Why isn’t everybody uploading H.264 codec?

daniel -- why do you avoid H264 knowing you can transcode to PR422 (or whatever) without a problem?

First of all many or most "Pro" photographers shoot photos in RAW to avoid loss and then if coded in TIFF it is basically lossless. Try re-saving a JPG many times and watch the artifacts show. So RAW to JPG then to TIFF .... why ... Yes Alamy used to require a 50MB TIFF but then sold a JPG version. This way a high lossless copy was maintained. I just sold a single photo for many thousands of dollars (more than a F-150 Ford truck costs). They bough it as a JPG but asked for the RAW or TIFF if available due to JPG limitations.

It really doesn't seem like this is a discussion that you want to listen to. h.264 has not been a standard and has artifacts and various other problems after saving multiple times ... just like a JPG photo. Feel free re-process the native file and again compress the hell out of your videos. You have heard from a high end buyer who told you that they avoided h.264 as much as possible. The difference between a buyer purchasing a 100mb h.264 or a 500mb PJPEG due to size is negligible. Most internet has a much faster download than upload. I do not believe buyers really care about buying a smaller file size.

Up to this point h.264 and up-coming h-265 has limitations and problems that are more apparent than PJPEG. It appears that you want to use h.264 so go for it. I'm sure for general stock it is plenty good enough. Artist that worry too much about file sizes, storage costs etc etc are really in the wrong business. Trying to save storage is not what most of us worry about. Sell one or two 4K clips and that pays for a 4tb hard drive and a months worth of internet.

Nothing wrong with h.264 or JPEG. No discussion here will change minds of those who have already decided. Good luck.

RAW, which I use extensively when I shoot stills, is a great thing, as is 422, but my measly cameras only shoot h.264 And I shoot with my iPhone (stills) which I then take the images and upload to Getty and make sales from. I've done the same with stills from a GoPro. I don't much like editing those shots, but, hey close enough. It all depends on the market. Taking those shots and up-resing them, like I used to do for other agencies from smallish RAW files, doesn't do anything to them for quality except make them bigger. Transcoding files from the inferior ones they were shot on to better ones doesn't do anything for the clip, except perhaps make it appear nicer for those who like it that way. Digital lipstick is all it is.

@RekindlePhoto exactly my words :-)

One cannot simply say h264 has artifacts. H264 has many compression levels and profiles. It is not a simple codec as many of you often speak of. When one says h264 has artifacts one should mention the level and the profile one is speaking of. Many people are afraid of h264 because they do not know much about it. Worth mentioning is that Adobe products have an export limitation in regard to the level: 4.2(although it might be a Quicktime limitation).

http://en.wikipedia.org/wiki/H.264/MPEG-4_AVC


p.s: you mention h264 has having artifacts and say PJPG is better but we all complain about PJPG behaving very bad with gradients even when you pump it up to 100% quality!

pps: please do not take personally my comment. I know, I am a bit acid without my will(partly because English is not my mother tongue) but I love you all.

Whatever a file from camera has H.264 artifacts, when you save it in PJPEG, the resulting file will inherit all the artifacts of the original file. Saving As PJPEG does not cure artifacts.

Ionescu like you said h264 comes in many many flavours.Too bad that we all cater to a business that doesnt like to make assumptions (and for very good practical reasons) .When dslr's came out and people started shooting video with them like it was the end of the world, the first people that rebelled were apparently the post people,and unfortunately we sell products for post production.

h264 can and will suck as a codec, depending on the occasion,and noone who spends money for footage likes to gamble (unless he/she is desperate). ;)

the Canon 5DM2 shoots H264 codex. And it appears those files sell VERY well. should one transcode to PhotoJPG before uploading? if yes, why?

Also, FCPX and Adobe Lightroom allow non-destructive editing so I don't see a problem with all that.

I am never timid when it comes to challenging the status quo especially when it may be based on antiquated thinking and legacy workflows. But the truth to this issue lies in technical facts regarding the quality of the various codex. Surprisingly, that appears to be a gray area.

...sorting out workflow... still confused... loving debate ;)

"So RAW to JPG then to TIFF .... why ... "

No, I meant JPG to TIFF, no RAW to start with. Or better yet just JPEG. After all, every agency I can think of sells just jPEGs regardless of the format that image came from. No buyer I can think of ever asks, "does this image originate with a RAW". Doesn't happen. Maybe it used to, but not any more. Yes there are some high end users who require much better imagery, but most of them are shooting specific to a job. People know that jpegs are lossy, they know that if the do multiple saves that artifacts are introduced. They know that if they want of avoid this they "transcode" the file to a lossless format like TIFF or PSD or some such thing. They don't sit there and make multiple saves on a jpg file. I listened to years of people arguing about the evils of jpegs only to see the market adopt them in their entirety.

H264 directly out of a HDSLR has artifacts when the information in the footage is too much and the bandwith limit is reached. Back then when I was shooting with an HDSLR I saw that all the time. Try for yourself shoot smoke or water ripples, waves... and see for yourself.

Two years ago I've also made 2 short tutorials on how to get the most out of a HDSLR:


How to get the most out of a HDSLR


Professional workflow

You cant just convert H264 to PJPEG or 422HQ and think you get a better image quality. You have to use a professional workflow "denoise, sharpen, dither, ... and then you get a better image. The usual client won't do this tedious process.

In example many scenes of Shane Hurlbut's Act of Valor were shot on a 5D Mark II. They didn't just convert the H264 to DPX files. They did a very intensive post process to get the most out of their files for the big screen.

Nowadays I shoot directly with high bitrate on C300, BlackMagic or sometimes with 5D Mark III Magic Lantern RAW and I don't have to do this 8Bit to 10Bit process again because the source material is already in 10Bit.