- Thread Starter
- #101
It is not going to show up in the context of an acoustic measurement in non-silent room.Curious if there's a measured difference between analog in and digital in.
It is not going to show up in the context of an acoustic measurement in non-silent room.Curious if there's a measured difference between analog in and digital in.
Unusual or not, both Andrew Jones and the founder of Hedd, Klaus Heinz, say it's quite awkward, and decide to do the opposite with their current speakers.
But then again, it it takes digital as you mention via that AES3, what kind of digital signal is that? it then has a DAC inside? the dsp thing is done with that dac, or after it's turned into analog?
You probably need to consider the target market for these speakers which is professional studio mixing and mastering. The advantage of DSP in the monitors is that the digital mixer audio interface allows for endless options when it comes to adding presets, templates, etc and ease of transfer projects from one studio to another. and that the signal is sent as digital purely prior to DSP that is calibrated to speakers and the studio room. The DSP built-in to the speakers makes sense to me in that scenario.
IMHO, Klaud Heinz's approach is what's awkward Due to his insistence on avoiding DSP in the speaker itself, his systems require the user to run the output from their computer source through an FIR filter before sending it to the speakers so as to achieve the superior phase performance that could have been achieved inside the speaker if Hedd simply adopted the standard practice of using DSP crossovers in the speakers. Not to besmirch Mr Heinz's skill and reputation as a speaker designer, which is formidable - I just strongly disagree with him on this point.
From a practical perspective, there's no difference between AES3 and SPDIF. They both transmit digital audio and clock information between components. It's preferred in pro environments for various reasons, some practical and some historical. There's no reason you can't use it at home.
Regarding the location of the DAC in the system, this is in essence how the signal path flows:
Digital source --> AES3 input (digital) --> DSP --> DACs --> amps --> speaker drivers
In other words, the signal remains wholly in the digital domain until after DSP processing. This is the optimal signal path primarily for the following reasons:
PS: a few home audio companies are now also using AES3 rather than SPDIF in their speaker systems. Kii is one that comes to mind.
- Only one DA conversion takes place.
- The DSP enables exponentially more precise control and complexity in the implementation of the crossover than analogue filters would allow.
- The user does not need to concern themselves with analogue gain staging, since the entire analogue chain is inside the speaker.
IMHO, Klaud Heinz's approach is what's awkward Due to his insistence on avoiding DSP in the speaker itself, his systems require the user to run the output from their computer source through an FIR filter before sending it to the speakers so as to achieve the superior phase performance that could have been achieved inside the speaker if Hedd simply adopted the standard practice of using DSP crossovers in the speakers. Not to besmirch Mr Heinz's skill and reputation as a speaker designer, which is formidable - I just strongly disagree with him on this point.
From a practical perspective, there's no difference between AES3 and SPDIF. They both transmit digital audio and clock information between components. It's preferred in pro environments for various reasons, some practical and some historical. There's no reason you can't use it at home.
Regarding the location of the DAC in the system, this is in essence how the signal path flows:
Digital source --> AES3 input (digital) --> DSP --> DACs --> amps --> speaker drivers
In other words, the signal remains wholly in the digital domain until after DSP processing. This is the optimal signal path primarily for the following reasons:
PS: a few home audio companies are now also using AES3 rather than SPDIF in their speaker systems. Kii is one that comes to mind.
- Only one DA conversion takes place.
- The DSP enables exponentially more precise control and complexity in the implementation of the crossover than analogue filters would allow.
- The user does not need to concern themselves with analogue gain staging, since the entire analogue chain is inside the speaker.
Sure, but there is not that scenario, there are several scenarios, each scenario is going to give slightly different results, and therefore, that is what I am interested about, how much is that slightly.
Then again, not sure how that AES input is used at home indeed ;-)
I think this is very obvious; if you are sending a digital signal to the speaker, the DSP is awesome idea, otherwise, it is not a good idea.
But then again, how is the DAC in the speaker then if you are sending digital to it??? if these speakers can take digital AES, where is the measurement of that DAC? is it good the dac, is it 50$ good, or 200$ good, or 2000$ good? Why 50$ dacs are measured but then these 4000$ speakers have a dac and we do not care about how it performs?
I'd say it's more a low tech assembly process, than not a high precision process, it was a simple jig they were using to get each pleat the same.I did see a video of a HEDD employee hand-folding a ribbon tweeter. It was not a high precision process.
But can you hear a difference?But, but, but, and then but again...
The jigs may seem low-tech but we must not forget the 20 years of experience of the staff making the pleats, apply damping and assembling the whole things. And note, perfect smoothness and geometry of the pleats to a single, static target doesn't neccessarily give best performance. AMTs are a can of worms, design-wise.I'd say it's more a low tech assembly process, than not a high precision process, it was a simple jig they were using to get each pleat the same.
But on the other hand they are drivers that can be self manufactured with a relatively small amount of investment and tools, that's why its very fashionable for many small high end loudspeaker manufacturers to build their own ones as a "unique selling point" while almost none is able to build their own cone or dome driversAMTs are a can of worms, design-wise.
For clarity, I wasn't implying low tech was bad, I like a relatively simple engineering solution to a problem.The jigs may seem low-tech but we must not forget the 20 years of experience of the staff making the pleats, apply damping and assembling the whole things. And note, perfect smoothness and geometry of the pleats to a single, static target doesn't neccessarily give best performance. AMTs are a can of worms, design-wise.
I think that there is an AES EBU consumer protocol, exactly identical to spdif, and an AES EBU pro version, with some more bits of info, excluding also the copy protection.Something worth mentioning, just so there isn't too much confusion. The AES3 and S/PDIF protocols are essentially identical.
IMHO, Klaud Heinz's approach is what's awkward Due to his insistence on avoiding DSP in the speaker itself, his systems require the user to run the output from their computer source through an FIR filter before sending it to the speakers so as to achieve the superior phase performance that could have been achieved inside the speaker if Hedd simply adopted the standard practice of using DSP crossovers in the speakers. Not to besmirch Mr Heinz's skill and reputation as a speaker designer, which is formidable - I just strongly disagree with him on this point.
From a practical perspective, there's no difference between AES3 and SPDIF. They both transmit digital audio and clock information between components. It's preferred in pro environments for various reasons, some practical and some historical. There's no reason you can't use it at home.
Regarding the location of the DAC in the system, this is in essence how the signal path flows:
Digital source --> AES3 input (digital) --> DSP --> DACs --> amps --> speaker drivers
In other words, the signal remains wholly in the digital domain until after DSP processing. This is the optimal signal path primarily for the following reasons:
PS: a few home audio companies are now also using AES3 rather than SPDIF in their speaker systems. Kii is one that comes to mind.
- Only one DA conversion takes place.
- The DSP enables exponentially more precise control and complexity in the implementation of the crossover than analogue filters would allow.
- The user does not need to concern themselves with analogue gain staging, since the entire analogue chain is inside the speaker.
In real life, the dynamic range of the speaker at the listening distance is the only "sinad" parameter that counts, and that is what should be measured even now.
So, max spl (with low thd) at the whole 3rd octave is needed, this is the limit. Background noise level is the other limit. Between these two, you can find the useful dynamic range of the speaker system.
The first 2 octaves belong to the subs.
Ah, big questions.
Maybe i can just give some facts, and then everything else is quite obvious.
First of all, I suppose that these questions can apply to all speaker measurements.
By definition, the perceived low level detail (this is the point) is above Fletcher and Munson's 20 phon curve (if we assume that a nice quiet listening room has the threshold of noise down to 20 dB's, which is rather underestimated), so it is frequently dependent - at different spl's.
The perceived usable spl is the following:
View attachment 50282
Which means, that the maximum possible usable dynamic range is between these lines, i suppose that you can calculate the difference in every frequency.
The problem in this diagram is nothing else but the extreme high spl's at low frequencies, which can be tolerated quite easily.
Of course, listening to reference level, doesn't need to cover all the available in this graph spl's, but then go figure the really usable resolution, without forgetting the facts of psychoacoustics, the loudness curves, the need for loudness filter at low levels and the average noise in a typical home.
So, before asking the measurements of the speakers, do you really know what is the absolutely best expected resolution in high fidelity reproduction?
Self generated noise, is quite easy to measure, this is not a problem in high quality active speakers, it is well below the noise floor of the room, or even under the phon 10 curve.
Any thoughts?