GH2 More important: The lens, or the hack?

[mpgxsvcd]

For shooting moderately-detailed static scenes, the GH2's unhacked 24Mbps bitrate in 24H video mode works quite well. Where the limited bitrate of the stock AVCHD encoder becomes painfully visible is when shooting highly-detailed, fast moving subjects, for example, a shaded stream of running water:

The real difference in shooting something like this isn’t that the subject is moving. It is that the entire scene is moving. In fact it is moving in a chaotic motion. That absolutely wreaks havoc on a codec trying to do compression. Noise does essentially the same thing.

The good news is that shooting an entire scene that is moving chaotically rarely ever happens in real world shooting. Sure if you are shooting Niagra falls or a flock of birds you need higher bit rates like this. However, if you are simply shooting an athlete doing fast action without aggressively panning it won’t tax the compressor that much.

The problem is that most people show a scene like this and then equate it to a scene where the subject is in motion but the surroundings are static. They are two totally different things.

 

[Gary Huff]

mpgxvcd, since your Vimeo page is filled with video after video of mostly static camera tests, I'm not sure what kind of shooting situation you're using the GH2 hacks (and are you even using the newer stuff like Sedna?) in that allow you to make these statements.

I have shot both hacked and unhacked with the GH2 in actual shooting situations, and the hack does make a ton of difference, especially in resolving finer detail.

Oh, I also watch everything I do on a calibrated 1080p monitor in full screen. Can you say the same?

 

[mpgxsvcd]

mpgxvcd, since your Vimeo page is filled with video after video of mostly static camera tests, I'm not sure what kind of shooting situation you're using the GH2 hacks (and are you even using the newer stuff like Sedna?) in that allow you to make these statements.

I have shot both hacked and unhacked with the GH2 in actual shooting situations, and the hack does make a ton of difference, especially in resolving finer detail.

Oh, I also watch everything I do on a calibrated 1080p monitor in full screen. Can you say the same?

Yes I use the Vimeo page almost solely to show the results of the many tests I have run through. I do those tests because so many other people would rather spew on about how pretty the noise is with high bit rate hacks than actually do a test case to show it.

I pay for the upgraded Vimeo page so that anyone can download the original files and see the evidence for themselves. So many other people are unwilling to do that.

I use my youtube page for the actual content I shoot. None of it is shot for anyone else. I just shoot things for myself and if anyone else is interested in the subject then they can watch it as well. However, most of it does involve the subject moving but the overall scene remaining static.

I would much rather reduce the noise by using a fast lens than to have it accurately represented with a slow lens. That may just be me. However, I really don’t think noise looks good in any shot whether it is like a “fine wine” or sour grapes.

I have said it before but I feel like people are still missing the point. The hack is great for a lot of things. However, it simply is not necessary for every shot.

Yes I watch the content on a calibrated 55" 1080 screen. Not all of my screens are calibrated. However, when I need to do critical viewing I use the large calibrated 1080 screen.

 

[Gary Huff]

Yes I watch the content on a calibrated 55" 1080 screen. Not all of my screens are calibrated. However, when I need to do critical viewing I use the large calibrated 1080 screen.

Your TV is not a calibrated monitor. And how are you playing back GH2 hack clips on your TV? That could also make a difference.

EDIT: And I should add the video evidence you submitted is a YouTube video that doesn't appear to have a counterpart on Vimeo, so it's rather pointless to see exactly what's going on, don't you think?

 

[mpgxsvcd]

Your TV is not a calibrated monitor. And how are you playing back GH2 hack clips on your TV? That could also make a difference.

EDIT: And I should add the video evidence you submitted is a YouTube video that doesn't appear to have a counterpart on Vimeo, so it's rather pointless to see exactly what's going on, don't you think?

My TV is calibrated when it is hooked up to a computer and playing the original AVC-HD files off of the hard drive. I calibrated using the H.264 DVE Blu-Ray in the same program that I play the GH2 videos with.

If you calibrated your display using anything but the program you play the GH2 files with then it is likely that it is not actually calibrated for those files.

You obviously did not read the description for the youtube video. It has a link to my Vimeo account that house the originals.
https://vimeo.com/user442745/videos

 

[Gary Huff]

My TV is calibrated when it is hooked up to a computer and playing the original AVC-HD files off of the hard drive. I calibrated using the H.264 DVE Blu-Ray in the same program that I play the GH2 videos with. If you calibrated your display using anything but the program you play the GH2 files with then it is likely that it is not actually calibrated for those files.

You're not exactly inspiring confidence that you can legitimately make statements concerning the quality of the AVCHD encoder hacks outside of personal opinion.

You obviously did not read the description for the youtube video. It has a link to my Vimeo account that house the originals.
https://vimeo.com/user442745/videos

You link to the video page, of which there is a lot of videos all of the same subjects. You need to link the specific video that you think demonstrates that the GH2 encoder patches do not help resolution in fine detail areas.

 

[mpgxsvcd]

You're not exactly inspiring confidence that you can legitimately make statements concerning the quality of the AVCHD encoder hacks outside of personal opinion.



You link to the video page, of which there is a lot of videos all of the same subjects. You need to link the specific video that you think demonstrates that the GH2 encoder patches do not help resolution in fine detail areas.

You have done absolutely nothing to back any of your claims up. A TV is nothing more than a monitor when hooked to a computer. If the output of the computer and the setup of the TV are properly calibrated then it is no different than using a smaller monitor instead.

Really you are starting to sound like you believe your last name qualifies you to be an expert without actually providing any sound advice.

The videos you requested are available at that link. If you are too lazy to click on them then there really isn't anything I can do about that.

 

[Kholi]

You have done absolutely nothing to back any of your claims up. A TV is nothing more than a monitor when hooked to a computer. If the output of the computer and the setup of the TV are properly calibrated then it is no different than using a smaller monitor instead.

A Television is not a calibrated monitor, and must be calibrated by a professional for each source it's being fed. Simply connecting a monitor to a computer does not mean it's calibrated. Calibration itself is an employable and tedious task.

There's a good reason why a Dreamcolor costs more than I suspect most of our televisions cost. And, there's a good reason why an engineer costs 150/hour to come out and calibrate plasmas, etc.

If you were running your monitor through an AJA card HD-SDI out into a Plasma or LCD with an HD-SDI interface, likely the things you see would change. That's how most of us in post houses work.

The videos you requested are available at that link. If you are too lazy to click on them then there really isn't anything I can do about that.

But, I think Ian and Gary have a very valid point.

What do you shoot?

As filmmakers, we often comment from a very narrow-minded place which is born from "things that we do". You made the statement that compression won't ruin your shot. Well, that's quite a statement. I wonder what the creators of several external recording devices to bypass sub par compression would think about that? Or the industry employees that spend time and money employing them to make sure that compression artifacts aren't something that they have to deal with?

Believe it or not, most working professionals that understand digital image acquisition and delivery do everything they can to avoid compression artifacts. Why? Because they feel that they ruin footage. You start with the very ground floor and work your way through the chain, which means start with eliminating them at the camera level.

Maybe you aren't doing anything where it actually matters, but I am.


As far as Resolution and Detail goes, no, technically the hack does not increase resolution or anything like that. However, it's pretty clear that it increases the amount of detail that remains through compression. That's pretty obvious, and that's why you use it or an external recorder, so on and so forth.

Whether that has value or not to one person is that person's deal.

 

[mpgxsvcd]

A Television is not a calibrated monitor, and must be calibrated by a professional for each source it's being fed. Simply connecting a monitor to a computer does not mean it's calibrated. Calibration itself is an employable and tedious task.

There's a good reason why a Dreamcolor costs more than I suspect most of our televisions cost. And, there's a good reason why an engineer costs 150/hour to come out and calibrate plasmas, etc.

Look I think the problem here is that you both don't really read what others say. You just jump to conclusions. I said I calibrated both the computer output and the TV for the actual program that I use to play the videos. I spent years studying calibration for my specific TV and while I am not a certified calibration technician. I know how to calibrate my specific display after spending all of those years learning from people who are.

If you are doubting the validity of the DVE Blu-ray then that is one thing. It isn't perfect but it was the best calibration disk I have access to. If you have better reasonably priced calibration material then please let me know about it.

Yes I agree we should always try to avoid compression. If the GH2 could shoot full uncompressed RAW we wouldn't have ever had this conversation. However, the fact is that it can't do that and in order to get it to shoot even high bit rate compressed video you have to introduce some other limitations.

I would love it if the GH2 could shoot 180 mb/sec GOP = 1 without ever failing. Then if you had the space to store it there wouldn't be any reason to use anything else. The fact is that there are still write errors and specific scenarios that those settings will not work for. There are far more circumstances where not getting an error is far more important than reducing compression artifacts to zero.

Let's see some real world examples where a moderate bit rate or even the stock 24 mb/sec bit rate is simply not sufficient for. All I shoot are real world scenarios. I don't shoot staged scenes professionally for other people. I understand that in some of those scenarios compression artifacts are not acceptable. However, write errors would not be acceptable there either. You have to balance between the two.

I think we are starting to get off of the original topic though. I don't really see any argument as to why the hack would supersede the lens choice. The hack simply does nothing to affect what the right choice of lenses would do. I can overcome the limits of non hacked footage for almost all scenarios. For those scenarios where the hack is required they simply do not occur that often in real world scenarios. If I had to shoot those scenes then I would just suffer through the compression before I would use the wrong lens.

Are you guys saying that you think the hack is more important than the lens choice?

 

[DBP]

To me, the lens is more important because the GH2s stock 24mbps is 'good enough' in most cases.

I think if the bitrate were capped at 5mbps, this would be a non issue. It'd be the hack all the way. GH2 + Voigtlander @ 5mbs VS GH2 + 14.42 kit @ 150mbps. Which would you choose then? (sufficient lighting for each)

At some point, the bit rate gets lowered enough that the quality loss is blatant and unacceptable, beyond any difference a lens can make. Where that ends up being is at least a bit subjective to individual tolerance of compression artifacts.

 

[Gary Huff]

I would love it if the GH2 could shoot 180 mb/sec GOP = 1 without ever failing. Then if you had the space to store it there wouldn't be any reason to use anything else. The fact is that there are still write errors and specific scenarios that those settings will not work for. There are far more circumstances where not getting an error is far more important than reducing compression artifacts to zero.

Funny, I shot a 13-minute short film in January and two music videos since then, pretty much all close to that 180Mbps scenario you mention (like 147-ish, but what's 40Mbps amongst friends?) and have had zero issues with write errors. Have you even shot with such patches as Quantum X or Sedna?

I think we are starting to get off of the original topic though. I don't really see any argument as to why the hack would supersede the lens choice. The hack simply does nothing to affect what the right choice of lenses would do.

You are going to see the biggest difference, right off the bat, with a quality encoder patch, moreso than any lens. I don't even understand why this is an issue, since it's not a question of one or the other. You can easily have both.

 

[mpgxsvcd]

Funny, I shot a 13-minute short film in January and two music videos since then, pretty much all close to that 180Mbps scenario you mention (like 147-ish, but what's 40Mbps amongst friends?) and have had zero issues with write errors. Have you even shot with such patches as Quantum X or Sedna?



You are going to see the biggest difference, right off the bat, with a quality encoder patch, moreso than any lens. I don't even understand why this is an issue, since it's not a question of one or the other. You can easily have both.

Yes I have shot with both of those and I get write errors when using m4/3s lenses with AFC and ETC. I understand most people shoot with MF glass and those settings work fine for those scenarios. However, I have not encountered any high bit rate settings that work 100% of the time for all resolutions with all lenses.

I am always open to suggestions though. I haven't tried all of the versions of the settings. If you know of any high bit rate settings that will work with all scenarios then I will definitely try them.

I don't understand why the question was asked either since you can have both. However, that is the question we are responding to. I simply don't agree that the quality encoder patch is more beneficial.

 

[Gary Huff]

It seems like your poor personal experience is leading you to a personal opinion that the encoder patches are not worth the effort. And that is absolutely fine. However, to state unequivocally that those patches don't make that much of a difference is a statement that you have not shown you are qualified to make in a professional sense, and is not borne out by others experiences who have actually used these patched cameras on paying jobs for clients.

Again, outside of the typical bugs that are well-documented and affect hacked and non-hacked users alike, I have encountered no problems. But that's probably because I purchase higher-end SD cards since this is what I do for a living.

 

[mpgxsvcd]

It seems like your poor personal experience is leading you to a personal opinion that the encoder patches are not worth the effort. And that is absolutely fine. However, to state unequivocally that those patches don't make that much of a difference is a statement that you have not shown you are qualified to make in a professional sense, and is not borne out by others experiences who have actually used these patched cameras on paying jobs for clients.

Again, outside of the typical bugs that are well-documented and affect hacked and non-hacked users alike, I have encountered no problems. But that's probably because I purchase higher-end SD cards since this is what I do for a living.

I have never said they are not worth the effort. I do believe that lens choice is more important but they do have benefit. The problem is that some people show extreme scenarios and say see look what happens if you don't use this. What they fail to mention is that it fixes one thing but causes so many other issues.

Can you point me to specific settings that will work with any lens? I think that is what this whole thread is about. If you have the choice to use lower bit rate settings that work with any lens or use higher bit rate settings with only a specific lens that it works with which one would you choose.

If you can provide high bit rate settings that work with any lens then this whole question won't matter.

 

[Gary Huff]

The problem is that some people show extreme scenarios and say see look what happens if you don't use this.

Can you be more specific?

What they fail to mention is that it fixes one thing but causes so many other issues.

Only if you are not prepared. If you want to shoot 140Mbps on your $20 Class 6 card, well, you're out of luck.

Can you point me to specific settings that will work with any lens?

I have used Quantum X, Sedna AQ1, and Canis Major (just throw-away tests so far), on Nikon F, EOS, and Lumix G glass (both ETC and non) and have never found any of the hacks to be unstable on the recommended Sandisk 64GB 95MBps SD card. I just used Sedna AQ1 on a long-form shoot using Nikon and Lumix G glass and 24L Sedna AQ1 on the Sandisk and then a Transcend 32GB Class 10 card and a Patriot LX 32GB Class 10 card and clips spanned without error. These involved takes of 30-40 minutes in length.

 

[Lpowell]

LPowell’s example does show a good use of the high bit rates. If I am shooting flowing water I would definitely consider those settings to be necessary for a good shot. However, I can’t recall ever wanting to shoot water like that.

Ever notice what happens to a leafy tree in a stiff breeze? A flock of birds startled by a dog on a beach? A stadium crowd rising to cheer a touchdown? Some of us take the GH2 outside the studio on occasion...

If you can provide high bit rate settings that work with any lens then this whole question won't matter.

Yes, any lens, in all video modes and options. Stay tuned, reliability testing is almost complete...

 

[mpgxsvcd]

Ever notice what happens to a leafy tree in a stiff breeze? A flock of birds startled by a dog on a beach? A stadium crowd rising to cheer a touchdown? Some of us take the GH2 outside the studio on occasion...


Yes, any lens, in all video modes and options. Stay tuned, reliability testing is almost complete...

Is that the Flow Motion Patch in your sig? I will try that one now as well.

 

[mpgxsvcd]

Only if you are not prepared. If you want to shoot 140Mbps on your $20 Class 6 card, well, you're out of luck.

No I don't expect it to work with a class 6 card. I would expect it to work with a good 30 MB/sec class 10 card though. I simply can't afford to pay $160 per card and have more than one of them. I know I am not alone in that respect.

 

[stefancolson]

Dan i suspect is too much of a gentleman to respond to your comments,
Regarding Pnasonic lenses i have the 14-42 as well as the 14-140 which you say is better than the EVIL 14-42 and is nonsense.
Mine at the same settings f stop/ shutter/ picture settings give identical video,as do my other Pana lenses and olympus lenses,it is in lower light that i can see differences.

I don't see how I did anything other than praise Dan. I said that he obviously has talent and skill, only that I though the lens he was using was the weak link in the chain. If lauding somebody for their ability but criticizing an inanimate object that they happen to own can be considered a personal attack then I don't know what f'ed up alternate reality I've stumbled into. I had nothing but positive things to say about Dan and his work, and I don't make a habit of bad mouthing people in public forums, only crummy lenses.

I have a 14-42. I have compared it extensively to other lenses (from Panasonic and other manufacturers) in both controlled and real-world scenarios. I do not like it. At all. This is my opinion (which the OP asked of all of us). Deal with it.

 

[Gary Huff]

I simply can't afford to pay $160 per card and have more than one of them. I know I am not alone in that respect.

Then you can't expect to get reliable 180Mbps recording when you don't want to pay the price to get there.