denoir wrote:
Wayne, yes - those are examples of the 'behind the scenes' LAB processing. When you set the layer mode in RGB to be 'color' it does the processing in LAB mode. It just does it behind the back of the user. Same thing with luminosity.
As for LAB<->RGB conversions affecting IQ, I don't see how that would be possible. LAB mode is orders of magnitude larger than RGB so any transitions should be lossless. You can of course go out of RGB gamut with LAB, but that's a different thing.
Yes, I agree with you, just have heard from those who are against.using lab mode say that the conversion in and out of lab mode has a negative affect on the image.
I never saw them explain how that is so.
denoir wrote:
As for LAB<->RGB conversions affecting IQ, I don't see how that would be possible. LAB mode is orders of magnitude larger than RGB so any transitions should be lossless. You can of course go out of RGB gamut with LAB, but that's a different thing.
Maybe you can get some degradation if you go back and forth between RGB and LAB (especially on an 8 bit file), due to quantization errors or something like that? I think going to a larger CS and then back to a smaller one isn't always lossless (clipping isn't the only loss that can occur).
carstenw wrote:
Bringing this thread back to the top, I have a question:
has anyone here played much with Silver Efex Pro vs. Silver Efex Pro II? I have the former and am wondering if the latter is worth the upgrade?
Great thread, just found it. Thanks denoir for sharing your schemes!
Carsten, I have the same interest having used Silver Efex for quite some time - but not 2.0. Looking through the specs there are a few differences - if they're worth it or not I am not sure.
(sry of I missed anyone else adding this info)
Some new refined controls are there, added to the "structure"-control (ie local contrast/clarity) there is something called "fine structure". Similar enhancements have apparently been done with contrast and also some alterations (?) on the highlight and shadows controls that are called "amplify black" / "white". There's a history palette added which could be really useful. And a new border engine is added - on top of current vignetting/burn edges etc. Engine is now 64-bit. And there's also a (to me) quite useless "spare a color" function.
In all it seem like a gradual improvement. The beauty of the previous version is the simplicity and therefor power of the controls. But it is not without glitches. History (and perhaps implementation of the now lacking ctrl-z) would be a massive improvement to me. Simple things like scrollbuttons on the mouse might also be implemented but who knows. I like the 1.0 so much so I will probably upgrade..
Luka: Yes, LAB is powerful in the way it separates color and contrast into different entities.
The a,b curve steepening you showed in your examples is a classic LAB 'move' and it works very well for certain subjects. Canyon walls in dull light are prime examples for this. And you are quite correct that this is well suited if you want a nice way to address color separation.
^^^I believe Flickr addresses that problem somewhat by gently sharpening the automatically generated smaller sizes for you. (Not that I recommend using flickr. )
^^ you sure? my photos look like crap compared to yalls... I upload my full rezzers to flickr b/c its easier for family/friends to see them there....(plus its a cheap backup) but it looks like I may need to have two folders, one for full size and another for web sized 1024 :/
Yes, flickr sharpens. If you use Lightroom to upload and apply screen sharpening in the standard upload process to already perfectly sharpened images they look light crap on flickr.
The solution is to upload resized images and not to sharpen for screen during upload.
crazeazn wrote:
^^ you sure? my photos look like crap compared to yalls... I upload my full rezzers to flickr b/c its easier for family/friends to see them there....(plus its a cheap backup) but it looks like I may need to have two folders, one for full size and another for web sized 1024 :/
I upload max 1200 wide photos to flickr. Flickr will sharpen the smaller sizes but not the full size.
For the cost of a Flickr membership, why not spend a bit extra and get proper web hosting with your own domain? This way you can upload whatever size you want to show and link directly to that file without any site imposed processing happening along the way. If you don't want to deal with a web host, Amazon's S3 is an example of a pay as you go option at around 10¢/GB, though you do pay for data transfers as well (up and downloads), at around 10-15¢/GB. If you're just linking photos to forums, it shouldn't be significant. I believe a lot of the 'archive to the cloud' services use S3 but repackaged with a more user friendly interface. A benefit of dealing directly with Amazon is they're well established and unlikely to run into operational issues, like going bankrupt, which is some peace of mind if you're looking for a cloud backup option for existing physical backups.
rscheffler wrote:
For the cost of a Flickr membership, why not spend a bit extra and get proper web hosting with your own domain? This way you can upload whatever size you want to show and link directly to that file without any site imposed processing happening along the way. If you don't want to deal with a web host, Amazon's S3 is an example of a pay as you go option at around 10¢/GB, though you do pay for data transfers as well (up and downloads), at around 10-15¢/GB. If you're just linking photos to forums, it shouldn't be significant.
I'll second the suggestion for using Amazon S3 for data hosting. I've been using it for a few years for all my FM image hosting, along with occasional photo galleries for sharing with friends or distributing miscellaneous files, at a cost of ~25¢/month (though I have been hit with a bill of over fifty cents after a couple rare instances of unusually heavy use). I started when the service was fairly new and you had to be fairly computer-nerdy to use it, but now there are lots of pointy-clicky-draggy interfaces available that should make it pretty easy for anyone to use.
I use flickr because I want people to see my images. I guess that's what we all want, thats why we are here and posting images. I do have webspace (more than enough), but prefer flickr.
^^Yeah, that's a good reason, but personally I care more about having a nicer/cleaner presentation of my pictures. I just don't like the flickr interface and don't really care about getting all the oohs and aahs of the general flickr public. It feels like it's mostly a social site of people collecting images to me. 9 out of 10 times, the standard sizes of the pictures are too small for my taste, which makes it annoying to browse people's galleries.
Well, there is a php script that lets you display your flickr images on your own website without actually hosting the images on your website. It is very nice and clean. And if you resize and upload with care, image quality is not affected. Here is mine (keep in mind that I have to re-process some images that where uploaded before I realized the sharpening issue):
This is a great thread! I need to develop my PP techniques and this seems to be a good starting point, with some nice ready to use recipes and technical background to understand the recipes.
I just do the trivial curves/levels, some USM and the always needed with a 5D erasure of dust spots.
I'm bookmarking this thread and would like to thank all the contributors to it.
My try in "sharpening" department.. I was using similar to all my photos, just a lot weaker. I didnt knew that 5DMK2 is so so soft.
Btw. that LAB discussion is pretty interesting. I suspect that C1 is using LAB mode too, cause it allows you to do some magic with color channels. Though its nowhere specified..
@denoir: Interesting, although I'm not sure how relevant the test image is for real-world subjects, except for certain man made things like the blinds in morfeus' image perhaps.
I'm seeing and measuring a green cast in your version of the image, whereas the bicubic version is perfectly neutral (grayscale). Something going on with the conversion to Lab and back to RGB?
Btw, do you normally use "remove: Gaussian blur" for sharpening? I find the lens blur setting much more effective, but perhaps you specifically chose Gaussian for your M9 files?
AhamB wrote:
@denoir: Interesting, although I'm not sure how relevant the test image is for real-world subjects, except for certain man made things like the blinds in morfeus' image perhaps.
Very relevant. It simply includes a bunch of spatial frequencies going from coarse detail 10 lp/mm to very fine detail 72 lp/mm. The big question is what resize & sharpen method to use in order to:
1) Keep the detail in the resized image (i.e maintaining a high MTF)
2) Keeping the internal relations of the detail constant.
3) Minimizing aliasing if possible
The test image is really a very simple case. What would be interesting would be to construct a more general one that includes multiple directions (horizontal, vertical, sagittal & tangential) with patterns of different frequencies and then try to find the transfer function (i.e sharpening method) that will yield the best results.
The current resize & sharpen method that I use is an improvised one. It took me a very long time to discover a major flaw in my previous version and there's no guarantee that the current one even approaches anything optimal. I'm getting better results than with bicubic, but that's really trivial to accomplish as bicubic resize is terrible.
I'm seeing and measuring a green cast in your version of the image, whereas the bicubic version is perfectly neutral (grayscale). Something going on with the conversion to Lab and back to RGB?
Heinz (morpheus) image or the test image? In Heinz image it's the LoCA in the blinds that gets preserved while it is lost when using bicubic. The test image is grayscale so there can't be any cast.
Btw, do you normally use "remove: Gaussian blur" for sharpening? I find the lens blur setting much more effective, but perhaps you specifically chose Gaussian for your M9 files?
I find the lens blur setting to be too aggressive - I use smart sharpen primarily to fine tune the highest frequencies (small radius) and don't want the effect to be too strong.
denoir wrote:
Heinz (morpheus) image or the test image? In Heinz image it's the LoCA in the blinds that gets preserved while it is lost when using bicubic. The test image is grayscale so there can't be any cast.
The test image. In PS both are indeed neutral so it looks like Firefox is rendering your image test_denoir with a green cast here for some reason...
Regarding the relevance of the test image: "line pairs" are not often seen in nature, so I wonder how useful it will be to be looking at the appearance of set of line pairs processed/resized in different ways. Are you hoping to find some "golden mean" conversion procedure that creates least artifacts while retaining most detail?
)
