Pro Member
- Joined
- Oct 17, 2023
- Posts
- 898
- Solutions
- 3
- Likes Received
- 1,679
- Points
- 93
- Name
- Ed
- City/State
- Ottawa
- CC Welcome
- Yes
HEIF. Why should it gain more traction?
Follow along with the video below to see how to install our site as a web app on your home screen.
Note: This feature may not be available in some browsers.
I don’t know anything about HIEF. I’ve heard about it. I’ll look it up.So why doesn't HIEF gain more traction? All modern Canon cameras have it. Yes I use C-Raw. For long term storage I use JPG. Almost three decades of pictures. RJ
Because HEIF is not generally supported by legacy OS's, and other programs without converting them. Lot's of folks are still using Windows 10 and probably some XP users out there as well.So why doesn't HIEF gain more traction? All modern Canon cameras have it. Yes I use C-Raw. For long term storage I use JPG. Almost three decades of pictures. RJ
JPG has been around forever. 8bit, Hief has 10 bit . A promise of better JPG type file..HEIF. Why should it gain more traction?
Thanks. That’s enough info for meJPG has been around forever. 8bit, Hief has 10 bit . A promise of better JPG type file..
JPG is already good enough as an end product. For shooting, it's raw. JPG and raw. Where is the need for HEIF?JPG has been around forever. 8bit, Hief has 10 bit . A promise of better JPG type file..
Is that the same test results as would come from in camera shots, Archibals? I don't think so. My comparisons have been a whole lot closer in quality than the CR3 to JPG show, but I appreciate your test result thanks.Hmm, yes, that could be relevant. The JPG is 1,531 KB.
To find out the quality of the JPG, I converted the CR3 to JPGs of different qualities. At a quality of 70% the file size matched that of the embedded JPG. So I surmise the quality of the embedded JPG is 70%.
The next question is what JPG quality the camera saves, and I don't have the answer for now.
I Googled this and this is what it returned:
Key Differences
- File Size and Compression: HEIF uses modern compression that makes files about 50% smaller than JPEG while keeping the same or better image quality.
- Color Depth: HEIF supports 10-bit or higher color (over a billion colors) for smoother gradients and less color banding, whereas JPEG is limited to 8-bit color (16.7 million colors).
- Compatibility: JPEG opens on virtually any device, web browser, or platform. HEIF works well on modern Apple and Android devices, but older systems or specific web tools may require a file conversion.
- Advanced Features: HEIF can store multiple images in one file, depth maps, and transparency, which JPEG does not support.
Which Should You Use?
There are some footnotes associated with this, but I didn’t bring them across. My search was “HEIF vs JPEG” if you want more.
- Use HEIF if you want to save phone storage space and capture richer colors on modern devices.
- Use JPEG if you need to share files instantly across older computers, print labs, or websites without compatibility hiccups.
I don’t quite understand the 10-bit and 8-bit part in terms of gradient and banding. Maybe some with more knowledge can chime in. I’m happy with craw and jpeg. Maybe heif is better for phone cameras.
It would be interesting to see the original with EXIF. Was the original overexposed?Ed,
Thanks for explaining that. I have a photo that has been bugging me for a while. There is some banding in the blue sky, and I can’t figure out why. Your explanation makes sense. I have messed around with it in Lightroom, and no luck. What can I do, if anything, to avoid it or fix it? I took the photo with my R50 and my 18-45 kit lens shortly after I got them, if that helps. See the blue sky below.
View attachment 49562
It looks like the sky was overexposed. There are quite a few pixels that are maxxed out and don't have faithful data. Given that it is a JPG, it will be difficult to fix.The original is a jpeg. Can you recover EXIF info from a jpeg?
I have the sky saved as a mask, and the only instance where I can remove the bands is when I jack up the exposure in the sky, and that just turns the sky white. How do you add noise? I went to the denoise slider, and took it to extremes either way, and can’t tell much difference. In retrospect, this photo may not have been the best example to use.
Thanks for your expertise and patience.
OK!Thanks for the analysis. Looks like I need to pay more attention to exposure. This particular photo was taken a while back before I was aware of the histogram, and what it does. Again, thanks for the advice.
The iPhones IQ is quite good. in 2015 (I think) I got a Canon G11 for shooting an event. I never used it as it is just too small for me. Two years ago we went on a short trip and only I took it and my iPhone. I preferred the iPhone over the G11 for night shots even though it was set to RAW. I sold it. One day when I can't hold up a full system I'll have a look at modern P&S.And then there Apple:
If you take a photo on an iPhone, it uses HEIF technology to compress the image, but saves the final file to your storage with a .heic extension.
Great review. I've been using CRAW (but without in-depth testing) for many years, and am enjoying the smaller file sizes. Based on what you say, I will continue doing that, as my subject material is mostly birds and bugs.Disclosure first: I make a Mac RAW editor, RevelRaw, and the "RAW or CRAW?" thread here in September made me want to measure this rather than guess. I took all 110 Canon CR3 files in the public raw.pixls.us archive (CC0, 26 cameras) and compared C-RAW against RAW from the same camera, session and ISO, on the 20 cameras that have both.
Size. C-RAW came out a median 45% smaller, from 31% (PowerShot SX70 HS and EOS M6 Mark II) to 59% (EOS R5), and every C-RAW was smaller than its RAW partner. Canon's own manual tables say 42 to 57%, a median of 49% across 18 cameras. In practice it's close to twice the frames per card: our R6 Mark III pair was 34.61 MB RAW and 17.10 MB C-RAW, which is about 3,700 versus 7,500 frames on a 128 GB card.
Speed. What C-RAW buys is buffer depth, not frames per second; no Canon manual I read says it changes the frame rate. From Canon's manuals: an R5 Mark II on a UHS-II SD card goes from about 95 RAW frames to 580 C-RAW in High-speed continuous plus, and an R8 at 40 fps from about 56 to 100.
Quality. In mid-tones and highlights I found no difference my test could detect (it resolves roughly 5 to 10%), on eight cameras with well-aligned same-scene pairs. Where C-RAW does differ is near black. In the covered strip at the sensor's edge, which never sees light, C-RAW stored 24 to 39% less pixel-level variation than RAW on 6 of 8 cameras (ISO 100 to 200). That isn't an improvement: those pixels get no light, so a smaller spread means the stored values were changed. Brightened four stops, the R6 Mark II's C-RAW shadow has softer grain that clumps into slightly coarser blotches; in a comparably dark R5 Mark II crop I couldn't see a difference. It fits Canon's own warning that if you badly underexpose and then brighten, "noise levels will typically increase more in C-RAW files". Canon Europe also says outright that C-RAW is lossy.
So: C-RAW for sports, birds, events and travel; RAW for astro, night and landscapes, or if you often rescue underexposed frames. Two limits worth knowing: you can't record RAW to one card and C-RAW to the other, and I know of no Canon tool that turns existing RAW files into C-RAW later.
Most cameras had one pair and one scene, mostly at low ISO, and I didn't test highlight clipping or high-ISO pairs. The files are all CC0 on raw.pixls.us if anyone wants to check. Per-camera table here.