Why isn’t everybody uploading H.264 codec?

So at what average bitrate do artifacts are things of the past.

I wish a cure-all bitrate would apply.Sadly it's about the codec itself not the workarounds of it.One could encode at 20-30mbps at the highest level of quality as determined by his/her encoding software to feel better, but it's about other factors as well,such as vbr or cbr, profiles (levels), and many low level functions that most people won't bother with even with professional encoders.

And it all comes to the nature of the footage,the motion,colours,detail,camera source (cameras also have their own profiles for encoding that favour certain types of footage while disregarding others) that makes this rather complicated and time consuming.Very often, when i see stock footage at h264, color corrected and the lot, and i see a screenshot at 1080p, i literally cringe.

So, since noone really will do tests for various types of shots to determine the best level of compression for any given shot,professionals tend to go with proven high bitrate wavelet encoding such as proses,dnxhd,cineform etc.Better yet if these are implemented in camera.

I have seen footage fom my C100 and i can really see where the avchd codec falls apart (and it does a LOT despite its excellent implementation) .By exporting as prores the only thing i accomplice is achieving some sort of damage control,and this is highly regarded in a professional environment.I am literally counting the days to receive my atomos to record on a serious codec for that matter.

PS.For example look at pond5's compression at this clip of mine to see the huge amount of artifacts present,and then open a screenshot to see the source (avchd->prores hq) .You can still see some artifacts and these were present and maybe amplified a bit by the grading but no further compression artifacts like in the preview clips.

https://www.pond5.com/stock-footage/41133683/silhouette-man-fishing-sunset-3.html

great post gcrook! i wish someone would make a DVD or online tutorial that would sort all this out -- if it is possible. Something like, "The Simplified Guide to Formats and Codecs for Selling Stock Footage". Hell, I would pay $100 for that. Instead, I have spent over 100 hours researching this in order to determine my work flow and still not sure about it.

I bought into the higher bitrate thing with a Atomos Ninja, later became a Sound Devices PIX220. You have to quadruple the bitrate (and file size) to even approach the quality of super lossy 25 MB/s HDV and 28 MB/s AVCHD, and you don't really see all that much more detail, except, as I noted, sensor noise. Yep, disappointing in most cases - as you so adroitly point out, gcrook, it is subject dependent....in some cases you'll see great improvement, in others, it's just not worth the cumbersome external recorder, extra battery headache, and higher file sizes. I've come to believe that unless you are shooting true RAW, it is the in-camera processing that means the most, and if it is a higher end camera, the results you will find in many situations are much better than a lower-end camera shooting the same resolution with the same codec.

FS700 shooting AVCHD results better than RX100m2 shooting AVCHD

Not yet tested: RX100m3 shooting 50MB/s XAVC-S vs RX100m2 shooting AVCHD

Noted: RX100m3 seems a lot better, but in driving POV shots, with lots of colorful autumnal birch trees whizzing past, the XAVC-S is challenged. Static shots reveal it's almost as good as the AX100 for fall colors, but lacks the crispness (which is both good and bad, depending)

I really don't know about all this stuff. These are just my thoughts as I try to make some headway in the quest for sustainable quality improvements, and also try to make my expenditures make more sense, for longer.

Thanks gcrrok, it's kind of like I thought. There are better cameras with better sensors that require less post. There are clients who have higher requirements that are willing to pay more. You can shoot more and upload more if you slack off on the tech side a bit. Take your choice.

Yes it's all a matter of time and effort.I mostly produce timelapses,and having a huge files workflow doesnt have the same impact as it would have if i would upload 1000 clips in 5-6 months.H264 and all these interframe codecs can be pretty good actually if the camera operator has done his homework and knows how to expose properly and grade accordingly.

I've seen pjpeg is fine most of the time but sometimes when you have banding even h264 does better so in these cases i use h264 or prores...

I propose an experiment:

Let's take several good clips.
Save them in H.264 and PJPEG.
Let several honorable members of this forum upload these files, with the same metadata.
Some will upload only H.264 and some only PJPEG.
You can get a clip only if you contribute a clip.
People with two accounts on Pond5 can exchange with themselves.
After one year we will look into the number of sales in each format.
This will settle the debate.
I contribute three clips, which I will upload in H.264. Who wants to upload them in PJPEG?
Each seller will get the profit from the sales in his account.

In any way i don't understand this discussion!
I remember when starting here 2010 only H.264 codec in .mov was allowed by P5!
In between without P5 saw the helpful need to inform us by email newsletter things changed and we can upload also mts and mp4. My camera AX100 shoots 4k in mp4 with around ~60.000 kbps. As my 64bit software doesn't provide .mov export (which works only with 32bit render software) i have only the possibilities to render 4k in .avi or .mp4, so i render in mp4 with best settings of around ~70.000 kbps.
Is there anything wrong with it? Don't know what is PJPEG, i have no choise for this!

Beckhusen:
Nothing wrong but it's well known that pjpeg is more editable codec than h264,so potentially sells better..
I don't know which software you use to edit but usually istalling quicktime player on your system gives you the possibility to export pjpeg from your nle even if it windows 32or 64 bit doesn't 't matter...