this could be a REALLY interesting recorder:
https://www.youtube.com/watch?v=MGebCVb1CUM
thumbs up.
rob
smalltalk productions/nyc
https://www.youtube.com/watch?v=MGebCVb1CUM
thumbs up.
rob
smalltalk productions/nyc
NAB release: Zoom F6 - a recorder where no need to set the gain! 32bit float-really?
- Thread starter robfilms
- Start date
That's pretty awesome- no clipping and no need to set gain!
The same concept can be applied to video or any analog to digital conversion process. For audio 24 bits (3 bytes) becomes 32 bits (4 bytes); only a small increase in file size. If going from 16-bit raw video, it's only 2x to get to 32-bits, with efficient compression (e.g. wavelet) could completely change video shooting too: dual ADC sensors prevent over/under exposure providing 32-bit float per Bayer pixel (would need some kind of auto ISO + electronic ND to prevent sensor overload + metadata could provide all the information needed to get perfect exposure (since ISO/ND ramping would happen during what would normally be over exposure events).
The same concept can be applied to video or any analog to digital conversion process. For audio 24 bits (3 bytes) becomes 32 bits (4 bytes); only a small increase in file size. If going from 16-bit raw video, it's only 2x to get to 32-bits, with efficient compression (e.g. wavelet) could completely change video shooting too: dual ADC sensors prevent over/under exposure providing 32-bit float per Bayer pixel (would need some kind of auto ISO + electronic ND to prevent sensor overload + metadata could provide all the information needed to get perfect exposure (since ISO/ND ramping would happen during what would normally be over exposure events).
Bassman2003
Veteran
Very cool. Now that somebody is "going there" I am sure the industry will adapt. 32bit for cameras is probably been planned and on hold, waiting in the wings as they need to sell us something in the future... 
Matt Gottshalk
Resident Preditor
Audio "Raw".
Noiz2
Sound Ninja
Mixed feelings though probably great for the R&G crowd. FCP6 and beyond used 32bit float and I mostly hated it for post, but that might just have been their interpretation. The problem for post was that you kind of had control taken away and left in the hands of the algorithm. But I can see a lot of upside for it as a capture format.
As the programers explained it, it was fairly confusing. Where is the clip point? "there isn't one", But where is zero? "well it moves around", how do I set levels? "you don't have to", but what if I want to? "you can't" "well you can set relative levels but it's always adjusting the absolute levels so you never clip". But what if I want to clip something? "you can't", why not? "it won't let you".
So for in the field I can see it being a boon, but in post, and even for FX not always great. To be fair for FX recording if you intentionally clipped you almost always want that on the analog side so then you would do that with a preamp and the recorder would do a great job capturing. But you know if everything goes 32 bit float then there will be plugins that will do any clipping you want so I'm probably complaining about the demise of buggy whips.
As the programers explained it, it was fairly confusing. Where is the clip point? "there isn't one", But where is zero? "well it moves around", how do I set levels? "you don't have to", but what if I want to? "you can't" "well you can set relative levels but it's always adjusting the absolute levels so you never clip". But what if I want to clip something? "you can't", why not? "it won't let you".
So for in the field I can see it being a boon, but in post, and even for FX not always great. To be fair for FX recording if you intentionally clipped you almost always want that on the analog side so then you would do that with a preamp and the recorder would do a great job capturing. But you know if everything goes 32 bit float then there will be plugins that will do any clipping you want so I'm probably complaining about the demise of buggy whips.
Bassman2003
Veteran
I see your point. Kind of like shooting a wide shot in 4k and cropping in post. In some ways it is less work to just have a person frame it properly live. I see this as a great capture format with a limiter driven final delivery. So whatever you come home with it can be reflated or dialed back to the delivery output levels. Easy in the field but maybe more work in post.
For almost everyone who works in the sound department then the Zoom F6 is not relevant for them, as it gives them almost no benefits (except maaaaybe as a drop recorder), is a nonstandard workflow to handover to post, and is otherwise a very big step backwards from the already very affordable Zoom F8n (the F6 is even a step back for us from the F4).
Do I see a place for the Zoom F6 for ultra low budget / casual solo shooters and the like who are giving low priority / attention to audio and will handle the entire workflow themselves? Yes, in fact if all the details work out and the F6 lives up to its promise then the F6 might just be perfectly tailored for them.
Do I see a place for the Zoom F6 for ultra low budget / casual solo shooters and the like who are giving low priority / attention to audio and will handle the entire workflow themselves? Yes, in fact if all the details work out and the F6 lives up to its promise then the F6 might just be perfectly tailored for them.
Jim Feeley
Veteran
IME, the Zoom F-series preamps are OK. I think the F6 preamps are pretty much the same as the earlier ones, right? And the dual A/D conveters is at least clever (and perhaps more than). The 32-bit format sounds useful, but I'll want to hear how the preamps and A/D converters perform before I get too excited. Cool ideas, though!
They're more than "ok"! Zoom F series is indistinguishable by people in a blind A/B test from Sound Devices.
And yes, the F6 has been confirmed to have the same pre amps as the F4/F8/F8n
It works as expected- setting the gain only affects the floating point scale. Provided the mic itself doesn’t clip, it’s unclipable. If set too low, won’t add noise when scaled in post:
Also cool how small it is compared to the F4!
While I prefer the AKM ‘Velvet Sound’ converters in SD products, this is indeed a game changer level innovation (I have a Zoom F4 and SD USBPre2 and can hear fuller sound on the SD (702 series audio topology). What converter is Zoom using?).
Looks like a Sony L battery (I and most AV people have many of these)- if no issues with power consumption, an advantage over the SD MixPre series (L sled has some issues).
EDIT: 32-bit is limited to 96kHz; most video related audio is 48kHz so not an issue for dialog (Foley perhaps). At 192kHz** the 10dB look ahead limiter is still available. I recently had a chance to 'test' this limiter at the limit with the F4, and after 10dB is passed, it's not nearly as nice as SD's analog limiter (after 10dB is straight up clipping). For pure sound quality SD still has the edge (to my ears*), however for versatility and not having to ever set gain (that's truly awesome), compact size, and easy power options, the F6 is a breakthrough product for audio.
* curious if that 'edge' can be matched with EQ? The F4 sounds a bit thin compared to the USBPre2, in other words sounds less full (missing low and mid info) and less detailed (less high info). Like comparing a good dynamic headphone to an electrostatic headphone for the same material. I realize in a world where the average listener is using perhaps a sub $1 driver (or much less if using ear pods) that the difference in quality between these ADCs may not be apparent, however as a 'performance enthusiast' I appreciate the higher quality listening experience. That said, I only use e.g. my Stax electrostatics occasionally. The ear-brain adapts and pretty soon what was once high quality becomes normal, and e.g. ear pods on my phone sound like garbage. Switching back and forth thus keeps things interesting. Or, if you want to go off the deep end and constantly buy new gear, check out Z Reviews lol: https://www.youtube.com/user/ZeosReviews
** On B&H there was a post about a bug with MixPre's at 192kHz- not clear if that was ever resolved. I've only briefly tested 192kHz (all recorders owned) for voice and used occasionally for instruments / audio effects (which ultimately got (very high quality e.g. Ideal Bandlimited Sinc Interpolation) resampled to 44.1/48kHz: goal was 'closest to analog' capture as possible).
Zoom vs. SD: I can hear a clear difference in fullness & smoothness (Schoeps CMC641- we use this mic too), less harsh/brittle compared to the Zoom on my MBP speakers! https://youtu.be/PM6VOEYpm4g?t=102 (queued to the demo start). See the comments- matches my experience too. I think I found the 'blind test' on a forum from a guy in Japan- I couldn't hear much difference in his tests... I have both devices, no bias, SD has better preamps. That said, there's some flaws with the MixPre series that prevented me from purchasing.
Also cool how small it is compared to the F4!
While I prefer the AKM ‘Velvet Sound’ converters in SD products, this is indeed a game changer level innovation (I have a Zoom F4 and SD USBPre2 and can hear fuller sound on the SD (702 series audio topology). What converter is Zoom using?).
Looks like a Sony L battery (I and most AV people have many of these)- if no issues with power consumption, an advantage over the SD MixPre series (L sled has some issues).
EDIT: 32-bit is limited to 96kHz; most video related audio is 48kHz so not an issue for dialog (Foley perhaps). At 192kHz** the 10dB look ahead limiter is still available. I recently had a chance to 'test' this limiter at the limit with the F4, and after 10dB is passed, it's not nearly as nice as SD's analog limiter (after 10dB is straight up clipping). For pure sound quality SD still has the edge (to my ears*), however for versatility and not having to ever set gain (that's truly awesome), compact size, and easy power options, the F6 is a breakthrough product for audio.
* curious if that 'edge' can be matched with EQ? The F4 sounds a bit thin compared to the USBPre2, in other words sounds less full (missing low and mid info) and less detailed (less high info). Like comparing a good dynamic headphone to an electrostatic headphone for the same material. I realize in a world where the average listener is using perhaps a sub $1 driver (or much less if using ear pods) that the difference in quality between these ADCs may not be apparent, however as a 'performance enthusiast' I appreciate the higher quality listening experience. That said, I only use e.g. my Stax electrostatics occasionally. The ear-brain adapts and pretty soon what was once high quality becomes normal, and e.g. ear pods on my phone sound like garbage. Switching back and forth thus keeps things interesting
. Or, if you want to go off the deep end and constantly buy new gear, check out Z Reviews lol: https://www.youtube.com/user/ZeosReviews** On B&H there was a post about a bug with MixPre's at 192kHz- not clear if that was ever resolved. I've only briefly tested 192kHz (all recorders owned) for voice and used occasionally for instruments / audio effects (which ultimately got (very high quality e.g. Ideal Bandlimited Sinc Interpolation) resampled to 44.1/48kHz: goal was 'closest to analog' capture as possible).
Zoom vs. SD: I can hear a clear difference in fullness & smoothness (Schoeps CMC641- we use this mic too), less harsh/brittle compared to the Zoom on my MBP speakers! https://youtu.be/PM6VOEYpm4g?t=102 (queued to the demo start). See the comments- matches my experience too. I think I found the 'blind test' on a forum from a guy in Japan- I couldn't hear much difference in his tests... I have both devices, no bias, SD has better preamps. That said, there's some flaws with the MixPre series that prevented me from purchasing.
Last edited: