Canon small sensors cannot capture reds and purples accurately.

Here's an example of a very, very, very RED rose in daylight taken with my S3 using auto bracketing at + - 1 stop and a -1/3 compensation already dialed in:



--
My gallery: http://www.flickr.com/photos/inghramjp

I'm guessing somewhere between the middle one and the darker one.
--
Lee Jay
(see profile for equipment)

Out of gamut reds can be a problem on many cameras. Some other options that will give better results than underexposure are:
1, set the saturation level to "low" or "(-)"

2. shoot in aRGB colour space rather than sRGB, if your camera has this available
Both options work better with RAW

The image below contains two 100% crops, from jpegs, showing the effect of the saturation setting on the red channel.

Note that the out of gamut reds have caused the pixels to bloom and seriously degrade the sharpness. BTW, both images were taken with the camera on a tripod, so it's not due to camera shake.
No sharpening or any other PP was carried out on either image.



regards... Sangio

--

'Twelve significant photographs in any one year is a significant crop' Ansel Adams
http://www.pbase.com/isabel95
http://www.pbase.com/digipets (not only for pet digital photography!)

Here's a saturation test sequence with my S3 from -2 to +2 with 0 in the middle. Wish the S3 had AdobeRGB, or better yet RAW, but as it is I just leave saturation down -1 just to help a bit without cutting too much color out.



--
My gallery: http://www.flickr.com/photos/inghramjp

I am impressed by the group knowledge of photography and cameras. Thanks for demonstrating solutions to problems I encountered. It is all appreciated and helpful. I can't wait to have another go with those red and purple flowers.

No, both pictures were taken in the sRBG colour space as indicated. The only difference is that for the second picture the saturation in the camera was set to low, i.e. (-)
I didn't try aRGB for this particular picture.
regards...Sangio

Go check out the Sony forum on here. They come out just perfect.
I may switch soon.

Janice

I managed to try DCF Full Spectrum a little while back and didn't think it did a bad job for this kind of stuff.

Probably doesn't top what someone can do on their own with a lot of patient tweaking, but seemed much better than I had expected it to be.

It's come up on the forums before if a search is done, and can be found at the following:

http://www.tribecalabs.com/

Thanks for the link. I have looked at their site and am tempted to buy the product. I will research it more. Have you or anyone else been successful in applying this software to flower images and seen a big difference?

I was starting to think the red and purple problem was mine when in fact it is something that all digical cameras have trouble with. Surely this technology can be built into cameras and make such software unnecessary. I hope Sony and Canon are reading these forums.

OvinceZ said:

I tried several times to get the correct colour with my G6 to no
avail. See the samples on my gallery.

Click to expand...

you mean these are not good enough ?
The reds are many magnitudes more intense than any rose could ever get

(sorry I do not have any others available right now, resized to 600x.. not best of quality)





--
if needed, email me at : [email protected]
Horum Omnium Fortissimi Sunt Belgae !
(CanFT-QL)CanG6-SOON-SonH5CanA520-M3358-DH1758

Rotareneg said:

Here's a saturation test sequence with my S3 from -2 to +2 with 0
in the middle. Wish the S3 had AdobeRGB, or better yet RAW, but as
it is I just leave saturation down -1 just to help a bit without
cutting too much color out.

Click to expand...

try that in full sunshine and you'll get another thing coming

--
if needed, email me at : [email protected]
Horum Omnium Fortissimi Sunt Belgae !
(CanFT-QL)CanG6-SOON-SonH5CanA520-M3358-DH1758

OvinceZ said:

Thanks for the link. I have looked at their site and am tempted to
buy the product. I will research it more. Have you or anyone else
been successful in applying this software to flower images and seen
a big difference?

I was starting to think the red and purple problem was mine when in
fact it is something that all digical cameras have trouble with.
Surely this technology can be built into cameras and make such
software unnecessary. I hope Sony and Canon are reading these
forums.

Click to expand...

Funny .. I posted this info on sensor color sensitivity months ago together with the Tribeca link, both in the Canon & Sony forums and it was laughed away .. In fact a US CTF member owning an S2 and myself did some comparative tests (using pantone color ref cards) with the G6 and Sony H1 and the "purples" department" got us a serious problem .. confirming what is populating this thread now .. oh well .. history is a constant repeat ..

--
if needed, email me at : [email protected]
Horum Omnium Fortissimi Sunt Belgae !
(CanFT-QL)CanG6-SOON-SonH5CanA520-M3358-DH1758

I did a search of the product and it sure is controversial. Some dismiss it outright but a few suggest it might be useful. Hardly anyone has actually tried it but of the few who did they say they got good results. Heck if it solves the deep purple and dark red rose problem it will be worth it for me. Plus the other violet colours that are nowhere to be found in the images on my computer screen. Sure cameras show purple and violet flowers but not accurately like you see in real life. I mean, they are not even close. That beautiful dark red rose is a travesty of the beauty seen with the eyes. The deep dark violet pansy is impossible to depict in my images. I am going to try different things but apparently this phenomenon is common and several people on other discussion boards on this and other sites have been trying to solve this problem with mixed results. Some are downright perplexed and confused and I agree, it is a big problem. It surely is not good enough to have to tweak images and then buy special software to correct this problem. I think it will be worth 50 US dollars just to find out for myself. Otherwise spending all that money on the many digical cameras I have is a bit of a waste because I am just fooling myself that I have taken accurate photos of the flowers I so admire and appreciate.

When experts have different theories trying to explain a phenomenon then a problem exists. We will see what eventuates about violet and red. In the meantime the so-called fixes don't seem to always work. I have some testing to do myself and will attend to that as soon as it is sunny again. The thing is questions about matters of fact are either right or wrong or partly right. I don't think anyone has explained this problem away although many seem to dismiss it. In the meantime we are fooling ourselves that these cameras can capture accurate flower colours in violet and deep or bright red. I will try different saturation settings and white balance settings and see if they make a difference. This is getting to be like witchcraft! What a bunch of nonsense. The sensors suck and that is the truth. Or that and the software. Someone suggested instead of RGBG we have VEYR. (violet, emerald, yellow and red) He could be right. It is possible the sensor designers have made a mistake. Their choice is great for general landscapes but sucks when it comes to flowers. If you take heaps of flower photos you have to be honest and admit the sensors are not accurate regarding the colours.

OvinceZ said:

Thanks for the link. I have looked at their site and am tempted to
buy the product. I will research it more. Have you or anyone else
been successful in applying this software to flower images and seen
a big difference?

Click to expand...

It was actually purple flowers a friend let me try it on. There was a brief bloom of some unusual ones locally a couple months back, and I was very frustrated with them coming out either too red or too blue no matter what I did. Really frustrating to see wonderful colors in person and not be able to capture them.

With whatever "mapping" the plug-in did, the flowers ended up a lot closer to the actual shade of violet, and certainly a color that my camera could never put out on its own (I was using an SD600 at the time).

It was decent. It was nice to suddenly say "ah ha! that's much more like the color." I actually liked what it did to greens, too. The portrait setting seemed like it might be useful. I think it was the vivid setting that seemed most effective on the plants.

Whether or not it's worth paying $50 for the plug-in, I'm not sure. I recall liking what it did to colors a lot more than anything something like Photoshop did automatically to colors.

I've since gotten a bit more into using a "selective color" mix by hand for anything that matters to me a lot, and/or getting my in-camera settings better set-up for post-processing.

I'll see if I can dig up a before and after example of the plug-in....

All I have around from before that's easily accessible is a shot from after the DCF plug-in was applied. The original is archived away somewhere.

This is resized and cropped from the original, with no other post-processing done other than DCF on the Vivid setting.

I know it's not the best thing without the original for comparison, but it's all I can come up with at the moment. I'd have pulled it aside because I recall it being one of the better jobs DCF did....

The image has a variety of purple and violet shades so that is impressive. I appreciate your posting this for us to see.

Nikon D200 - 2.3/180 Angenieux APO - RAW - Silkypix Developer Studio 2
--
Peter M.C. Werner
http://www.leicaphoto.net

Can I just say what a change it is to see such a good thread...

Interesting problem posed.

Several theories postulated.

Theories evaluated without recourse to put downs, name calling and other bad manners.

Oh if all threads could be like this.

Regards