Editor's note: Following this shootout there was a lot of controversy surrounding the results from the Sony a7 III. Our team tested a second a7 III body to verify the results and discovered some important things that could impact the rankings – depending on what microphone you use. Please watch the video update below.
We invited a professional audio engineer to test the microphone pre-amps in cameras from Canon, Nikon, Sony, Panasonic, Fujifilm and Olympus. Find out which ones deliver the best sound.
What about music? Well, you are not going to hear many examples because of copyright. Bu you can record outdoors performers playing classics or improvising (unamplified, so natural). Here is a Nikon Z 6 live music video using the cheap external stereo Sony ECM CS3 mic. Manual mode audio: https://youtu.be/w-8XPp1qU6w
One thing that really annoys me about Sony Alpha's external microphone input, and you can see it proudly printed next to the jack, is Plug-In-Power. That's 3v DC to power the simple electret mics. There's no option in the camera to turn it off. In theory, you can't hear DC (zero Hertz), but in practice, DC is the worst thing you can to to a potentiometer (simple volume control on your feeding device), or the voice coil of a dynamic mic (slams it against the stops), or even a slightly corroded connector (crackles when you bump it). I wonder if the feeding devices in the test suffer from having DC across their output?
As far as the audio inputs on the Sony Alpha's hot shoe, there's no volume control for that input like there is for external mic jack. The $600 KLR K2M box has volume control but comes with a shotgun mic none of us want. ECM-XYST1M mic through the hot-shoe is too hot; volume control in the camera requires the 3.5mm patch cable (included) to mic in (crackle). -- Mike
The XLR-K3M microphone adapter has a level control for each input - the included microphone and 2 XLR inputs. There is also a mic/line/mic+48V selector switch for each input.
DC applied won't "slam" the coil of a speaker unless the voltage exceeds the capacity of the speaker. Plugging a speaker into a microphone input is probably not the best choice anyway.
AC (ie capacitatively coupled) inputs crackle if the connection is poor just as badly as a DC coupled input. Power is applied equally to the audio supply and return (tip, ring) of Sony electret microphones, with the DC return through the signal ground (sleeve). Variations in the voltage are balanced out of the audio signal.
AFIK, all trim and fader pots in mixers are DC circuits, and seem to work just fine with that arrangement.
Boy, Ed, let's start with "speaker" ... nowhere in my post did I use that word. I wrote "dynamic mic". I see that you are an expert in misinformation, or at least misunderstanding. It's just that I don't like the *idea* of DC across the voice coil of a dynamic microphone, who's signal output is millivolts, and the DC is 3 volts.
I get my information from direct experience with my Sony a7Sii, which I regularly feed external stereo audio into the mic in from a Radial Pro-ISO DI, which has a volume pot on it's unbalanced stereo mic level output. Just touch the pot while recording, and the noise clips!
You also seem to be of the opinion the camera's mic in is balanced mono, but it's unbalanced stereo. And just touch the patch cable to the mic in while recording and you get noise.
I don't own an XLR-K?M (? == 1, 2, or 3). The only hot-shoe audio input device that I own is the ECM-XYST1M, which has no volume control. Trim pots, faders are cap coupled to block DC.
I shoot a lot of loud bar bands video with a Sony a7Sii fed from an external mic/audio recorder setup and most of the time I'm happy with the video file's audio track. The only time I use the audio from the recorder is when the camera's audio is noisy from cable movement when panning, or when I botch the levels on the camera. I never notice noise on the camera's external mic feed when the camera and cables are still.
If the a7Sii's mic level is set way high, then it might start showing noise, but I usually run at a mic level of 5 out of 31 since I feed it pretty hot.
A interesting experiment would be compare noise levels with sequence of decreasing mic gains along with increasing feed levels. I've never used an a7iii, but I'm guessing you can get adequate signal to noise from the external mic input with proper gain staging.
Yes as they said in the video Canon’s audio used to be on the bad side but with the 1dx3 they made a big leap forward. Hope it will be the same with lower end cameras
Me too. I think it'd be interesting to get a comparison against the GH5 or GH5S (maybe same pre-amp?), G9 and whatever else they have. I can't keep their naming straight.
I saw Zaxcom, Lectro, Sennheiser logos. What was used exactly? Was there wireless going on? Were they all the same? Did each get an XLR and a wireless test?
Boom - Sennheiser MKH50 Lav - DPA 6060 in to a Zaxcom TRXLA3.5 Mixer - Zaxcom Nomad 12 Signal to camera - Zaxcom UHF receivers, outputting a mic-level audio
Look at the specs of any of the latest Sony cameras. Wise people don't consider them for video anymore and you just saw another reason why. Another (of MANY) example of how far Sony is behind. In 2014 I shot video with a data rate double the max rate all Sony cameras use in 2020. Sony is still behind cameras from 6 years ago for video. Some of today's cameras use bit rates 4x what Sony does with codecs more than twice as efficient (so 8 times the data), while others use Pro Res (which is like RAW vs. highly compressed jpegs in Sony's case).
Wow, you really embarrassed yourself. Those are NOT codecs and I did't say what you falsely claimed. I said "In 2014 I shot video with a data rate double the max rate all Sony cameras use in 2020. " I didn't mention what codec nor did I mention compression for that example (i did elsewhere). I was 100% correct.
And Panther Fan doesn't have a clue what he is talking about! Ignore the toll. :D The codecs (which I didn't mention in the 2014 case) are XAVC S and H.264. All-I and IPB are type of compression used. The problem with Sony is the bit rate and inefficient codec do not allow for much data compare to today's larger bit rates and h.265 or Pro Res. XAVC S is Sony's version of H.264, but is only a 'consumer' version of their XAVC that is not as good.
“I didn't mention what codec nor did I mention compression for that example (i did elsewhere). I was 100% correct.”
Not taking codecs into account is exactly what panther fan is accusing you of. How is that in your defence?
“All-I and IPB are not on that list because they are NOT codecs like our troll friend thought.”
They are not considered codecs per se, but for our purposes here, we can definitely consider “Panasonic’s H.264 encoder that produces All-I bitstreams” and “Sony’s H.264 encoder that produces IPB bitstreams” as different codecs that just happen to produce outputs that can be read by the same decoders.
spider-mario, my point about about 2014 was there was at least 1 camera with double the bit rate of what Sony uses in 2020.
That is 100% true. No trolling or falsely changing a quote will change that. I didn't mention compression bit depth, Chroma Subsampling, or compression. Don't try to change what I said.
"They are not considered codecs per se,"
No they are NOT codecs PERIOD. It is ignorant to claim they are. Compression type and codecs are two different things. No trolling will change this.
@Kyle Uhm IPB and All-I are flavors of codecs. The major part of H.264 compression is temporal compression. Something H.264 All-I doesn't use and therefore is less efficient
Heck even H.265 All-I is less efficient than H.264 IPB
200mbps All-I has worse IQ than 100mbps H.264 IPB. That's why the GH4 also has IPB ;) All-I is a relict out of times where you wanted easy to decode codecs for NLEs on slow computers
PS: Already the GH2 could shoot at 176mbps but without taking codec efficiency into account that is worthless information
Panther fan, You are humiliating yourself even more. codecs don't have "flavors". (Unless you are talking about how Sony uses a stripped down version of their proprietary XAVC. ) That's like saying codecs have different flavors of bit rates.
No matter what you say. It won't change the facts. 1) IPB and All-I are NOT codecs. 2) in 2014 other cameras used double the bit rate of what Sony uses today. 3) in 2020, Nikon, Panasonic, Fujifilm and others have vastly better video specs than the outdated stuff Sony uses today in their A series cameras. Sony has the worst video specs and now we find the the worst audio of those cameras.
Have it Sony trolls. You can't walk back those humiliatingly wrong comments.
“No they are NOT codecs PERIOD. It is ignorant to claim they are. Compression type and codecs are two different things. No trolling will change this.”
I am not claiming that “All-I as an independent concept is a codec” or anything to that effect, and to my knowledge, no one here has. But the very beginning of the Wikipedia article that you linked to is: “Α video codec is software or a device that provides encoding and decoding for digital video […]. Most codecs are typically implementations of video coding formats.”
Which means that under that definition, two different encoders are considered different codecs *even* if they produce compatible bitstreams by implementing the same format. As a stretch (which I am personally happy to make), one encoder used with different sets of compression settings could be considered as several codecs, in particular if they make use of different sets of compression features, as they do here.
Also, no one claimed that it was not true that “in 2014 other cameras used double the bit rate of what Sony uses today”. The claim that was made was that it was irrelevant.
Please read up on codecs. Not all H.264 is created equal. Which compression features you use matters. In fact not even all H.264 IPB is created equally. There are vast differences in encoding quality/ efficiency, but that would go too far for this thread.
Also I don't deny there are better codecs out currently. But they were not in 2014 in ILCs
As an aside, I find it somewhat funny that you would call me a Sony troll when I have never owned a Sony camera, as you can see in my gear list which is public. Or that you would call me wrong or ignorant when image compression is literally my job (which admittedly does not automatically make me right, but still).
panther fan You really have no clue. "Not all H.264 is created equal. " Do you know what XAVC and XAVC S are? (Hint: H.264) Clearly not. I tried to tell you above but then you think All-I is a codec! lol!
I don't know which is funnier, your post about All-I being a codec, or your comment AFTER I mention Sony XAVC S is inferior to XAVC, that not all H.264 is the same! WOW.
Talking to Panther Fan is like talking to a box of rocks.
About 5 posts AFTER I told him, "XAVC S is Sony's version of H.264, but is only a 'consumer' version of their XAVC that is not as good."
Panther Fan comes up with, "Yes xavc is H.264 that doesn't mean it's equivalent to all other H.264" DUDE, I just said several posts up there are different types that are not equal! ROTFL!!!!!
But what do you expect from a guy who thought All-I was a codec?
We expect this ignorant post by Panther Fan.... Panther Fan said about AVC (H.264) "Also I don't deny there are better codecs out currently. But they were not in 2014 in ILCs" Except the NX1 used H.265 in 2014, and yes H.265 is a better codec than H.264/XAVC just as USB 3.1 is better than USB 2.0
Which reminds me, guess what Sony uses in all their latest A6x00 cameras. Hint: not USB 3.1
And how good was the H.265 encoder in the NX1? Just as all H.264 are not equal, all H.265 are not equal either. It’s even possible for them to be worse than a good H.264 implementation. I am not claiming that it is the case here, but it is something to keep in mind.
He said codec. H.265 is a better codec than H.264, just as USB 3.1 is better than USB 2.0. Its that simple. He was laughably wrong to say newer/better codecs were not in use in 2014. btw, Sony uses their worst version of XAVC in their A cameras. It is XAVCS, and is what they call 'consumer grade'. Its never been a great implementation of H.264.
The very list of codecs that you cited earlier mentions H.265 not as a codec but as a coding format, of which x265, for example, is one implementation. The codec, then, is x265, not H.265. The reference encoder at https://hevc.hhi.fraunhofer.de/svn/svn_HEVCSoftware/trunk/ would be another. Likewise for whatever was in the NX1. Just because it produces an HEVC bitstream does not automatically make it a “better codec” than, say, x264. So no, it is not, in fact, that simple.
@Kyle Give up. You started by claiming the 200mbps H.264 All-I in the GH4 was superior to Sonys 100mbps H.264 IPB.
It is not - fact
Then you claim H.265 is better than H.264 -no it's not it all depends on the implementation
I can write you an H.265 compliant encoder that encodes at 800mbps. But IQ will still suck, because my encoder won't be very efficient.
H.265 and H.264 etc... are coding formats. They define and allow what tools can be used to compress footage. But you are free to use however much or little of those tools you want. So a bad H.265 codec is worse than a good H.264 codec like XAVCS
H.265 All-I for example is barely better than motion JPEG. Just because it has H.265 in the name doesn't mean it's a good codec
So compare apples to apples and you will see that the Sony 100mbps IPB was class leading in 2014 and except the X-T4 and S1H is not far behind other cameras today.
The GH5 is at 150mbps H.264 ipb The EOS R is at 120mbps H.264 ipb So neither are miles ahead
//The GH5 is at 150mbps H.264 ipb The EOS R is at 120mbps H.264 ipb So neither are miles ahead//
Bit rate is relative to source. The A7III's 4K is downsampled 6K, so the 100mbs of the Sony contains far more information than the either the R or the GH5, despite the higher bit rates - that is why the A7III video is so detailed.
The quality is behind the XT-3 but beats everything else on the market, despite endless trolls wishing it was otherwise.
The 4K from the A7III is even more detailed than the Panasonic S1H's, because of that camera's strong anti-aliasing filter, although of course that camera can output 5.9K if required.
Surprising what results you get when an unbiased professional testing cameras vs people that are seeking confirmation bias. Probably a good move that Sony is moving away from the camera industry to concentrate on the PS5. The first generation Nikon Z6 is better in every way than a 3rd gen Sony a7iii. I am really disappointing in Jordan, I use to value his opinion, but now he clearly just makes things up. He stated how awful the Nikon preamps were and they out beat his s1h. We need more tests like this.
I definitely didn't say that the Z6 pre-amps beat the S1H pre-amps; quite the opposite in fact! The only thing holding back the S1H were the limiters, which might not be a deal breaker for some people.
Not all limiters are equal. Furthermore, non-professional input limiters are generally inserted after the preamp. While they prevent digital overload, they don't keep the preamp from splatting.
@AlexCMitchell I apologize for paraphrasing, but I guess my point is Jordan is outspoken about how awful the preamps are on the z6, watch the video when he reviews it, and praises the S1H. In your video, true the s1H may be have a more pleasing overall preamp, but they are more simular than different, when compared to the a7III. I owned an a7III for almost a year with the z6. I also had similar results between the two.
worldaccordingtojim I think that DPreview is drifting perilously close to the level of the youtube clickbait shills. Especially with regards to Sony .Take a look at the Sony 35mm vs the Nikon 35mm , yay Sony is the winner despite having a significantly narrower FOV , and PF so bad it looks like a cheap special effect
It is amazing when you look at the results from actual photographers and independent reviewers just how different the conclusions are. Take a look at this 2 part review of the Z50 from Maarten Heilbron and compare it to the Sony fan club reviews including DPreview . Sony makes some excellent stuff { I still use a couple of their lenses on my Z7 with the TZE-01} . But their true mastery is marketing and promotion https://www.youtube.com/watch?v=cXDw5wAtUYM
I agree, hopefully this video marks a change towards fair comparisons. Or at least demonstrating what your findings are. There is a huge difference in saying Nikon preamps are good or bad, or Sony preamps are good or bad, and actually doing this type of demonstration. I would rather them give me the results and let me decide and if they want to add their opinion also that's great.
But I always stress also, most cameras need a little bit of tweaking to get the best results, Canon and Nikon have an over processed standard picture profile, and so many judge their video quality on that, I would prefer they find the flattest profile, then compare. Just like this test, Nikon preamp on auto turns up the gain too high and amplifies the hiss, as soon as you adjust it manually it is fine. These cameras are not an iPhone, they need some settings to make them perform. Learn the camera, then review, and then compare them.
If I heard right, some newer Sony DSLRs have an external audio interface that can attach digitally, which would hopefully give users the option for a cleaner input. I haven't tested it but would love to in the future!
What about the X-T3 which costs $600 lens and "soundly" beat the Sony? lol!
btw, there are better external recorders that can be used on any camera for less than the proprietary Sony junk. Remember that OLED VF Sony made for NEX cameras? It was nice until Sony stopped supporting it in the newer cameras like the A5100.
So the S1H did win until the dude fed 10dB too much in to it. 10dB is a lot! Don't do that. You can always boost it post. Generally your meter should run from 50% to 75% on the scale leaving 25% for headroom.If you go past 100% the limiter will kick in to save your a$$. The Canon won't even come close if you add the XLR1 option to the S1H, and if you give a #$%^ about your work and your client you should. It also makes life easier on site and in post.
Hey Jeffe! I elaborated on the "hot" test on another comment but I'll reiterate it here:
Regarding overdriving the signal, I actually had a reason for testing this like we did! Many cameras reference unity at -12dBFS, whereas motion picture sound gear usually references at -20dBFS or 0dBU. We set the input of the camera so that a -20dBFS signal registered at -12; a mistake I see a lot of folks make when they're starting out. Knowing that some cameras could survive this common mistake was something I thought would be good for people to know, even if we didn't say that in exhaustive detail.
I wonder why the Nikon Z6 preamp got such a bad rating when that camera was tested by DpReview earlier. And why the result was nearly the opposite this time. Could you provide an explanation please? Your previous result was decisive when I bought a new camera two months ago. I boughr a Panasonic GH5. Now I think I should have bought the Nikon Z6 because I have Nikon lenses.
We fed in a mic level signal from a UHF audio receiver, which itself does not draw power from the 3.5mm TRS. It's possible that using an un-powered mic like a VideoMic Go could introduce more noise by stressing the circuit, but I didn't test that. Either way, I always encourage users to extensively test gear before they buy if they can! There are so many use cases out there and it's hard to test them all.
We do. It’s in the queue for review. We’re working through how to finish a number of reviews with limited access to our studio (where much of the testing is done).
You shouldn't buy a hybrid camera for its sound. You have many ways to fix the sound, but with video, what you get is all you get. I didn't buy a Sony A7iii strictly for its video either (4,2,2 8-bit), but what would I do with $20K in lenses? Actually 8-bit is okay if you don't make drastic grading adjustments. One more reason to get it right in the camera.
the canon is big enough to hold a pci sound card. Costing the most it should be good. The sony did sound the worst.
In the past I used a rodeo shotgun mic for d7000, not sure if it was a true preamp. Just got a Zoom H5 4-Track Portable Recorder and impressed recording somebody at a restaurant/ bar with all the background noise for the z6.
Canon seems intent on taking back the crown, in every area, including video.
Canon and Nikon both are showing that they can bring to video the same high quality standards they have brought to stills photography for decades.
Sony now ranks last in video for ILC's. It is apparent that in their rush to get new models to the market, they cut some corners. Even their IBIS is now worst in class.
It doesn't really matter to me, an external pre amp will solve most sound issues and do more. But for as-is to get by, better is good, but at what cost is people standard. Personally i wouldn't use the built in mic unless it is unplanned spur the moment.
I absolutely agree! Double system sound will almost always give you better results than recording audio in camera. Not everyone has the gear for that though, so I figured it was worth testing camera pre-amps like this.
I agree to a point, but half the technical budget of a commercial film is sound. Even at my level, it's 1/2 to 3/4 of the investment in gear and setup time.
I think the guy made a mistake and switched Nikon and Sony's files or something. I have them both and believe me, the Sony sounds much better than the Nikon.
How do your Nikon and Sony compare aside from the noise? I'm no sound buff so am probably just missing something, but I thought compared to the other cameras, the Nikon sounded like it was lacking bass tones?
Why did you and others always mentioned the Sony pre-amps as "solid" and Nikon pre-amps as "weak" in all earlier tests? I mean that has been consistent for many reviews and years now. Any idea what went different there?
Also how do the hotshoe inputs compare to the 3.5mm mic inputs?
As someone who didn't really know what he was listening to, I thought all of them sounded perfectly acceptable except for the A7III, which sounded like hot garbage.
I'll agree. I just listened again with my volume cranked. The Fuji, Nikon, Olympus, Canon all sounded very good through my stereo. I think he got the rankings spot on. The Panasonic sounded the richest and fullest of them all through my equipment. The Sony was bad just plain bad.
The Canon and Panasonic are my favorites. The Sony sucked. The Nikon sounded a bit pinched, the Fuji is not bad but not well isolated from processor or power supply noise. The Canon is a big surprise to me too. Sony needs to step it up. Bad audio is never forgiven.
The Canon costs the most, but still surprising. The Z6 fared better than I would have guessed. I would say the top 5 are very usable, all very close. They all offer high bite rates (or pro res) and C4K IIRC. And the Oly has the best IBIS which I think is often overlooked.
Why am I not surprised about Sony in a distant last? Years back I was disappointed with their limited bit rate (more motion artifacts and less color information) and nothing has changed while everyone else improved. If video is a priority, those other 5 manufacturers make noticeably better choices...just as we saw in this video.
In the official list, Panasonic shared the 2nd place with Nikon Z6.
In controlled environment, Panasonic was declared the clear winner. The problem with Panasonic is that the audio limiter is not very good. This should be easy to overcome by lowering the levels and amplifying them later. The quality should still be as good as with 1DX III. However, there is no way to make 1DX III as good as S1H in controlled environment.
edit.: Sony is by far the worst, all the others are at least OK.
Well, this is a comparison of a video-centric feature which cannot be directly experienced in still media. They could have embedded audio clips in a written comparison I suppose, but that wouldn't really be less clunky than a video. It's not like we're comparing image quality here; being able to actually listen to the sound from these cameras conveys things in a way that words cannot.
Why was it so hard to give the credits to the Nikon Z in this video it was fully deserved .Instead they let the Canon 1Dx win.... However for that camera you could buy yourself 3 Nikon Z’s.
I was amazed to see Sony being so worse as they are also known to make videocameras.
It would be better to use a decent external recorded for the price. The article is about how the cameras perform, not extra devices which cost upwards of $600.
Good point, sillen, they compared the brand new $6500 Canon with the 3 year old $1500 Sony. They should have used at least the A7RIV for this comparison to be fair. Also, remember that Sony 7RIV and A9II has unlimited video recording, nor more 30 minute limit on recent Sony cameras. The other point is that Sony videographers typically use the Sony XLR adapter in the MI shoe. I think this comparison is total nonsense from semi-professional videography perspective.
Sillen, don't worry so much about the mic jack and pre-amp. Depending on how you use it, you should get acceptable results.
I just plugged a powered Rode Video Mic Pro into the a6600 mic jack with +20 (Camera gain 1) and the results are cleaner than I got on my older Canon DSLR. At +0 (gain ~10) it still sounds fine.
If you are concerned, consider doing a rental before you buy and try it with your own mics and equipment. I had a good experience with Lens Rentals.
These are just cameras and as a user (vs investor), there is no importance to competition between brands- they are just tools to get the job done.
great to see Alex Mitchell on with Chris and Jordan if your at home with time to kill have a listen to the show they did when they were still the camera store TV its lots of info and great fun
Here are the two episodes we've done with Alex:
https://youtu.be/qeHOJMG1Lk0
https://www.youtube.com/watch?v=as3TmHJWunk
Thanks for reminding me of the live show!
In the stillness of the pandemic I recorded birds chirping and no noise problems with my Z7 and the built-in microphone. I think it is amazing that the Nikon Z6 ties with the Panasonic just below the - somewhat more!!!! expensive Canon D1x-III. That said I am surprised that something as ancient and tested and tried as a microphone pre-amp should in any way pose a problem to implement in a camera. To me the sound is of utmost importance. I had great sound on Sony Video Hi-8 decades ago, so what gives Sony?
Thanks. No real workaround- I just use a mic that's self-powered with a relatively hot signal and get it near your subjects. That works fine for all DSLRs, Sony included. My Canon Rebel pre-amps were worse when I increased gain, but it didn't stop me from using it..
Glad to see (hear) camera audio covered. Poor audio has been an annoyance to me for years. Many of us use a lower end camera to these and some have horrendous audio. That Sony was a distant last, but believe it or not, many cameras are worse. Some things that annoy me are: *Poor frequency response. *Limited control over audio settings. *Hiss, clicks or other noise in the audio. *Noise reduction or compression that damages the sound. *High dynamic compression that squishes the sound.
I made a prediction before watching that Sony would be the worst. Will this be the case until pre-amp quality will be defined in the specs-sheets?
Each manufacturer has certain traits: -Sony has the best specs, but may be disappointing otherwise -Canon is very robust with the basic things, but the implementation of latest tech maybe disappointing -Nikon provides decent quality, but the specs may be uninspiring -Fuji provides quality, but they have an obsession to differentiate from the others, just for the sake of it (not to mention Sigma) -Panasonic products are well refined, but It may take long for them to abandon their own tech and switch to another -Olympus is a question mark, you never know what to expect. Usually they are good
My how things have changed over the years. Sony has cemented themselves as last on the list when considering video. I thought the other 5 cameras were fairly close, but wow, Sony sounded bad. You can get an external recorder, but that won't help the major Sony video issues, such as too low of a max bit rate to prevent major motion artifacts and macro blocking, their unpleasant video colors, no 10 bit or pro-res option, and the 4:2:0 limit.
Next I'd like to see a comparison of IBIS during video. As bad as Sony's IBIS is, they will beat Canon.
It would be nice to show the test rig and methods. Most engineering and scientific papers spend more time on methods than results and conclusions combined. That's the only way to evaluate the process, and to reproduce the test if desired.
With S1H the levels could be kept a bit lower than normally. This way there would be less need to handle clipping audio (something that S1H is not very good with). I guess this way the audio quality would be more or less equal to 1D X mk III, considering all the three tested scenarios? The benefit of S1H would then be that in controlled environment, the audio quality would be even higher than of 1D X mk III.
Thanks for noting that I was unclear. In meant that in situations when there is danger that the audio levels go beyond the limit, the levels can simply be lowered.
The internal microphone is something you would use only out of desperation. One (small) step up would be an unbalanced, high impedance microphone tiypically purchased for use with DSLR or MILC video. Any microphone will need 40 to 60 dB gain, which also amplifies any noise in the electronics. A better approach is to use an external preamp with balanced inputs. The output only needs 0 to 10 dB gain and can take a studio-quality microphone if needed. These preamps are small enough to fit in a pocket or shoulder bag, or on the camera (preferably with a rig).
The best solution is to record separately and sync the sound in post, using the internal recording as a template. This is my approach, even with a professional camera (Sony FS5), which has good preamps and balanced inputs.
A professional quality run-and-gun rig typically uses a a small wireless pack on camera and wireless mics for the talent. or a short shotgun with a dead cat silencer on camera for man-on-the-street stuff.
They weren't using the built-in mics. This wasn't about the quality of the cameras' internal microphones, it was about the preamps that the cameras use to amplify signal coming from external mics prior to recording the data.
If you read more carefully, I quickly dismissed the internal microphone for more suitable alternatives, with or without the internal preamp.
Sony has a digital adapter, XLR-K3M, which has two XLR inputs with phantom power, and fits in the smart shoe of the A7iii, other alpha models, the FS5 cinematic camera, and various handycams. Since the FS5 already has high quality XLR inputs, the smart shoe is best used with a Sony wireless receiver.
Nothing wrong with reminding everyone to be cautious during the current pandemic... however the repetitive insertion used in this video was totally frustrating for me. I stopped the video after the third insertion. If you must, place it in the beginning of the video... don't interrupt the content. Look, it's on TV nearly every time I turn it on... I don't need to hear it during a technical/training special video. Overboard... JMHO, Bob
Chris’ mic as usual is clipping. Next time let’s see measurements with known values. At least establish a base line by plugging an UNPOWERED mic such as a Rode VideoMic Go or VideoMicro in to each camera to get a true representation of the built in mic preamp’s performance. If you are testing with a wireless system understand preamps are BUILT IN TO THE TRANSMITTER and the (already boosted) output signal from the receiver actually has to be attenuated to match the nominal input level of the camera. Your crime scene is contaminated! And overdriving the input level way past clipping and relying on the limiter is just plain bad practice. Set a proper level THEN engage the limiter or ALC to protect an accidental and occasional peak. It’s the same as overexposing your image and then wondering why there are no detail in the sky or wedding dress in post.
A professional quality preamp will bypass the mic pre for a line input. The cheap approach is to attenuate the input signal 20 dB or so before the pre, and hope for the best by adjusting the trim.
It is not possible to bypass the mic pre on a DSLR/mirrorless type of camera. camera. A few of them have a “line” setting but all that does is insert a pad in the circuit to attenuate the voltage from a high level signal down to a mic level which can be accomplished with an external device such as cables from Sescom. If we are using an external preamp or wireless receiver the best starting point is to set your camera input level to it’s lowest setting and keep it there. Turn off auto level control. Speaking of ALC there are two different types. Some of them only limit the peak level (preferred) and some of them will boost levels in quiet situations and reduce in loud situations. The latter will greatly raise the noise level and produce very unnatural dynamics.
- It would be super cool to do a more scientific test but I definitely wanted to keep this as "practical" as possible. It's fun to do a lot of the numbers stuff but at the end of the day when someone asks me what they should buy, how it performed in an anechoic chamber doesn't help as much as I'd like. ;)
- Regarding overdriving the signal, I actually had a reason for testing this like we did! Many cameras reference unity at -12dBFS, whereas motion picture sound gear usually references at -20dBFS or 0dBU. We set the input of the camera so that a -20dBFS signal registered at -12; a mistake I see a lot of folks make when they're starting out. Knowing that some cameras could survive this common mistake was something I thought would be good for people to know, even if we didn't say that in exhaustive detail.
Loved your feedback btw! You clearly know about audio. :)
There is a fuzzy line between terrible and tolerable. The latter is about as good as you can hope for in the camera.
20 dB boost in post is not unusual, with one caveat. Digital is usually recorded at -20 db from the maximum, but peaks typically reach -6 db, which limits the final gain unless you use compression and limiting. With digital, you record for the peaks.
I feel like the Canon was out of place compared to the other cameras in the review. It is simply not in the same category as any of these other cameras. Perhaps a better model to compare would be the EOS R or RP or even the D5m4. Another option would be to add in comparable cameras like the new Nikon D6 (when it's available) and/or the Sony A9m2.
Understand, but it does not say anything about being a hybrid camera as a requirement. Just says "Mirrorless and DSLR microphone pre-amp shootout", so you can imagine the PENTAX loyal feeling left out.
Anyway, I was having a little fun with my comment over you guys forgetting them in editing on a previous video. As always, this was an excellent video and I appreciate the work you guys do. Your humor really lifts my spirits during these times.
Is the same pre-amp used in the different cameras by the manufacturers? Does the video specific camera made by Sony (A7sII) sound the same as the lowest priced Sony (A7III )?
Cool video! Thank you! Thoughts: 1, How about the built in mic quality?
2, Shame on you Walkman guys!
3, In the era when the 3.5 jack was still exist on every phone I checked just with my ear some phone to hear the audio quality and shockingly the Sony was the lowest volume always. That time Samsung was also quite low volume. Of course it is more than that low or loud but as an average user with an earphone it is pretty much the most important thing. HTC was the best audio experience than and Apple also was always good. Now when no 3.5 jack I can just buy an HTC adapter and make any phone sound more or less like my beloved HTC One M8... So when people say bring the jack back I say nope, bring more USB c instead!
LG V20 has one of the best audio on ANY cellphone. Shame they didnt continue it with latest versions :( Great Audiorecorder , too. Hardly use my TASCAM DR -05 any more, its so good.
The A7iii finished dead last in every sound category. You will definitely want to capture sound separately which means this cannot be used as a run-n-gun solution.
I trust this test and I wished Sony did better, but lets not forget that it still record audio good enough to hear what people are telling. And in a run and gun situation many other things will be audible, like appliance noise, wind noise, people around and the bad audio from the A7III might not be a real problem. Again I hope Sony will do better in a next camera, hope they learn from this.
It seems like Sony might be providing a second-rate "legacy" mic input with the preferred interface being the multi-interface shoe (as Orim_70 suggested). That would be disappointing, but not really surprising.
The other possibility is that the plug-in power isn't matching the mic used -- apparently, the power isn't really well standardized, and this could easily explain the Sony needing to apply a lot of extra amplification to the signal. The same could happen for impedance mismatches. At least to me, the Sony actually sounded fine in the this video when supposedly given too loud a mic signal....
Kyle Style: Yeah, it's very noisy in the normal and low input level test, but it sounded more normal with the louder mic in. Definitely either Sony botched it (which seems unlikely, but maybe if they want you using the shoe input?) or there is something making the mic input level always too low, so it needs to amplify more.
Feed a decent signal to the Sony and you can use it for run and gun. I use the Rode Video Mic Pro (original) and the +20 feature keeps you far away from the noise floor. I also mount a h4N on top for stereo mic sound which also works fine. Camera used: a6500 and a6600.
Canon has released firmware updates for four of its mirrorless cameras, its flagship 1D X Mark III DSLR and its RF 50mm F1.2 L USM lens. The updates mainly address minor bugs and brings support for newer lenses to Canon's mirrorless cameras.
Panasonic has announced firmware upgrades for its S1, S1R, S1H and S5 full-frame mirrorless cameras, which extend the degree to which the four models share features and performance levels. The G100 vlogging-focused Micro Four Thirds camera also gains webcam compatibility.
Atomos and Olympus have announced their plans to bring Apple ProRes RAW video recording capabilities to the OM-D E-M1X and OM-D E-M1 Mark III when connected to an Atomos Ninja V recorder via HDMI.
The Panasonic S5 includes important upgrades to its autofocus system, but how much do those improvements extend to video? We put it to the test, shooting side-by-side against the S1H, to find out.
Fujifilm's 30mm F3.5 R WR is a super sharp 24mm-equivalent lens for the company's GFX lineup of digital medium-format cameras. Is it good enough to warrant a place in your camera bag? Find out in our field review.
The Insta360 One R is a unique action camera: it has interchangeable camera modules, including one with a large 1"-type sensor and a Leica lens. We show you how it works and ask, 'who's it for'?
Exposure X6 is the latest Adobe Lightroom competitor from Exposure Software. With great image quality, impressive speed and powerful features, it's a compelling option that doesn't require a monthly subscription.
Sigma's 35mm F2 DG DN designed specifically for mirrorless cameras is a compact, well-built lens that produces lovely images. Is it a good fit for you? Find out in our field review.
US manufacturer Really Right Stuff just released a new lightweight travel tripod, aimed at active and weight-conscious photographers that don't want to compromise on quality. Does its performance justify its high price? Find out in our initial review.
Whether you make a living out of taking professional portraits, or are the weekend warrior who knows their way around flashes and reflectors, you'll want a camera with high resolution, exceptional autofocus and a good selection of portrait prime lenses. Click through to see our picks.
What's the best camera for shooting landscapes? High resolution, weather-sealed bodies and wide dynamic range are all important. In this buying guide we've rounded-up several great cameras for shooting landscapes, and recommended the best.
What’s the best camera costing over $2500? The best high-end camera costing more than $2000 should have plenty of resolution, exceptional build quality, good 4K video capture and top-notch autofocus for advanced and professional users. In this buying guide we’ve rounded up all the current interchangeable lens cameras costing over $2500 and recommended the best.
What's the best camera for travel? Good travel cameras should be small, versatile, and offer good image quality. In this buying guide we've rounded-up several great cameras for travel and recommended the best.
There are a lot of photo/video cameras that have found a role as B-cameras on professional productions or A-camera for amateur and independent productions. We've combed through the options and selected our two favorite cameras in this class.
The Canon 200mm F1.8L may be over 30 years old, but the fact that it still keeps up with the newest high resolution sensors is a testament to its design. Featuring guest photographer Irene Rudnyk.
We teamed up with Canadian portrait photographer Irene Rudnyk to shoot a sample gallery with the legendary Canon 200mm F1.8L. Check out the photos and tell us what you think!
Is the SD card slot coming back to the MacBook Pro? Is the polarizing Touch Bar finally getting kicked to the curb? Bloomberg is reporting that there are many changes coming next-gen Mac computers and the changes all sound promising.
Today, B&W film photography remains popular both for its aesthetic appeal and its ease-of-use, whether you're a beginner, taking an intro to darkroom photography class or a seasoned pro. Here's everything you need to know about the medium.
The new limited edition Reporter version of the M10-P comes with a Kevlar jacket and a deep green paint — though you (probably) won't be taking it to war
Last year we covered PhotoStatistica, a macOS app that visualizes the EXIF data of your images and shows you the ways in which you capture photos. A new version was just released, introducing many improvements, including a new UI and new filtering tools.
Fujifilm's 30mm F3.5 R WR is a super sharp 24mm-equivalent lens for the company's GFX lineup of digital medium-format cameras. Is it good enough to warrant a place in your camera bag? Find out in our field review.
The next-generation AAT system can identify more objects in photos, perceive where each object is located relative to each other and provide more detailed descriptions.
US face recognition developer has been found to have used pictures from the Ever storage app without permission, and now has to delete all its algorithms.
Irix's new 45mm F1.4 Dragonfly lens is fully-manual and ready to be used with Fujifilm's GFX 50 and 100 camera systems. It's currently available to pre-order for $795.
The Tamron 17-70mm F2.8 is a fast, large aperture zoom for Sony E-mount APS-C cameras. Does it hit the sweet spot between price and performance for an everyday zoom lens? We tested it to find out.
If you're a Sony APS-C shooter in search of a versatile, walk-around zoom lens, the Tamron 17-70mm F2.8 should probably be on your short list. Check out our sample gallery and judge image quality for yourself.
Exploredinary has published a video tour of the Ilford photographic film and paper factory in Mobberley, England. The factory, operated by Harman Technology, which trades as Ilford Photo, has been operating on the same site since 1928. Ilford Photo traces its roots back to 1879.
Qualcomm has introduced its new Snapdragon 870 5G, a faster version of the aging 865 mobile platform that brings support for 200MP single cameras and 720p slow-motion recording at 960fps.
Is it really necessary to pay for photo editing software when it already comes included with your camera purchase? We test Nikon's own editing apps against the industry go-to.
The lens is optically identical to its black and silver siblings, but spices things up with a bright-red paint job and a custom lens cap to celebrate the Year of the Ox.
Join filmmaker John Webster and his team as they voyage into the beautiful Sawtooth Wilderness in Idaho with Manfrotto's Befree 3-way Live Advanced tripod.
The inclusion of in-body stabilization in Fujifilm's X-S10 means it's able to offer a lot of the features of the flagship X-T4. So, price aside, what are the differences between the two models, and how much of a bargain is the smaller camera?
Which high resolution mirrorless camera is best for you? This week, we compare the Canon EOS R5, Sony a7R IV, Nikon Z7 II and Panasonic S1R to answer that question.
As part of CES 2021, Canon launched a new website allowing users to view select locations on earth from the Canon CE-SAT-1 satellite. Using the onboard Canon 5D Mark III and Canon telescope, you can zoom in and see our planet from a fresh perspective.
The new Pro+ and Platinum+ plans cost $150 and $300 per year, respectively, and add additional benefits over the complimentary 'Pro' plan Nikon Professional Services offers. These NPS plans are limited to residents of the United States and U.S. territories.
The Insta360 One R is a unique action camera: it has interchangeable camera modules, including one with a large 1"-type sensor and a Leica lens. We show you how it works and ask, 'who's it for'?
Comments