Ugly Hedgehog - Photography Forum
Home Active Topics Newest Pictures Search Login Register
Photo Analysis
Raw vs jpg: Can you spot the difference?
Page <<first <prev 4 of 7 next> last>>
May 18, 2017 10:02:13   #
streetmarty Loc: Brockton, Ma
 
Shootist wrote:
DMan
Thanks for the work you have done. It does really show what the differences are. I personally shoot RAW+jpeg and for some type of images jpeg does just fine. However, as some here have written, RAW gives the PP latitude that some images really cry out for, landscapes for example. I realize that what will draw most folks into a print will be composition and colors, not technical perfection; this is as it should be. But for my own satisfaction I want to get the best technical photo I can produce, sometimes it is a jpeg other times a RAW. I do it for me, if others like what I do then that is a really nice bonus.
DMan br Thanks for the work you have done. It does... (show quote)


Whenever a "Raw vs jpeg" question comes up on this site, this is what should pop up and that should be the end of the thread!! Perfect!!

But no....we will just keep beating this dead horse day after day.



Reply
May 18, 2017 10:05:45   #
RichieC Loc: Adirondacks
 
This discussion keeps coming up. It takes on the atmosphere of ones home team you are rooting for- we also see this very same thing in politics.

JPEG stands for Joint Photographic Experts Group, which created the standard. It is simply a "lossful" adjustable compression algorithm whose primary objective is to compress file sizes in large part by pixel averaging. The range of color and value of individual pixels it chooses to flatten- or make the very same color, is adjustable for greater compression. Each time you re-save a jpeg image, the algorithm is applied again, averaging even more of a swath of pixels with a progressive degradation of detail from the original. That's the facts of what JPEG is. So why use it? JPEG is widely supported , and in fact does a pretty good job for all the reasons most people root for it. It is not superior past that in any way- never pretended to be.

This "rooting" of RAW vs JPEG is humorous to me because its specious- as every jpeg started its life as a raw image. SO in a nutshell, what is being compared is which is superior- Team negative or team print- trouble is they are different sports- the real trouble is some don't realize this and continue to root for their team anyways... The computer in your camera decided what was good enough for you, and ran the image through the algorithm, compressed the image and dumped detail- no argument about it- just what happened. Like dropping off your negatives at the pharmacy to get prints made- some are very happy with that. They don't really take much more than a second to analyze each image, they run them through as quick as possible. You pay extra to have them enlarge a single print and adjust the print settings-you'd expect a better print- so it is work flow that has changed.

RAW is the "negative", JPEG, TIFF, png, or any other final format are the "print" Some retain detail much better than others. Those who would process their own negative to control the final image as it will finally appear- should shoot in raw- in the max bit depth, if they understand the process.Detail can always be dumped, but can not be acutely recreated- even if using fractals etc that can sort of create believable detail. I can set Lightroom to make a my entire chips content of RAW DNG's into jpeg's pretty quickly. THEN we can get both teams playing on the same sport and compare my computers decision on how to handle the raw digital negative, over your in-camera computers decisions over your original, and now gone forever, RAW DNG's you yourself took... get it?

So the argument isn't which is the superior format, it's the potential that one work flow can deliver over another... but even that is a variable. If the individual capture is crap - the jpeg is going to be crap, you may have the possibility or resurrecting the image much more from the RAW DNG, but there may be nothing of value to work with... RAW does not a good photograph make. But even if the perfect image was captured, with no PP necessary, a tiff or other lossless format will enlarge and print better than a JPEG- very high end printers demand them, The ones the want JPEG- are beacuse they prefer them for their workflow- not because they are superior. To get the best out of them, I convert my tiff or PSD file to jpeg at final size, as the very final step before sending.

This boils down to only an argument of work flow and final intent. If you are shooting for large/high quality reproductions, and all prints are enlargements over the 35mm ( full frame sensor size) original capturer size... to be hung and displayed or be sold, good-enough-for-who-its-for will not compete well. JPEG is the end of a life of an image- not the best choice for a beginning.

There is no argument on this- but go ahead and root for your favorite team in images, football and politics- just don't pretend you are rooting for the best team or even the same sport. (Yankees are better than the Patriots- RAh Rah)... I'll sit back, sip my beer and watch LOL.

Reply
May 18, 2017 10:12:14   #
rehess Loc: South Bend, Indiana, USA
 
RichieC wrote:
This discussion keeps coming up. It takes on the atmosphere of ones home team you are rooting for- we also see this very same thing in politics.

JPEG stands for Joint Photographic Experts Group, which created the standard. It is simply a "lossful" adjustable compression algorithm whose primary objective is to compress file sizes in large part by pixel averaging. The range of color and value of individual pixels it chooses to flatten- or make the very same color, is adjustable for greater compression. Each time you re-save a jpeg image, the algorithm is applied again, averaging even more of a swath of pixels with a progressive degradation of detail from the original. That's the facts of what JPEG is. So why use it? JPEG is widely supported , and in fact does a pretty good job for all the reasons most people root for it. It is not superior past that in any way- never pretended to be.

This "rooting" of RAW vs JPEG is humorous to me because its specious- as every jpeg started its life as a raw image. SO in a nutshell, what is being compared is which is superior- Team negative or team print- trouble is they are different sports- the real trouble is some don't realize this and continue to root for their team anyways... The computer in your camera decided what was good enough for you, and ran the image through the algorithm, compressed the image and dumped detail- no argument about it- just what happened. Like dropping off your negatives at the pharmacy to get prints made- some are very happy with that. They don't really take much more than a second to analyze each image, they run them through as quick as possible. You pay extra to have them enlarge a single print and adjust the print settings-you'd expect a better print- so it is work flow that has changed.

RAW is the "negative", JPEG, TIFF, png, or any other final format are the "print" Some retain detail much better than others. Those who would process their own negative to control the final image as it will finally appear- should shoot in raw- in the max bit depth, if they understand the process.Detail can always be dumped, but can not be acutely recreated- even if using fractals etc that can sort of create believable detail. I can set Lightroom to make a my entire chips content of RAW DNG's into jpeg's pretty quickly. THEN we can get both teams playing on the same sport and compare my computers decision on how to handle the raw digital negative, over your in-camera computers decisions over your original, and now gone forever, RAW DNG's you yourself took... get it?

So the argument isn't which is the superior format, it's the potential that one work flow can deliver over another... but even that is a variable. If the individual capture is crap - the jpeg is going to be crap, you may have the possibility or resurrecting the image much more from the RAW DNG, but there may be nothing of value to work with... RAW does not a good photograph make. But even if the perfect image was captured, with no PP necessary, a tiff or other lossless format will enlarge and print better than a JPEG- very high end printers demand them, The ones the want JPEG- are beacuse they prefer them for their workflow- not because they are superior. To get the best out of them, I convert my tiff or PSD file to jpeg at final size, as the very final step before sending.

This boils down to only an argument of work flow and final intent. If you are shooting for large/high quality reproductions, and all prints are enlargements over the 35mm ( full frame sensor size) original capturer size... to be hung and displayed or be sold, good-enough-for-who-its-for will not compete well. JPEG is the end of a life of an image- not the best choice for a beginning.

There is no argument on this- but go ahead and root for your favorite team in images, football and politics- just don't pretend you are rooting for the best team or even the same sport. (Yankees are better than the Patriots- RAh Rah)... I'll sit back, sip my beer and watch LOL.
This discussion keeps coming up. It takes on the a... (show quote)
I believe exactly the same discussion would have occurred twenty years ago between those who used negative film which they processed in their own darkroom and those of us who used Kodachrome which was processed by automation.

Reply
 
 
May 18, 2017 11:35:24   #
camerapapi Loc: Miami, Fl.
 
All JPEG images are either processed in camera or processed by us. In both cases some quality will be lost depending on compression. Modern JPEGs shot at minimal compression are excellent files and once saved as TIFF they become lossless.
RAW data from the sensor requires editing and finally a compressed JPEG, especially for printing. Compression if done at a low compression rate will end up in an excellent JPEG file. Nobody can deny that the RAW data from the sensor has all the information recorded during shooting. When the file is edited and finally compressed we loose pixels and details. Keep in mind that a 14 bit per pixel file becomes an 8bits per pixel when compressed.
I shoot both files depending on my subject. I always go with the less compression that will give me the quality I need when using JPEG images. It is not the same to compress for the web as when compressing for a print if you know what I mean.
Because a JPEG looses data every time it is saved it is best to save it as a TIFF and when working with the original JPEG file work on a copy.
Some of my best shots have come from original JPEG files following the least amount of compression.

Reply
May 18, 2017 12:17:55   #
cdayton
 
Unless I missed it, no one mentioned that both images look very "overprocessed " and unnatural. That bothers me more than if under extreme magnification there is additional noise in the jpeg, which there is.

Reply
May 18, 2017 12:28:08   #
MichaelH Loc: NorCal via Lansing, MI
 
camerapapi wrote:
All JPEG images are either processed in camera or processed by us.


Correction: "All JPEG images are either processed in camera or processed by us." should be "All JPEG images are either processed in camera or processed out-of-camera by us."

You are the one who chooses the in-camera jpg settings so you are editing/processing in-camera. And creating a tiff seems to be a part of the work-flow that is not required if you keep the RAW. If you are one of those who fear that the ability to read the RAW file may be "lost" some day then save it as a "DNG". {When the day comes where the ability to read a DNG is "lost" I doubt that we will be able to read any file format as we will be back living in caves. Can anyone name a file format that can no longer be "read"?}

Reply
May 18, 2017 12:38:57   #
rehess Loc: South Bend, Indiana, USA
 
cdayton wrote:
Unless I missed it, no one mentioned that both images look very "overprocessed " and unnatural. That bothers me more than if under extreme magnification there is additional noise in the jpeg, which there is.
I tend not to comment on that, because while I prefer a natural look, preference for images which "pop" more seems to be a matter of artistic taste.

Reply
 
 
May 18, 2017 14:25:00   #
TheDman Loc: USA
 
Tracy B. wrote:
Thank you TheDman. That was a great comparison. I personally only shoot Raw. I did my own test once. I shot raw + jpeg. Now I only shoot Raw.


Thanks everyone. If course you're all right that this subject has been discussed to death, but i don't recall seeing many threads that actually put both formats to the test, and I wanted to give that a try. And after my adventures in raw+jpg yesterday, suffice to say the jpg format has been permanently disabled on my camera.

Reply
May 18, 2017 14:26:22   #
TheDman Loc: USA
 
cdayton wrote:
Unless I missed it, no one mentioned that both images look very "overprocessed " and unnatural. That bothers me more than if under extreme magnification there is additional noise in the jpeg, which there is.


That's rather odd, as there wasn't any processing done to these that couldn't have been done with a 3-stop gnd.

Reply
May 18, 2017 20:19:21   #
blackest Loc: Ireland
 
TheDman wrote:
That's rather odd, as there wasn't any processing done to these that couldn't have been done with a 3-stop gnd.


nice example, kind of a shame it got moved here, where it will hardly be seen and the same old argument will be played out over and over again in main discussion...

Reply
Jun 11, 2017 01:39:29   #
WILLARD98407 Loc: TACOMA, WA.
 
[quote=speters]Why would that matter in any way, I don't really understand that question, to

I may just be strange ( ask my wife), but I am not one who shoots raw with the final goal of a jpeg image. at least 90% of my work is done raw with the final goal of a tiff. I can always dupe the tiff for a jpeg if needed for something like email. I have had clients that want a jpeg sooc, but I always record raw to the other card in camera for additional work that may be desirable. shooting in tiff is not workable, as the files run 105 to 116 mb per.
as for backup, I use two 5tb hard drives that are dupes. have about 1tb space left on each, then I'll have to have another set.

Reply
 
 
Jun 11, 2017 01:43:47   #
WILLARD98407 Loc: TACOMA, WA.
 
TheDman wrote:
In one of the many previous threads on this topic, DeanS got to the crux of the matter with this question: how many shooters could distinguish between a well shot/well pp'd jpeg and a similar RAW photo? So I've decided to take a completely impartial whack at it: let's export a raw file at default settings as a jpg, process the raw file in Camera Raw, then make the exact same ACR edits to the jpg and see what we get.

Since I believe the biggest benefit of shooting raw is the ability to recover shadow and highlight detail, I used an example covering a wide tonal range. Here is our starting point, a shot straight into the sun from Northern Ireland:

http://www.ddphotos.com/orig.jpg

That is exported straight out of ACR before any edits. This was a bracketed series, and I chose this one because the next brightest exposure blew out the sky. Ordinarily my pp method would blend several exposures on a shot like this for reasons we will see later, but for this example I will use just this one raw. I started out by double processing the raw into a shot optimized for the sky and one for the land:

http://www.ddphotos.com/raws.jpg

I then blended the two together with a simple gradient on a layer mask. I then opened the jpg I had exported earlier and double processed it using the exact same ACR settings, and blended it using the exact same gradient mask. Here are the results. Can you tell which is which?

http://www.ddphotos.com/comparison.jpg

Kind of tough at that reduced size, but a sharp eye could tell. Now let's zoom in:


http://www.ddphotos.com/comp1.jpg

http://www.ddphotos.com/comp2.jpg

http://www.ddphotos.com/comp3.jpg


See it now? The raw image has more shadow detail, smoother tonal gradations, no noise around the sun as opposed to the jpg which has artifacts, etc. The jpg actually held up better than I thought, but I still would be horrified to print this large. Now if I'm just shooting portraits or pictures of my cat, I'm not taxing the tonal range of my sensor and therefore jpgs should be quite fine. But for landscape work I want the best quality possible, so it's raw all the way. On that note, you can see that the foliage on the rock in the closest foreground is a bit out of focus and noisy, which is due to the shadow recovery and f13 not quite reaching it. In my actual process of this scene I used a separate shot for the front foliage, exposed solely for the foliage and focused precisely on it:

http://www.ddphotos.com/foliage.jpg

So there you have it, an actual test. Read into it what you will.
In one of the many previous threads on this topic,... (show quote)


thanks,Dman. confirms what I've been telling myself since the day I started digital and gave up my

80,000$ darkroom. Raw, Raw, Raw.

Reply
Jun 16, 2017 19:04:53   #
speters Loc: Grangeville/Idaho
 
[quote=WILLARD98407]
speters wrote:
Why would that matter in any way, I don't really understand that question, to

I may just be strange ( ask my wife), but I am not one who shoots raw with the final goal of a jpeg image. at least 90% of my work is done raw with the final goal of a tiff. I can always dupe the tiff for a jpeg if needed for something like email. I have had clients that want a jpeg sooc, but I always record raw to the other card in camera for additional work that may be desirable. shooting in tiff is not workable, as the files run 105 to 116 mb per.
as for backup, I use two 5tb hard drives that are dupes. have about 1tb space left on each, then I'll have to have another set.
Why would that matter in any way, I don't really u... (show quote)

OK, than your final product is a TIFF, a lot of other folks final piece is a Jpeg - and as those, they should be far superior to the base Raw image!

Reply
Jun 16, 2017 20:52:13   #
WILLARD98407 Loc: TACOMA, WA.
 
speters wrote:
OK, than your final product is a TIFF, a lot of other folks final piece is a Jpeg - and as those, they should be far superior to the base Raw image!


not actually superior to the raw. the raw has all the unfiltered information that was available to the sensor (that it is capable of recording).

a jpeg is a file that has the filtering done to the image inside the camera, between the time the sensor "sees" the image and it is recorded to memory, according to what instructions you gave to the camera-I.E.: white balance, iso, contrast, sharpening, saturation, exposure compensation, etc.
then the stored image is compressed to save space.

a tiff is like a jpeg in that it also takes directions about what to do to the image before it is stored to memory. the difference is that the tiff is not compressed to save space, therefore it will be a bigger file. if I shoot Large, fine jpegs in my camera they run about 26 mb. tiffs taken in-camera run about 106- 112 mb.

you can open a tiff as many times as you want with no degradation of the image because there is no "re-compressing" going on.

with a jpeg, after it is opened a number of times (10, maybe), a loss of detail can be noticed, especially in a fairly clear area like a sky. imntbho, this is not a good thing.

if you shoot raw you can open it and tweek it and then make a tiff copy, then close the raw without the chancges so that you can have another go at it with different ideas in mind.

hope this helps

Reply
Sep 27, 2017 19:58:31   #
papa Loc: Rio Dell, CA
 
TheDman wrote:
In one of the many previous threads on this topic, DeanS got to the crux of the matter with this question: how many shooters could distinguish between a well shot/well pp'd jpeg and a similar RAW photo? So I've decided to take a completely impartial whack at it: let's export a raw file at default settings as a jpg, process the raw file in Camera Raw, then make the exact same ACR edits to the jpg and see what we get.

Since I believe the biggest benefit of shooting raw is the ability to recover shadow and highlight detail, I used an example covering a wide tonal range. Here is our starting point, a shot straight into the sun from Northern Ireland:

http://www.ddphotos.com/orig.jpg

That is exported straight out of ACR before any edits. This was a bracketed series, and I chose this one because the next brightest exposure blew out the sky. Ordinarily my pp method would blend several exposures on a shot like this for reasons we will see later, but for this example I will use just this one raw. I started out by double processing the raw into a shot optimized for the sky and one for the land:

http://www.ddphotos.com/raws.jpg

I then blended the two together with a simple gradient on a layer mask. I then opened the jpg I had exported earlier and double processed it using the exact same ACR settings, and blended it using the exact same gradient mask. Here are the results. Can you tell which is which?

http://www.ddphotos.com/comparison.jpg

Kind of tough at that reduced size, but a sharp eye could tell. Now let's zoom in:


http://www.ddphotos.com/comp1.jpg

http://www.ddphotos.com/comp2.jpg

http://www.ddphotos.com/comp3.jpg


See it now? The raw image has more shadow detail, smoother tonal gradations, no noise around the sun as opposed to the jpg which has artifacts, etc. The jpg actually held up better than I thought, but I still would be horrified to print this large. Now if I'm just shooting portraits or pictures of my cat, I'm not taxing the tonal range of my sensor and therefore jpgs should be quite fine. But for landscape work I want the best quality possible, so it's raw all the way. On that note, you can see that the foliage on the rock in the closest foreground is a bit out of focus and noisy, which is due to the shadow recovery and f13 not quite reaching it. In my actual process of this scene I used a separate shot for the front foliage, exposed solely for the foliage and focused precisely on it:

http://www.ddphotos.com/foliage.jpg

So there you have it, an actual test. Read into it what you will.
In one of the many previous threads on this topic,... (show quote)


That's an unusual way to do HDR; but you've convinced me of this, that there's always another way around the mountain. I think I'll try it next time, just for fun. Thanks.

Reply
Page <<first <prev 4 of 7 next> last>>
If you want to reply, then register here. Registration is free and your account is created instantly, so you can post right away.
Photo Analysis
UglyHedgehog.com - Forum
Copyright 2011-2024 Ugly Hedgehog, Inc.