Author | Thread |
|
10/15/2006 09:00:51 AM · #26 |
I'd keep all the originals in one place, and, if I wanted to work on one, I'd take a copy of that one file and work on the copy. That way no matter what your editing program(s) do, you've always got the original to go back to. |
|
|
10/15/2006 09:22:37 AM · #27 |
the first thing i do, when i open a raw file is to duplicate it. then i close the cr2 file and don't save it, therefore i have the original raw and the new psd version of the image.
|
|
|
10/15/2006 09:48:35 AM · #28 |
Interesting/informative conversation. I haven't been back in since posting and just got caught up...
Originally posted by GeneralE: Originally posted by glad2badad: A JPEG file is compressed (I believe, 8bit), the RAW file gives you all the details (I think 24bit). |
JPEG compression has nothing to do with the bit-depth (color range) of the image -- an "8-bit" image is actually 8 bits per channel or a 24-bit (6.7 million colors) color space. "16-bit" color is likewise 16 bits/channel, or 48-bits overall. |
Regarding the post on JPEG 8 bit vs RAW 24 bit - I was talking about the data that is available. If you have an original JPEG you have approximately 2/3 less data to work with in post-processing than you do with an original RAW file - yes? Once that JPEG is generated you're losing information that can't be regained.
From some other posts I've read here it sounds like I have it just about right, or did I miss something? Trying to keep it straight in my own head. Thanks! :D
|
|
|
10/15/2006 09:55:45 AM · #29 |
This is a very crude explanation, and I am sure that I am leaving many details out.
When a compressed image is made a software program goes pixel by pixel and compares each pixel to is nearest neighbors. The software will then decide, based on the level of compression, which pixels are similar enough to their neighbors to average the data they contain and discard the original. Once this is done the image file only needs to contain the distinct averaged data and the locations on the image where that averaged data is contained.
To see examples of this look at images on this site where the subject of the photo has the sky as a background. Where the subject and the sky transition you might find some areas that appear blocky. These blocks are examples of the software not getting the pixel comparisons and averages correct.
If you are willing to live with this missing data when 'manipulating' your images then shoot jpeg. If you want to work with all the details that you can when 'developing' your images then shoot raw. |
|
|
10/15/2006 10:07:01 AM · #30 |
Originally posted by e301: ... But what you really win is detail ... |
Good point! This possibly is the most important reason for choosing RAW.
|
|
|
10/15/2006 10:07:58 AM · #31 |
i would suspect if you turn your parameter settings to all 0's and shoot in AWB the differences between the JPEG and RAW files would be minimal.
i can't vouch for this as i haven't once looked at the embedded JPEG file from my 10d.
|
|
|
10/15/2006 12:05:23 PM · #32 |
I was a jpg shooter before i 'went pro' - to go and take 100 pics in a day or something is not a big deal in JPG, but I find the RAW workflow so much more efficient for large numbers of images. At a wedding I capture 800 or so.
I use Canon's DPP program in my workflow.
My Workflow:
Shoot and use a card reader to put the images on my HD.
Open the directory in DPP. In there is a Quick Check Tool - i use it to tag the good (or bad) images and cull the herd to the good one (throw out bad focus/technical, dupes you took to be safe, etc).
I then WB the images, all of them. Outdoor weddings make this easier :D
I then do any exposure correction
I then do any basic 'trimming' (DPP speak for cropping).
Next is RAW to JPG conversion - from start to now is 2-3 hours.
Shooting JPG is tons more work as there is no way to blanket correct for WB and altering exposure can destroy the image. The RAW file is never touched - so I can re-process at any time including the cropping.
There is more dynamic range and better color with RAW than with JPG. You have to find the workflow and/or software that lets you work the way you need /want to.
|
|
|
10/15/2006 12:17:45 PM · #33 |
Originally posted by glad2badad: Regarding the post on JPEG 8 bit vs RAW 24 bit - I was talking about the data that is available. If you have an original JPEG you have approximately 2/3 less data to work with in post-processing than you do with an original RAW file - yes? Once that JPEG is generated you're losing information that can't be regained. |
I don't think RAW is 24-bits/channel either; most likely 16-bit/channel. I don't think there are any (commercialy available) sensors sensitive enough to capture 24 bits/channel.
So, yes, the JPEG will have already discarded part of the color information, but it has little or nothing to do with the JPEG compression algorithm -- you'd have the same situation if you saved as 8-bit (per channel) uncompressed TIFF. |
|
|
10/15/2006 12:19:12 PM · #34 |
This last weekend I went camping in the Eastern Sierras (talk about freezing in the moring!!)
For the heck of it I decided to shoot RAW + JPEG. When I came home and started working with the pictures I took, I could not believe how much better the RAW files looked straight out of the camera vs the jpegs.
All I had to do really was adjust the raw settings (WB, exposure, shadows, contrast) a touch and it seemed no pp was needed with PS at all.
I was amazed at how much better the RAW images were right away and the minumal tweaking of exposure and wb, etc. did so much more for me than editing the jpegs in ps.
Is this unusual?
I thought the jpegs would look better than raw. The color differences were amazing! |
|
|
10/15/2006 12:30:16 PM · #35 |
i think RAW files are 16bit but only actually using 12bits.
this touches upon it
//www.luminous-landscape.com/tutorials/bit-depth.shtml
Originally posted by GeneralE: I don't think RAW is 24-bits/channel either; most likely 16-bit/channel. I don't think there are any (commercialy available) sensors sensitive enough to capture 24 bits/channel.
|
Message edited by author 2006-10-15 12:32:02.
|
|
|
10/15/2006 01:08:00 PM · #36 |
Originally posted by slickchik: ....I was amazed at how much better the RAW images were right away and the minumal tweaking of exposure and wb, etc. did so much more for me than editing the jpegs in ps.
Is this unusual?
I thought the jpegs would look better than raw. The color differences were amazing! |
Nope - RAW has more latitude and wiggle room, which some of us need :-) |
|
|
10/15/2006 01:10:40 PM · #37 |
Originally posted by Prof_Fate: I then do any basic 'trimming' (DPP speak for cropping). |
The 1 thing I wish RSE and/or Minolta's Dimage Viewer can do is crop. That would definately help.
|
|
|
10/15/2006 05:45:58 PM · #38 |
Originally posted by soup: i think RAW files are 16bit but only actually using 12bits.
this touches upon it
//www.luminous-landscape.com/tutorials/bit-depth.shtml
Originally posted by GeneralE: I don't think RAW is 24-bits/channel either; most likely 16-bit/channel. I don't think there are any (commercialy available) sensors sensitive enough to capture 24 bits/channel.
| |
Found some more here (2 articles) ...
JPEG Vs. RAW: The Advantages and Disadvantages Explained
RAW vs. JPEG: Which format should you shoot?
... does sound like 12-bit for RAW (I had 8-bit x 3 stuck in my head for some reason).
|
|
|
10/15/2006 05:55:12 PM · #39 |
That link does say 12 bits or more for RAW.
|
|
|
10/15/2006 06:29:50 PM · #40 |
Originally posted by stdavidson: I'm a RAW newbie. I'm interested in the salient differences between RAW and .jpg at capture time so want to compare the same picture captured simultaneously in RAW and in .jpg for comparison.
The extracted .jpg from a RAW+jpeg file is different from the RAW converted picture using "camera settings" with Adobe CS2's raw image converter.
Questions:
1-Why is the extracted .jpg different from the RAW converted file using camera settings?
2-For those that say, "RAW is 'better' than .jpg, I'd never go back", can you specify your MAIN reasons for saying that?
I understand the inherent technical advantageous of RAW over .jpg. For that alone I will use it for better enlargements, color transitions and prints. But I'm thinking those advantageous make very little difference in the type of web graphics we create for DPC challenges. Am I wrong? If so, why? |
As somebody who had these exact same questions, all I can tell you is try raw for a while. Once you do you will see the advantages pretty quickly. I have had some entries lately that have had little or no photoshop because I have been able to do all the adjustments in Raw. I used Raw Shooter on this and really only used photoshop for the border and the crop. The original is very green, a little on the over exposed side for my taste, and didn't fit the high contrast challenge. Because of the Raw file I was able to do tons with the exposure, hue, and contrast. I hated raw at first, but love it now that I am used to it. I won't go back because I really feel like I have more control over the images, even the ones where I might have made some errors in setting my camera.  |
|
|
10/16/2006 10:04:48 AM · #41 |
A big advantage to doing a lot in the RAW image converter as opposed to doing it in Photoshop is that converted 16-bit .tiff files get VERY large (100-200+ megs) very fast in PS.
Message edited by author 2006-10-16 10:14:38.
|
|
|
10/16/2006 10:58:10 AM · #42 |
Originally posted by stdavidson: A big advantage to doing a lot in the RAW image converter as opposed to doing it in Photoshop is that converted 16-bit .tiff files get VERY large (100-200+ megs) very fast in PS. |
Yes, but that depends on what you're saving in the TIFF file.
With my 10D, unedited 16-bit tiff files created with Adobe Camera Raw averaged about 36MB. With my 5D, they're about 75MB.
It's when you save editing data, such as layers and channels, that they start getting really big. I've had one or two files that got up over 500MB.
|
|
|
10/16/2006 11:29:49 AM · #43 |
Originally posted by micknewton: Originally posted by stdavidson: A big advantage to doing a lot in the RAW image converter as opposed to doing it in Photoshop is that converted 16-bit .tiff files get VERY large (100-200+ megs) very fast in PS. |
Yes, but that depends on what you're saving in the TIFF file.
With my 10D, unedited 16-bit tiff files created with Adobe Camera Raw averaged about 36MB. With my 5D, they're about 75MB.
It's when you save editing data, such as layers and channels, that they start getting really big. I've had one or two files that got up over 500MB. |
Yeah... that eats up disk space like it is going out of style and it slows down Photoshop as well. I did a family photo for Christmas cards for a niece and the converted file got over 300 meg with only a few standard layers I normally use for .jpg workflow.
The trick in RAW is to do as much of the 'standard' levels, curves, shadow/highlighting, contrast and color processing that you can in the conversion so there is much less to do PS. That keeps the final file sizes down.
If I could get a 50-70 Meg 16-bit converted .tiff image that had most of the post processing already completed then the master file is not significantly bigger than what I already do in 8-bit .tiff files from .jpg originals applying standard processing.
Personal Note:
Btw, Mick, you missed a nice little photoshoot last Saturday morning with DrAchoo, fas-ligand and jackryan. We shot Wahkeena and Fairy Falls.
Message edited by author 2006-10-16 11:36:44.
|
|
|
10/16/2006 11:51:59 AM · #44 |
Something I've always wondered with DPP is how much better straight out of the box raw files look in DPP as opposed to Photoshop or RSE. The contrast and tones are much better right off the bat in DPP, and images in the other programs are more flat.
I know the camera parameters set for jpegs don't have an effect on the actual raw images (even though I noticed the parameter data is still recorded in the exif), but it almost seems like DPP imports (or at least some of) the parameter data on the file and applies it to the raw file in the software. This is just a guess, nothing scientific or thorally reasearched. Has anybody else noticed this, or is it just me? |
|
|
10/16/2006 12:11:57 PM · #45 |
Originally posted by Telehubbie: Something I've always wondered with DPP is how much better straight out of the box raw files look in DPP as opposed to Photoshop or RSE. The contrast and tones are much better right off the bat in DPP, and images in the other programs are more flat.
I know the camera parameters set for jpegs don't have an effect on the actual raw images (even though I noticed the parameter data is still recorded in the exif), but it almost seems like DPP imports (or at least some of) the parameter data on the file and applies it to the raw file in the software. This is just a guess, nothing scientific or thorally reasearched. Has anybody else noticed this, or is it just me? |
It's not just you. Because DPP is Canon software, it fully understands the recorded parameters, and applies them to the image. The result should have WB, color, contrast and exposure almost identical to the in-camera JPEG. Other converters will not be able to lfully reproduce the in-camera settings. They may not even understand basic things like the as-shot white balance. Nikon, for instance, recently started to "encrypt" their WB information to make it more difficult for 3rd party converters to understand it. At least one 3rd party converter does understand it, but some do not, Adobe's products included. |
|
|
10/16/2006 12:40:58 PM · #46 |
Originally posted by kirbic:
It's not just you. Because DPP is Canon software, it fully understands the recorded parameters, and applies them to the image. |
Hey thanks Fritz! That's what I was thinking, but needed to hear it from a reliable source such as you. Thanks again. |
|
|
10/16/2006 12:41:14 PM · #47 |
Originally posted by kirbic: ...Nikon, for instance, recently started to "encrypt" their WB information to make it more difficult for 3rd party converters to understand it.... |
I saw that and it troubles me no end :-/ I can see a world where you are locked into the manafacturer tools set for RAW which in most cases is seriously lacking compared to some 3-rd party tools. At the least it's restrictive and shows a lack of respect for customers... but I guess it's a good idea for profit motives. |
|
|
10/16/2006 01:37:01 PM · #48 |
Originally posted by stdavidson: The trick in RAW is to do as much of the 'standard' levels, curves, shadow/highlighting, contrast and color processing that you can in the conversion so there is much less to do PS. That keeps the final file sizes down. |
For what it's worth... adjustment layers (such as curves and levels) don't chew up too much space. It's the bitmap layers that really make the file grow (and chew up memory and CPU at the same time).
So here's a trick to reduce the memory footprint of your edits:
If you are just doing a layer blending mode, such as Multiply or Soft Light, *without* actually touching the bits, then this will work:
Create a levels (or curves) layer and do nothing inside of it. Then set the blend mode of your adjustment layer to the desired blend mode. This has the SAME EFFECT as if you had "duplicated the bitmap layer" and set the blend mode! But since it's an adjustment layer, you haven't increased the memory footprint hardly at all.
Now if you want to get a little more creative, after you've set the blend mode - go back into the levels (or curves) layer you just made and try adjusting the settings. Sometimes this can be pretty cool - sometimes it's pretty whacked. But it's good to know that you can do it.
So the only time you really need to duplicate a layer is if you plan to edit the contents of the layer.
Message edited by author 2006-10-16 13:37:24.
|
|
|
10/16/2006 02:20:19 PM · #49 |
Yeah, it's too bad I can't be in two places at the same time. We've had several houseguests for the past week and they've been driving me nuts. Becky took 'em to Vancouver Saturday, so I had myself a nice, peaceful 'home alone' day.
|
|
|
10/16/2006 02:59:46 PM · #50 |
Originally posted by dwterry:
So here's a trick to reduce the memory footprint of your edits:
If you are just doing a layer blending mode, such as Multiply or Soft Light, *without* actually touching the bits, then this will work:
Create a levels (or curves) layer and do nothing inside of it. Then set the blend mode of your adjustment layer to the desired blend mode. This has the SAME EFFECT as if you had "duplicated the bitmap layer" and set the blend mode! But since it's an adjustment layer, you haven't increased the memory footprint hardly at all.
Now if you want to get a little more creative, after you've set the blend mode - go back into the levels (or curves) layer you just made and try adjusting the settings. Sometimes this can be pretty cool - sometimes it's pretty whacked. But it's good to know that you can do it.
So the only time you really need to duplicate a layer is if you plan to edit the contents of the layer. |
Wow! Great tip.
I tried it and it doesn't add a byte to the original .tiff file when added with an adjustment layer. In my test image it added 14 megabytes when added with a data layer. Big difference. I don't do that lots but that is something I'll remember.
They say you learn something new every day. I just got my allotment for today. :)
Thanks!
Message edited by author 2006-10-16 15:01:03.
|
|
Home -
Challenges -
Community -
League -
Photos -
Cameras -
Lenses -
Learn -
Help -
Terms of Use -
Privacy -
Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 07/19/2025 11:58:30 PM EDT.