Personal View site logo
Shotgun mics directly to the GH2 with the 448kbps hack. Choices and Advise
  • Hi everyone,

    I accidentally started a conversation on the Driftwood patches thread on audio straight into the GH2 with the 448kbps audio patch.

    I am a documentary photographer and recently joined the PV forum after deciding that a GH2 is the tool to have for my work.

    I had two queries regarding spanning and in-camera playback and they both have been answered in the Driftwood patch thread: 64GB SD Sandisk Extreme Pro 95mb/s is the only card that spans with Moon T7 reliably and the 448kbps audio patch will make any clip non playable in camera.

    Now, to the reason of this thread.

    My needs are to simplify the gear I carry with me. I used to work with Rode Videomics plugged to my 5DII and that was that. The audio from them wasn't great.

    Now, with the 448kpbs hack, the GH2 can be a great audio recorder without the need for an external recorder like an Edirol or a Zoom.

    The conversation as left in the Driftwood patch forum relates to these two tests I found for shotgun mics (most of which can be plugged straight to the GH2) which are outdoors and at two distances.

    The distance from where I work and from where the mics would be to my subjects is in between those two tests, both indoors and outdoors.

    So far, through my DT770 32ohm which is fed by a Dragonfly, the MKE600 is a bit bassy but very good and the AT897 and the MKH 416 perform similarly at a distance. Close up, the differences are far more obvious.

    Also, there has been a suggestion of using an ISK CM 60 which is battery powered:

    However, I am unable to find a sample from it. @oto02 said he is getting one and getting ready to test it.

    Also, @Jim_Simon mentions that the GH2 can record really good audio so I have hopes that I can work out this rather simple rig for documentary work (TV, festivals)

    Many thanks for all contributions in advance.

    EDIT: I just heard the tests again with my Sony V6 and still, the AT and the 416 sound roughly similar on both. I am not sure if the MKE 600 equalized would be similar to the 416 raw or if one can up the bass a bit on the AT and make it sound less flat. Both, At and 416, sound crisp.

  • 20 Replies sorted by
  • @hexagonal, where did you see the suggestion to use that pencil mic CM-60? I thought that looked like an interesting option...

    As for the shotgun tests you posted, I agree completely with your assessments of the mic results in there, but the problem is that this is a pitiful use of the mics, since they are pointing not only at the speaker, but also at noise sources behind him. If they would have boomed that set of mics this test would have been really helpful for me. I think the AT897 sounds really good in the test also, but a lot of people complain about it.

  • The GH2 is not a professional quality audio recorder even with the "Pasadena" audio hack, which I consider dubious and unnecessary, and a possible cause of instability. Even with the audio hack the GH2's internal audio recording is still compressed AC3 with always-on AGC, fine for home videos but not fine for pro work, which should always be linear PCM for recording and mixing, and remain uncompressed until final encoding for distribution. I always delete the audio hack from any PTool file I update my GH2 with as a matter of course.

    Yes, the GH2's internal audio is better than the wretched audio side of the various Canon DSLRs, but that doesn't make it great, just better than awful. You're always better off recording external audio, whether with the recorder's onboard mics or a different mic plugged into the recorder. Learn to sync audio in post like a real pro, because shortcuts on the audio side shout AMATEUR even louder than white sky.

  • @hexagonal

    I understand your situation but the reality is that you need different tools for different jobs. I have shotgun microphones, cardiod microphones, lavaliere's and a cheap $200.00 studio microphone for vocals. The hard part about tests on the internet in general is that there are so many variables that come into play. Sometimes people are pushing something with an ulterior motive of pushing a product by a manufacturer etc etc. The other problem is there testing methodology that they used might vary so in turn you will get different results.

    Most of the stuff that you are planning on shooting should determine what microphone would be better suited to your needs. However there is no one size fits all when it comes to microphones. Different tools for different jobs or needs. I saw a video recently were this guy shot a video for a DJ company, ironically the audio sucked big time and although the video was decent the audio just sucked ass and made the video seem really cheap.

    So I would suggest if you don't have the money to buy what you need right now rent it! I know it sucks to rent but at least the audio will sound good instead of trying to make a microphone work in an environment that it was not designed for. I also agree with Shaveblog in terms of recording audio in camera, you need an external audio recorder to have good sound.

  • As a dubbing mixer and previously boom swinger and recordist on film sets for years, Id suggest A: hiring a professional, as he'll lift your pictures far more than pixels or a grade will ever get you and B: speak to the guy mixing your audio and ask what he wants - saving 10 bucks on a mic a day may reduce your products overall perception by a HUGE percentage more - also use radios as last resort and C: do A: - did a time spent analysis for Warner Bros recently comparing time spent trying to roll the turd in glitter v just getting it right in the first place - was astounding the post $ they spent would have covered a tooled up pro on set for the whole shoot in the half day I spent trying to polish student film sound bodged by mate of yours or novice cameraman. You wouldn't expect your camera chap to vaguely get focus with a lens that was sort of O.K. so why would you expect your sound to be any different. Sound is > 50% of the picture - shit pics good sound = art, great pics shit sound = amateur. Would you skimp on lenses - reading VK's forum fkin ell no lol - you lot spend days twatting about with pixels, so why skimp on the most sensory supportive part of your project.Chris Nolan used to come to CTV to blag sound dubs with Dan Gable for his early forays into his films, knowing sound would polish his camcorder turds so ... I blame the lecturers (who didn't) forwarding the old meme of sound don't matter... they didn't do you ?

  • @hexagonal

    the 448kbps audio patch will make any clip non playable in camera.

    Running the GH2's Dolby AC3 encoder at 448kbps will also flagrantly waste space on your SD card, but it's your bandwidth to allocate however you please. The technical reason why it's overkill, however, is because the GH2 can record only two tracks of audio at once. Having worked professionally on the AC3 audio codec, I don't advise boosting the bitrate at all. The 448kbs bitrate is designed for recording 5.1-channel surround sound, not for the 2-channel recordings of the GH2. Decades of Mushra tests have shown that it's perceptually useless to push the AC3 audio encoder's 2-channel bitrate beyond its stock setting.

  • Agree with @LPowell . If you look around we had discussion about it, I think in audio hack topic.

  • Thanks everybody for the words of wisdom. I will try and address a few things here that are mentioned above.

    First @lpowell and @vitaly_kiselev , thanks for that clarification. I wasn't completely aware of the reasons for the 448kbps design other than to "improve audio input". How about the variations in the volume levels seen below? Would it be worth tweaking those to improve audio intake? I see @thepalalias addresses the Rode Videomic Pro which I don't like but its a mic that its often used. I wonder then whether the tweaking of those AGC settings can improve the use of the shotgun mics I mention.


    Yes, horses for courses. I understand. Directional is what I need and it has to be mounted on the camera. I can't use lavaliers (I will explain below) and ideally I would want to just record straight into the GH2. I can say that if I was to rent, I might as well sell gear to buy gear but I can't do that either. That money doesn't exist because I don't have it. The project I am working requires me to do an hr train commute (from one city to another) 2-4 times a week for the next 5 to 6 months. That alone would be £300-500 a month. I am as independent as it gets because the work I am doing doesn't want to be touched by NGOs or charities. I even had to sell gear to afford doing the commute besides the little work I get every month to pay for living.


    For the sheer reason that I need to work with people that are very vulnerable and many of my "shifts" go from 8am to midnight, for the time I specify above, I can't hire a professional. It is not only the cash...but also the way I work. I, literally, live with the people I work and their situation is not one where I can have a second individual there. I know that without details it sounds superficial or maybe arrogant even, but I need to maximise with minimal equipment. After November, when my lease is up where I am living now, I may be able to move places to this other city and reduce my living expenses, but that alone is a few months away and I starting to work now.

    I am sure, and I agree, that using an audio professional is the best idea. I would love that if I could afford it, both in money and in the situational environment that I work, but I can't. I can also see how my text above makes me sound like a person that is closing down on options but that's how the situation I am in works. I am a photojournalist by trade. Like my colleagues, we get great access to difficult places because we work on our own and we have no other character to defend but our own when dealing with people that doesn't trust us. I am bringing video to this mix and it is very important, but do mind that the situation I encounter myself in truly rings the bells of "less is more". Recording straight to camera is a way of skipping something else that can add pressure to the situation I am in. Also not carrying batteries so much is a nice plus.

    My reason for not using lavaliers is because the subject I am working with encounters "itself" (lets say it has no sex) in some situations with others and the using a lav will give me a quality to the subjects voice and another quality to everyone else.

    So: mounted on camera, directional (if I can increase volume at distance without increasing the noise floor much, that would be nice), battery operated is a major plus since I don't have anything that can feed phantom and record at the same time.

    Thanks again,

  • i donot agree at all with the statement that recording audio on the GH2 is inferior compared to PCM. with the pasadena pulse hack and lowest recording setting there is only some limiting above -5db. I made a recording of a drum solo CD using my SQN field mixer. i used this once famous CD originally used by audiophiles for testing loudspeakers (Charlie Antolini "Knock Out") to find out how the GH2 limiter works. But doing comparisons to the original I did not hear a difference. If so, it would have taken a long time to find out and still differences would be less than negliabe. We are not recording conversations with bats here but voice and ambience and even sometimes live music (sometimes straight out of a mixer) and for this purpose the internal amps (if kept low) and AC3 is more than sufficient even for heavy filtering in post. You will only have problems with very low level recordings like accidentally doing an interview @-30db. Too low for proper AC3 quantisation. But also too low for 16 bit PCM anyway. So avoid the hassle of syncing image and sound and bwt run when someone wants to give you a first generation ZoomH4, it goes out of sync after a minute

  • Alright. After much research, I have decided to try the following:

    AT897 -> iRig Pre hacked-> GH2 AT897 -> iRig Pre hacked -> Edirol R09

    I will post the results here when I get them I am just waiting for the XLR -XLR cable.

    Apparently you can power the AT875r with the GH2 alone according to this thread:

    Now, the tests he made are no longer available. However, reading a bit online it seems that the AT897 has more noise than the AT875r so I am going to test that too with the iRig Pre and get also a cable that will convert the balanced to unbalanced signal and do what @balazer did with both the AT897 and the AT875r.

    I tried an MKE 600 from a colleague who was using it with his DR100 and I didn't like the sound very much. He said he got that because someone else recommended that to him. I am going to go with my ears and see whether the ATs have the sound I am looking for.

    Also, I was thinking of using the Edirol mounted on the rig on its own and record stereo sound and then mix in post whenever needed.

    I'll come back with some samples.

  • OK.

    I haven't hacked the iRig Pre yet. Tested the iRig with the AT897 plugged into my Blackberry (to make sure everything works as intended) and...I don't like the sound.

    Ok. If I am looking and talking straight to the mic...its brilliant. But, doing documentary and a lot of it indoors, when people may not be talking towards me...its not good. There is echo/can type of sound.

    So, I need a more versatile mic. I know I requested a "battery operated mic" in the original post, but since I got the iRig Pre, I can do with a mic that needs phantom power (info for anyone considering the AT897...the noise is higher when using the AA battery than when using phantom power)

    If anyone can help here for me looking for a mic (@matt_gh2 still hasn't returned with his impressions on the ISK CM 20) that needs to be mounted on camera with a good angle consistency (of say 14mm lens on a GH2)

    Cheers again,

  • It is the nature of very directional mics that their sound is colored when it arrives off axis. The frequency response is not uniform. Choose a less directional mic, like a cardioid, and this is improved. But of course a less directional mic rejects off-axis noise less. That is the trade-off. Compensate by getting the mic very close to the source, e.g. an omni lavalier.

  • @balazer

    Thanks for the advise. I am now looking into:

    AT875r (it appears that the pick up is slightly wider) Beyerdynamic MCE 72 CAM (sadly enough, the Phantom Power version comes with TWO XLRs so it has to be battery) Q Mini Shotgun (still trying to find samples)

    Most of these mics will have not only on axis audio but also the subject speaking to the mic. I am trying to find a mic that can pick up good sound when the subject speaks at least 45* away from the mic.

  • I think the AT875R might still be too directional for you. Maybe a cardioid. Or can you use a pair of on-camera mics aimed to hit two subjects?

    It's really impossible for you to mic up two subjects with lavs? You can even wire them to the GH2 to avoid using an external recorder and having to sync.

    Have you tested the Rode VideoMic on your GH2?

  • @balazer

    The subject is one person, who lives in a house with 3 others in a relatively dynamic environment (people walk in and out of the room, along with the subject going to the kitchen/living room but also being on its own for a while, sometimes with the partner) so using lavs is something I would either have to supply to everybody and try to turn on/off the signals as they are in the room or out and make sure they don't mess them up while they go along their lives (they are all uni students)

    The documentary needs the interaction of the subject with others as well as the quiet moments. Because I am dealing with something sensitive, I can't ask to "show me again". It is either a in or out type of thing.

    I have heard samples from the Videomic and videomic pro (the latter on my former 5D2) and I still keep my stereo videomic (non-pro) but I hate the sound of those. They are very narrow, especially after 1.5m or so from the mic. They hiss too much too.

    So, if the AT875r would still be too directional, then I can only so far come up with this list:

    Beyer MCE 70 Senn MKH 40 Senn ME 64

    I am going to scour Audio Technica's and Rode's menu next...

    What I am not liking about the Beyer (although is the cheapest of the three listed) is its self noise it has although the quality is very clinical, non coloured, which I think would allow for a bit of a boost of character in post.

    All beyer samples I have heard so far come from German users and sites...why is it that they have not been more prominent in the mic business? They are great in the headphone business. is so much easier.

  • I think most of your problem is getting the mic close enough to the subjects. Do whatever you can to get the mic closer. Then the distance and the angles will dictate which type of mic is best, with a trade-offs between natural response and off-axis noise rejection. Maybe you can rig up something to put the mic in front of the camera, above or below and just out of frame. Use a wide-angle lens so you can get up close.

    Try the AT875R, I guess. It might be the best for your situation, depending on the distance and angles. I found it unsuitable for music, but I suppose it is ok for dialog.

    Lavs will be a big leap in sound quality. Omnis have the best frequency response, the most consistent response at different angles, and the best rejection of wind noise and handling noise - even cheap omnis. But of course omnis have no rejection of off-axis noise, so they need to be placed close to the subjects. I'm thinking of reality TV like The Real World and such - I'm pretty sure they use lavs for most of that. There are some topics here on p-v about small clip-on recorders, as an alternative to wireless body transmitters or larger recorders. You'd still want on-camera audio as a backup.

  • Alright.

    At last, after getting all the cables (amazing how difficult it is to get short XLR cables) I managed to do some tests.

    I am linking three vimeo videos here. The title is the set up and settings. After much playing around, I found these three to be the most useful to demonstrate what I have, what I am looking for and what I can get with this super cheap set up.

    First of all, the AT875r is a great mic, I think. I don't have the cash to try an MKE 600 so I cannot compare to that or the Rode NTG series. That being said, it is considerably more quiet (hiss wise) than the AT897. I would say that the only thing going on for the AT897 is the battery power. I would imagine, from the tests I hear, that the MKE 600 could do better in the self noise department but I didn't like the rather bassy signature of the Senn. I do like the MKH 416 from what I have heard but I think that its very direction capability makes it unusable for my needs.

    Worth noting the AT875r is not as directional. It can pick up noise from the sides much better (I see why people would find this to be negative in this mic category) so it works better for me for off-axis pick up.

    As I speak, you will know the distance I have from the mic/camera (starts at 2.5m in front) and how I move away from the centre area of the pick up.

    The room has some echo, so mind that when you hear the samples.

    The "videos" are downloadable and with Premiere Pro's settings I have set them to AAC 320kbps. If there is a better setting, please let me know (I am still learning)

    First link: AT875r -> hacked iRig Pre -> Hacked GH2 448kbps AGC 0: 235

    Second link: AT875r -> hacked iRig Pre -> Hacked GH2 448kbps AGC 0: 203

    Third link: AT875r -> hacked iRig Pre -> Edirol R 09 imput level 70

    My take on the above:

    1) GH2 internal recording distorts too quickly. Am I hearing this right? It appears to be that if you have a peak in speech, you get a distortion on the upper end. The graphs inside premiere appear to show a constant level at which they peak. So, it more or less looks like a long mesa-type of mountain whereas the Edirol's graphs show a more dynamic take (and it sounds to me like so as well)

    2) Edirol seems to win.

    3) Hiss is still there but with a bit of noise reduction I reckon it gets to a very good level (I did a tiny test with Premiere Pro CS6's denoiser and reducing the noise floor by 4db, freezing the variable "green line" so it is a constant reduction)

    It appears then that my solution here will be to use the Edirol after all. Since I don't have money to invest in a separate recorder with phantom power (and I read that the Zoom and DR 40 introduce hiss) then this would be it. If I do manage to get money for a compact portable recorder, apparently the DR 100 is the one to go for.

    Unfortunately @balazer 's links from his self powered AT875r don't work anymore but I wonder if you could make them available again, please? Just to see how it sounds?

    Thanks again.

  • @balazer

    Sorry for the delay in getting back here. Thanks for uploading those again. I can see how less power affects the mic. It is a great mic, either way, but the self noise on your samples would make it unusable for me.

    I am going to carry on with the iRig-> Edirol combo and see how it goes in the long run. I happen to have met a sound mixer last weekend so I will get some hands on supervision as I go.

    Thank you everyone for the advise. I will dehack the sound on the GH2 so I can regain in camera playback.

    My next test is to make sure the higher ISOs perform well (1250 and above) as I read that NR -2 may not be good for the higher ISOs compared to using NR 0.

    Thanks again!

  • Self noise? Do you mean noise coming from the mic?

    I think the noise in my samples is probably more from the GH2 than from the mic. I recall that the mic has very good noise specs, and mic noise is usually not affected by the power supply voltage. The SPL before distortion is usually the thing that suffers when a mic is powered at a lower voltage.

  • I think you may be right. It may be that the audio gain on the GH2 kicks in if the signal is too low even when you have the manual audio volume settings. I guess what I meant to say is that there is more hiss on your recordings than the ones I made with the iRig feeding into the GH2.

    Distortion is something I notice on earlier tests (not heard here, I think) when I would clear my throat and even though the volume doesn't peak (red) on either the GH2 levels or Premiere Pro's levels, I can hear a limit as if the camera couldn't reproduce that loud sound. Again, volume is not high (no peak) but GH2 somehow seems to limit the amount of loudness received maybe? or is it the range of sounds it can gather? since it distorts the sound. I am not sure because my understanding on that is limited at the moment.

    On the Edirol recordings, however, I can hear a defined and clean sound and even when volume peaks (red) there appears to be no distortion.

Start New Topic

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Sign In with Google Sign In with OpenID

Sign In Register as New User

Tags in Topic

Top Posters