[garth1948]

Agree, you can't use the file size difference as a valid argument.
After all, the aim of photography is to record light AND reproduce the capture using a computer monitor and hard copy print.
These output media have very finite dimensions - a monitor displays 8 bit colour (cf. 14 and 12 bit Raw data) and a common printer has a resolution of 600 - 1200 dpi with a monitor at far less. Jpeg compression strips off unusable data in an output sense. Like Mp3 does to music and ECW does to scientific imagery.
I have today, after reading this thread, shot the same picture in Raw and Jpeg on my K100D. Final output of the Silypix processed Raw and the in ca,era Jpeg are almost identical.
I will go back to shooting Jpeg and taking care of the camera settings to get the desired result.

Garth

[simons-photography]

fact is that with raw versus Jpeg you have

no lossy compression so its tiff quality
you have 16 X the amount of levels per channel (making that 4096 X the amount of colours per pixel meaning instead of 16 million colours you have nearly 70 billion colours)
the added information allows for exposure correction that done on a Jpeg would yeild disastrous results

put simpley there is just more to play with

[garth1948]

This 'numbers of levels' argument is partly true and is available if you are considering a section of the captured image in terms of dynamic range. There is an advantage if you crop the image to say a fairly dark section and process that separately from the rest of the image - the so-called HDR image) but when joined back to the remaining data the result is highly unnatural looking. Fact is the sensor as a whole has a fairly finite dynamic range when compared with the way the human eye processes images.
I totally agree with Pentax K20D advances in increasing pixels, decreasing noise AND making a far less compressed JPEG option.
Garth

[simons-photography]

well from what I read 1/2 the levels are in the first stop of the sensor, the next stop contains 1/4 the total levels, the 3rd stop contain 1/8 the total levels, the 4th stop 1/16 the total levels, and so on. so basically the darker the image the less levels available to represent the image.

what I do as suggested by the article that explained this is to shoot as brightly as possible this means less noise and more levels captured as I'm using the brighter end of the sensor that can capture more levels, then I lower the exposure in post processing, as the 16 bit file can contain more information than the 12 bit raw and I think it is more linear things fall back into place and there is much less noise.

the thing is its not really the sensor that has this exponential reaction but our eyes have an exponential curve so a CCD sensor that really is linear has to have its information converted to the exponential scale that our eyes work on, this is why the stops scale is exponential.

[garth1948]

This 'stop' talk is very confusing. What do you think is the relationship between the size of the hole that lets the light in (stop) and the energy of the light at particular wavelengths?
Garth

[dylansalt]

Ken Rockwell (love or hate him) does offer very relavant scientific data regarding above issues and he very very seldom shoots raw and he is a highly proficient photographer.

Imagine after a modeling shoot with lets say 500 images and a deadline to meet and everything was shot in RAW - you would have a nightmare of pp on your hands.

I have taken numerous test shots in RAW and after pp can detect a very slight increase in sharpness on the monitor compared to the jpeg (if you have tweaked your camera settings that is)

My friends with Canon's seem to find it very difficult to get a decent jpeg straight off out - and tend to shoot raw - just to get a decent jpeg a very tiring process

But the fantastic K10D has Sooo many tools/options to choose from its really hard to get a crappy shot in Jpeg format

But for sure I will definately have a raw backup file for a shot which could have decent stock/magazine viability, as they tend to need large file sizes (I convert them to Tiff after pp)

d

[garth1948]

And that is exactly why I now shoot Pentax after years with Canon.
The Jpegs straight from the camera are very good.
Garth

[simons-photography]

This 'stop' talk is very confusing. What do you think is the relationship between the size of the hole that lets the light in (stop) and the energy of the light at particular wavelengths?
Garth

what exactly do you mean ? I just described or attempted to that our eyes are not linear but the image sensor is so for us to see a "normal" image from the sensor we have to put its information into the same scale that ours eyes work on.

[garth1948]

what exactly do you mean ? I just described or attempted to that our eyes are not linear but the image sensor is so for us to see a "normal" image from the sensor we have to put its information into the same scale that ours eyes work on.

Our brain creates an image from many 'snapshots' recorded by our eyes. Each snapshot has its own f stop and dynamic range. A normal image in our brain is a 'normalised' multi-image.
I would say that it is impossible to reproduce a human brain image with a photograph - only one aspect of the image and that is the art in photography to produce that aspect.

[attack11]

Jpeg compression strips off unusable data in an output sense. Like Mp3 does to music and ECW does to scientific imagery.

unusable data? macro blocking, inaccurate color and aliasing. that's what jpeg offers. mp3 removes so much data it's unreal. i'm guessing you've never compared cdda to mp3 on quality hardware; there's a huge difference in sound stage and fidelity.

i'm kinda shocked people still think jpg is better than raw. if you have shoot raw and need to pp 500-5000000 images, use an app that'll batch things as needed.

[aleonx3]

I shoot RAW because I don't have to worry the WB which is hard to correct in JPEG; besides, batch conversion of RAW files to JPEG is fast from Photo browser - no need to worry. After the conversion, I use the JPEG files to scan the pictures and discard the bad ones (RAW+JPEG). I only backup the RAW files and the JPEGs I tweaked.

[simons-photography]

I've saved many a picture with WB ajustment

[patrickt]

I don't get it. With JPEG, the camera is doing the processing using the same settings for every photo. If you shot 500+ pictures and are on a tight deadline, you could batch all the RAW files to produce 500+ JPEGs using the same settings. I don't see a significant savings in time.

I really don't care what anyone uses and if I shot in a controlled environment I might get away with letting the camera settings decide. I don't shoot in a controlled environment, ever, so I shoot raw.

No one has ever made the arguement that the results from raw are in any way inferior to JPEG.

[stewart_photo]

Yep! It takes a person with BIG cahoonas (or an ignorant one) to shoot important pictures in JPG only!

Not all that big, or ignorant, considering photographers using slide film have been doing something very similar for decades (nice images to show for it too). Since slide film directly from the camera was often the final product (slides), there were very limited options to correct mistakes later. Their solution was to depend on skill and experience.

stewart

[stewart_photo]

RAW is fine if your not shooting 3 or 4 hundred shots a day.

Since memory cards are cheap, my solution to this whole JPEG versus RAW issue is to shoot using the RAW+JPEG setting. If the JPEG images are technically fine (99%), I don't have waste even one moment messing with RAW files (instant trash). Otherwise, I clean up the few relevant RAW files and save in JPEG format for additional editing within my graphics editor.

stewart