Is the ADI 2's EQ Digital or Analog? Is it different that EQing in EQ APO or a DAP?

[Fraxo]

Just trying to understand degradation of sound via EQ and different types and implementations so any info about it (direct or externally linked) would be appreciated.
Specifically in common devices though, DAPs, Dac-Amps, via Computer etc... What are the technical aspects, (perhaps beyond phasing influence) that allow frequency manipulations?

Thanks in advance

@amirm

 

[Jimbob54]

Ill answer the headline question- its digital. As to the technicalities- above my paygrade.

 

[tuga]

A link to The Complete Guide To Parametric Equalization/EQ

The first 3 chapters will help.

 

[DonDish]

The ADI-2 PEQ is digital and non degrading. I think this just work by doing calculations to the signal.

By non degrading I mean. In theory in a digital PEQ you apply calculations to the signal, and those calcualtions could be reversed to maintain an original signal. Not like its done like this or that it makes sense. A thought experiment. With analog equipment I think this would be impossible. Information would be lost to noise and whatnot.

 

[theREALdotnet]

Just trying to understand degradation of sound via EQ and different types and implementations so any info about it (direct or externally linked) would be appreciated.

The point of EQ is improving the sound, not degrading it. When applied correctly, the gains to be had through EQ far outweigh the asymptotes left to be polished at the edge of measurability.

 

[Jimbob54]

I'm thinking you should be the one giving the answers? You seem pretty proficient in the field of digital manipulation of audio . Or perhaps Ive misunderstood the main thrust of the topic?

https://www.audiosciencereview.com/...does-it-translate-to-music.32188/post-1136838

 

[Fraxo]

A link to The Complete Guide To Parametric Equalization/EQ

The first 3 chapters will help.

Thanks a lot, I'll definitely check it out.

 

[Fraxo]

The ADI-2 PEQ is digital and non degrading. I think this just work by doing calculations to the signal.

By non degrading I mean. In theory in a digital PEQ you apply calculations to the signal, and those calcualtions could be reversed to maintain an original signal. Not like its done like this or that it makes sense. A thought experiment. With analog equipment I think this would be impossible. Information would be lost to noise and whatnot.

Thanks. So in this case - do you give credit to RME for achieving something rather difficult with a non degrading EQ or is it rather simple to implement with any DSP even on cheap DAPs etc?

 

[Fraxo]

I'm thinking you should be the one giving the answers? You seem pretty proficient in the field of digital manipulation of audio . Or perhaps Ive misunderstood the main thrust of the topic?

https://www.audiosciencereview.com/...does-it-translate-to-music.32188/post-1136838

You slightly missed the more refined point of the post but a great opportunity for me to further clarify
You are correct, digital manipulation via EQ is something I'm very proficient with, probably more than I should be haha.
Now, since I usually do it via Plugins on a capable PC, I was wondering just how efficient and "transparent" are some of the DSPs (digital signal processors) in physical devices such as DAPs, Sound cards, dedicated EQs who do this digitally. For instance - even most TWS earphones introduce DSP, so I wonder what it takes to do it well without degrading the sound, since hardware isn't my strong suit.

Does that make more sense?

 

[Fraxo]

The point of EQ is improving the sound, not degrading it. When applied correctly, the gains to be had through EQ far outweigh the asymptotes left to be polished at the edge of measurability.

But there must be some specifics about the Digital Audio Processing that could determine a good EQ implementation from a bad one no..?
I mean, what about a cheap DAP, is it as easy to implement a clean EQ on it just as if it's a 1300$ device?

Assuming there are differences in quality - what are they in the scientific sense of the matter, is it computational limitations? Components' quality? Other audio hardware issues?
That's the info I'm struggling to find.

 

[DonDish]

Thanks. So in this case - do you give credit to RME for achieving something rather difficult with a non degrading EQ or is it rather simple to implement with any DSP even on cheap DAPs etc?

I dont know about how difficult that is. All this is RME dsp secret juice. Read adi-2 dac manual that will give u some clues. If u dont own this I can tell u it works flawless up to 768khz.

 

[Fraxo]

I dont know about how difficult that is. All this is RME dsp secret juice. Read adi-2 dac manual that will give u some clues. If u dont own this I can tell u it works flawless up to 768khz.

I've looked heavily into the ADI 2, but I'm highly interested in what that "secret juice" is.

 

[KeithPhantom]

I was wondering just how efficient and "transparent" are some of the DSPs (digital signal processors) in physical devices such as DAPs, Sound cards, dedicated EQs who do this digitally.

If they are recent devices and they use the right sampling rate and bit depth for DSP, it should be all the same. Recent devices are pretty powerful for audio processing for the most part (with some edge exceptions).

 

[theREALdotnet]

Assuming there are differences in quality - what are they in the scientific sense of the matter, is it computational limitations? Components' quality? Other audio hardware issues?

Yes, there can be (and are with some devices) computational limits. They determine how many digital filters you can stack while still being able to process music in realtime. However, if your DSP device is capable of applying the filters you’ve got without falling behind then you’re set.

A faster or more expensive device isn’t going to sound “cleaner”, a device that lacks computational power will simply start to stutter or stop playing altogether.

 

[Fraxo]

Yes, there can be (and are with some devices) computational limits. They determine how many digital filters you can stack while still being able to process music in realtime. However, if your DSP device is capable of applying the filters you’ve got without falling behind then you’re set.

A faster or more expensive device isn’t going to sound “cleaner”, a device that lacks computational power will simply start to stutter or stop playing altogether.

Thank you so much for this information! I'm starting to get a better understanding of this thanks to you...
So not a big relevance to audio quality in regards to performing/processing/computing as long as the DSP runs consistently with no obvious stutter. now:

1) what about quality factors regarding bit rate and sample rate?
a) Would it not be of high importance to output the signal "cleanly" and match the correct bit/sample of the input?
b) Won't matching variable bit\depth rate (or handling higher ones) generally require much more processing power?
c) Is there a way to find DSPs specs in that matter?

2) I'm also thinking maybe quality in power supply components might make a difference in audio quality(?), but perhaps instead of my inexperienced ass trying to fish for it - could you perhaps mentions any common potential degradation factors regarding DSPs throughout their chain of Being Powered > Receiving Input > Digitally Processing (for EQ) > Outputting a high enough Bit Rate and the correct Sample Rate - all in attempt to get the most transparent sound possible with the least amount of electric and processing power... I have no way to gauge if this is difficult for reaching Hifi level sound (24 bit 48-92khz) or if rather cheap DSP chips could do it just as cleanly.

I might have asked some of these questions wrong, which is why I'm highly appreciative of the patience and knowledge sharing here so many thanks.

 

[Dro]

The EQ is digital, but still minimum phase, rather than linear phase (something that can only be done with digital EQ). It it substantially similar to any other minimum phase PEQ (if it doesn't tell you it's linear phase, it is safe to assume that it is minimum phase), but the interpretation of Q can be different.

 

[DonDish]

Thank you so much for this information! I'm starting to get a better understanding of this thanks to you...
So not a big relevance to audio quality in regards to performing/processing/computing as long as the DSP runs consistently with no obvious stutter. now:

1) what about quality factors regarding bit rate and sample rate?
a) Would it not be of high importance to output the signal "cleanly" and match the correct bit/sample of the input?
b) Won't matching variable bit\depth rate (or handling higher ones) generally require much more processing power?
c) Is there a way to find DSPs specs in that matter?

2) I'm also thinking maybe quality in power supply components might make a difference in audio quality(?), but perhaps instead of my inexperienced ass trying to fish for it - could you perhaps mentions any common potential degradation factors regarding DSPs throughout their chain of Being Powered > Receiving Input > Digitally Processing (for EQ) > Outputting a high enough Bit Rate and the correct Sample Rate - all in attempt to get the most transparent sound possible with the least amount of electric and processing power... I have no way to gauge if this is difficult for reaching Hifi level sound (24 bit 48-92khz) or if rather cheap DSP chips could do it just as cleanly.

I might have asked some of these questions wrong, which is why I'm highly appreciative of the patience and knowledge sharing here so many thanks.

I dont see the relevance of these questions, but Im gonna take a shot.

1) what about quality factors regarding bit rate and sample rate?
What does this even mean? your content is 44/16 48/24 192/24 or whatever. The DSP can either handle this or its not suited for audio
a) Would it not be of high importance to output the signal "cleanly" and match the correct bit/sample of the input?
In the adi 2 dac/pro its ultimately levels thats being manipulated in the DSP. Theres no added space or echo. Nothing like that. Since ADI PEQ is probably derived from TMfx. It goes like this. 16bit input - processing off - 16bit out. This is done in a bitperfect manner. If you add effects or even adjust a level output is 24 bit. Bitperfect is destroyed. So if bitperfect is what you want, dont buy an eq or any other DSP. Any brand. Well ADI-2 xxx can do bitperfect too.
b) Won't matching variable bit\depth rate (or handling higher ones) generally require much more processing power?
These are two questions 1: variable bit/depht rate: Is this MP3? 2: ADI-2 xxx can lay on PEQ to two channels at 768 khz, but most other DSP effects are shut off, maybe bass/treble is active. ADI-2 xxx have a section in manual about DSP LIMITATIONS thats right up your alley. I think dsp effects are calculated per sample. More samples, more computational power is required.
c) Is there a way to find DSPs specs in that matter?
If you can identify the chip you can find spec on the web. In the end is not about the power of the dsp, that limits what you can do at all. You would want a chip that is adequate to its tasks. This is about coding and code optimization, the dsp only does what its told. So with smooth and intelligent code and a DSP thats powerful enough this could all be toast and butter. A DSP is just a computational unit. You feed it, it does its calculations to the bitstream, and outputs the result which goes to the DAC chip.

2. Trust me this is a dead end. What you want? Some quad GPU arcitecture to put on an EQ? These are engineering design choises to make a product that does what its supposed to. If you are into minimalist audio stuff. Why would you want any prosessing at all? RME is prob the only manufacturer of consumer products for this with their ADI-2 xxx there is no competition. There are many other pro manufacturers though and nowaday much is made in software through DAW plugins and such. Analog PEQ is abundant just look at some studio supplier webshop.

 

[DonDish]

I wouldnt worry to much about these things. We pay others to do that. When we can dissect it here and at other places to see if the technology is valid. ADI-2 dac/pro, I think has stood some tests. TMfx(Totalmix FX), which I suspect most of ADI technology is derived from, has been developed over two decades. So thats what ADI is. 25 years of company experience boiled down into two products, soon to be three.

Before I got into this I thought just like you. Is this "pure" or "audiophile" or whatnot. I just read up on ADI-2 DAC and RME technology, and other comparative stuff. So I was impressed, and found these guys trustworthy enough to have my ears and my money. I really cant tell you how this stuff works down to the core of things. Some of it is RME boasting about their technology, theres small bits of information all over the web. These are also trade secrets. Eventually and probably inevitable, someone will take a shot at their success. Better to not let competition know exactly what to take aim for. Do your google searches for better understanding. RME manuals from all of their products is also interesting reading. Check out DDS and Steadyclock. Hightech stuff compared to the simple baseclock technology of lesser products. Thats just a taste. Highly competitive stuff So I just feel Im in good hands with these people. One certainely doesnt have to have the competence of designing a DAC circuit, to understand or to muse over these things.

A simple user case: Sennheiser HD650 frequenzy respons. In short; low bass, strong mid and spiky treble. Goto AutoEQ, find your phone in the list. Dial in the numbers given. 5 bands of PEQ, Gain, quality and frequenzy. This is a correction curve for your phones to flatten the frequenzy respons. For me it almost immediately made the phones sound "right", and after a while, turning PEQ off, it just sounds weird. This goes beyond subjectiveness. Thanks to Jaako Pasanen at AutoEQ for measuring and collecting response curves for hundreds of phones. These are the things you gain with these products. With a not to expensive headphone rig and software you could even make your own measurements. Just for the fun of it!
If you want a straight line from your mediaplayer to ADI-2 dac, RME has bittests for that. Its in yor PC thats everything gets messed up. Very handy tool. I use it everytime I alter or try some new player. This stuff is so highly addictive!, Surgeon General warning should be applied!

So dont worry over quality of DSP and FPGA chips. These are design, cost and engineering desicions. No use if no one can afford it. This isnt a supercomputer, it just does what its supposed to. Up to 768khz, and thats strong. Two channel PEQ at 768khz is somewhat a clue to the efficiency and optimization applied. Code is probably stripped to the bones, highly efficient. Its the implementation of components thats crucial to the product.

Best wishes
Mikael