GH3 comparing gh3 videos with ac90?

keinan

Member
Hi folks!
I have curiosity to compare the quality of videos from new gh3 (all-i/avchd) with the ag-ac90 (avchd).
I suppose that ac90 have a robust codec for recording and better quality overall but im not sure.
Anyone have any video comparison or link for clear my doubt?
Thanks in advanced
 
Hi folks!
I have curiosity to compare the quality of videos from new gh3 (all-i/avchd) with the ag-ac90 (avchd).
I suppose that ac90 have a robust codec for recording and better quality overall but im not sure.
Anyone have any video comparison or link for clear my doubt?
Thanks in advanced

The AC90 is limited to 28Mbps AVCHD, where the GH3 shoots up to 72Mbps (MOV). You might say, the GH3 50/72Mbps codec is much more "robust" as it can withstand more intense editing.

Also, broadcast quality is usually considered 50Mbps or higher. 28Mbps AVCHD does not qualify.

Good luck.
 
The AC90 is limited to 28Mbps AVCHD, where the GH3 shoots up to 72Mbps (MOV). You might say, the GH3 50/72Mbps codec is much more "robust" as it can withstand more intense editing.

Also, broadcast quality is usually considered 50Mbps or higher. 28Mbps AVCHD does not qualify.

Good luck.

You have to be a bit careful with a statement like that. No two encoders are a like and bitrates do not always give the same look based on numbers alone. There are some instances of a good hardware encoded 28mbps looking better then much higher software encoded bitrate levels.

Not saying the GH3 will not look better just that you cannot just assume it does because the bitrate is higher. The 72mbps mode in fact does not look better since it is intra frame encoded and can have a lot of noise and artifacts compared to a solid hardware encoded 28mbps.

If I can recall correctly the GH cameras use a software encoder which is why they are easy to hack. Video cameras I believe typically use a hardware encoder. Not that it makes them automatically better but something to think about.
 
Also, broadcast quality is usually considered 50Mbps or higher. 28Mbps AVCHD does not qualify.
That's an oversimplification. The 50mbps you're talking about has another very important aspect -- it requires 4:2:2. So a GH2 or GH3 at 10,000mbps still wouldn't qualify as "broadcast quality" under that standard.
 
Correct me if im wrong, isn't that a BBC regulation? I have seen lots of video here in the states that were nowhere near the 50mbps, or 4:2:2 color space. I remember reading some of the reality shows use dslr's (canon) and before the 5d3.
 
Outside of Europe the only thing people care about is "Does it look good?". Lots of great TV content has been shot at 35mb Sony EX codec.

By the time its compressed and distributed to your TV is is a fraction of its original bit depth anyway.
 
Correct me if im wrong, isn't that a BBC regulation?
The BBC specifies that long-GoP acquisition has to be at a minimum of 50mbps and 4:2:2 color sampling, yes.

I have seen lots of video here in the states that were nowhere near the 50mbps, or 4:2:2 color space. I remember reading some of the reality shows use dslr's (canon) and before the 5d3.
We all have seen such, yes. The facts are that the BBC, PBS, Discovery, etc all have their standards and specification. Other networks may have higher or lower standards, and some may have no standards whatsoever. Lots of footage on reality shows ends up being from GoPros or the like, because that's what's necessary to get the shot...

The point was -- if you want to reference the 50mbps broadcast requirement, you have to reference the entirety of the broadcast requirement. 50mbps @ 4:2:2. Under that scenario, no DSLR would be acceptable, regardless of the bitrate.

As far as what cameras they find acceptable, that's all up to the broadcaster. The shooter doesn't get to make that call. You can't go out and shoot the next great series on an iPhone and expect them to accept it, arguing "but ... you used GoPro shots in "The Amazing Race", and the iPhone shoots better than a GoPro!" That kind of argument simply won't hold water. If you don't meet Discovery's specifications, you have no right to expect that Discovery will accept your footage. They MAY, but if they do, that's up to them. The safest route of action is to find out what your target network's specifications are, and meet them. That way you're guaranteed your programming won't be bounced due to technicalities.
 
Outside of Europe the only thing people care about is "Does it look good?". Lots of great TV content has been shot at 35mb Sony EX codec.
That's a gross oversimplification... you're essentially saying that nobody in America cares about standards? I'd dare say that there are many broadcasters here who would disagree.

Now, truth be told there's lots of places outside Europe that are still shooting in standard-def, so there is some truth in what RatLab says, but to ignore broadcaster standards on the basis of "well, it looks good" is poor business planning, imo.

By the time its compressed and distributed to your TV is is a fraction of its original bit depth anyway.
Yes, which is EXACTLY why you MUST start with a higher-quality source in the first place! It is specifically because of the compression problems that the BBC ruled that Super16 film was no longer acceptable for HD programming, and why they have their requirements set so high in the first place. The only way for something to survive the compression/broadcast torture chamber is for it to have been of exceptional quality to start with, so that acceptable quality will emerge from the broadcast.
 
I have my suspicions that in the eye of the BBC that 50mbits still isn't good enough but they had little choice since many broadcast formats have gone that direction as opposed to higher bitrate tape formats. Limiting to say DVCPROHD would have made them look like they were favoring only one camera manufacturer which could have caused issues. Personally I consider most mpeg2 formats pretty bad even at 50mbits. That specification typically is referring to mpeg2 and AVCHD at 28mbits can look as good as mpeg2 at 50mbits.

It is also my belief that part of the logic behind the limitation of 50mbits has part to do with the cameras usually fitted with that codec and not so much the codec itself. The codec is of course important to hold details and color conversions but the BBC probably figures a 50mbit camera is going to be a pretty darn good camera to begin with. A GH1 can be hacked to shoot at 100mbits but that in no way puts it near a $20,000.00 professional head. A lot of cheaper cameras recording live via hdmi to uncompressed would still make the BBC cringe. A lot has to do with the quality of the camera itself that usually sits behind that 50mbps.
 
That's a gross oversimplification... you're essentially saying that nobody in America cares about standards? I'd dare say that there are many broadcasters here who would disagree.

Of course its a simplification but it is also the reality out there. If you are selling a great looking show no one is going to turn it down because it wasn't 50mb/s 422. Yes broadcasters have standards but they aren't as rigid as Europe which is why I said "If it looks good they don't care really". Yes many broadcasters would say they have standards but what you see on air might be a different thing. Again it comes to how well it looks and the content more than the tech specs in most cases. Not something I would get hung up on as a producer. DSLRs, GoPros, pencil cams, EX1's, FS700 hi-speed etc make appearances in a lot of broadcast shows and are great to have in the tool box.

I have much less experience with American broadcasters than Canada but I do work for the largest broadcaster in Canada as a producer and have yet to encounter issues with MB/s and colour space unless it is a poor looking material.
 
Of course its a simplification but it is also the reality out there. If you are selling a great looking show no one is going to turn it down because it wasn't 50mb/s 422. Yes broadcasters have standards but they aren't as rigid as Europe which is why I said "If it looks good they don't care really". Yes many broadcasters would say they have standards but what you see on air might be a different thing. Again it comes to how well it looks and the content more than the tech specs in most cases. Not something I would get hung up on as a producer. DSLRs, GoPros, pencil cams, EX1's, FS700 hi-speed etc make appearances in a lot of broadcast shows and are great to have in the tool box.

I have much less experience with American broadcasters than Canada but I do work for the largest broadcaster in Canada as a producer and have yet to encounter issues with MB/s and colour space unless it is a poor looking material.

That's because it is to cold up in Canada too notice the details. Lol
 
You have to be a bit careful with a statement like that. No two encoders are a like and bitrates do not always give the same look based on numbers alone. There are some instances of a good hardware encoded 28mbps looking better then much higher software encoded bitrate levels.

Not saying the GH3 will not look better just that you cannot just assume it does because the bitrate is higher. The 72mbps mode in fact does not look better since it is intra frame encoded and can have a lot of noise and artifacts compared to a solid hardware encoded 28mbps.

If I can recall correctly the GH cameras use a software encoder which is why they are easy to hack. Video cameras I believe typically use a hardware encoder. Not that it makes them automatically better but something to think about.

Then, the ac90 avchd is better by the hardware encoder?
 
Then, the ac90 avchd is better by the hardware encoder?

I don't really know. That was my whole point. We can't just say for sure by reading about the bitrates or if the encoder is software or hardware. You have to actually compare side by side. Even the same bitrate between the GH2 and GH3 can look different just imagine how different encoders between different manufacturers can be.
 
What I've always wondered is, how do they know what your source is if you've converted it to something else?

If I shoot a series with an HMC150 (not approved), instantly take the source footage and convert it to DVCPRO HD with panasonic's old freeware tool, then lie and say it was shot with an HPX170 or HVX200, would they really know any better? Will they examine it at a pixel level or just take your word for it?
 
You may be able to fool them sometimes but these guys are good at telling. They will look for motion artifacts in complex movement to see if it macro blocks. They will be able to tell. Plus it can be fairly easy to tell if something was shot 4:2:0 and converted to 4:2:2 unless you really know how to do it well. Even then I'm sure they can tell.

Hey I used to do the same thing with SD. I would shoot on my Canon XL1 which was of course DV. I would then edit and master to Beta SP for a TV station once a week. Some of them could tell something was slightly off but they could never figure out what and they accepted it anyway. Considering the stuff at this small cable station shot stuff on Beta SP that looked like utter vomit worthy trash it was hard for them to say anything. Did the same thing with two TV spots as well. That was a more mainstream cable station and they fought me so hard on that one but the client in the end said show it because he loved the content compared to the crappy TV spot the station would have shot.

Of course looking back at that material now makes me want to vomit. It is amazing what we considered awesome at one point and now wouldn't show it to our dogs.
 
Its ironic that the Beeb have these standards but only put out around 9mbps for us to actualy see,watching bbc HD channels and switching over to either my GH2 or XA10 looks far sharper.
 
If your content is compelling enough they won't care what it is shot on. If you happen to see the inside of a Tornado with your handy cam then Discovery channel would air it. However, if you shoot the next "idiot" reality show it better at least meet their standards for quality.
 
Back
Top