Editing jpegs

[imagemaker46]

^^^if the most you want is snapshots and phone cam quality picss with very limited post flexibility. "Rush job"

So basically what you're saying is that if you shoot jpegs the quality you can expect is similar to having snapshots from a cell phone? The only reason someone would shoot jpegs is because they want to rush the job and quality isn't important?

I can see you're leading right up to your canned, "I'm a perfect photographer and I get it right every time because I'm that damned good" spiel.

Why you have to come off sounding like a total dick. So I shouldn't make a comment on this forum? I asked 2WheelPhoto a question that in no way had anything to do with you.

 

[KmH]

I believe people are better served by trying to give them accurate information.
Dumbed down explanations are to often so dumbed down they are inaccurate, counterproductive, and do way more harm than good.

 

[AaronLLockhart]

I believe people are better served by trying to give them accurate information.
Dumbed down explanations are to often so dumbed down they are inaccurate, counterproductive, and do way more harm than good.

"If you can't explain it to a six year old, you don't understand it yourself." - Albert Einstein.

 

[pixmedic]

Editing jpegs is really the same as editing raw files once you open them in lightroom. you still go through all the editing motions, you still export the file, nothing really changes. if you are editing in lightroom anyway, you might as well shoot raw and get the extra data wiggle room. timewise though, there really isnt any difference between editing a jpeg compared to a raw file. plenty of people shoot jpeg, edit in lightroom or photoshop (or whatever program they use) and turn out a great picture. the advantage to the extra data in a raw file however, is that you have a better chance of salvaging a photo taken with less than ideal camera settings. I shoot raw because like my golf game...I need the handicap.

 

[HughGuessWho]

I thought the point of shooting jpg was to avoid the time spent editing later. That's not to suggest that jpg is inferior to raw, just that the advantage is the lack of editing and pp, which is part and parcel of shooting raw.

That depends really on who you are,.... However, they are so large in file size, it's harder on your computer hardware.

So, let me ge this straight... the reason to shoot jpeg or raw depends on who I am?
Also, raw files are hard on my computer hardware, but jpegs aren't?

Are you sticking to those statments as well?

 

[AaronLLockhart]

I thought the point of shooting jpg was to avoid the time spent editing later. That's not to suggest that jpg is inferior to raw, just that the advantage is the lack of editing and pp, which is part and parcel of shooting raw.

That depends really on who you are,.... However, they are so large in file size, it's harder on your computer hardware.

So, let me ge this straight... the reason to shoot jpeg or raw depends on who I am?
Also, raw files are hard on my computer hardware, but jpegs aren't?

Are you sticking to those statments as well?

Yes, you have obviously never tried to load a RAW file on a P4 Machine with 1GB of RAM and a 256mb graphics card, verus a Jpeg. I'm a computer WIZ! I build them, take them apart, tweak them, water cool them, overclock them, game on them, etc.



But hell, what would I know



Secondly, yes, it does matter who you are. Because if you've never used photoshop or lightroom a day in your life, or use it very little, you're not going to "save any time" editing jpeg over Raw, or vice versa.

If you've got more photoshop experience, non-photographic (such as myself) you can edit jpegs much faster than you can RAW, because you know where all of the tools are, how they work etc.

So, yes, I will very PROUDLY stand by all of that that I said... and you may stick your foot in your mouth.

 

[Majeed Badizadegan]

It has nothing to do with "who you are".


If you care about your photos, you shoot RAW. And if you're running 1gb ram and 256gb graphics card in 2013, you have bigger problems.

 

[AaronLLockhart]

It has nothing to do with "who you are".

If you care about your photos, you shoot RAW. And if you're running 1gb ram and 256gb graphics card in 2013, you have bigger problems.

1. That's a hugely biased opinion. Really the truth of the matter is that RAW allows for a larger margin of error. The quality difference isn't strong enough to tell by the eye on a standard sized print. So, that's simply matter of preference. I actually prefer most of the time to shoot in Jpeg, because I have e VERY extensive knowledge base of Adobe Photoshop. I can edit white balance in photoshop using hue/saturation, and color temperature pallets just as easily as you can in LightRoom or ACR with a white balance slider.

2. I'm not running that. I'm running an AMD 4.0 ghz 8 core with 32GB 1600 ripjaws and twin 7850's in crossfire, giving me a total of 4gb gddr5 dedicated video memory. I also have a 2.9 i5 quad core 27" imac with 8gb 1333, and a 1gb geforce gpu upgrade. I was simply saying its rather ignorant to think that a raw file doesn't stress hardware more than a jpeg. After all, that's why compressed files exist in the first place, to make things faster, easier, and less stressful on the machinery.

 

[imagemaker46]

I shoot jpegs because it is my choice and because of the type of work I do. I don't always have time, and I don't need to tweek every single pixel out of an image. I have never had any problems shooting jpegs. If people want to shoot raw because they feel it is the best for them, terrific, if they are planning on hanging poster size images on their walls and need every little detail, then shoot raw. Does it matter to me what any other photographer does with his camera, not in the slightest, it has no affect on the images I need to do my job. It has everything to do with the photographer that is shooting and what his or her requirements are.

Shoot raw or jpeg, both work, there is no right or wrong to this discussion.

 

[TCampbell]

Fundamentally you cannot have more stops of dynamic range per channel than the bit depth per channel. So an 8 bit per channel JPEG has 8 stops of dynamic range.

If we pick on, say, just the green channel, then the following is true:

0000 0000 - value 0: that's black (no green light whatsoever)

0000 0001 - value 1: that's a very dark green - near black. There is no variability within that level.

0000 0010 - value 2: that's still a dark green but it is technically "twice" as bright. There is minimal variability.
0000 0011 - that's not a doubling ... that's a half-tone brighter than the above.

0000 0100 - value 4: that's actually twice as bright (finally) as the "0010" color. Since there are two more bits... you can have quarter tones within this "stop" of light.
0000 0101 \
0000 0110 -> these are quarter tones within this stop
0000 0111 /

0000 1000 - value 8: this is now twice as bright as the "0100" color. Since three bits trail, you technically get 1/8th tones within this "stop" of light.
0000 1001 \
0000 1010 \
0000 1100 \
0000 1101 /
0000 1110 /
0000 1111 /

0001 0000 - value 16: this is now twice as bright as the "1000" color. Since 4 bits trail, you get 16 levels of variability within this stop.
0001 0001 \
0001 0010 \
0001 0011 \
0001 0100 \
0001 0101 \
0001 0110 \
0001 0111 \
0001 1000 >
0001 1001 /
0001 1010 /
0001 1011 /
0001 1100 /
0001 1101 /
0001 1110 /
0001 1111 /

By now you should be noticing the pattern... you get 1 stop per bit... 8 bits = 8 stops. The difference is the trailing bits behind the significant bit allows for sub-levels within that stop. Remember a stop is a "doubling" of the light level... its not linear system... it log base2 system. I wont continue to detail out the stops because each time I do, I'll need twice as many rows and this will be a very long post.

But because the trailing bits create some variability of range within a stop, really only about the top 5 stops have latitude and the lower 3 stops have no latitude, half-tone latitude or quarter-tone. This is why dark areas in a JPEG often look a bit "blotchy" with digital artifacts.

I think of JPEG as having about 5 "useful" stops of dynamic range and I suppose one could debate whether a generous person would be willing to give it 6 stops of dynamic range.

Meanwhile... raw usually has 14 bits per channel. So even if you strip away 3 stops because the level of granularity within that "stop" isn't good enough, you STILL have 11 stops left over! The latitude in raw files is amazing. This, btw, is one of the key reasons why, over and over, you'll see comments about by landscape photographers that you should be shooting raw for landscapes. It's because you'll almost certainly need more than 5 stops of dynamic range and JPEG wont do very well with that.

BTW, this has nothing to do with getting the image right in the camera vs. post adjustment. This assumes that both the raw and jpeg photographer got it "right" in the camera... the raw simply has better dynamic range for those images that require it.

If you shoot in JPEG then your camera is performing adjustments to your image (in the camera) even before you unload. The camera always actually records a raw image. The question is... does it save the raw data to the memory card or does it process it, convert it to a jpeg, and THEN save it to the memory card. JPEGs almost always have (at a minimum) white balance, sharpening, and de-noising... and may have additional processing as well (lens de-vignetting, etc.). A raw has no 'destructive' processing performed (nothing that would result in a loss of data). That means when you import a raw into a computer and look at it next to a jpeg, the jpeg may actually look "smoother" and "sharper" because it's had some processing performed. Many raw workflow tools automatically perform commonly required raw processing merely because the image was imported into the computer -- making it just as easy to work with raw as with jpeg (you don't have to know the "standard" things... the software does it all for you automatically.) The difference is, in jpeg, an arbitrary adjustment was performed by the camera and the original data was tossed away. In raw, an arbitrary adjustment was made by software but the original data was NOT discarded and you can override those default adjustments. Basically you have control.

The downside of raw? I can think of two. (1) it occupies more disk space and (2) you can't necessarily just send someone a raw file and expect that they can view it. Raw isn't a "standard" format... it's a concept for a format. Each camera has a slightly different raw format but they all follow the same rules (if they don't then they aren't technically 'raw'.) A recipient would need to have a raw viewer/importer/converter for your specific model camera.

It's not so much that it's 'wrong' to use JPEG. Lots of people shoot in JPEG. Even many top pros use JPEG. Just be aware of the difference. There are many situations for which JPEG is not only "fine" -- it's preferred. But if you need the latitude for higher dynamic range, the ability to preserve detail... particularly near the limits of the dynamic range (in the whites & blacks... HINT: wedding dresses are usually "white" and tuxes are usually "black") then probably your situation is a very strong candidate for shooting in raw.

 

[Ysarex]

Raw is an uncompressed file source. What you get when shooting a raw photograph is an interpretation of exactly the sensor saw. When you get into Jpeg, you start losing some of the quality in detail because of data compression. You edit them the same. However, raw images will always give you a higher range of alteration, because of the fact that none of the data was lost from the sensor to the file.

It really doesn't matter which one you shoot in, as long as you get the settings right in camera.

This is wrong. It really can matter quite a lot regardless of any settings that may be available in your camera and how you use them.

However, before providing a clear example of why this is wrong I want to stress another point. I noticed that Scott posted in this thread and I know he shoots camera JPEGs. I would do the same if I were in his business. There are many appropriate reasons to make the best possible JPEG and I can't think of a more appropriate one than shooting sports journalism when you're working under deadlines so extreme that some photos get web published before the event is over -- not a lot of time there for raw processing. Take a look at the photos on Scott's website and be properly humbled about your own skill using a camera.

I just bought a camera that is wi-fi capable so that I can take a photo and upload it directly from the camera before taking a step away from where I was standing; that's if I upload the JPEG, not the raw file. We have all seen how this capability in cell phone cameras has literally changed our world. Nobody's going to be uploading raw files anytime soon.

With the above said; there are circumstances where a raw file skillfully processed is the only way to get the shot and that no matter what you set on the camera JPEG engine that software will crash and burn and leave you with nothing but cr*p that you won't be able to salvage by editing. Camera JPEG engines work reasonably well when they're not stressed by challenging lighting conditions. Even so, they are engineered to produce mediocrity as their target -- that's what the word automatic means.

An example: Over Spring Break I took a short drive down to Big Oak Tree State Park and I took this photo:

That scene has just begun to become backlit. Backlighting is challenging lighting. You can use ADL on your Nikon or HTP on your Canon or any combination of camera contrast, backlight and exposure settings till your blue in the face and no camera JPEG engine will give you the photo I produced from the raw file. AND no matter what you do you will not be able to salvage the camera JPEG in editing to produce the photo I got from the raw file. This is a backlit landscape. If I'm shooting camera JPEGs I should have fixed the lighting right? Yeah right. I should have come back after the sky went fully overcast an hour latter? I should have waited until it wasn't backlit or arranged to take the photo a couple months later when the sun wouldn't be so far into the southern sky? Yeah right. I don't have to be limited by the restrictive capabilities of my camera's rather crude processing software.

I wanted the backlight and I knew how to handle it with a raw capture and post processing. Most critically I didn't want an exposure that would clip the highlights in the clouds and I wanted as much color in that hazy sky as I could get. I took the photo with a -1 exp. comp. Here's that camera JPEG unedited:

There's nothing that you could set on the camera that would have coaxed it to take my version of this photo. The shadows in the JPEG are blocked up and you can't recover them now in editing. Of course the shadows are blocked up, I set a -1 exp. comp. what do you expect. But if I hadn't set that -1 exp. comp. the camera would have clipped the highlights -- this is a case of pick your favorite mistake. And you can't edit the JPEG now to salvage the photo. The data isn't there any more. If you don't believe that here's a higher-res version of the JPEG; go ahead and try: http://fc08.deviantart.net/fs71/f/2013/087/3/7/untitled_by_skoparon-d5zk2rj.jpg

If you shoot JPEGs you can get some good photos and they can be tweaked within rather constrictive limits to be made better. Photographer's who can do that and do it well -- Scott for example -- have a very respectable (and profitable) skill. It is a different skill than mine. Different is good. What we need to do is recognize when and why different is good and put things appropriately in their place. You can't tell me it doesn't matter that I shoot raw and that if I set the camera right I can get the same result shooting JPEG -- that's total nonsense. It's likewise nonsense to tell Scott the next time he photographs a sporting event where he's expected to deliver results 30 minutes after the event ends that he should be taking all his photos in raw.

Joe

 

[AaronLLockhart]

This is wrong. It really can matter quite a lot regardless of any settings that may be available in your camera and how you use them.

However, before providing a clear example of why this is wrong I want to stress another point. I noticed that Scott posted in this thread and I know he shoots camera JPEGs. I would do the same if I were in his business. There are many appropriate reasons to make the best possible JPEG and I can't think of a more appropriate one than shooting sports journalism when you're working under deadlines so extreme that some photos get web published before the event is over -- not a lot of time there for raw processing. Take a look at the photos on Scott's website and be properly humbled about your own skill using a camera.

I just bought a camera that is wi-fi capable so that I can take a photo and upload it directly from the camera before taking a step away from where I was standing; that's if I upload the JPEG, not the raw file. We have all seen how this capability in cell phone cameras has literally changed our world. Nobody's going to be uploading raw files anytime soon.

With the above said; there are circumstances where a raw file skillfully processed is the only way to get the shot and that no matter what you set on the camera JPEG engine that software will crash and burn and leave you with nothing but cr*p that you won't be able to salvage by editing. Camera JPEG engines work reasonably well when they're not stressed by challenging lighting conditions. Even so, they are engineered to produce mediocrity as their target -- that what the word automatic means.

An example: Over Spring Break I took a short drive down to Big Oak Tree State Park and I took this photo:

That scene has just begun to become backlit. Backlighting is challenging lighting. You can use ADL on your Nikon or HTP on your Canon or any combination of camera contrast, backlight and exposure settings till your blue in the face and no camera JPEG engine will give you the photo I produced from the raw file. AND no matter what you do you will not be able to salvage the camera JPEG in editing to produce the photo I got from the raw file. This is a backlit landscape. If I'm shooting camera JPEGs I should have fixed the lighting right? Yeah right. I should have come back after the sky went fully overcast an hour latter? I should have waited until it wasn't backlit or arranged to take the photo a couple months later when the sun wouldn't be so far into the southern sky? Yeah right. I don't have to be limited by the restrictive capabilities of my camera's rather crude processing software.

I wanted the backlight and I knew how to handle it with a raw capture and post processing. Most critically I didn't want an exposure that would clip the highlights in the clouds and I wanted as much color in that hazy sky as I could get. I took the photo with a -1 exp. comp. Here's that camera JPEG unedited:

There's nothing that you could set on the camera that would have coaxed it to take my version of this photo. The shadows in the JPEG are blocked up and you can't recover them now in editing. Of course the shadows are blocked up, I set a -1 exp. comp. what do you expect. But if I hadn't set that -1 exp. comp. the camera would have clipped the highlights -- this is a case of pick your favorite mistake. And you can't edit the JPEG now to salvage the photo. The data isn't there any more. If you don't believe that here's a higher-res version of the JPEG; go ahead and try: http://fc08.deviantart.net/fs71/f/2013/087/3/7/untitled_by_skoparon-d5zk2rj.jpg

If you shoot JPEGs you can get some good photos and they can be tweaked within rather constrictive limits to be made better. Photographer's who can do that and do it well -- Scott for example -- have a very respectable (and profitable) skill. It is a different skill than mine. Different is good. What we need to do is recognize when and why different is good and put things appropriately in their place. You can't tell me it doesn't matter that I shoot raw and that if I set the camera right I can get the same result shooting JPEG -- that's total nonsense. It's likewise nonsense to tell Scott the next time he photographs a sporting event where he's expected to deliver results 30 minutes after the event ends that he should be taking all his photos in raw.

Joe

Well, that being the case, I don't shoot in those types of environments. I specialize in portraiture, with manipulated lighting. I can't tell the difference between a RAW or Jpeg in my field of interest. That is all.

 

[Ysarex]

Well, that being the case, I don't shoot in those types of environments. I specialize in portraiture, with manipulated lighting. I can't tell the difference between a RAW or Jpeg in my field of interest. That is all.

Thanks, that's much better. I agree that when you can manipulate the lighting the advantages of raw capture are substantially obviated.

Joe

 

[imagemaker46]

Fundamentally you cannot have more stops of dynamic range per channel than the bit depth per channel. So an 8 bit per channel JPEG has 8 stops of dynamic range.

If we pick on, say, just the green channel, then the following is true:

0000 0000 - value 0: that's black (no green light whatsoever)

0000 0001 - value 1: that's a very dark green - near black. There is no variability within that level.

0000 0010 - value 2: that's still a dark green but it is technically "twice" as bright. There is minimal variability.
0000 0011 - that's not a doubling ... that's a half-tone brighter than the above.

0000 0100 - value 4: that's actually twice as bright (finally) as the "0010" color. Since there are two more bits... you can have quarter tones within this "stop" of light.
0000 0101 \
0000 0110 -> these are quarter tones within this stop
0000 0111 /

0000 1000 - value 8: this is now twice as bright as the "0100" color. Since three bits trail, you technically get 1/8th tones within this "stop" of light.
0000 1001 \
0000 1010 \
0000 1100 \
0000 1101 /
0000 1110 /
0000 1111 /

0001 0000 - value 16: this is now twice as bright as the "1000" color. Since 4 bits trail, you get 16 levels of variability within this stop.
0001 0001 \
0001 0010 \
0001 0011 \
0001 0100 \
0001 0101 \
0001 0110 \
0001 0111 \
0001 1000 >
0001 1001 /
0001 1010 /
0001 1011 /
0001 1100 /
0001 1101 /
0001 1110 /
0001 1111 /

By now you should be noticing the pattern... you get 1 stop per bit... 8 bits = 8 stops. The difference is the trailing bits behind the significant bit allows for sub-levels within that stop. Remember a stop is a "doubling" of the light level... its not linear system... it log base2 system. I wont continue to detail out the stops because each time I do, I'll need twice as many rows and this will be a very long post.

But because the trailing bits create some variability of range within a stop, really only about the top 5 stops have latitude and the lower 3 stops have no latitude, half-tone latitude or quarter-tone. This is why dark areas in a JPEG often look a bit "blotchy" with digital artifacts.

I think of JPEG as having about 5 "useful" stops of dynamic range and I suppose one could debate whether a generous person would be willing to give it 6 stops of dynamic range.

Meanwhile... raw usually has 14 bits per channel. So even if you strip away 3 stops because the level of granularity within that "stop" isn't good enough, you STILL have 11 stops left over! The latitude in raw files is amazing. This, btw, is one of the key reasons why, over and over, you'll see comments about by landscape photographers that you should be shooting raw for landscapes. It's because you'll almost certainly need more than 5 stops of dynamic range and JPEG wont do very well with that.

BTW, this has nothing to do with getting the image right in the camera vs. post adjustment. This assumes that both the raw and jpeg photographer got it "right" in the camera... the raw simply has better dynamic range for those images that require it.

If you shoot in JPEG then your camera is performing adjustments to your image (in the camera) even before you unload. The camera always actually records a raw image. The question is... does it save the raw data to the memory card or does it process it, convert it to a jpeg, and THEN save it to the memory card. JPEGs almost always have (at a minimum) white balance, sharpening, and de-noising... and may have additional processing as well (lens de-vignetting, etc.). A raw has no 'destructive' processing performed (nothing that would result in a loss of data). That means when you import a raw into a computer and look at it next to a jpeg, the jpeg may actually look "smoother" and "sharper" because it's had some processing performed. Many raw workflow tools automatically perform commonly required raw processing merely because the image was imported into the computer -- making it just as easy to work with raw as with jpeg (you don't have to know the "standard" things... the software does it all for you automatically.) The difference is, in jpeg, an arbitrary adjustment was performed by the camera and the original data was tossed away. In raw, an arbitrary adjustment was made by software but the original data was NOT discarded and you can override those default adjustments. Basically you have control.

The downside of raw? I can think of two. (1) it occupies more disk space and (2) you can't necessarily just send someone a raw file and expect that they can view it. Raw isn't a "standard" format... it's a concept for a format. Each camera has a slightly different raw format but they all follow the same rules (if they don't then they aren't technically 'raw'.) A recipient would need to have a raw viewer/importer/converter for your specific model camera.

It's not so much that it's 'wrong' to use JPEG. Lots of people shoot in JPEG. Even many top pros use JPEG. Just be aware of the difference. There are many situations for which JPEG is not only "fine" -- it's preferred. But if you need the latitude for higher dynamic range, the ability to preserve detail... particularly near the limits of the dynamic range (in the whites & blacks... HINT: wedding dresses are usually "white" and tuxes are usually "black") then probably your situation is a very strong candidate for shooting in raw.

I was lost at "Fundamentally"

 

[HughGuessWho]

That depends really on who you are,.... However, they are so large in file size, it's harder on your computer hardware.

So, let me ge this straight... the reason to shoot jpeg or raw depends on who I am?
Also, raw files are hard on my computer hardware, but jpegs aren't?

Are you sticking to those statments as well?

Yes, you have obviously never tried to load a RAW file on a P4 Machine with 1GB of RAM and a 256mb graphics card, verus a Jpeg. I'm a computer WIZ! I build them, take them apart, tweak them, water cool them, overclock them, game on them, etc.



But hell, what would I know



Secondly, yes, it does matter who you are. Because if you've never used photoshop or lightroom a day in your life, or use it very little, you're not going to "save any time" editing jpeg over Raw, or vice versa.

If you've got more photoshop experience, non-photographic (such as myself) you can edit jpegs much faster than you can RAW, because you know where all of the tools are, how they work etc.

So, yes, I will very PROUDLY stand by all of that that I said... and you may stick your foot in your mouth.

By your own comment, it has NOTHING to do with WHO YOU ARE, rather what your experience is.

Though the pictures of your Super Dooper computer demonstrate that you can build a computer with pretty neon lights and water cooled processor with all the bells and whistles, it means nothing in reference to your statement that raw file are larger and harmful to computer hardware. I have been in the IT field since 1985 and managing truly talented IT people for over half that time, so a self-proclaimed computer “Wiz” impresses me none.

Both of your statements are wrong, so I wont be sticking my foot anywhere.