This is why no RAW on the G7

[howiesmith]

According to Canon USA, so is the 1Ds MkII.

You'll find it in "Consumer Products Home :: Camera :: EOS (SLR) Camera Systems".

They're also calling their SLRs "advanced professional looking".

However, they do call their own software "professional level".

As for the G7, they say:

"Canon's high-end PowerShot digital cameras incorporate the creative performance of a professional digital SLR camera and the compact convenience of a point-and-shoot."
[a href=\"index.php?act=findpost&pid=90139\"][{POST_SNAPBACK}][/a]

You win.

[Ray]

I can get an additional stop of DR from the highlights in ACR over the camera JPEGs. And the ability to apply my own tone curve in ACR is much preferable to applying a comparable curve on the JPEG; less posterization/banding and color shifts. And of course NR is more effective when one can use the tools on 16-bit data that hasn't been mucked up with JPEG artifacts and 8-bit conversion, and one can adjust the noise reduction settings to match the specific image.

The magnitude of difference I see between Olympus camera JPEGs and RAW is comparable the magnitude of difference between Canon DSLR RAWs and JPEGs, but with smaller format cameras, taking every possible technical advantage is more important, because smaller-format camera files have more technical shortcomings than those from DSLRs; higher noise, less DR, etc. You have to work the digicam files harder to get a good result (technically) than the DSLR files.
[a href=\"index.php?act=findpost&pid=89752\"][{POST_SNAPBACK}][/a]

Jonathan,
I just poked my 5D through the hotel window here in Bangkok and took several shots of a cityscape against a bright sky with a few clouds, shooting RAW+jpeg at 1/3rd stop intervals with contrast, color saturation and sharpness set at a minimum.

Comparing the jpegs with the RAW images converted in ACR with up to -2 EC and brightness, contrast and shadows all at zero, it seems I've possibly underestimated the 2/3rds of a stop DR advantage of RAW. It might be closer to one stop, but not quite as great as one stop. The jpeg with one stop less exposure shows marginally greater detail in the brightest cloud.

Your points about slight image degradation due to jpeg compression prior to upsampling to 16 bit for processing purposes, are of course quite valid, as are the other advantages of RAW you mention. I'm not trying to say such advantages do not exist, just that they are probably not worth the effort. In your own situation with the Olympus SP-350, if you are truly getting one whole stop of additional DR from RAW then that alone might make using RAW worthwhile with that camera.

But a 10 second delay between shots doesn't make sense to me.

That might be okay for Howie who might take a whole afternoon to set up a shot, but I can't understand why someone whose output is as prolific as yours would bother with such a handicap. I'd rather carry a camera that's just a bit heavier and more bulky and I think that companies like Canon or Olympus would better serve the public with an ultra compact APS-C format camera with the usual RAW option but no mirror and no optical viewfinder. I'd rather sacrifice a through-the-lens optical viewfinder for the sake of less bulk and lighter weight, than feel compelled to work extra hard to extract the last iota of image quality from a tiny sensor.

[Jonathan Wienke]

Jonathan,
I just poked my 5D through the hotel window here in Bangkok and took several shots of a cityscape against a bright sky with a few clouds, shooting RAW+jpeg at 1/3rd stop intervals with contrast, color saturation and sharpness set at a minimum.

Comparing the jpegs with the RAW images converted in ACR with up to -2 EC and brightness, contrast and shadows all at zero, it seems I've possibly underestimated the 2/3rds of a stop DR advantage of RAW. It might be closer to one stop, but not quite as great as one stop. The jpeg with one stop less exposure shows marginally greater detail in the brightest cloud.

Your points about slight image degradation due to jpeg compression prior to upsampling to 16 bit for processing purposes, are of course quite valid, as are the other advantages of RAW you mention. I'm not trying to say such advantages do not exist, just that they are probably not worth the effort. In your own situation with the Olympus SP-350, if you are truly getting one whole stop of additional DR from RAW then that alone might make using RAW worthwhile with that camera.

But a 10 second delay between shots doesn't make sense to me.

Let's say I'm not real excited about it, but I still managed to capture some good stuff while in Iraq. It's not that much worse than a film camera with manual frame advance, like my old Argus C-4 rangefinder. It's a limitation I'm willing to accept when I'm already carrying 60 or more pounds of gear, and weight and size of additional crap to carry means DSLRs are not feasible, regardless of their many advantages.

Per your request a few posts ago, I did some comparison shots between JPEGs and RAWs with my Olympus SP-350. The RAWs exhibit about 2/3 of a stop greater highlight latitude than the JPEGs. I've attached 2 samples, one RAW, and one JPEG. The JPEG is as-is from the camera except for resizing, and the RAW has had a click white balance and a small curve tweak in ACR to get the black point to a sensible level. I've done no noise reduction or other processing on either image. As you can see, the camera got the white balance wrong (I had it on Auto with fluorescent lighting) but even taking that into account, the colors are still not as accurate as the RAW. Contrast is also lower; the camera was set to its default contrast setting of 0, and raising it would reduce JPEG highlight latitude even further. I left the EXIF data intact, so you can poke through it with Bridge or whatever to your heart's content. Enjoy!

[bjanes]

Per your request a few posts ago, I did some comparison shots between JPEGs and RAWs with my Olympus SP-350. The RAWs exhibit about 2/3 of a stop greater highlight latitude than the JPEGs. I've attached 2 samples, one RAW, and one JPEG. The JPEG is as-is from the camera except for resizing, and the RAW has had a click white balance and a small curve tweak in ACR to get the black point to a sensible level. I've done no noise reduction or other processing on either image. As you can see, the camera got the white balance wrong (I had it on Auto with fluorescent lighting) but even taking that into account, the colors are still not as accurate as the RAW. Contrast is also lower; the camera was set to its default contrast setting of 0, and raising it would reduce JPEG highlight latitude even further. I left the EXIF data intact, so you can poke through it with Bridge or whatever to your heart's content. Enjoy!
[a href=\"index.php?act=findpost&pid=90238\"][{POST_SNAPBACK}][/a]

Contrary to popular belief, one can adjust WB in JPEG captures. Here is Jonathan's JPEG picture edited in Adobe Lightroom with the WB adjusted on one of the neutral squares and the black point reset to a more suitable value. The JPEG is on the left and the raw capture on the right. The edit took about 10 seconds.

[attachment=1354:attachment][attachment=1355:attachment]

However, it is better to make these edits from the raw file, because you have more bits to work with and will not lose as much information in the edit.

Bill

[Jonathan Wienke]

However, it is better to make these edits from the raw file, because you have more bits to work with and will not lose as much information in the edit.

And fixing the JPEG white balance still leaves a greater degree of color inaccuracy than the RAW from the calibrated converter. Does Canon's software have a tool similar to Lightroom's JPEG WB repair?

[howiesmith]

Does Canon's software have a tool similar to Lightroom's JPEG WB repair?

[a href=\"index.php?act=findpost&pid=90289\"][{POST_SNAPBACK}][/a]

I have no idea.  But perhaps the important things to remember here are:

1.  Does the usual target market for the G7 have Lightroom?  Or even want to have it if they bought  G7?

2.  Does the usual target audience care that much for superaccurate about white balance?

3.  Is Canon trying to sell Lightroom's white balance or cameras with their software?

4.  Canon is a publiclly held company with no requirement to meet the needs of every potential camera customer, but a very real requirement to make a fair profit and return on investment for their shareholders.

From Canon's website:

"Canon's acclaimed G-Series has a new top-of-the-line model, and serious photography enthusiasts have a lot to be excited about. A matte black, retro-hip design houses impressive capabilities and top core specifications, including 10.0 megapixel resolution and a powerful 6x optical zoom. Loaded with Canon's latest and most advanced technologies, including an Optical Image Stabilizer for clear shots throughout the zoom range, PowerShot G7 is a camera to be reckoned with. "

Note the camrea is aimed at "serious photography enthusiasts," not professionals.  Also note the camera is "[l]oaded with Canon's latest and most advanced technologies."  It is not loaded with Abobe stuff nor is it marketed to require more than Canon's software.

[aaykay]

Look at the below link that compares an in-camera JPEG to a RAW processed JPEG - both from a 10MP 1/1.8" CCD:

Eyeopening, from a detail extraction perspective:

http://forums.dpreview.com/forums/read.asp...essage=21079381

This leads me to believe that even a perfect in-camera JPEG, with all the technical settings set perfectly, will still suffer from factors like noise reduction that are beyond the control of the photographer.

[howiesmith]

Look at the below link that compares an in-camera JPEG to a RAW processed JPEG - both from a 10MP 1/1.8" CCD:

Eyeopening, from a detail extraction perspective:

http://forums.dpreview.com/forums/read.asp...essage=21079381

This leads me to believe that even a perfect in-camera JPEG, with all the technical settings set perfectly, will still suffer from factors like noise reduction that are beyond the control of the photographer.
[a href=\"index.php?act=findpost&pid=90307\"][{POST_SNAPBACK}][/a]

I have some trouble with the examples.  They are shown at 100%.  That is a huge print.  Zoom out a couple clicks and the differences aren't as obvious, to me anyway.  Why not make the comarison at a more reasonable size and viewing distance.  Maybe compare real prints instead of computer enlargements.

One would expect RAW to be better than JPEG.  But maybe the comparisons should be more realistic.

[aaykay]

Something to ponder: Michael apparently chose the "No RAW, but well build" option over the "6sec RAW, but Fuji build" option. No doubt there are other significant factors, but they may come down to the same conundrum: "No RAW + advantage" over "6sec RAW + disadvantage"...

Nothing to ponder here.  Michael chose the G7, ***inspite*** of the lack of RAW, rather than ***because*** of the lack of RAW.  He chose it because of other factors in the camera and from his perspective, the camera would have been a lot better, if the RAW option was avaiable (whether 6 secs or not).

"6 sec RAW + disadvantage" etc are interesting asides to a camera that is otherwise uninteresting.  In that camera, even the presence of RAW was not a compelling persuasion, due to its other lacks.

We are talking of 2 completely different things here.

[Jonathan Wienke]

Note the camrea is aimed at "serious photography enthusiasts," not professionals.  Also note the camera is "[l]oaded with Canon's latest and most advanced technologies."  It is not loaded with Abobe stuff nor is it marketed to require more than Canon's software.

That's highly schizophrenic on Canon's part (and a bit disingenuous on yours); "serious photography enthusiasts" (as opposed to casual snapshooters) are exactly the kind of photographers who appreciate the benefits of RAW. Canon is saying that the G7 is designed for photographers who are beyond the casual snapshooter stage, but then treats those "serious photography enthusiasts" like a bunch of clueless newbies by refusing to offer the one feature that advanced-level photographers use most to control of their image processing: RAW. It makes no sense at all; if Canon won't add RAW support, then they should market the camera to the clueless newbies who don't know RAW from a hole in their butt, not the "serious photography enthusiasts" who are much more likely than the clueless newbies to know what RAW is and really want it.

[bjanes]

And fixing the JPEG white balance still leaves a greater degree of color inaccuracy than the RAW from the calibrated converter. Does Canon's software have a tool similar to Lightroom's JPEG WB repair?
[{POST_SNAPBACK}][/a]

The key to WB is to be working in a linear (gamma 1.0) space. The internal working space of Adobe Camera Raw and Adobe Lightroom is gamma 1.0 with the chromaticities of ProPhotoRGB. Once you are in a gamma 1.0 space, WB merely involves multiplying the blue and red channels by a constant. For example, the table whose link is listed below gives the coefficients for various Nikon cameras:

[a href=\"http://www.pochtar.com/NikonWhiteBalanceCoeffs.htm]http://www.pochtar.com/NikonWhiteBalanceCoeffs.htm[/url]

If you are working in a gamma 2.2 space, the conversion is nonlinear. and more complicated. Lightroom is the only program I know of that has the ability to WB from a gamma 2.2 JPEG, but I have not worked with Canon software.

For example, if you are using a Nikon D2X and photographing at 2700K, the blue multiplier is nearly 4.0. This means that the blue channel full scale would be 0..64 rather than 0..255 and you would lose a lot of tones. With 12 bit raw, you would have 0..4095 tones in the green channel and 0..1024 tones in the blue channel

[aaykay]

I have some trouble with the examples.  They are shown at 100%.  That is a huge print.  Zoom out a couple clicks and the differences aren't as obvious, to me anyway.  Why not make the comarison at a more reasonable size and viewing distance.  Maybe compare real prints instead of computer enlargements.

One would expect RAW to be better than JPEG.  But maybe the comparisons should be more realistic.
[a href=\"index.php?act=findpost&pid=90312\"][{POST_SNAPBACK}][/a]

The point in the whole post was not whether the differences between RAW and JPEG are observable at 100% or not.  It was that in a camera processed JPEG, critical, fine detail is **irretrievably** lost from within the picture, even if you can salvage things like white balance etc from a jpeg in post processing.  The difference between a watercolor painting and a fine photograph.

What if you really want to print big ?  Sorry, the detail does not exist. Gone ! Poof !  ***Even though*** the shot itself was taken perfectly, with the exposure nailed, the ISO tailored to the situation and the white balance perfectly done.  Till the end of eternity, even if printer quality improves in the future, you simply cannot recreate the original details, as when you took the photograph, since it was irreversibly destroyed by the camera's processing engine.

Bottomline, RAW as an ***option*** can be turned off, if you want to shoot JPEG exclusively.  Many people do so.  But not providing that ***option*** is what we are so riled up about.

[Jonathan Wienke]

I have some trouble with the examples.  They are shown at 100%.  That is a huge print.  Zoom out a couple clicks and the differences aren't as obvious, to me anyway.  Why not make the comarison at a more reasonable size and viewing distance.  Maybe compare real prints instead of computer enlargements.

One would expect RAW to be better than JPEG.  But maybe the comparisons should be more realistic.

Given that well made prints can show much more detail per area unit than any monitor, the comparison is not meaningless. Perhaps the difference wouldn't be visible in a 4x6 print, but it would certainly be visible in anything 8x10 or larger. And the larger the print, the more obvious the difference will be.

[howiesmith]

That's highly schizophrenic on Canon's part (and a bit disingenuous on yours); "serious photography enthusiasts" (as opposed to casual snapshooters) are exactly the kind of photographers who appreciate the benefits of RAW. Canon is saying that the G7 is designed for photographers who are beyond the casual snapshooter stage, but then treats those "serious photography enthusiasts" like a bunch of clueless newbies by refusing to offer the one feature that advanced-level photographers use most to control of their image processing: RAW. It makes no sense at all; if Canon won't add RAW support, then they should market the camera to the clueless newbies who don't know RAW from a hole in their butt, not the "serious photography enthusiasts" who are much more likely than the clueless newbies to know what RAW is and really want it.
[a href=\"index.php?act=findpost&pid=90315\"][{POST_SNAPBACK}][/a]

Jonathan, perhaps we could end this winless back and forth if I say maybe Canon screwed up, and you could say maybe Canon is not a bunch of arrogant fools and know Canon's (their) business better than you do.

Parting shot though.  Canon does have access to a lot of insider information that maybe you know nothing of.  Maybe?

[bjanes]

The point in the whole post was not whether the differences between RAW and JPEG are observable at 100% or not.  It was that in a camera processed JPEG, critical, fine detail is **irretrievably** lost from within the picture, even if you can salvage things like white balance etc from a jpeg in post processing.  The difference between a watercolor painting and a fine photograph.

What if you really want to print big ?  Sorry, the detail does not exist. Gone ! Poof !  ***Even though*** the shot itself was taken perfectly, with the exposure nailed, the ISO tailored to the situation and the white balance perfectly done.  Till the end of eternity, even if printer quality improves in the future, you simply cannot recreate the original details, as when you took the photograph, since it was irreversibly destroyed by the camera's processing engine.

Bottomline, RAW as an ***option*** can be turned off, if you want to shoot JPEG exclusively.  Many people do so.  But not providing that ***option*** is what we are so riled up about.
[{POST_SNAPBACK}][/a]

High quality JPEGS have detail that is virtually indistinguishable from non-compressed files, and I doubt if you would see much difference in a large print. When uploading digital files for large scale digital prints, many people use JPEG with excellent results to save on upload time.

[a href=\"http://www.dpreview.com/learn/?/key=jpeg]http://www.dpreview.com/learn/?/key=jpeg[/url]

[aaykay]

High quality JPEGS have detail that is virtually indistinguishable from non-compressed files, and I doubt if you would see much difference in a large print.

Just retyping what was shown in the above post (JPEG from the camera vs JPEG processed from RaW) - both from a 10MP 1/1.8" sensor:

http://forums.dpreview.com/forums/read.asp...essage=21079381

[bjanes]

Just retyping what was shown in the above post (JPEG from the camera vs JPEG processed from RaW) - both from a 10MP 1/1.8" sensor:

http://forums.dpreview.com/forums/read.asp...essage=21079381
[a href=\"index.php?act=findpost&pid=90323\"][{POST_SNAPBACK}][/a]

One anecdotal report from one camera under unknown conditions proves nothing at all. Why don't you look at Thom Hogan's report of JPEG with the Nikon D50, which was mentioned earlier in this thread or do a few comparisons yourself?

[aaykay]

One anecdotal report from one camera under unknown conditions proves nothing at all. Why don't you look at Thom Hogan's report of JPEG with the Nikon D50, which was mentioned earlier in this thread or do a few comparisons yourself?
[a href=\"index.php?act=findpost&pid=90324\"][{POST_SNAPBACK}][/a]

I will go through the article later today but from what I recall, Nikon uses 3 different RAW formats in their cameras.

Their higher-end cameras like the D2X (and probably the D200 too), uses the data that comes right off the sensor and prior to the processing engine of the camera having a go at it.

A watered down version of the "RAW" is present in cameras like the D80, where the data is NOT 12-bit but an emulation of it (or something along those lines).

An even further watered down version, which they also called "RAW", is present in cameras like the D40/D50 etc., which throws away even more information that the above D80 "RAW".

If the above is true, I would not be surprised to see very little difference between a high quality JPEG and this watered down version, termed by them as "RAW".

My definition of RAW is the data that came off the sensor, prior to any kind of processing from the camera - nothing more, nothing less.

The D40/D50 "RAW" is not a true RAW in my opinion, even though they may term it as such, for marketing convenience.

As far as the "anecdotal report" is concerned, you are right.  This is just one anecdotal instance.  But keep in mind, that these are small sensored cameras with a lot of MP stuffed into them.  The noise reduction systems (specifically at any ISO greater than the lowest ISO) in such cameras, would be ruthless in eliminating detail and noise and there is nothing anecdotal about this fact.  The attached link is highly representative of what a typical small sensored camera does, at 400 ISO, when it creates a JPEG....nothing anecdotal or surprising about it.

[BernardLanguillier]

I will go through the article later today but from what I recall, Nikon uses 3 different RAW formats in their cameras.

[a href=\"index.php?act=findpost&pid=90349\"][{POST_SNAPBACK}][/a]

Could you please provide a link to that article?

Thanks,
Bernard

[howiesmith]

I just read an article here on LL about RAW and JPG files and their differences.  Interesting reading.  Of note, it said in-camera RAW converters, except for Nikon and Kodak, range from poor to acceptable.  Could it be that Canon's in-camera RAW converter for the G7 was poor and that is why JPG files are nearly the same, and RAW deleted by Canon?

The article also likened RAW to the latent image on a piece of film.  Is converting RAW yourself like custom wet processing film yourself, and JPG more like sending your film to the corner drugstore to be processed?

IF that is the case, it seems to me that some folks might feel the G7 was lacking because they didn't get a latent image, but something more akin to a Polaroid print.  Some folks, on the other hand, are quite happy because they have save a trip to the drug store, and got a processed negative they can print or fiddle with if they want.  I have no data, but it seems to me many many more folks had their film processed by a third party (one dunk fits all) than got their hands wet.  Maybe another factor for Canon's decision.