Sound Devices MixPre-6 II (second generation)

Thanks, Brad. It was the MKH416 that I used when I tested the close mic approach and liked it. Good to know it’s a candidate anyway instead of the CM-4 mics that are best a few feet away and used as a pair, ORTF.

It's worth experimenting with an individual CM-4 as well, though. I used my CM-3s recently to record a podcast interview (one mic for me, one for my interviewee) and they sounded fantastic. I used the foam windscreens and placed the mics on the desk between us, a couple of feet away and roughly at chest level.

Yes, that’s one of the use cases that suits the CM-4 Macs I have. If I find that lacking, Rob’s suggestion to use the Shure mic is good too; I could go that way if needed.

And here is Andrew Huang’s fun video (he also uses the Shure SM7B). The $3600 mike is a Neumann U87.

$22 MICROPHONE VS $3600 MICROPHONE

What large condenser mics would be a good choice?

This is by the fellow from Toronto who iZotope uses for its videos. In the first part he talks about setup, including the Shure SM7B he uses. The SM7B is a US$400 dynamic mike...
I think that this is the most common sense video on voice over on the internet.

Much appreciated, Rob; I’ll view these sources. It would be rather tasty to have an alternative to my condenser mics for some purposes.

Andrew Huang’s video is also a good endorsement for the Shure SM7B that he regularly uses.

Edit:
CloudLifter? Seriously? Do you think the Mix-Pre Series II will handle the Shure without an external boost?

I think that the issue with high quality small diaphragm, end address condenser microphones is their sensitivity. Addressing plosives is straightforward. Either don’t talk straight into the end of the microphone or be willing to get rid of plosives, and indeed other undesirable human sounds, after the fact. In an urban environment, I find, unless one is recording in a dedicated studio, that the bigger problems are things like street noise and sounds from old electrical systems. It’s amazing what a Schoeps CMC641 can pick up.

Sometimes a good miniature/lavalier mike is the answer. Where to place it? If it’s a DPA, DPA says that the best spot acoustically is on your forehead. Seriously. If you don’t have a headband, it just so happens that URSA makes a head strap: https://ursastraps.com/product/headstraps/

:)

CloudLifter? Seriously? Do you think the Mix-Pre Series II will handle the Shure without an external boost?

Try it. Not hard to find a Shure SM7B in Victoria/Vancouver. I strongly suspect that the answer is yes. Can’t think of a source right now, but I’m pretty sure that there’s info on the internet about MixPre preamps and dynamic mikes. If I recall, Sound Devices itself has info on this. Sounds like Andrew Huang does not use a Cloudlifter.

I have a bunch of dynamic mics here, mostly Shure Beta 58a and SM 57, and just tried a couple in the MixPre 6. No problem at all in terms of enough gain. I don't see any need for a Cloudlifter.

Can’t resist commenting on URSA’s head strap...

Recently, I had to decide how to record the voice of someone walking on busy New York streets. I wanted a good balance between street sound and voice. Tried a Schoeps omnidirectional CMC62. Too much street. Tried a supercardiod CMC641. Still a bit more street than I wanted. A shotgun would have been too much voice, not enough street. A DPA omnidirectional 4060, wired, provided exactly the balance that I was looking for.

I picked up an URSA head strap so that I could easily mount the 4060 on the forehead, which is the best place acoustically and avoids the whole issue of clothing sounds. A headband in summer and a toque in winter can disguise the URSA so that the person being recorded doesn’t look like a member of the Borg.

I’ve since discovered that the URSA head band is great for managing cable for binaural recording, mounting a DPA 4060 to an Apple AirPod in each ear :)

I have a bunch of dynamic mics here, mostly Shure Beta 58a and SM 57, and just tried a couple in the MixPre 6. No problem at all in terms of enough gain. I don't see any need for a Cloudlifter.

Odd he was concerned about the sound of silence so much but then puts a CloudLifter inline. I’m not familiar with his recorder and it might be a good one, but my first thought was try the mic directly feeding the MixPre!

... Not hard to find a Shure SM7B in Victoria/Vancouver. I strongly suspect that the answer is yes. Can’t think of a source right now, but I’m pretty sure that there’s info on the internet about MixPre preamps and dynamic mikes. If I recall, Sound Devices itself has info on this. Sounds like Andrew Huang does not use a Cloudlifter.

I’ll look in the manual but the website may be a better place. Thanks!

Odd he was concerned about the sound of silence so much but then puts a CloudLifter inline. I’m not familiar with his recorder and it might be a good one, but my first thought was try the mic directly feeding the MixPre!

The mic into the MixPre works fine. Turn off phantom power first on that channel (the Shure dynamic mics won't be damaged by phantom power, unlike most ribbon mics for example, but you might as well save battery power by turning it off when you don't need it). To get an average level of -18 dbfs I found I could set the gain knob roughly at noon, which is not much higher than I set it with my CM-3 mics. I have my MixPre set up in custom mode, where everything is in advanced mode except the front knobs, which I use to set channel gain rather than to serve as faders for mixing. It's easier for me to set the gain for the ISOs using the front knobs than to fiddle with the wee headphone encoder knob.

Odd he was concerned about the sound of silence so much but then puts a CloudLifter inline. I’m not familiar with his recorder and it might be a good one, but my first thought was try the mic directly feeding the MixPre!

Manchester uses an Apollo 8 QUAD. It’s a pretty high-end interface, although as he says in the first video there are people who diss it. I don’t doubt that there are rough and tumble discussions about everything that Apollo makes on Gearslutz.

Manchester briefly explains the reason why he purchased a Cloudlifter with the mike in the second video above. Note that Andrew Huang, who doesn’t appear to use a Cloudlifter, also uses an Apollo interface.

I have my MixPre set up in custom mode, where everything is in advanced mode except the front knobs, which I use to set channel gain rather than to serve as faders for mixing. It's easier for me to set the gain for the ISOs using the front knobs than to fiddle with the wee headphone encoder knob.

Lots of people apparently do that, but here’s another view.

If I recall, the default in Advanced Mode is that that “wee headphone encoder knob” is used both to set gain and headphone monitoring level. That is also what it is for on Sound Devices’s high-end field recorders. Why? Because it is significantly more precise than the mixer knobs, and far less likely to be turned accidentally. Indeed, when the small knob controls gain, it isn’t even possible to change gain accidentally. It is also well placed ergonomically regardless of whether one is using the recorder over a shoulder or in carry bag.

My Sound Devices 702T doesn’t even have mixer knobs. When I got a MixPre-3, I tried setting gain using the mixer knobs, and decided pretty quickly to stay with the wee knob on the side.

But as always, to each their own :)

Thanks, Brad and Rob. I use the front knobs as the mix faders and the side knob to set the preamp gain. I am getting used to it and must admit I found it all confusing until I actually started working with the recorder. I’m sure I’ve still got a lot to learn.

For example, it doesn’t matter how many times I read about re-recording a track later at a different sound level. I’ve never managed to pull it off. Maybe it’s because I’m using 32bit or maybe I have a mental block, but I can’t do it. With 32bit recording, why would I need that anyway? Unless I re-recorded as 24bit for convenience?

I have an ElectroVoice RE20. On my OG MixPre3 using a Cloudlifter made a big difference in the sound. When I initially got the mic I was thinking that it should be fine but I was needing to push the levels higher than I anticipated and it was a little noisy. After conferring with Sweetwater, I got the Cloudlifter and the recording quality was much better. In fact, I forgot the CL for my group's last podcast and the person who does the editing noticed straight away. Since I just got the new generation Mixpre 3 I'll have to test it out to see if it's still necessary.

Chris, that would be very good feedback; thanks. And you might consider also using 32bit audio and boost in post if needed… perhaps? I am feeling the CloudLifter isn’t need by anyone using 32bit audio as it’s very easy to rescue low levels as well as peak levels in post.

I do not think the 32 bits will help on boosting if the mic output is too low. It is question about preamplifier.

I remember building a mic preamp many, many years ago (35), it used high quality transformer to match the low impedance of dynamic mic. I still have that somewhere.

What is the CloudFilter? some kind of preamplifier or transformer?

I use also transformer with moving coil cartridge on record player. The discrete phone preamp can handle the cartridge just fine as it is, but using less gain on it and transformer just provides much better result.

Interesting Kim. Let’s see what Chris concludes with his test if he also uses 32bit on the MixPre series II.

For example, it doesn’t matter how many times I read about re-recording a track later at a different sound level. I’ve never managed to pull it off. Maybe it’s because I’m using 32bit or maybe I have a mental block, but I can’t do it. With 32bit recording, why would I need that anyway? Unless I re-recorded as 24bit for convenience?

Are you talking about remixing a stereo mix? I don’t need this function, but useful if you have to distribute a mix quickly, are unhappy with the mix and don’t want to spend time bringing it into a DAW.

Curtis Judd’s video on remixing in the recorder:

BTW, I must say so far my experience with SD customer service is pretty good. The 10T was missing the camera attachment screw that is at the center of top cover. Contacted them and now they are sending me one.

Thanks, Rob! I was mixed up about the sequence as the manual talks about selecting the isolated tracks which I did, but I didn’t realize that had to be done before you hit the Re-Mix LR button as hitting that button starts the re-recording immediately. You really want to know how you want the ‘faders’ set on each isolated tract and do that first. Hope I’ve got that correct. I tried a couple of variations on a test and it went fine.

Well. I guess that’s one less mystery in life… I must confess i have one more mystery:

The captive ¼-20 screw in the top of the MixPre… I can try to pull it out but it snaps back. How do I use that screw to attach a camera on top? There’s an Allen wrench but I’ve no idea where to shove it.

Edit:
I see you push the wrench through the bottom hole and the top screw is revealed. Since I had the MixPre already on a stand, this wasn’t obvious!

BTW, I must say so far my experience with SD customer service is pretty good.

Sound Devices built its business on customer service. Before the MixPre, most of its clients worked in film, television and radio production. Lots of professional sound people who can’t afford gear failures in the middle of a shoot. The company’s field recorders were expensive, and its clientele was demanding.

Last year, I was at a trade show in New York and got talking to the guy at the Sound Devices booth. No name tag, no hint of his role in the company. I knew who he was. It was John Tatooles, co-founder of the company, a long way from Wisconsin manning a trade show booth. Just chatting with customers.

When Sound Devices released the MixPre recorder/mixer/interfaces, which cost a lot less than the company’s core field recorders, I wondered whether customer service would suffer. Seems like it’s doing fine.

On this one CloudLifter tested 11:30

I have a lot of time for Julian Krause, who in this video talks about the Shure SM7B specifically:

Do you really need a Cloudlifter or FetHead (ft. Shure SM7B)

Kim, thank you for that video that is a very good comparison of the Electro Voice RE20 and the Shure SM7B.

YMMV but IMHO, I immediately felt the CloudLifter coloured the voice. It may still be a good idea if it helps keep noise levels under control and allows you to be a little further from the mic, but I personally would not order the CloudLifter until I was certain it was a necessary option.

I don’t know how far the MixPre-II pre-amps will allow me to raise levels on a dynamic mic, but the MixPre in general has a good reputation for boosting the audio. If needed, I would use the dynamic mics close to the voice which might reduce the appeal of the CloudLifter.

Interesting how well the RE20 stayed relatively flat even used up close. I think it would take some practice for an actor using the SM7B to speak at an optimum distance without resulting in too much bass. Using the RE20 would be easier in that regard as you can be assured the voice over or ADR will match the film.

My Sennheiser MKH416 probably adds some colour too used up close. When I tried it, it sounded interesting, but not flat. It would work well for Voiceover but need care when used close if doing ADR. When I did my actual ADR it wasn’t up close and was fine. Perhaps it was a foot or two away... which is close for a shotgun usually meters away.

Rob, the noise floor on the MixPre is -130 dBV (-128 dBu) which is similar to the CloudLifter and lower than the Zoom H5 so following Julian’s flowchart, he’s concluding a CloudLifter is not going to be a noticeable improvement. I agree with his demonstration of the Shure SM7B recording to the H5, the CloudLifter was helpful to reduce noise levels.

I use also transformer with moving coil cartridge on record player. The discrete phone preamp can handle the cartridge just fine as it is, but using less gain on it and transformer just provides much better result.

Exactly, my understanding is that the sensitivity of the RE20 and Shure SMB7 are near the limits of certain recorders. I read a lot on various forums (jwsound, gearslutz, tapersection) about this and a Cloudlifter seemed to be what was used to give you extra gain. I talked to Sweetwater and they agreed and I bought the Cloudlifter. It worked great. We had three people in a room about 3-4 feet away and our podcast sound had less hiss and background noise and I wasn't riding the levels near their top end any longer. But that was the original Mixpre3 which I've replaced with the new version which I think is superior to the old version in this regard. I'll do some tests and see whether I think the Cloudlifter is still needed. I don't mind using it and I don't think it colors the sound but yeah, if I could sell and not use one piece of gear, that would be great.

I tried the Sennheiser MKE600 and a couple other cheap podcast mics, but they didn't sound great and picked up lots of background noise. The RE20 was much better in that regard. I love the sound of the RE20.

And yes, my Fidelity Research phono cartridge requires me to use a step-up transformer to my tube phono stage! Maybe that explains why I like the RE20...

Rob, the noise floor on the MixPre is -130 dBV (-128 dBu) which is similar to the CloudLifter and lower than the Zoom H5 so following Julian’s flowchart, he’s concluding a CloudLifter is not going to be a noticeable improvement. I agree with his demonstration of the Shure SM7B recording to the H5, the CloudLifter was helpful to reduce noise levels.

I think that Krause’s conclusion is that the SM7B is likely to sound better with a Cloudlifter unless, in the case of talking, one is speaking quite close to the mike. That said, I think that the takeaway from his video is to try the mike without and with one. I think it’s interesting that Manchester uses one, but Huang appears not to. I think that they both know what they are doing. Also, note that Huang is a singer and instrument player as well as a presenter.

It would be very good selling argument if the second series has better gain, lower noise floor, but based on the technical data this is not the case, they both have same equivalent input noise, this is 10 II

10 II.png (57.55 KiB) Viewed 2135 times

and 10T.

10T.png (59.96 KiB) Viewed 2135 times

A bit surprising they both have 32 bit ADC. Maybe they are utilising just 24 bit of that on 10T ?

PS. It would be nice if someone would actually explain this data with understandable terms, without any sales speech and dual 32 bit ADC floating +-700 db Hambo Mambo.

As I would imagine it works is,

- in second series they have 2 ADC per channel, similar as the single one on first series.

- one of them is working on lower gain and the other on higher gain, it would be nice to know these gains.

- If the higher gain one clips they automatically use the data from the lower gain ADC.

- On second series the data from the two ADC is converted to 32 bit and for some reason to float, maybe to avoid the data rounding errors?

Basically as I understand the noise floor remains just the same, but the low level signals have much better resolution. The only advantage really is this, better resolution on low level signals, this makes it possible to bring back all the details even when the gain is set so low that the second ADC will never clip.

The disadvantage of the 2 ADC would be, that ideally it works great, but on real life there sure is problems matching and combining the signal perfectly form two ADC to one 32 bit value. The temperature changes, ageing of the device etc. could cause problems. This is probably the area where the innovations are on this. The lack of explanations on this area is the reason I decided to go with first generation device for now.

With the fist generation device it should be possible to get just the same level of details at low signal levels. It is just that then there is risk that high input levels would cause clipping. The gain needs to be adjusted correctly, as there is not the second ADC automatically backing up.

The 120 db vs. 142 db dynamic range comes from as on the second generation device it is possible to have high signal levels and low level details at the same time*1. On first generation device it is possible to have both just as good, but not at the same time. With first generation device one needs to select high levels or low level details, by adjusting the gain.

It probably would be good idea to adjust the gain also on second generation device, so that there is no constant switching between ADC1 and ADC2. i.e. that the recording would happen with ADC1, except if the sound unexpectedly gets very loud. I do not know, maybe the switching between ADC1 and ADC2 is so good that it does not affect the sound in any way in any conditions, that would be awesome.

*1 This is not exactly true if it works as I assume, i.e. that a single 32 bit sample can have only data form one ADC, i.e if the low gain ACD is clipping the data is only form High gain ADC and if the low gain ACD is not clipping the data is only from it. The same goes for the 142 db dynamic range, second generation single sample has only 120 db dynamic range. However as the following samples can have different gains, in practise I agree with their marketing. Note though the dynamic ranges are given at 10 db gain. At higher gain I would assume the dynamic range difference be less (gen. 1 vs. II)


This is just how I am trying to understand it, please correct as I really would like to understand it, as it is evident from how many times I edited this after I started thinking about it.

The Mixpre first series uses the AsahiKasei AKM557x series ADC.
The ADC support both PCM and DSD 24/32-bit output.
SoundDevices went with 32-bit internal precision for the ADC but 24-bit PCM and channel summation to gain dynamic range for recording and USB I/O.

The 10T seems to have 32 bit available when working as USB DAC.

The 10T seems to have 32 bit available when working as USB DAC.

Ahh ok. Thats great,
Perhaps something have changed along the way.

I believe the first series only did 16 & 24-bit up to 96KHz while the II-series did up to 96 as well but managed 16, 24 and 32-bit Int and float over USB I/O.

I really would like to understand it, as it is evident from how many times I edited this after I started thinking about it.

Sound Devices’s patent is a public document, available on the internet. You may find the following exchange interesting. Note that one of the participants, Paul Isaacs, works for Sound Devices: https://jwsoundgroup.net/index.php?/top ... amic-range

Would you have link to the patent, though those are usually almost non understandable. I really hope they would explain with common sense, without using words that do not support* the understanding of the principle, limitations and benefits. I am sure that would be less than one page, if someone who understands what she or he is writing does that.

* The terms I do not like to hear on the explanation how it works.
- kasmir preamp, very good marketing term, but has nothing to do with the principle
- >700 db, this is just distraction, yes, the 32 bit is capable to this, no the preamplifier or 2 ADC is not. The dynamic range of the device is 142 db (and that is very good).
- etc.

I post earlier image chart extracted from their patent PDF (this is for 3xADC design). I really would like to know if they use 2x or 3x ADC in their current products.

I believe Paul stated 2 ADCs in some comment over at GS or JW.
But their patent is for multiples ie three but "at least 2". This is to protect themselves from other kind of patent applications which might find a way to create a something close to their solution.

You would be more then welcome to open your up and take some photos of it for us too look at
I've been searching the web but never found any II-series internal PCB photos.

Some photos of MixPre (first gen) internal components where posted here
viewtopic.php?f=2&t=97842&start=300#p553982

Got curious about that Apollo 8 recording interface used to produce all that Grammy award-winning music in a small bedroom. It might be a small room, but the equipment is respectable.

https://www.uaudio.com/audio-interfaces/apollo-8.html

Got curious about that Apollo 8 recording interface used to produce all that Grammy award-winning music in a small bedroom. It might be a small room, but the equipment is respectable.

https://www.uaudio.com/audio-interfaces/apollo-8.html

You obviously don’t spend enough time reading the expert opinions on Gearslutz :)

Rob, that’s true, I’m not on GearSlutz, just Taper. Care to share the consensus opinion on GearSlutz re the Apollo 8?

Rob, that’s true, I’m not on GearSlutz, just Taper. Care to share the consensus opinion on GearSlutz re the Apollo 8?

Gearslutz doesn’t do consensus :)

Jonathan Wyner is a terrific educator who has mastered music for top artists. He does a video series for iZotope. This one, uploaded yesterday, is the best effort that I’ve seen at explaining what loudness standards are and how they relate to metering. Everything that he says applies to film soundtracks. The discussion from 17:55 - 19:40 deals with a feature specific to iZotope software, but the rest of the video is of general application.


“Loudness Metering and Visualizations in Mastering”

Just a few minutes ago the MixPre-10 II, a Schoeps Colette, Ambient Quickpole QSM and a set of Tentacle Sync devices arrived in my studio. Hope I'll get some time for testing things out soon.
For now I am struggling to add the straight cable kit to the Ambient Quickpole

Just a few minutes ago the MixPre-10 II, a Schoeps Colette, Ambient Quickpole QSM and a set of Tentacle Sync devices arrived in my studio. Hope I'll get some time for testing things out soon.
For now I am struggling to add the straight cable kit to the Ambient Quickpole

Cool. Which Tentacle Sync devices did you get and how are you going to use them?

I'm also looking forward to hearing your experiences with the Tentacles (I think Rick uses them too).

I'm planning to get a set of Sync E boxes for multicamera: one for each camera and one for my (first generation) MixPre 6, which doesn't have internal timecode. And while I'm not planning to use lav mics very much I may get the new Track E when it comes out, since I can also use it for binaural recording. Given the limitations of my cameras and recorder, I'll have to keep the Sync E devices plugged into the cameras and recorder fulltime to ensure sync. That's expensive (three Tentacle Syncs=more than $1,000 Canadian dollars) and I've been wavering about just sticking to my usual technique of a clapperboard (or even clapped hands), but there are some situations such as concerts where that's not an option.

Cool. Which Tentacle Sync devices did you get and how are you going to use them?

It's the Sync E standard set + a single set (so 3 units) + some brackets, bodypack Y-adapter, microphone Y-adapter, tentacle to iphone setup, tentacle to BNC and DSLR cables

We are planning to put the whole audio package into our rental service and for our own productions when there are multi-cam shootings.

Congratulations, Robert. Brad, I agree the prices are high, but they do a very good job and last more than a day. The Tentacle app is free and essential. Since the iPhone resets its clock every time it boots, and the Tentacle app will reset the Tentacle clock to your iPhone via Bluetooth, it’s a very easy to be very accurate.

One way to reduce the cost of the Tentacle Sync E: I have proven the clock in the MixPre-6 II firmware release 6 is very good and once you have jammed it to a Tentacle vía Aux1, you can remove the Tentacle knowing the MixPre will hold accuracy for hours. I’d recommend a resync at a lunch break.

I do agree that each BMD camera should have a dedicated Tentacle driving the Timecode so there is no camera drift.

Once you start using these, whether for multicam or single camera with dual recording you’ll never go back.

I feel the same way about the MixPre. If I want ambient sound, I can put the Sennheiser MKH on the camera, but when I need quality, all my own mics feed the MixPre with, because I’m a stupid lazy loner, 32bit audio.

I have proven the clock in the MixPre-6 II firmware release 6 is very good and once you have jammed it to a Tentacle vía Aux1, you can remove the Tentacle knowing the MixPre will hold accuracy for hours. I’d recommend a resync at a lunch break.

I do agree that each BMD camera should have a dedicated Tentacle driving the Timecode so there is no camera drift.

Did you ever use a timecode generator to jam sync your original MixPre? If so, and if the timecode generator was removed, did the MixPre maintain accuracy? If there was drift, how much? Did the MixPre have to be jammed again if it was turned off and restarted?

I know what’s recommended, but if you tested I’d nevertheless love to know the answers to those questions.

Same questions for jamming the Pocket 4K, if you’re using that camera.

Thanks

Not a rigorous test, but I set the time in my Pocket 4K to a time reference several days ago, and the camera has continually lost time, maybe 2 seconds per day. I would expect that after a jam sync my camera would lose maybe 2 frames per hour. But as I said, this has not been a rigorous test, offered FWIW.

Not a rigorous test, but I set the time in my Pocket 4K to a time reference several days ago, and the camera has continually lost time, maybe 2 seconds per day. I would expect that after a jam sync my camera would lose maybe 2 frames per hour. But as I said, this has not been a rigorous test, offered FWIW.

May not be rigorous, but it’s a pretty clear message :)

Rob, I only have the MixPre-6 II, not the original. I tested turning the MixPre off and on again and it maintains Timecode. Again I would still reset it after several hours. I’ll turn it Off for a few hours now and check later in the day.

I concur with Michael’s test on the camera. It will drift a little over time and best to leave the Tentacle on the camera and have perfect results.