••• ok so it's MUD free but what good is it if the VBR still remain low? •••

Coco Bermudez

Coco Bermudez

Well-known member
I might be missing something here...so yeah...I have to agree that the idea of MUD free video is awesome...so much that I had to re-purchase a GH-1 after previously selling mine. I did the hack and I can whip that cam and no MUD.

What has me baffled is that in certain scenes shot with AVCHD the Mbps skyrockets to 35Mbps...but in most scenes it stays around 13-19 Mbps (using C Settings)

Am I missing something? It doesn't give me much latitude to color correct at such low Mbps.

Aside from this...I love it.
 
That's the nature of the VARIABLE AVCHD codec. We have been hoping for a minimum bitrate patch to keep the rate on low detail scenes at around 20mbps, but it may not be needed as first thought.

Still lot's of work being done, early days yet.

Y
 
and I don't mean to be dissing the hard work put into the GH13...I love it. I just want to know if there is anything that can be done to disable maybe the variable bit rate? I might be talking out of my arse here....
 
I don't think bit rate has much to do with color correcting. It would seem to me that bit depth would be more pertinent. What is probably even more important is how the codec handles the information in the dark areas. If the codec is designed to sacrifice shadow information because it isn't very "visible", then the codec itself isn't friendly to color correcting and no amount of bit rate enhancement is likely to improve the situation more than a little bit.

You might want to experiment with how well MJPEG tolerates color correction.
 
Apparently, this guy stopped worrying about the Apocalypse, which we all know to be revelation of all kinds of strange inner detail about your footage (what kind of blurry wash out it must be at 10 Mbps anyway?), and is shooting a feature with his GH13! THE END IS NIGH!!!
 
svecher said:
Apparently, this guy stopped worrying about the Apocalypse, which we all know to be revelation of all kinds of strange inner detail about your footage (what kind of blurry wash out it must be at 10 Mbps anyway?), and is shooting a feature with his GH13! THE END IS NIGH!!!
Click to expand...

From what I've seen from the AVCHD, I'd be inclined to shoot with my camera settings as flat as possible, bringing shadows up. I'd expect to record more overall detail this way - especially in the shadows. I'd then crush the shadows as necessary in post to get the look I want. If the shadows are crushed before encoding, I fear that you will never get back the detail that you might want that was once there.

That said, I haven't actually tested this. But this is where I'd start my testing before shooting a feature.

Other than this, I suspect the best answer is to simply get it "right" in camera to begin with so you have very little correcting to do.

BTW, after frequenting these discussions and beginning to look at GH1 footage more closely, I'm coming to the inescapable conclusion that the picture broadcast by DirecTV is essentially unwatchable. I see mud, pulsing and blocky artifacts far worse than those generated by my hacked GH1. ;^)

And if you think this is bad, don't get into the habit of trying to analyze lens bokeh by looking at out of focus lights in movies. If you do, you may find that you've sat through an entire movie and missed the whole thing - but have some interesting insights into the apertures that were being used. ;^O
 
Not to start a war of anything...but after revisiting the GH-1 with the new hack...the GH-1 with a crappy M42 50mm f1.8 blows away my 7d with an L Series lens...same scenario, same subject, same light. Jaw dropping...

I know about getting it right...thats what I aim for...but always clients want to "punch" this and that. Would be nice to have always high numbers rather than variable ones. I guess I could try MJPEG.
 
Super8 said:
Not to start a war of anything...but after revisiting the GH-1 with the new hack...the GH-1 with a crappy M42 50mm f1.8 blows away my 7d with an L Series lens...same scenario, same subject, same light. Jaw dropping...

I know about getting it right...thats what I aim for...but always clients want to "punch" this and that. Would be nice to have always high numbers rather than variable ones. I guess I could try MJPEG.
Click to expand...

Sure. I like the GH1 a lot also. But it uses the AVCHD codec which was originally intended for consumer use. And the GH1 implementation seems to be weaker than what the standard can ultimately support, not stronger. The hack helps by cramming more bandwidth at a weak codec implementation. There's a reason for other formats like AVC-Intra.

And to be clear, I wasn't suggesting that you don't know how to "get it right" in the camera. And I understand how clients can be fickle. But trying to get a $1200 consumer camera to do certain more professional things will present some fundamental problems and limitations.
 
wturber said:
And to be clear, I wasn't suggesting that you don't know how to "get it right" in the camera. And I understand how clients can be fickle. But trying to get a $1200 consumer camera to do certain more professional things will present some fundamental problems and limitations.
Click to expand...

I know you did not suggest that...I was just trying to make the point that no matter what you think looks good...someone always wants to pee on it.

I just wish I could get my GH-1 to behave like my HPX2000...but when lit well the GH-1 puts all the big boys to shame
 
well since the whole point of compression (and especially AVCHD) is to get a smaller file size. You can only do that by lowering bitrate.

AVCHD is pretty darn efficient for what it is. You can pack more information in AVCHD than you can MJPEG for the same amount of space without really "losing" anything. We all know you lose something when you compress but all things being equal, you can/will have lower bitrates in AVCHD but still have comparable picture quality to higher bitrate MJPEG.

To that end, there really isn't any reason to "force" AVCHD to have a higher bitrate on it's lower end. You'll just be encoding garbage/noise IF you could do this. The problem though is that the codec was designed from the ground up to get the lowest bitrate possible while still keeping quality. You'll simply be wasting space for no reason but to make you feel like you are getting better quality.

In other words, the codec uses the bitrate it NEEDS not what you think it should have. Before the hack, the camera was intentionally limiting the bitrate with poor results, MUD, etc. Now that we can lift that upper limit, the codec can go to town and we reap the MUD less benefits.

There is just no reason to do what you ask about.
 
You must log in or register to reply here.
Back
Top