Use Raw?

[Arch]

Another thing i have thought (but again could do with confirming) is that subtle tones in the 16bit RAW file can be adjusted more accuratly without loosing tone graduation data. If you shot the image as a jpeg you start at 8bit and further editing can reduce tonal data further (especially if your not using layer masks) and of course re-saving causes information loss.

The idea is that although you may save a copy of the 16bit file as an 8bit jpeg, you dont have the same amount of data loss because the editing of the tonal graduations and other alterations like sharpening, have taken place.

Try to imagine if the effect on sharpening pixels of a 16bit image, then saving it at 8bit, would be as severe as sharpening an 8bit image.

 

[Digital Matt]

I don't have time to write a long reply now, but notice when you tweak the exposure of a raw file, the entire histogram shifts left and right. You can't do that in photoshop.

 

[fotogenik]

Curiousity has gotten me here. How does one tell how much to adjust the exposure slider bar to get the exposure up or down 1 or 2 stops excatly? I have been guessing for the most part but am curious whether there is a guide to tell you how much adjustment is needed to lower the exposure by 1 f-stop.

 

[markc]

If you shot the image as a jpeg

I don't know if you are still replying to me, but I'm definitely not talking about JPEG. As far as I know, RAW files use 12bit color data, and 16bit TIFFS can easily store it all. The difference is that RAW files are linear at 1.0 Gamma and get boosted to 2.2 (I think it's 2.2) when converted to TIFF. Since the brighter areas are getting boosted, you have to be careful as to how much so they don't get blown.

 

[markc]

I don't have time to write a long reply now, but notice when you tweak the exposure of a raw file, the entire histogram shifts left and right. You can't do that in photoshop.

If you mean in the PS RAW converter, it looks it does behave differently than I first thought, but it is very much like using a levels adjustment. I think "exposure" is a lousy label for it, and seem to be there for the layman. What it's doing is moving the white point, and you can see the effect on the histogram live, which you can't when you do a levels adjustment. "Shadows" moves the black point in the same way.

In levels, if you move the white point, that position becomes the new white point and everything to the right of it becomes white (right in my configuration, anyway). The black point stays the same and the histogram "stretches" in the view, because the right edge is always white, not somewhere in the middle. The opposite happens with setting the black point. It looks like the same thing is happening in the RAW converter, except that the histogram stretches "live". That, and moving the exposure slider to the right is picking a lower white point rather than higher. Moving the shadow slider to the right picks a higher black point like PS though. You can also pick a really high white point when you move the slider all the way left, which is interesting. I'm assuming this is because the RAW data is linear at 1.0 gamma and so gives you a lot of breathing room there. I really do have to read up more on this.

 

[markc]

I was thinking that moving both sliders all the way to the left would show you how the RAW data is stored linearly, and finding this makes me more confident in that.

http://www.dpreview.com/learn/?/key=sensor+linearity

 

[markc]

I compared this to a linear converstion, so it is the same. So what this means is that those sliders are just a levels adjustment on the linear data to bring it into a gamma that looks right for our eyes. I'm going to play with some linear 16 bit TIFF conversions and see what I get.

--edit--

Crap. No it's not, but it's close. There are too many auto settings. Still playing. I'll shut up until I my brain is working again.

 

[markc]

Okay, I'm guessing that some other things are going on in that PS RAW converter screen, but it give you a good idea of what linear looks like. You still need to do an actual linear conversion elsewhere.

I do believe that letting a converter auto batch convert every image may not be the best idea, but I think it only comes into play when you have blown out areas. If you don't, you're fine. It looks like there are a group of people that prefer to convert linearly and do all editing in Photoshop.

http://www.outbackphoto.com/workshop/photoshop_corner/essay_12/essay.html

This avoids the blown out issue completely so that you can do all processing in PS, and the only way the TIFF will have a blown spot is if the RAW data is blown to begin with. If you use the RAW converter to make adjustments without blowing out a spot (or compressing the black, but that's unlikely unless you move the slider), it's like doing a PS adjustment on a linear conversion.

And it looks like "brightness" is the one that is the same as the middle slider in levels for setting the mid grey.

 

[Digital Matt]

I've exhausted my expertise in this area. I'm no expert.

There are thousands of different workflows, and in the end, who can really see the difference in a print?

For me, it's a labour of love. I love taking the photos, processing the raw file, and making the print, and I've found the workflow that works best for me, ie: gives me the look I want in my prints, and allows me to enjoy the process from start to finish.

 

[markc]

I've exhausted my expertise in this area. I'm no expert.

There are thousands of different workflows, and in the end, who can really see the difference in a print?

For me, it's a labour of love. I love taking the photos, processing the raw file, and making the print, and I've found the workflow that works best for me, ie: gives me the look I want in my prints, and allows me to enjoy the process from start to finish.

That's cool, and I have no problem with that at all. I just think it's good to know exactly why something is the way it is, or if it really is the way someone says it is before repeating it as fact. If it's repeated often enough, it becomes a fact in the mind of people hearing it. I think we do that a lot, especially if it sounds good and makes sense to us.

Here's one I've repeated: "A good prime lens will take a better picture at a certain focal length than a good zoom lens, becuase the zoom lens has to make compromises across it's focal range." It makes sense to me, and was often repeated years ago, but does it still hold true today? Has modern manufacturing caught up? Have any of us done the research to check?

Ultimately, whether it involves simply trusting someone's deductions or looking at data yourself, you have to trust the source, so it's hard to call much of anything a "fact". Anyway, I certainly learned something new, so I'm happy, and it's prompted me to change my workflow for some images. Maybe every image.

 

[Arch]

I don't know if you are still replying to me, but I'm definitely not talking about JPEG.

Na, i wans't..... just making sure all bases were covered.... this thread should be a good link for people asking about RAW in future.

As for your latest findings, I have to say there is little info out there to question the linear conversion method. Im not overly convinced by the 'outback' website.... seems you need an action to get the most out of thier method..... and that action is at a price ....... but the theory is sound...... I am unable to find any further information which says otherwise. Nor can I find any more info that implies that thier are any other benefits for altering an image in RAW software as apposed to PS.

It does seem to me that as long as thier is no information loss or compression from the transition between RAW software and PS, then there is no real benefit to sticking with editing using RAW software....... it all comes to personal preference...... and in a similar way to matt, im use to my method...... and the chances of being able to notice a difference in the finished artical is slim to none.

It does beg one question tho...... why some of the more expensive RAW converters claim to produce better results?........ this could just be the fact that they have more options for the people that wish to edit more in RAW software...... but they can also be misleading....... the following is from a magazine about using RAW....

"While the results from the Adobe Camera Raw plugin are quite acceptable, for example, professional level software like Capture One Pro produces the best results of all - though at 287 pounds (thats $540), you might expect it to."

So is this not rubbish?..... if you can make the same adjustments in ps without damaging the image file, how can this software be better?

 

[Digital Matt]

"While the results from the Adobe Camera Raw plugin are quite acceptable, for example, professional level software like Capture One Pro produces the best results of all - though at 287 pounds (thats $540), you might expect it to."

So is this not rubbish?..... if you can make the same adjustments in ps without damaging the image file, how can this software be better?

I get very different results from different raw software, with my 20D files. C1 handles noise and sharpening differently than any other software, including photoshop. At work, we use Adobe Bridge and CS2, and I've occasionally shot stuff on the way to work and processed it at work with CS2. I found there to be more noise, and I was unhappy overall with the look of the file. Upon bringing it home and using C1, I was able to get a much smoother look, which is what I'm accustomed to getting from my 20D.

Every piece of software using it's own algorhythms to process the files, and all of them, except for the software that comes from the camera manufacturer itself, are "reverse" engineered, so they cannot be all created equal.

 

[markc]

Na, i wans't..... just making sure all bases were covered.... this thread should be a good link for people asking about RAW in future.

S'cool.

As for your latest findings, I have to say there is little info out there to question the linear conversion method. Im not overly convinced by the 'outback' website.... seems you need an action to get the most out of thier method..... and that action is at a price ....... but the theory is sound...... I am unable to find any further information which says otherwise. Nor can I find any more info that implies that thier are any other benefits for altering an image in RAW software as apposed to PS.

Yeah, that's the same boat I'm in. I'm going to try out the linear method myself, though. I think the action is just to give you an automated and quick way to get to a visually "usable" image.

It does seem to me that as long as thier is no information loss or compression from the transition between RAW software and PS, then there is no real benefit to sticking with editing using RAW software....... it all comes to personal preference...... and in a similar way to matt, im use to my method...... and the chances of being able to notice a difference in the finished artical is slim to none.

That's exactly my thought, and until there's some proof otherwise, I didn't want to see other workflows labeled as inferior.

It does beg one question tho......

Sort of...

why some of the more expensive RAW converters claim to produce better results?........ this could just be the fact that they have more options for the people that wish to edit more in RAW software...... but they can also be misleading....... the following is from a magazine about using RAW....

"While the results from the Adobe Camera Raw plugin are quite acceptable, for example, professional level software like Capture One Pro produces the best results of all - though at 287 pounds (thats $540), you might expect it to."

So is this not rubbish?..... if you can make the same adjustments in ps without damaging the image file, how can this software be better?

It could be the number of adjustments, or how the adjustments are made. There are certain adjustments in PS that will kill your data. Again, this is repeating info I've heard elsewhere, but as far as I know, using the dodge and burn features, or contrast and brightness, can lose data. You lose distinction between one pixel and the next. If two pixels next to each other (in b&w to make it simple) are at 240 and 250, they can be changed to 242 and 244 and still have a distinction, even if that distinction is visiably smaller. You can make more changes and they will still be treated as different colors. However, if they both become 242, then that distinction is lost and they will forever after be treated as the same color. Data is "lost".

If the conversion in the PS RAW software isn't good, then maybe it isn't keeping all of the various levels of color distinct. That's one of the benefits of working in higher bit color, like 16bit vs. 8bit, and why you should use 16bit TIFFs when converting linearly. There are more levels of distinction, and you might compress the blacks otherwise.

Or maybe people just don't like how the conversion looks, and there really isn't a data problem. There doesn't seem to be a lot of hard info out there, just people sharing their workflows.

 

[markc]

I get very different results from different raw software, with my 20D files. C1 handles noise and sharpening differently than any other software, including photoshop.

That's a good point. There's a lot more going on than just the color conversion if you have the other options checked. I could see those easily coming into play. I guess we each need to do some comparisons and find which software does what better. Maybe our RAW converter removes noise better than whatever PS plugin we are using, but there may be a plugin out there that does a better job if we weren't happy with it.