8bit JPEG in 2017 ... WHY?!

Become a member of the CGSociety

Connect, Share, and Learn with our Large Growing CG Art Community. It's Free!

REPLY TO THREAD
 
Thread Tools Display Modes
  01 January 2018
8bit JPEG in 2017 ... WHY?!

Just a general rant... I dont get it.
Why would ppl still use 8bit + jpeg compression in their workflow.

I know there are huge projects out there where texture sizes can get absolutely ridiculous and if you're an indipendent artist in those cases I can understand it... just a little.


But I see so many ppl using 8bit and jpeg in absolutely cringe worthy ways ...

Just an example of a fairly known 3D artist after randomly watching a recent youtube video recording of his workflow (I shall name no names):
Rendering to textures saved as jpeg and then continues to process that texture in photoshop (fairly heavy levels adjustment + other stuff) resaves it as jpeg and uses that for the final render
... seriously

And this is not an isolated incident ... so many ppl seem to do this.
If you look at the 'professional' models sold out there for example... so many jpeg textures...
Even so-called 'professional' texture packs!

First of all... you lose so much info when saving as 8bit, especially if you are going to boost the contrast and/or crush the blacks and whites.
Then, as if thats not enough, in most video's I see ppl tend to save as jpeg... which is just terrible... even if you set it at maximum quality you still lose so much chroma (color) information (jpeg saves chroma channels at half the resolution).

Combine the two and... well yeah...

Pls do yourselves (and your clients!) a favour and at least use 16bit + lossless formats (png, tiff, ...) for textures/render ... especially if you're going to process/edit them!
Float .exr would be overkill for most textures, except for displacement maps or when you need the over exposed data.

I personally only use jpg's for previews, online posting/mailing or test renders...


*end of rant*

Last edited by ACiD80 : 01 January 2018 at 11:07 PM.
 
  01 January 2018
This is what you get when people become badly overspecialized from day one, learning only some 3D operations in one 3D software for example.

Anyone who comes to 3D from a web design, multimedia, interactive media, game development or similar background knows all about PNGs, TIFFs, alpha channels, video codecs, compression techniques and so on.

So yeah, if you ONLY learn a few key aspects of a 3D software from Youtube tutorials or online courses, and you restrict yourself to "I'm gonna become a character modeler for games", you are probably going to be pretty damn ignorant of the wider ecosystem of technologies that support CG work, even really basic stuff such as image file formats and how video CODECS work.

I've noticed again and again in online discussions that people with no programming background, for example, think that a lot of 3D stuff that can be developed in a few weeks or in under 3 - 4 months is "amazingly innovative" and "must have taken a long time to develop".

Most of those 3D technologies really are not quite as amazing under the hood as one might think. A lot of the techniques used have also been explored in detail in academic computer graphics papers published many years earlier.

One problem this creates is that when CG artists ask CG software developers to implement a given feature, they have no idea whether it is an easy or hard feature to implement.

A related problem is that the 3D software marketers can then market the sh*t out of pretty ordinary 3D soft features that really aren't all that complex under the hood, making them appear to be something extra-special that took a lab full of geniuses many years to develop and implement.

The ideal background for really knowing how EVERYTHING works in CG would be a Computing Science or Software Engineering degree followed by a Computer Graphics degree.

Then you could really look at CG software objectively and understand how everything really works down to the technical details.

You'd also be in a position to create your own scripts, functions and plugins for doing new stuff that the 3D soft cannot yet do out of the box.
 
  01 January 2018
I'd venture to say most people in the 3d world, their work doesn't go too much further than their own computer, maybe their office, rarely outside sources. In many cases, I personally like to apply the "if it stupid but works, it's not stupid" check. Whatever pays the bills in my opinion.
 
  01 January 2018
@skeebertus: its pretty damn basic knowledge if you ask me...
 
  01 January 2018
I agree.
As soon as an artist puts out a working demo reel they start to see limitations and fix them if they don't want their reel looking like shit!

But maybe they are worried about stolen content?! You can use or steal my shit jpegs while I keep my 16 bit working assets secure for 'me' only...

Last edited by circusboy : 01 January 2018 at 04:45 PM.
 
  01 January 2018
I teach quite a bit, and I always make it a point to check what formats people work in and then fill in the gaps and correct them when needed. Honestly, teaching everyone from new people to those who have been doing PS/AE/3D for 10+ years, the majority simply don't know anything about file formats or bit depths. Some rough estimates from my experience;

50% don't know about lossy/lossless and will gladly work with jpg files
95% don't know about 16/32 bit and what they're for, theres rarely any distinction between them knowing about 16bit for smoother gradients or 32bit for dynamic range.
90% will render straight to video formats, and even then dont know about formats or bitrates, just accepting the defaults.
100% have never heard of EXR
__________________
Matthew O'Neill
www.3dfluff.com
 
  01 January 2018
This is a topic I think a lot of artist that move from hobby to working student have no info on or never comes up. Anyone in this thread want to do an article for it for the forum or front page? I think it would be useful to discuss the reasons and times to use it etc
 
  01 January 2018
I left my former 20 year career as Graphic Designer for print
in the Wash DC metro market, partly due to having to explain
,for the seven trillionth time, that a 72 DPI, index color jpeg Logo
that the clients insists.... "looks perfect on their website",


CANNOT BE USED on a 6 x 4 foot poster for their trade show booth
with a viewing distance of less than one meter



Alternatively however, people who make 3D content for DAZ
use 4k image based texture maps on everything and
Do not understand the Advantage of using native render engine
procedurals despite Having Nvdia Iray bundled in Daz Studio for years now.


They post endless threads about how IRay keeps kicking their
Scenes over to the CPU When All that Image Data exceeds the capacity
of their Graphic cards.


It is said "time equals money"
also that "Education is expensive"
For those who complain that Education is expensive
I suggest they consider the devastating cost of ignorance in terms of time & money.

Last edited by ThreeDDude : 01 January 2018 at 01:54 PM.
 
  01 January 2018
Im talking from my POV in my country in Latin America, USA may be different.

Cant say I agree ((on final render)). I meant, sure more information is better for compositing, but I have worked in advertising industry back when flat screen tvs didnt existed, and no matter how hard you worked and how pretty the final image looks in a controlled environment, once its in the air, you have no control if the target audience is gonna notice the banding in the quality of the picture in their TVs between eating, reading the newspaper or the lightning in the room, and the compression of the TV Network (I mostly watch netflix with low bandwith, so there you go), or the compression of Youtube, Twitch, and so many variables. Really, if your not gonna edit them, the TV network will compress it anyway...

Even this forum cant accept images beyond a certain Kb size and no one seems to notice it when looking at the gallery of finished work.

edit.. also, from experience in my country, most -if not all- clients asks for JPGs and compressed MPG / H264, with separated alpha, and its also a requirement in TV networks including the one I currently work for. Almost all my work end up in an AVID or in a Vizrt / Xpression (both systems uses their own codecs to compress the image), and everything is recompressed when sending the signal towards the sattelite (its too much data to travel uncompressed).

second edit.. EXR takes a lot of space (wich we dont have in the first place), and while its great to have control over the contrast and levels of the image, plus AOVs, I have to explain my coworkers how to open it and why it has so much info inside, and I do that everytime I share a project with someone else (wich fortunatedly, I dont do often). AE should by default add the EXtractoR plugin automattically everytime is added into a comp. My humble point is, you guys dont know what is the workflow and rules / policy of the company people work for, so don't assume is the artist's fault. If its a small company, great, it can be discussed and the workflow can be established before starting a project... but In a 500+ company... not so much.
__________________
artstation
Close, open-relationship: C4D / Zbrush
Hate / love: Maya / Houdini
Former gf: XSI

Last edited by luisRiera : 01 January 2018 at 04:08 PM.
 
  01 January 2018
Originally Posted by luisRiera: Im talking from my POV in my country in Latin America, USA may be different.

My humble point is, you guys dont know what is the workflow and rules / policy of the company people work for, so don't assume is the artist's fault. If its a small company, great, it can be discussed and the workflow can be established before starting a project... but In a 500+ company... not so much.
FYI Its not a 'country thing'.
Its a 'Film production Vs TV production thing'.

Film image size and formats are overkill for most TV production.
EXRs can be so heavy because they are designed for *complex*compositing (for film most of the time).

Meanwhile DV for TV quality soon looks like shit if everything else is at film spec.
It is a big mistake to try and use TV spec/formats for anything but simple tests if you are working on film VFX or Animation!

Last edited by circusboy : 01 January 2018 at 06:20 PM.
 
  01 January 2018
I must admit working with Print graphics 20 years and later 3D /CG graphics
I sometimes get frustrated with the ignorance of laypeople
when I try to explain concepts that seem simple to us.

I know a fellow who's cousin bought a 4K TV and all he ever watches
are Old B&W DVDS of American sitcoms from the 1970's("Sanford and son" etc)
he swears they look "so much better" on his new tv and wont hear
anything about source resolution vs viewing device resolution.


I told this same fellow (not the cousin) how much drive space was being
consumed by the optimus prime rig scene file for "revenge of the fallen" on ILM's
servers (30 terabytes)


He thought it was "not possible" because the entire transformer movie"revenge of the fallen"
that the got off torrent was only 950 megabytes on his laptop.
 
  01 January 2018
Originally Posted by luisRiera: Im talking from my POV in my country in Latin America, USA may be different.

Cant say I agree ((on final render)). I meant, sure more information is better for compositing, but I have worked in advertising industry back when flat screen tvs didnt existed, and no matter how hard you worked and how pretty the final image looks in a controlled environment, once its in the air, you have no control if the target audience is gonna notice the banding in the quality of the picture in their TVs between eating, reading the newspaper or the lightning in the room, and the compression of the TV Network (I mostly watch netflix with low bandwith, so there you go), or the compression of Youtube, Twitch, and so many variables. Really, if your not gonna edit them, the TV network will compress it anyway...

Even this forum cant accept images beyond a certain Kb size and no one seems to notice it when looking at the gallery of finished work.

edit.. also, from experience in my country, most -if not all- clients asks for JPGs and compressed MPG / H264, with separated alpha, and its also a requirement in TV networks including the one I currently work for. Almost all my work end up in an AVID or in a Vizrt / Xpression (both systems uses their own codecs to compress the image), and everything is recompressed when sending the signal towards the sattelite (its too much data to travel uncompressed).

second edit.. EXR takes a lot of space (wich we dont have in the first place), and while its great to have control over the contrast and levels of the image, plus AOVs, I have to explain my coworkers how to open it and why it has so much info inside, and I do that everytime I share a project with someone else (wich fortunatedly, I dont do often). AE should by default add the EXtractoR plugin automattically everytime is added into a comp. My humble point is, you guys dont know what is the workflow and rules / policy of the company people work for, so don't assume is the artist's fault. If its a small company, great, it can be discussed and the workflow can be established before starting a project... but In a 500+ company... not so much.

That sounds like a frustrating situation.
Sure if the clients demand that you deliver in jpeg there's not much you can do after you tried to inform/educate them why it's a bad idea.
I think its important to educate clients about this. It can get tiring having to do this over and over, but its worth it.
I've sent such clients 16bits anyway in the past and when they saw the difference in post afterwards they were very thankful I did.
You can even design a little informative folder about it that you can give to clients that don't get it.

You mention youtube... Even youtube support 10 bit HDR for quite some while now. Even 10bit gives a world of difference (4 times more smooth gradients vs 8bit so it prevents banding).
I also have to mention this; it's not quite correct to think just because youtube/twitch quality is bad that its not worth sending/streaming as good quality as possible to them.
A good source will still en up looking much better than a crappy quality source sent to youtube/twitch/etc...
Its like compressing a wav/flac to mp3 vs recompressing an mp3 to mp3... the last one will sound noticably worse... same thing counts for video.

Though using 8bit for the 'final' output is not ideal... it's still not as bad as using 8bit + jpeg for images that will get processed and then resave them as jpeg and use that in your workflow.
While it takes absolutely no extra effort to save as an 16bit tigg/tga/png/whatever... (or 32bit exr where needed).
Diskspace should not be a big issue anymore with current prices (just get a decent NAS). Unless you're working on extremely big projects...
 
  01 January 2018
Originally Posted by circusboy: It is a big mistake to try and use TV spec/formats for anything but simple tests if you are working on film VFX or Animation!
True, but the OP didnt mentioned film VFX, he talked about animation in general terms.. why would we assume he was talking about movies if the problems he refers dont exist in the movie industry because over there everyone follows protocols?
__________________
artstation
Close, open-relationship: C4D / Zbrush
Hate / love: Maya / Houdini
Former gf: XSI
 
  01 January 2018
Here are some quickly googled examples of what is different about 8bit vs several higher bit depths.
I think several people will recognize these artifacts, especially after processing an 8bit image in comp...
I hope this helps showing why you should never again use 8bit (except for sending a preview, for use on the web)






Last edited by ACiD80 : 01 January 2018 at 06:24 PM.
 
  01 January 2018
Originally Posted by ACiD80: Its like compressing a wav/flac to mp3 vs recompressing an mp3 to mp3... the last one will sound noticably worse... same thing counts for video.

Though using 8bit for the 'final' output is not ideal... it's still not as bad as using 8bit + jpeg for images that will get processed and then resave them as jpeg and use that in your workflow.
While it takes absolutely no extra effort to save as an 16bit tigg/tga/png/whatever... (or 32bit exr where needed).
Diskspace should not be a big issue anymore with current prices (just get a decent NAS). Unless you're working on extremely big projects..
Mp3 is a compressed format.. sometimes is better to work with whats more practical to use because if not, we would still be using phonograph records -many claim the quality is better- and a lot of storage space. Although, I agree, I use png / exr whenever I can, but as part of my work is News, I tend to use JPG from google a lot ( I often search for photos of presidents around the world or soccer players for example ).

Dont get me wrong.. im not talking about working with a mediocre quality.. its just... fine details are lost in television.. For example, often clients procure the designers an specific pantone for the color of the brand we are working for.. but, that also is very subjective when its in the air.. do a quick search for the coca-cola logo on google, and you'll notice, not every logo is the same red color.
__________________
artstation
Close, open-relationship: C4D / Zbrush
Hate / love: Maya / Houdini
Former gf: XSI
 
reply share thread



Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

vB code is On
Smilies are On
[IMG] code is On
HTML code is Off
CGSociety
Society of Digital Artists
www.cgsociety.org

Powered by vBulletin
Copyright 2000 - 2006,
Jelsoft Enterprises Ltd.
Minimize Ads
Forum Jump
Miscellaneous

All times are GMT. The time now is 05:52 AM.


Powered by vBulletin
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.