl Smith
<p>
Couldn't find color chart samples comparing your Ricoh with the 20D, but the tiny Canon S500 5 megapixel has the same size ccd (and an even smaller body) as your Ricoh. Comparing the color charts for the Canon S500 at<br>
and the Canon 20D at<br>
I don't see ANY difference in dynamic range in a straight shot.
<p>
With the lower noise of the 20D, you could raise the levels more to bring out detail in the shadows because of the lower noise compared to a P&S. (That's true for my Canon 300D and Canon 3 megapixel S30), so you can salvage shadow detail better with postprocessing. Perhaps you could call that "usable dynamic range".
-
<i>"My camera body is a Digital Rebel, so I like to limit myself to the ISO 400-800 range indoors to avoid signal noise." </i>
<p>
Use Noise Ninja (or Neat Image) for a 2 stop decrease in noise. ISO 1600 shots end up looking like ISO 400, with an 8x10 print from it as good as from 35mm ISO 100 Fuji Reala film. You can batch process the images, so you don't have to wait at your computer for it.
<p>
You'll never hesitate to use 1600 again.
-
Jean-Baptiste's final suggestions are in line with my observations after many "pixel peeping" trials, at least at low ISO.
Of course, if you're not doing any levels/curves/sharpen manipulations, it doesn't make any difference. But if you're doing much manipulation, it does - and the more extreme the manipulation, the more difference it makes.
I have the S30 (3 megapixels) as well as Canon's Digital Rebel (300D). While most other small sensor digital cameras certainly are noisy enough for Jean-Baptiste's initial comment "doesn't matter", that's why I picked Canon - the noise was noticably lower. After extensive testing with the S30, I was impressed enough with the 16 bit versus 8 bit TIFF from RAW that I changed my Photoshop workflow to do as much as possible in 16 bit before switching back to 8 bit.
-
Jean Baptiste, 9000 (base 10) = 10001100101000 (base 2), so it takes 14 bits to accurately enumerate those 9000 photons. (I'm taking your figure of 9000 on faith - I have no idea of the actual number and you offered no evidence.)
So how do you get only a 7 bit requirement?
-
Use a regular pen instead and sign a piece of paper. Take a picture of it with a digital camera, or a film camera followed by scanning the neg/slide.
Or use a tablet/stylus instead of a mouse. Don't have one? Find someone that does, use their computer for a few minutes, then save the image file.
Either way, save the image of the signature same as any other image file.
Add the image with the signature to another image by adding a layer, then flattening the layers (if desired).
Nikon vs Sigma 55-200mm for D70
in Mirrorless Digital Cameras
Posted
The Sigma is slightly better than the similar Canon 55-200. See <a href="http://www.popphoto.com/assets/download/PP0505_SigmaLensTest.pdf">http://www.popphoto.com/assets/download/PP0505_SigmaLensTest.pdf</a> and <a href="http://www.popphoto.com/pdfs/2002/0502/Canonlens_sqf.pdf">http://www.popphoto.com/pdfs/2002/0502/Canonlens_sqf.pdf</a>.
<p>
Don't see the Nikon review on Pop Photo's web site, though I seem to recall it in the magazine not long ago, and I think it was no better than the Canon.
<p>
Detailed reviews and samples at <a href="www.dpreview.com">www.dpreview.com</a> of the Sigma 55-200 have been right in line with Pop Photo's test.
<p>
Looks like you got lucky!