I think slowing down mainly concerns 5K-Retina-iMac-users (without the best hardware)
better stick to 16 bit
Fastest performance is still going to come in 8-bit mode (from what I can tell).
If you do use a higher bit depth for editing, I think 32-bit float linear should be faster?
This may be highly memory dependent, I haven’t had a chance to hack at it to test more definitively.
I just know that all of the GEGL operations will be done in 32-bit float linear.
What you want to use higher bit-depth for is to delay making the “damage” you do to your image visible, especially if you’re doing a lot of changes. Every Single Change you make to your image after the shutter closes does a bit of damage, and doing it in a higher data precision helps to keep the accumulation well within a sub-visible range. In 8-bit precision, a few edits can easily push these artifacts into what you can see.
Most camera raw files deliver the raw image as 16-bit unsigned integer values, so continuing at that precision avoids a transform. Keep in mind there are 256 16-bit values between every to adjacent 8-bit values, so there’s a good bit of room within which to work in 16-bit. Native floating point (32-bit float and 64-bit double) provide more precision within which to work, and the performance difference vice 16-bit integer is not significant. Some softwares will provide a 16-bit float, but that’s implemented in software and can introduce a noticeable performance difference.
But a significant reason to consider floating point is that the image storage convention, black=0.0, white=1.0, allows an edit that causes particularly white to go past 1.0 to actually retain a meaningful value, rather than be clipped as it would be in 8- or 16-bit integer format.
Probably a bit more than you want to know. The essential thing is, if you’re going to change stuff in your image (sharpen, saturate, LUT-it-up, etc.) , it’s a good idea to do that in at least 16-bit, and defer the crush to 8-bit for saving to the output file. However, if you’re always satisfied with the JPEG straight out of the camera, with maybe just a crop for good effect, you can probably live in 8-bit land.
Oh, if you shoot for Reuters, you’ll have to live with 8-bit. They don’t accept anything other than OOC JPEGs…
3 Likes
smart decision for a press agency. And way less trouble (= less labor cost), too.
1 Like
most stock photography agencies accept only 8bit jpegs in srgb, too. probably has to do with file size
Ten years ago, “work in 8-bit integer unless you know that creates problems” might be sensible advice. And if you know the image will be edited once in non-linear colorspace such as sRGB, used once and then discarded, the advice is still okay.
But when an image is edited in linear colorspace, or might be re-purposed, or re-edited in the future, or we simply want the maximum headroom without having to worry, 16-bit integer is the minimum that I consider sensible. Some of my work needs 32-bit floating point, and that’s almost become my usual working practice, though it is often overkill.
I will add a note of caution about floating point: digital arithmetic sometimes gives results like 65535.999999 or 0.000001 or even -0.0000001. With integer arithmetic, these are rounded as we would expect, but floating-point files will retain these weird numbers, and operations like “turn all non-black pixels white” don’t work as we would expect.
Integer values in the range [-16777216;16777216] are perfectly matched in 32-bit float space (including the value of 65536 you mentioned as 65535.99999)
Yes, the result of a calculation may be exactly 0.0 or 65536.0 and so on. My point is that sometimes the result of a calculation may not be exactly so. We can demonstrate the effect:
$ echo 'scale=15; 1/49*49' | bc
.999999999999994It’s a long time since I played with bc, but
if you change to scale=14, isn’t there too big
a jump between the results?
16 bits is not more info, it’s more progressive transitions, which limits posterization and quantization artifacts… The key concept here is 8/16 bits integers, so every pixel manipulations are rounded to the closest integer, which ends up being quite a gap in low-lights. That’s why clever softs input and output 8/16 bits integers, but work internally in 32 bits floats (where no rounding happens until the very end of the pipe).
That instructor should dig a hole and shot himself in it, that’s beyond stupid. You don’t know if it’s going to be overkill until you run into a banding effect and discover you can trash your whole layers stack and start over in higher bit-depth, or unless you work on gamma-encoded data instead of linearly-encoded, which is – again – stupid. And I’m pretty sure Photoshop doesn’t use 32 bits float internally, but keeps whatever you feed it as is (open a 32 bits float TIFF in it and try to use the healing brush : PS will issue an error saying the healing brush only works in 8 or 16 bits, so my guess is it works on integer RGB).
Doing crap for 20 years only makes you an expert at crap. Use clever, reproductible, systematic workflows. No assumptions, no guesses, no nasty workarounds (except when you can’t avoid them).
Exactly… 32 bits/8 bits sums up to RAM use (and CPU cache use, which is less and less an issue since AVX processors). No definitive answer here, it depends of what your CPU is and how the soft uses its optimizations possibilities. Basically, 8 vs. 32 bits means bigger memory chunks to move between the RAM and CPU cache, so a bit more I/O latencies. But there are several ways to overcome this, especially on modern CPUs (and GPUs), so the difference is neglictible.
I think this has more to do with ensuring the authenticity of the information and avoiding manipulated pictures and montages. See the World Press Photo and McCurry controversies in the past few years.
4 Likes
Do we really need statements like this? Not only are they mean-spirited, but this kind of hyperbole only serves to undercut the rest of your statement(s).
5 Likes
Yes we do. The issue of “experts” using their authority to propagate fake knowledge should be adressed with the utmost firmness.
You may still address it with firmness while not saying those kinds of things. In fact, omitting those types of communications make people more likely to believe what you’re saying, improve their own practices, and engage with you further.
On the other hand, being condescending and hyperbolic only makes people avoid you.
5 Likes
Yes, I’m well aware that the 2010’s are all about the form, and little about the content. Being nice is more important than being accurate or competent. The thing is, when people ask for disdain, disdain I give them. It’s one thing to be wrong, that happens to everyone. It’s another to be wrong your whole carreer and not discover it. But, again, that can happen. What is unforgivable is to teach your wrongness for money where you could maybe have done some research before and get some clue that your beliefs were not backed up by evidence. That, is asking for my wrath. Comm’on, Internet… You want to be a teacher ? Show some due-process. You are only one click away from Google Scholar.
You speak about form and content as if they’re mutually exclusive. They are not.
Nobody asked for disdain.
Still no, nobody is asking for wrath. Or disdain.
I thought the same thing when I read the first sentence of your last post.
Can you please stop those kinds of statements? They’re unnecessary.
4 Likes
I think I will stop posting fair and square instead. I don’t live in a state of mind where everyone is nice and trying to improve, and everyone deserves respect for just beeing alive. This world is becoming “Idiocracy” for real, Internet is merely a scope of that, and having to be polite with idiots is making it ok to stay an idiot. I have nothing to win here except grey hair and ulcers, and repeating the same things over and over.
You can be accurate/competent and nice or at least polite. In fact, I’d say it’s a failing if one is unable to muster some tact and politeness when communicating.
3 Likes
If that is what you want to do, best of luck to you.
I’m sorry you live in such a sad state. I hope it will improve for you.
1 Like
The dynamics of bit precision is one of those many things in imaging that is rooted in a simple concept with complicated implications. The challenge here is to craft the prose that successfully communicates the important concepts to folk who are not computer scientists, folk who want to just create compelling images.
@anon41087856, I also have little tolerance for the propagation of misconception. However, I’ve also found there is little to be gained by injecting vitriol into the discourse. Indeed that then becomes the focus, to the distraction from the essential ideas. I find your perspective in the recent topics to be substantial, an imager who has taught himself C++ to improve the state of things. That is no mean feat. And, it has allowed you to see things for what they are, rather than just blindly accepting the status quo. But all that is obscured by the disparagement.
You have authored a darktable module that made it into a release. Don’t think people here aren’t listening…
5 Likes
For the record, I think your comment about the guy digging a hole to shoot himself in is funny. No complaints here.
I think it also demonstrates a difference in mindset between theoreticians and users.
Ultimately, my goal of learning to use photo manipulation software better is to make more powerful images. That is the point of these software programs…to help people make more emotionally powerful images.
Often, when I edit, I’ll try tweaking one parameter in the effect I’m doing, or maybe tweak a little bit the color here or there, and I’ll want to compare my last version vs. the one with the tweak.
If I have to sit and wait 10 seconds on modern hardware/16 GB RAM + decent video card, for the screen to redraw because I’m using ultra technical 1 millien bit floating point, that is an impediment in my creative process. Because it gets in the way of me SEEING the image and comparing one version vs. another.
So when I say, “Does it REALLY matter if it’s 8- or 16- or 32- or 1 billion bits?” I’m looking at the final image I can make with the software. That’s the criteria. Because if I sit twiddling my thumbs while the computer computes every little nuance on its 16 cores every time I touch a tone curve or whatever, that interrupts the creative process.
All I want to do is make the best pictures I can, in the amount of time that I can afford to spend on each picture. If I can process in 8-bits and have instantaneous on-screen updating and the final image doesn’t suffer (much) from this shortcut, then that’s the way to go…in my world.
In your world you seek maximum precision and purity of theory.
The real world gets messy.
It’s a fundamental difference between the technician mindset and the user mindset. I’m not trying to talk shit on technicians. There are a bunch of optics geeks sitting around at Canon right now arguing over some crap about how to design the next lens. I have no idea about their formulas for light diffraction this or that, and I don’t care. But they will do what they do and in a couple years they will release a new lens that enables people like me to make better pictures, and I’m grateful for that and for their work.
Same thing with these softwares. At what point is it OK to start cutting corners? The “technician” doesn’t want to admit that it’s OK to sometimes throw away bit-depth information, even if the user can’t tell a difference using his eye. One person evaluates in the abstract ideal, and ones evaluates in the concrete. Somewhere we need to meet in the middle.
It’s a big world, and there’s room for all of us. The people you say should dig a hole and shoot themselves, have been getting paid to make powerful images for years, and they are successful at it, whether you like it or not. These are the sort of people that our FOSS image software is purportedly written for.
Here’s the portfolio site of the instructor who said 8 bits does the job most of the time. When I look at his site, I’m not seeing 8 bit photos. I am seeing lighting, color harmony and emotions. I see light, colors and emotions like when I look at your website, also. I can’t tell what pix were processed in 16 bits or not. Can you?
I think it’s interesting that I posted a question higher in this thread about if I could use lower bit depth in gimp to make processing faster. I got responses that were several hundred words, which I read a few times to try and make sense of. And I still don’t know the answer to my question.
1 Like