Personal View site logo
Sound Devices MixPre-3 and MixPre-6
  • 71 Replies sorted by
  • alcomposeralcomposer

    Tascam DR70d implementation of dual level recording is very limited (no pun intended):

    This function cannot be enabled if more than two channels are selected for recording or if the two selected channels prevent the recording of a stereo pair (for example, CH1 and CH3).

    Seriously getting annoying to have features that don't use the whole device, (4 channels etc). So many stupid limitations to remember.

    What is nice about MixPre is that the limiter works flawlessly on all 4 channels. Simple to remember. Especially in stressful onset situations, we need gear we can just turn on - press record - and get usable material from.

  • Vitaliy_KiselevVitaliy_Kiselev

    This function cannot be enabled if more than two channels are selected for recording or if the two selected channels prevent the recording of a stereo pair (for example, CH1 and CH3).

    But it fully proves my theory.

    It also explains why video approach is wrong. As he uses loud voice well above initial level clipping, it is above 12db such limiter has. In most situations were levels are set right initially it won't be possible to tell difference if signal is within this 12db.

  • alcomposeralcomposer

    So your view is that the MixPre's analog limiter isn't that exciting? Better to go with Zoom F4/8? (I'm kind of over the Tascam for a few 'firmware' related reasons.

  • Vitaliy_KiselevVitaliy_Kiselev

    So your view is that the MixPre's analog limiter isn't that exciting? Better to go with Zoom F4/8?

    You must think yourself here and consider your needs and check more info about limiters implementation behavior.

  • alcomposeralcomposer

    I know that the DR100MKii has an analog limiter, and I loved using that device, especially unmanned for performance audio. So personally an analog limiter is very reassuring that the quality of the audio will be maintained in a bad situation.

    So on that level the MixPre would be very useful for me, as I do lots of productions like that. MixPre looks like the best of both worlds (analog and digital) until digital limiters are designed correctly as you pointed out VK.

    Also the volume knobs on the DR70d are digital as well, so you can't simply change volume during recording as it will cause audible changes in volume (volume terracing) which I am assuming as the MixPre is analog volume then that would not be the case.

    Once again, this probably and should work well on a digital system, but its just not implemented in a way that is 'analog'. Pity. 99% of everything I do audio wise is digital (in DAW) just don't understand why that can't work on portable device?

  • Vitaliy_KiselevVitaliy_Kiselev

    I know that the DR100MKii has an analog limiter, and I loved using that device, especially unmanned for performance audio. So personally an analog limiter is very reassuring that the quality of the audio will be maintained in a bad situation.

    It depends on that you understand under 'analog limiter'. As such recorders use Digital variable-gain amplifier https://en.wikipedia.org/wiki/Variable-gain_amplifier or digital volume control paired with amplifier. So it can be just some form of marketing term.

    Also the volume knobs on the DR70d are digital as well, so you can't simply change volume during recording as it will cause audible changes in volume (volume terracing) which I am assuming as the MixPre is analog volume then that would not be the case.

    I am not sure if MixPre pots are just resistors and not digitally control gain. Needs checking. But considering DR70d it is not issue of such general design, but issue either of specific hardware solution or firmware. If you have big enough steps you won't be able to tell easy digital control from analog, no way. And in the long run digital can be made cheaper and be much more stable.

  • alcomposeralcomposer

    Looks like you can also connect Keyboard to MixPre and log metadata. Could be useful onset for more complex productions (from manual):

    The use of standard USB computer keyboards is supported. When using the MixPre-6 in situations requiring extensive metadata entry, titling and wired remote control capability, a USB keyboard may be desirable.

    http://cdn.sounddevices.com/download/guides/MixPre6-UG_en.pdf

  • alcomposeralcomposer

    Looking at comparison of zoom f4/8 series, there is also a similar limitation to track count and safety -12db track:

    When using safety track record feature the unit will only be able to use half its inputs. So the f8 gets reduced to 4 input etc. So if you are using this "feature" there is little difference between the specs of the f8 and mix pre 6.

  • Vitaliy_KiselevVitaliy_Kiselev

    @alcomposer

    When using safety track record feature the unit will only be able to use half its inputs. So the f8 gets reduced to 4 input etc. So if you are using this "feature" there is little difference between the specs of the f8 and mix pre 6.

    We always use Tascam safety track feature on exhibitions. Very useful thing.

  • GrimorGrimor

    Sorry if i make offtopic here, but whats is the correct workflow when you have a safe -12db channel and main channel have clipped peaks. Remplace only the clipped peaks with non clipped parts of the safe track? Or change the whole clip and match gains ?

  • Vitaliy_KiselevVitaliy_Kiselev

    @Grimor

    Depends on content.

  • alcomposeralcomposer

    @Grimor if you have the backup -12db then I would simply match gains and replace the region in question. Replace as large a region as possible (as there may be intra-sample clipping on other parts of the hot signal also).

    I actually don't know of software that would be able to analyze the file to replace only the clipped waveforms, and anyway, if its clipped then the level is above 0dbfs, so you would need to reduce the level - or limit it anyway.

    Also remember that limiting is only one aspect of audio recorders, Amplifier quality is also very important, and gain staging.

  • Vitaliy_KiselevVitaliy_Kiselev

    @alcomposer

    Sometimes if levels where way off or just one guy periodically talks very loud you can be forced to use all backup file, just faster. If you are not recording in studio or quiet place no one will even notice.

  • alcomposeralcomposer

    Yup, hence my comment 'as large a region as possible'. Its not worth the time trying to fix peaks, only if there is also very quiet content between issue areas.

    On side note, I really want a MixPre in white:

    image

    http://www.mixonline.com/news/news-articles/vatican-selects-sound-devices-preferred-audio-partner/429856#.WTax8oe1TTw.twitter

    vatican2.jpg
    400 x 300 - 46K
  • alcomposeralcomposer

    Good news @Vitaliy_Kiselev, looks like from this video Gain is digital. (Suppose it should be ok if it is smooth transition between volume levels?)

    Checkout beginning of video and gain value setting.

    More info about knobs and settings, Sound-devices has already saved a life:

    Already bugs in firmware, SD will fix (stated by SD):

  • GrimorGrimor

    Thanks @alcomposer. Last day, shooting with childrens, some of them shouted at the microphone for no apparent reason. kid stuff. Record with my pmd661 in dual stereo mode.One of the channels was -12db. I´ll try to fix.

  • alcomposeralcomposer

    Yeah, I have had that issue before, (with actors actually). Sometimes even -12db can be distorted, if only mildly. Use a repair software if it is - like RX6. There is an 'De-Clip' function, the only one I have found that actually makes the audio usable afterwards, its magic.

    I haven't looked at the pmd661 @Grimor. File encryption? Must be good for police / military usage?

  • alcomposeralcomposer

    I have contacted Sound Devices and they assured me the Rotary encoders are smooth regardless of what mode they are set to.

    So hopefully that should mean that if we change levels during recording (riding the gain depending on audio situation) there should be no problem with terraced volume or glitches.

  • Vitaliy_KiselevVitaliy_Kiselev

    Sometimes even -12db can be distorted, if only mildly. Use a repair software if it is - like RX6. There is an 'De-Clip' function, the only one I have found that actually makes the audio usable afterwards, its magic.

    Example can be Canon C200 interview audio from CG, source had not large but many clippings even at safe track.

  • GrimorGrimor

    My pmd661 is Mk1 so no file encryption.

  • alcomposeralcomposer

    @Grimor ah. Yeah I only mentioned it as when I looked it up that was a "major feature". I don't quite know when that would be useful, but for police interviews yes it probably would.

    @Vitaliy_Kiselev I believe that the Zoom F8 allows setting independent gain for safe tracks. So user can set level much lower than -12. Or in another situation one could use safe dual gain to "cheat" and record soft material at same time as loud material. (As long as they are not happening at same time obviously) for example if an actor had dialogue that went from whispering to yelling.

  • Vitaliy_KiselevVitaliy_Kiselev

    I believe that the Zoom F8 allows setting independent gain for safe tracks. So user can set level much lower than -12. Or in another situation one could use safe dual gain to "cheat" and record soft material at same time as loud material. (As long as they are not happening at same time obviously) for example if an actor had dialogue that went from whispering to yelling.

    Btw whole thing with safe tracks, instead of 32bit WAVs comes from Zaxcom guys, as I believe. Plus it is more easy for software guys to do :-)

  • alcomposeralcomposer

    Do Zaxcom have exclusive 32bit WAV license? (I never thought I would one day need 32bit, but here we are) :-$

    24 bit should allow 144db dynamic range. ( 24 bit * 6 db) Limiting factor is analog side. However with "safe track" if the zoom f8 actually does use a seperate analog stage for recording, should give better result than 32 bit, as that one 32 bit AD will still be sampling the same analog gain stage.

    (Obviously if equipment becomes 32 but it may actually have more DR anyway due to having better analog side)

    Problem is just like camera sensors. 120db is from clipping to noise floor. Just because we have 120db doesn't mean we can use all that range.

    Now, MixPre does have 32 bit AD, but dithers to 24. Considering many 24 bit AD's are not even real 24bit- (more like 20 bit) that could be interesting itself.

  • Vitaliy_KiselevVitaliy_Kiselev

    @alcomposer

    Do Zaxcom have exclusive 32bit WAV license? (I never thought I would one day need 32bit, but here we are) :-$

    I believe they have patent on so called clipless audio recorder, this means multiple ADC and algorithm to mix them to produce single file.

    As for 32bit WAV, it must long be standard for recorders and intermediate editing format. With 16 and 24bit as delivery formats.

  • alcomposeralcomposer

    So with a zoom f8 you get a bit of the clips less audio recorder tech (dual ADC) then we simply need software to make a HDR file out of two audio files. (Like in photography)

Add a Comment