My main point is that often, if you can see banding in the blue sky, it's because the jpeg compression is set to compress too much. Many people ignore that setting when they save their edits, so the final JPEGs look bad.
Yes and no, compression can cause banding, particularly in images with subtle gradients like skies, shadows, or smooth skin tones. JPEG reduces file size by compressing pixel data, discarding color information, and forcing smooth, gradual transitions into distinct, blocky steps. It also happens when values within a gradient get pushed so much that there is no color/value in the file to actually represent the mathematical change you’ve applied post. Having a fully exposed sensor image such that you don't have to push post stops banding.
What does "...push post..." mean?
Push post - increasing exposure, contrast, saturation, etc. In a JPEG you only have 256 tones per channel, banding is less likely if you're editing the RAW file with 4096 - 16,384 (12bit;14 bit) tones per channel and then exporting as a JPEG. However, if you push (edit) a wide gamut color space RAW file to far, then convert to a JPEG it could create banding. Also repeated editing and re-saving a JPEG results in additional compression.
I usually shoot manual.
That depends on what example you're talking about. If it's this one: RAW vs JPEG - Freedom Or Limiting then read the post and you'll see there's nothing unfair about presenting that SOOC JPEG. You think it can be edited to match the result from the processed raw file? Why don't you show us. You edit JPEGs so download the SOOC JPEG and edit it. The SOOC JPEG in that post was presented to make the point that creating (or editing) the JPEG was not an option.Interesting thread and the never ending debate. I appreciate the interesting ways that someone can make a RAW image better than in camera JPGs. I don't disagree. If I ever have banding problems with the sky, I'll know that I should adapt and use RAW.
The other parts are also interesting, editing, and technology, the people who are pixel peepers, vs the people who are equipment junkies. Someone can be both or neither or use a combination of some of each. My only point is, it's a personal choice. How deep someone needs to get, to get a high quality image. If I wanted "good enough" there's always the letter "A" on the camera.
One example here, someone posted an edited RAW image, vs a essentially, in-camera JPG. Hey, unfair.
JPEG is an archive format not designed for editing. Making tone/color changes to a camera JPEG degrades the image. It may be that you're better off with the changes but it's always double edged and so editing the JPEG necessitates degrading the image (avoidable by editing a raw file).
Wow, so you're getting a lot of underexposed sensor shots and then you're editing the JPEGs.
You always have a choice, everyone has a choice and there's noting wrong with that. And if you're happy with the choice you've made that's good. Having and making a choice however doesn't make the objective differences in technical IQ go away. In a forum like this some folks have an interest in learning about those objective differences to help them make their choice -- so it's worthwhile presenting them. In this post: RAW vs JPEG - Freedom Or Limiting you can see an example that compares Canon's SOOC JPEG lens profile with the one used by DXO PL-9. That object difference will not go away if you chose to shoot camera JPEGs -- you get Canon's sucky lens profile baked into every JPEG and you can't fix it, but yes, you can certainly chose it.
Yes the results are what matters, and the objective technical differences in IQ are real and they show in the results.Is it necessary to have a full frame camera? I owned one once. Nice, but the buffer wasn't enough for what I want. It was a nice Canon 1Ds. It was "wow". Same lenses and same photos as my crop cameras. Just beautiful and better all over. But not for what I shoot. A 1Dx mark something less than the last one, could still be in my future. I don't like the electronic viewfinders. Stick in the mud kind of person that hasn't adapted to change yet.
My view is also, the best lens I can afford, makes me look better than I am. Yet I still shoot DSLRs. I have a feeling that the opinion also goes back to film days when we had less choices and editing or printing was much more complicated and took much more time. Processing of the negatives made a difference, before I ever got to making a print. The film I chose had an effect on the resulting images. Digital sure makes life easier, doesn't it.
So you folks who choose to, keep shooting RAW, and I'll keep shooting JPG. I'll edit and get what I need and want, and so will you. And if I have an unnecessary 300mm f/2.8 Canon lens, you don't need to go find one, to get your best images. You can go get a f/4. Do you need the f/2.8 or is f/4 just fine? I question my own choices sometimes, but the brightness and color and saturation are just much more and visually noticeable, even if I'm going to be shooting at f/5.6. So there. I guess I'm an equipment guy? LOL but I shoot crop cameras, so I guess not?
If I need to shoot RAW to make my images better, and have the latest full frame mirrorless, will it make my images better? Another, this is personal thing. I'm 78. Aside from weak eyes, do I have an eye? Does what I shoot actually matter o need all the fancy and high tech enhancement, to make them better. Or they "are what they are" and nothing is going to make me better. (I'd like to call it the Popeye Theory of artistic impression and photography)
Individual choices, there's no law or absolute right way or answer to RAW or JPF. Only the results matter.
Your JPEGs are not underexposed -- that doesn't mean the sensor isn't. Here's a sample SOOC JPEG image from a Canon 90D in the sample set at DPReview: https://s3.amazonaws.com/files.prod...ccfdeb5ab4ee5418c730148b4d8f346b7e0e9ae8856d9 How's the exposure? Is it under or over or just right? Would you expose it differently and if so how?
I get
Try this: https://www.dpreview.com/sample-gal...eos-90d-review-samples-dpreview-tv/2814771673 In the DPReview samples for the Canon 90D it's a photo of a hand holding seeds for a small bird. It's a well exposed camera JPEG. To make that JPEG the camera underexposed the sensor by 1.3 stops. Less than 1/2 of the sensor was used. And if you make good SOOC JPEGs with your Canon camera then for some of those photos the same thing is happening. Using less than 1/2 of the sensor's recording capacity has real objective effects on image quality. Saving SOOC JPEGs often forces that result.
Of course. I've never said otherwise. What I am saying is that regardless of which option you chose there are and will remain objective differences in technical IQ between the two options.
Saving SOOC JPEGs is an appropriate choice for a lot of people. That choice is available and it's always been my opinion that people who want to or need to should do that. However when people make that choice it doesn't throw a magic switch that changes this: there are and will remain objective differences in technical IQ between the two options. In terms of technical IQ the two options are objectively not equal. And bottom line, you're not happy with that. In an earlier post in this thread you said this: "Where the argument of RAW over JPG fails is, JPGs are not so inadequate that they can't produce a comparable image." That you got wrong -- SOOC JPEGs can't produce a comparable image. That's an objective reality. That reality doesn't take away your choice or force how you should take photos, but it's there and once you make your choice it's still there.
You're going to need to explain that to me, please? It says JPG + Raw file. But now something new, and if there's an article to go with this, sorry I missed it. Why did the camera underexpose 1.3 stops? And why would less than half the sensor be used? Interesting.
Oh Oh, I missed that the first time past. I often save at 80% instead of 100%. Quality 10 in Photoshop, (is that 83%?) because Adobe is somewhat like Spinal Tap, their quality goes up to 12, not just 10.
Am I making that same mistake and creating potential errors? Is 80% acceptable? I hate to use the term Good Enough because to me that means, substandard, but "good enough". Should I be saving at 90% or 9, to avoid sky banding issues? I don't shoot that much nature and open sky, so this would help in the future.
There's no doubt Joe knows the technical side but sometimes the way he explains it is confusing. Here's a few thoughts to clarify. I'll condition these as Pentax only, but others are mostly the same i think.
There's no article -- just the sample DPReview images for the Canon 90D -- I picked one based on the JPEG appearance. Here's the raw histogram and stats for that image's raw file. Note that I've marked the unused sensor area as wasted. Note the numeric scale underneath the graph. A camera with a 14 bit ADC can store a max value of 16384. Half of 16384 (8192) is half of the sensor's total recording capacity. In the stats you'll see that the highest value recorded in this raw file is 6558 -- so over half of the sensor's total recording capacity is left unused.
Yes it does. So your JPEGs are a little light, but if that Canon 90D above underexposed the sensor by 1.3 stops then you're adding +.3 would still be leaving half the sensor unused. Exposing for a good JPEG will typically leave the sensor underutilized while exposing the sensor to full capacity will typically nuke the JPEG. So I ignore the SOOC JPEG and expose the sensor.
Shooting RAW allows achieving a better SNR than is possible shooting JPEG.
Create an account or login to comment
You must be a member in order to leave a comment
Create account
Create an account on our community. It's easy!
Log in
Already have an account? Log in here.
