The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Another new SD Recorder - 10T Recorders, Players & Tape Machines
Old 6th October 2017
  #31
Lives for gear
 
tourtelot's Avatar
Quote:
Originally Posted by TMetzinger View Post
I'll agree that 32bit conversion doesn't offer much practical value
That's correct. I agree and that's NOT what Plush said. Seems to have a grudge.

D.
Old 6th October 2017
  #32
Lives for gear
 
Plush's Avatar
Quote:
Originally Posted by tourtelot View Post
Why do you say that? You keep uttering the same thing and have shown no reason for anyone to believe it. Cite your source please. Oh and being derogatory for no reason. Bush league.

D.
If you read my post instead of skim it, you would not defend the indefensible. My critique is about how they promote the new machine. Unit has the same performance as any other well designed modern recorder. My point is obvious since audio performance is still around 20 bit encoding. Consult dynamic range figures.
Old 6th October 2017
  #33
Lives for gear
 

Quote:
Originally Posted by tourtelot View Post
Plush-

Just curious about you comment on 32-bit A to D convertors. A quick search shows many chip manufacturers at least claiming that their convertors are 32-bit. Here is but one example.

D.
Did you check the spec sheet? The dynamic range figures were quoted at 61 samples per second and 4000 samples per second. This is for a dc/low frequency ac measuring instrument, not audio. They can way oversample and then trade oversampling for higher resolution.

Now audio-wise, the new Benchmark DAC claims the use of a 32 bit converter - basically because on its own it gives the 3dB improvement in noise performance that they used to get by averaging the output of 4 24-bit DACs. So maybe the "32-bit" technology delivers some small improvement in the audio range, but as Plush rightly points out, we are still between the "rock" of maximum input before clipping, and the "hard place" of inherent/irreducable thermal noise* in the analog input components (ie., up to the sample and hold circuit). And right now, that adds up to a maximum dynamic range of around 123dB or about the inherent quantising noise of 20bit conversion.

But if the IC guys are making chips with "32-bit technology" at affordable prices and they perform as well or better than "24-bit" chips, then use 'em and flaunt it. Not sure that at that level the ADC's contribution to "sound" is significant in contrast to all the other design factors involved. The market (bless its heart!) will decide.

(* If you have the time and a cryogenic setup, you can always run the analog circuitry at close to 0° Kelvin. That will almost eliminate the thermal noise, but you will still have all the other "imperfection" noises to deal with. It's a hard life ... )
Old 6th October 2017
  #34
RPC
Gear Maniac
 

Quote:
Originally Posted by RPC View Post
Okay, I'm curious: what are the three little holes next to the channel 8 knob? Slate mike?
Just listened to the intro video - it is indeed a slate mike, engaged via the */** custom function switch above the joystick.
Old 6th October 2017
  #36
Gear Addict
Quote:
Originally Posted by hughesmr View Post
On-board backup. Thank you!
Just to clarify, after talking with Sound Devices, the 10T has an Auto-Copy feature that will automatically copy
a recording to an attached
thumb drive ( SD recommends a "Cruzer" because of the speed) any time the machine is stopped. If you start recording
again before the copying is finished it remembers its place and resumes again
when next in stop mode. It's a nice feature BUT it's not the same safety
as recording simultaneously to 2 different media with the 788, e.g.
Bill
Old 6th October 2017
  #37
Lives for gear
 

Thanks. Was thinking of getting one of these for minimal kit location recording, but the lack of preroll buffer is a huge deal breaker for me. Probably the most loved feature of my Nagra VI. Hopefully this will change with a firmware update.
Old 7th October 2017
  #38
Gear Addict
 

Quote:
Originally Posted by tourtelot View Post
Plush-

Just curious about you comment on 32-bit A to D convertors. A quick search shows many chip manufacturers at least claiming that their convertors are 32-bit. Here is but one example.

D.
Maybe look at the rest of the specifications? See anything that indicates that this converter isn't for audio?
Old 8th October 2017
  #39
Lives for gear
 
tourtelot's Avatar
Yeah, okay. But are there no 32-bit audio convertors? Is Sound Devices lying or do they use a 32-bit convertor?

Anyone from SD care to join in?

I just have a problem with name calling. Maybe I'd be better off just ignoring such inappropriate posts. I do so on other topics but bad behavior seems to have become acceptable in a certain part of the world these days. I'd like to see less of it.

As always, just my $.02 (and as we know, that and five bucks will buy you a cup of coffee.)

D.
Old 8th October 2017
  #40
Gear Addict
 

Quote:
Originally Posted by tourtelot View Post
Oh and here is a TI chip stated as being a 32-bit audio A to D. Maybe this one isn't "the one" but it says it's an audio A to D. FWIW.

D.
Well this certainly isn't the one .... as it's a DAC, not an ADC.

I don't really care one way or another about what SD claims .... but if you're going to go through the trouble of looking through data sheets and posting screen shots ... best to read more carefully, and understand them.
Old 8th October 2017
  #41
Lives for gear
 
tourtelot's Avatar
I'll just shut up now. Sorry all, to go on. :(

D.
Old 8th October 2017
  #42
RPC
Gear Maniac
 

Well, Asahi Kasei (AKM) do bill the AK5397 I mentioned upstream as a 32 bit audio ADC, though actual performance is in the 21+ bit range. A real 32 bit converter would have 192dB dynamic range, which would pretty much cover everything from hearing two molecules bump into each other to a Saturn V liftoff at 100 feet!
Old 8th October 2017
  #43
Lives for gear
It seems to me that the proof is in how it functions in actual use. So evidently the auto copy doesn't function as a real time second backup? Even my cameras backup to a second internal card. And you have to buy a sled to use an L battery?
Old 8th October 2017
  #44
Gear Addict
 

The whole discussion on 32 bit is rather futile. Tourtelot is right that it contains a 32 bit converter, because it can and will indeed output 32 bit data. Plush is right that these 32 bit long data words will contain something like 21 useful bits that actually give a faithful representation of the analogue signal. The 11 remaining bits present just superfluous nonsense, but they will be put out by the dac as well. They have no real impact on what we record or what we hear. As technology progresses, the number of meaningfull bits will progress further, but with very small steps.

People like Plush want to underscore that the 32 bit figure can be misleading, and that one should not expect a real 32 bit resolution present within that 32 data stream from the adc.

The same goes for all the dac chips that are billed as 32 bit chips. It sounds impressive, but they are not necessarily better than 24 bit dac chips...
Old 9th October 2017
  #45
Gear Maniac
 

And, as you are careful to say, even a 24 bit ADC does not give you 24 bits of useful (non-random) audio data.

Just too much "down in the weeds".

DG
Old 9th October 2017
  #46
Lives for gear
 

Question for the experts: are there fewer numerical errors maintained with 32 bit DAC, even though the final output is not 32 bit?
Old 9th October 2017
  #47
Gear Addict
 

Quote:
Originally Posted by aracu View Post
Question for the experts: are there fewer numerical errors maintained with 32 bit DAC, even though the final output is not 32 bit?
Well I certainly wouldn't consider myself an expert . . . but in short, no.

Virtually all modern audio converters are NOT "really" 32-bit, 24-bit, 16-bit or whatever . . . the actual converter is of the delta-sigma (a.k.a. sigma-delta) design. This means (for an ADC) that PCM data of the standard word-lengths we're discussing is a result of the decimation and digital filtering applied after the converter. In it's simplest form a delta-sigma converter is a 1-bit design . . . the data before the decimation/filtering would be a form of DSD, and the data afterwards is usually PCM. In a commercial integrated-circuit ADC, all of this happens on the chip itself, and the output data is available to the circuit designer in one of several standard (inter-IC) formats of PCM and/or DSD (i.e. I2S, end-justified, TDM, etc.). In reality the exact internal form of the converters themselves may be multi-bit delta-sigma, have more than one modulator/converter operating in parallel, or some really clever proprietary scheme . . . so there's likely to be some internal processing required even to achieve a standard DSD format.

It's important to remember that there has yet to be a commercial audio ADC chip developed, the extent of whose basic converter performance cannot be expressed in PCM with a succession of binary words of 21 or 22 bits. So while the market is dominated by "24-bit audio ADCs" . . . the "24-bit" part really has nothing to do with its internal conversion architecture, only the chosen output data format. The presence of a few extra bits in the output data are redundant. A "32-bit audio ADC" is of course exactly the same, just with more redundant bits (i.e. a "00032-bit" converter) . . .

The stages of decimation and filtration do indeed contribute very much to the sound quality of the ADC, but its execution likewise need not have any correlation to the converter architecture or the chip's output format. For example the PCM4222 has a 6-bit delta-sigma converter, with on-chip decimation and filtering for PCM-formatted output, as well as processing for a 1-bit DSD output. It also has a 6-bit parallel digital output directly from the converter . . . which a designer could use to i.e. pass to an external FPGA and design their own decimation and filtering. The output of such an FPGA could be expressed in any desired form . . . so if they chose to format it as 64-bit left-justified serial data . . . you'd have a "64-bit audio ADC" every bit as much as the AK5397 is a "32-bit audio ADC" used by itself . . . even though the datasheet for the PCM4222 specifies it as a "24-bit" part.

That's not to say that there might not be some circuit-development and format-consistency conveniences to the AK5397's 32-bit output format. I can envision a device with multiple DSPs, processing and mixing together multiple audio tracks, parts of the chain could certainly benefit from being able to exchange 32-bit audio, and perhaps it might streamline the design and layout if ADC chips could share a common TDM interface bus with such data. But provided there's proper design and arithmetical diligence, there's again no direct audio-quality benefit for a an ADC's output format to be expressed in a 32-bit format, if it doesn't have enough signal resolution to require such numerical resolution.

In fairness to Asahi Kasei, their use of "32-bit" in nomenclature is absolutely consistent with other manufacturers' use of "24-bit", and they don't appear to be resting their laurels for the AK5397 purely on it being a "32-bit" product. On the other hand, ESS certainly leveraged some marketing cachet with the "32-bit" Sabre DAC . . . and as a company don't seem to be as forthcoming with data and eval fixtures as the rest of the industry. The Sabre is a great-sounding DAC for sure . . . but IMO it's not clearly ahead of some of the other premium parts from other manufacturers . . . and certainly not to the extent that being "the first 32-bit audio DAC" would imply.

As far as Sound Devices is concerned . . . I think the use of "32-bit" as a marketing bullet point for audio professionals feels a bit cheap. It's similar to expressing their preamp EIN in dBV to make for a bigger number than if they used dBu. But on the latter at least they still used a 150 ohm source instead of a shorted input, and didn't just say "-130dB", both of which are also ploys in specsmanship. Oh well . . . I still own and generally like their stuff . . .
Old 10th October 2017
  #48
Gear Addict
 

Quote:
Originally Posted by kirkus View Post
Well I certainly wouldn't consider myself an expert . . . but in short, no.

Virtually all modern audio converters are NOT "really" 32-bit, 24-bit, 16-bit or whatever . . . the actual converter is of the delta-sigma (a.k.a. sigma-delta) design. This means (for an ADC) that PCM data of the standard word-lengths we're discussing is a result of the decimation and digital filtering applied after the converter. In it's simplest form a delta-sigma converter is a 1-bit design . . . the data before the decimation/filtering would be a form of DSD, and the data afterwards is usually PCM. In a commercial integrated-circuit ADC, all of this happens on the chip itself, and the output data is available to the circuit designer in one of several standard (inter-IC) formats of PCM and/or DSD (i.e. I2S, end-justified, TDM, etc.). In reality the exact internal form of the converters themselves may be multi-bit delta-sigma, have more than one modulator/converter operating in parallel, or some really clever proprietary scheme . . . so there's likely to be some internal processing required even to achieve a standard DSD format.

It's important to remember that there has yet to be a commercial audio ADC chip developed, the extent of whose basic converter performance cannot be expressed in PCM with a succession of binary words of 21 or 22 bits. So while the market is dominated by "24-bit audio ADCs" . . . the "24-bit" part really has nothing to do with its internal conversion architecture, only the chosen output data format. The presence of a few extra bits in the output data are redundant. A "32-bit audio ADC" is of course exactly the same, just with more redundant bits (i.e. a "00032-bit" converter) . . .

The stages of decimation and filtration do indeed contribute very much to the sound quality of the ADC, but its execution likewise need not have any correlation to the converter architecture or the chip's output format. For example the PCM4222 has a 6-bit delta-sigma converter, with on-chip decimation and filtering for PCM-formatted output, as well as processing for a 1-bit DSD output. It also has a 6-bit parallel digital output directly from the converter . . . which a designer could use to i.e. pass to an external FPGA and design their own decimation and filtering. The output of such an FPGA could be expressed in any desired form . . . so if they chose to format it as 64-bit left-justified serial data . . . you'd have a "64-bit audio ADC" every bit as much as the AK5397 is a "32-bit audio ADC" used by itself . . . even though the datasheet for the PCM4222 specifies it as a "24-bit" part.

That's not to say that there might not be some circuit-development and format-consistency conveniences to the AK5397's 32-bit output format. I can envision a device with multiple DSPs, processing and mixing together multiple audio tracks, parts of the chain could certainly benefit from being able to exchange 32-bit audio, and perhaps it might streamline the design and layout if ADC chips could share a common TDM interface bus with such data. But provided there's proper design and arithmetical diligence, there's again no direct audio-quality benefit for a an ADC's output format to be expressed in a 32-bit format, if it doesn't have enough signal resolution to require such numerical resolution.

In fairness to Asahi Kasei, their use of "32-bit" in nomenclature is absolutely consistent with other manufacturers' use of "24-bit", and they don't appear to be resting their laurels for the AK5397 purely on it being a "32-bit" product. On the other hand, ESS certainly leveraged some marketing cachet with the "32-bit" Sabre DAC . . . and as a company don't seem to be as forthcoming with data and eval fixtures as the rest of the industry. The Sabre is a great-sounding DAC for sure . . . but IMO it's not clearly ahead of some of the other premium parts from other manufacturers . . . and certainly not to the extent that being "the first 32-bit audio DAC" would imply.

As far as Sound Devices is concerned . . . I think the use of "32-bit" as a marketing bullet point for audio professionals feels a bit cheap. It's similar to expressing their preamp EIN in dBV to make for a bigger number than if they used dBu. But on the latter at least they still used a 150 ohm source instead of a shorted input, and didn't just say "-130dB", both of which are also ploys in specsmanship. Oh well . . . I still own and generally like their stuff . . .
Hi Kirkus, personally, I too have always been a bit disappointed with the Sabre dac. The 32 bit hype gave them more publicity than they should have earned by their sound quality. Strangely enough the new 9038 sounds phenomenal. I consider its sound a breakthrough, although they are essentially a package of a bunch of 9018 chips... I hope we will see a similar breakthrough with ad chips.
Old 14th October 2017
  #49
Lives for gear
 
iksrazal's Avatar
Can the 10t overdub? Seems to lack outputs, preferably 8 outs?

Those are the two features I really need for a recorder with decent time code support. I don't use computers in my work flow ymmv.
Old 14th October 2017
  #50
Lives for gear
 
jnorman's Avatar
Mmmmkay, pretty weird thread, guys...
Old 14th October 2017
  #51
Lives for gear
 
Plush's Avatar
Quote:
Originally Posted by jnorman View Post
Mmmmkay, pretty weird thread, guys...
Not weird if one wants accuracy and freedom from spurious claims.

Studio Sound Magazine used to bust companies like this for inaccurate technical claims. John Watkinson, the author of The Art of Digital Audio would never tolerate the marketing claims offered by the Wisconsin company.
New Reply Submit Thread to Facebook Facebook  Submit Thread to Twitter Twitter  Submit Thread to LinkedIn LinkedIn  Submit Thread to Google+ Google+  Submit Thread to Reddit Reddit 
 
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump