A house shot in JPEG. I had forgotten how much you can do in RAW

[JBW]

Brian,

Despite the lack of response on my diagram, it gives answers on many questions here.

I'll try to explain once more. First of all focus on the RGB raster image in that diagram. That's the file you see on your monitor, that's the file you edit. If you do so, than you can ask yourself: where does that file come from?

Take your example of shooting jpg and raw at the same time and ask yourself that question.
That jpg is a diskfile created from an in-camera converted RGB rasterfile. It's saved to disk,memory card, with a 8 bit tonal depth and a compression.
The raw file is a file with the sensor info. You need a converter to create a RGB raster file.
If the converter and the settings are equal as in the camera, then the result will be equal. Otherwise you will have 2 different images. And don't forget that viewing a raw file in a viewer will show you the embedded jpg, not the raw-file.

So coming to the title of this post, your question. As long the converter is not falling back on the raw-data, there's no difference. It depends on the converter and/or editor. I use CaptureNx2, Nikon. I once played with jpg in that program. WB and exposure corrections where disabled. And that exposure correction isn't even an exposure correction.
The main difference is: where does that RGB raster file comes from? Is it from the converter, than you start with a 12,14,16 bit tonal depth RGB raster file. If it comes from a jpg, than your RGB raster file is extracted from a 8 bit compressed file, if it comes from a 16 bit tiff, well then you've nearly the same as from a raw file.

Another question was: do you need a raw file. Manfred answered that.

What converter do you use? You started with Capture One, than Capture One Sony and now Capture One Pro Sony.

George

My converter, as you call it is Capture 1. Capture 1 Sony Express (the free version) has the same engine as Capture 1 Sony Pro but less bells and whistles. Capture 1 Sony Pro has the same engine as Capture 1 but only works with Sony Cameras. Or so I'm told.

[Stagecoach]

Seems to me that everything considered shooting in RAW or JPEG depends upon personal taste. As in how much do I want to manipulate the shot.

Yes you can look at it that way Brian.

To my way of thinking my computer has a lot more power than my camera and Cap.1 has a lot more power than the program in my camera. Seems reasonable to me to get the best shot I can in the camera and then have fun in Cap.1 and RAW.

Exactly. And in 99.99999999% of the shots you take (based on what we see) there is absolutely no reason whatsoever not to shoot in RAW + Jpeg fine.

And if you are worried about storage space delete unwanted RAW after your selection process.

[george013]

My converter, as you call it is Capture 1. Capture 1 Sony Express (the free version) has the same engine as Capture 1 Sony Pro but less bells and whistles. Capture 1 Sony Pro has the same engine as Capture 1 but only works with Sony Cameras. Or so I'm told.

I hope the link is good.
I still don't know which you have: the pro or the express.

Did you understand what I wrote about that diagram?

George

[JBW]

Yes you can look at it that way Brian.



Exactly. And in 99.99999999% of the shots you take (based on what we see) there is absolutely no reason whatsoever not to shoot in RAW + Jpeg fine.

And if you are worried about storage space delete unwanted RAW after your selection process.

There is one small dolwnside to shooting in J+R... I'll be deleting twice as many not so good shots I do delete the unwanted RAW.

[george013]

Never delete the raw. If your converter gives the same result as the in-camera produced jpg, keep the raw. You can extract a jpg, tiff or what else always from it.

@Tony(cognito),
I only shoot raw. The reason is I want to have the possibility to fall back on an original. My editing skill are changing. With a raw I can start over any time without being afraid to lose the original.
And there's another reason, based on the old nef file structure and CaptureNx2. Whithin the nef there's an embedded jpg file with the dimensions of the sensor, 100%. The first one is low quality. After editing and saving that file is replaced with a high quality jpg. Even when I don't need to do some editing, I will do some to force that higher quality jpg being embedded. Viewing with Iview I can resize the image and export it as a single jpg.
In 1 file the raw, the jpg and even multiple versions of editing.
I even can see on the thumbnails in IView which are edited and which not.

George

[JBW]

I hope the link is good.
I still don't know which you have: the pro or the express.

Did you understand what I wrote about that diagram?

George

Pro. Which has the same engine just more bells and whistles. I do have trouble with the science of photography.

[Mike Buckley]

Completely agreed with Grahame in his discussion with you about the color and monochrome images.

[Mike Buckley]

based on the old nef file structure and CaptureNx2. Whithin the nef there's an embedded jpg file with the dimensions of the sensor, 100%. The first one is low quality. After editing and saving that file is replaced with a high quality jpg.

I'll only write one post about this because I don't want Brian's thread to become hijacked, but I'm reasonably certain that is true only when using a Nikon camera model that is relatively old. The older models do not embed JPEGs that are of the highest quality, whereas the later models do. I just now completed a test about that proving that there is no difference between the JPEG embedded by my camera, which was discontinued years ago, and the JPEG embedded by Capture NX2 after editing the image. So, the camera model that does not embed a high-quality JPEG must be even older than mine.

[JBW]

I'll only write one post about this because I don't want Brian's thread to become hijacked, but I'm reasonably certain that is true only when using a Nikon camera model that is relatively old. The older models do not embed JPEGs that are of the highest quality, whereas the later models do. I just now completed a test about that proving that there is no difference between the JPEG embedded by my camera, which was discontinued years ago, and the JPEG embedded by Capture NX2 after editing the image. So, the camera model that does not embed a high-quality JPEG must be even older than mine.

Too late to be concerned. This thread has taken on a life of it's own. And I'm hoping others than myself are learning new things.

[JBW]

Completely agreed with Grahame in his discussion with you about the color and monochrome images.

Not arguing with either of you. Just wondering what advantage there is to shooting in both RAW and Fine JPEG?

[Manfred M]

Not arguing with either of you. Just wondering what advantage there is to shooting in both RAW and Fine JPEG?

Speed is the main advantage. If I shoot raw + jpeg, I can prepare an image for posting in just a few seconds whereas when I shoot raw, I have all the raw conversion work to do, lens correction, colour space, white balance, sharpening. These are things that my camera has already done for me. Often I can just post with little or no work (usually straightening and / or cropping).

This is why so many wedding photographers and event photographers shoot jpeg rather than raw; speed to have a good final product.

Don't forget that the camera manufacturers have spent a lot of effort ensuring that the camera creates high quality jpegs.


If I print, then I will use raw 100% of the time.

[Mike Buckley]

Not arguing with either of you. Just wondering what advantage there is to shooting in both RAW and Fine JPEG?

My thinking is that there are only two advantages to creating the JPEG when you are also creating raw files. The first advantage has to do with workflow, such as just one example when wanting to quickly display the images on the Internet while away from the computer setup in your home and finding JPEGs easier and faster to work with in that situation. The other advantage has to do with educational value; if you see that your camera settings are automatically creating a JPEG that has characteristics that are better than the characteristics of the images that you manually adjust using your raw file, you can strive to improve your post-processing capabilities. If neither of those reasons apply, I see no benefit to shooting both file formats in all but the very rarest situation.

[John 2]

There is a further advantage if you are forced to use a computer not equipped with your preferred PP App. Windows and basic Android won't handle anything other than the jpg format. When we are on our travels and relying our Sony Tablet for instance, jpg + raw allows me to review and discard as we go along and still have the raw to work on when we touch base again. I know you can do that in camera but a 10" screen makes life a bit easier.

[Cogito]

For the record, my posted comparison image had the EXIF deliberately stripped, so how you saw any EXIF in my post is beyond me . . .

Not your image in this thread, but your posts in general

[xpatUSA]

I'll only write one post about this because I don't want Brian's thread to become hijacked, but I'm reasonably certain that is true only when using a Nikon camera model that is relatively old. The older models do not embed JPEGs that are of the highest quality, whereas the later models do. I just now completed a test about that proving that there is no difference between the JPEG embedded by my camera, which was discontinued years ago, and the JPEG embedded by Capture NX2 after editing the image. So, the camera model that does not embed a high-quality JPEG must be even older than mine.

A lot of talk here about JPEG "quality" but lacking in numbers. So, I'm curious as what "highest quality" means for today's cameras, see:

https://en.wikipedia.org/wiki/Chroma...nd_subsampling

EXIF readers such as ExifToolGUI tell us the JPEG quality in the form of the above-referenced code.

The highest quality is 4:4:4 (1x1). If it says 4:2:2 (2x1) that's next best. And 4:2:0 (2x2) is least best.

The amount of compression is the other number of interest. Normally expressed as a percentage where more is less (this is photography after all), I prefer the ratio format but I'm not sure if it shows in EXIF but the app JPEGsnoop shows it for sure.

A less precise measure is bits per pixel (not bit depth). You converter the file size from Kb or Mb to bits and divide that by the image size in pixels. Anything over 6 should keep most folks happy.

[george013]

I'll only write one post about this because I don't want Brian's thread to become hijacked, but I'm reasonably certain that is true only when using a Nikon camera model that is relatively old. The older models do not embed JPEGs that are of the highest quality, whereas the later models do. I just now completed a test about that proving that there is no difference between the JPEG embedded by my camera, which was discontinued years ago, and the JPEG embedded by Capture NX2 after editing the image. So, the camera model that does not embed a high-quality JPEG must be even older than mine.

I've the D700. Compare the file sizes between the nef just out of the camera and after editing a simple thing.
I don't know what and how you tested. It would be appreciated telling what and how you did. And of course using CaptureNx2.

I only explained why I use pp in CaptureNx2, even when not necessary.

George

[george013]

A bit pissed about Mike's post saying he did do some tests. He posed some assumptions in it without clarifying what he meant. A wrong way of discussing.
So here some tests of mine with the explanation.

I use 1 nef file and explore that before and after it had been opened in CaptureNx. I opened it in CaptureNx did something and restored it again, just to set a flag inside the program, to force a save action.
The camera used is a D700 with 4256x2835 pixels.

And some explanation for Brian and others.

USING IVIEW
First using IView.
When opening a raw file in IView, it loads its embedded jpg in memory. Brian, going back to that diagram, that diskfile jpg is decompressed and placed in memory as a RGB raster image. All editing is done on that image. I can do some editing with IView, I only use the crop and resize tools. Than saving the picture as a jpg.
When a RGB raster file is saved as a jpg, it's compressed at that moment. The disksize is mentioned.
DSC_3614-d700
The embedded jpg straight from memory card and exported as a jpg. 3341 kb



Second I opened the nef in CaptureNx and forced a save. The low quality jpg is now replaced by a high quality jpg. Doing the same as before you see the disksize is over 2x as big
DSC_3614-nx
The embedded jpg in a nef after been in CaptureNx and exported by IView. 6818kb


USING EXIFTOOL
For every time I load the embedded jpg in memory and save it and compressing it again, the figures are not reliable. I can extract the jpg directly out of the raw with Exiftools. A program many viewers and tools are using.
exif-d700
The embedded jpg in a nef directly from memory card and extracted using Exiftools. 1436kb


And after being in CaptureNx. See the differences in size.
exif-nx
The embedded jpg in a nef after been in CaptureNx and extracted by Exiftools. 7004kb


FROM NEF TO JPG
Exported as jpg from CaptureNx quality 100, 7067kb



As an extra I used CaptureNx to save the image, once with quality 85 and once with quality 100
DSC_3614-converter85
Exported as jpg from capturenx quality 85, 3132kb



The size of the embedded jpg looks like quality 100.

Brian, do you see differences in these images? They're all 100%, no resizing. I think when you start editing you will see differences.

And some more for Mike, after been in CaptureNx the nef increased with 5614 kb, the embedded jpg with 5568 kb.

Maybe Ted can figure out what compressions where used?

George

[xpatUSA]

A lot of talk here about JPEG "quality" but lacking in numbers. So, I'm curious as what "highest quality" means for today's cameras, see:

https://en.wikipedia.org/wiki/Chroma...nd_subsampling

EXIF readers such as ExifToolGUI tell us the JPEG quality in the form of the above-referenced code.

The highest quality is 4:4:4 (1x1). If it says 4:2:2 (2x1) that's next best. And 4:2:0 (2x2) is least best.

The amount of compression is the other number of interest. Normally expressed as a percentage where more is less (this is photography after all), I prefer the ratio format but I'm not sure if it shows in EXIF but the app JPEGsnoop shows it for sure.

A less precise measure is bits per pixel (not bit depth). You converter the file size from Kb or Mb to bits and divide that by the image size in pixels. Anything over 6 should keep most folks happy.

Here's what Panasonic thinks is "high" quality out of their GH1 model:

Y'CbCr 4:2:2 (2x1)
Compression Ratio: 9.45:1
Bits per pixel: 2.54:1

Not bad but not the "highest quality" possible. One wonders what Canikons dish out these days? Anybody?

[Mike Buckley]

A bit pissed about Mike's post saying he did do some tests. He posed some assumptions in it without clarifying what he meant. A wrong way of discussing.

If I had been inclined to provide more information, the tone of those comments was a complete disincentive.

[Black Pearl]

Not arguing with either of you. Just wondering what advantage there is to shooting in both RAW and Fine JPEG?

I find it handy to have an instant jpeg file available to Wi-Fi over to my iPhone when I'm out and about and I sometimes like to have a bit of fun tweaking the film sims to produce in instant image that I will deliberately not fiddle with - but - should I want to and feel the extra data available in the raw file will be useful its always good to have both.