@itimjim I used pretty much standard workflow for both conversions... everything was set to max in Premier. My system's pretty beefy, so I leave everything on max all the time.
Also, keep in mind that neon reds are the absolute worst-of-the worst scenario for 8-bit 4:2:0 color space. I've never seen them not look bad in 4:2:0 camera. I've also never seen them reconstructed so well!
Most footage will not look this bad... the neon red was just a good extreme test I wanted to try.
If there are suggestions that might lead to better results with Premiere, let me know... I'll try them.
Again, thanks for the test, it shows a lot. I'm sure the doubters will have an answer for you, but it seems obvious to me. You only need to compare the native footage to transcoded to see the difference. The Premiere Pro alleged 4:4:4 destroys it.
@bhwitz in Premiere? the true test is the final render. Render out the 5DtoRGB clip and the Premiere MTS clip to identical codecs and then see. It may be a display issue in Premiere.
As I've said, I love 5DtoRGB, and would rather work on footage in ProRes or DNxHD as they are more stable and robust, but I've seen little discernable difference in AE between 5DtoRGB flat gamma transcodes and the raw MTS files. And I've looked pretty hard, at not only the visible differences but on the scopes as well. Both hold up very well to grading in AE. BUT in a FCP pipeline 5DtoRGB is the clear choice.
@shian Yes, this is a final render to Prores 4:2:2HQ.
And you're right, usually the differences are small... but little neon red needles are like the death chart for destroying red detail in 4:2:0 footage. So this is an extreme example, but also the easiest to test.
@itimjim "You only need to compare the native footage to transcoded to see the difference."
Yes, exactly. You don't even need to look at the Premiere stuff. You can see the chroma reconstruction from the native footage to the 5DtoRGB sample.
@bwhitz Wow... I guess that's an even better argument for finishing in AE.
One thing for sure, we all don't know one another to a point of understanding each person's thoughts on their work flow and understanding of camera and post work.
We certainly should respect each others thoughts. Having said that from what I've seen and heard about, @shian definately has defined his work flow well.
I loved the line "No, I should definitely move back to Phoenix, and drive a cab for a living, I'm clearly in over my head"
LOL!
I live in Phoenix Arizona and have been to the airport many times!
taxi....taxi....
@rajamalik Sedna is also all i-frames. All driftwood "high-end" 24H patches are GOP1 (all i-frames)
Thanks @bwhitz As you can imagine, we all hate fucking blocky reds... damn 4:2:0 LOL! BTW, great test with differences that are easy to see.
@proaudio4 I lived there for 13 years. Worked on my first feature and Directed my first 2 short films there, until John Lowry asked me to come out to Hollyweird and help him restore the color, and pull all the scratches and dirt out of some old film called "Star Wars", or something. I wuz like... "Sure, whatevs, bro."
Did you use cs5? I found similar jagged reds in cs5, much improved in cs5.5..
Guys please let's stick to the plan and ONLY discuss Sedna, Mysteron on this topic etc and no 5dtorgb and what is this or that.
Today i shot the party scene from the project i'm busy with. sedna hbr. I loved my footage , ty driftwood
@RRRR Yes, I used cs5. It would be cool if they fixed it in 5.5. I look forward to trying it when I upgrade!
"Guys please let's stick to the plan and ONLY discuss Sedna"
Well, I did shoot the footage with Senda. And I think the way that different tools interpret the footage is pretty important, since we're working with high-quality settings that people have spent a tremendous amount of time developing. I think we all want the best possible work-flow for these patches...
@bwhitz if you care to share the .mts from above I can make a cs5.5 export for comparison!
I agree with @bwhitz on using 5dtoRGB, and I give him a additional +1 because it appears he is a fellow Z owner, now lets please move all the 5dtoRGB talk to the correct thread.
I also went ahead and took the liberty of attaching 2 screen shots straight out of After effects CS5, same clip, just a little off on the time line. The first is straight import of the .MTS, the second is Prores4444. Which one do you think holds up better to color correction?
@Reckless Hmm.... let me guess...., the one on right? ;) LOL!
@shian +1 Great work.. to say the least!
Haha, yes I just went ahead and pushed it all the way, 100sat 1000sharp but in the end it is those large blocks that fall apart first when you push the colors. To bring it back around to the correct thread discussion I would like to add that it seems to me Sedna A and Mysteron both behave very similarly in post and are close in detail as well, Sedna seems to have just a little more leeway when coloring but not much.
How do I know if I succesfully hacked my GF2?
@Reckless can you please post that in the 5DtoRGB thread, and then CC onionhead on it, so he can angrily tell everyone how its an outrageous April Fool's Joke.... :)
@Reckless "because it appears he is a fellow Z owner"
haha... that I am. Nice catch. I was wondering if anyone would... :)
@klem007 great detail mate.
Sedna A AQ1 has really made noticeable quality in detail, especially as it rolls into the blacks.
@proaudio4 I am sorry I couldn't understand that. Could you re-phrase?
:-) I suggested that last week and onion jumped all over me. I really miss how the forum used to be - it was the place to go for direct info without all the fluff. Now you've gotta wade through a mess of off-topic posts to find the info you want. Starting to feel like dvxuser. Please, everyone post in the appropriate threads so we can keep things organized. Much appreciated.
It looks like you're new here. If you want to get involved, click one of these buttons!