Regarding UMIK and other USB mics, I suggest you contact Uli Brueggemann at Acourate (from the Audiovero website) and ask him about it. Some years ago he didn't recommend this with Acourate (because of the built-in ADC), but later releases might have solved this. I never explored this avenue because I followed his indication back then and purchased a good analog measurement mic and a mic pre (with phantom power) to input into Hilo.
An analogue mic and reasonable quality interface is the better way to do it for sure, then you get complete electrical loopback. The acoustic timing reference is the only option for USB mics and it is very good at removing clock drift issues and allows time locked measurements but it is still a workaround, Uli would no doubt recommend to do it properly with an analogue mic. If you are willing to pay the price for Acourate the cost of an analogue mic good enough is quite small.
I agree, Fluid.
Moreover, in my opinion and in the context of a $10k system and up, Acourate is not expensive either in light of the improvements it brings - provided it's the right tool for you. I purchased it 6 or 7 years ago and have used it non-stop and have gotten great support.
The learning curve is steep, though, and it requires time to learn. The e-book by Mitch Barnett is worth every penny to get a great guide and reference.
If you use analog sources the routing gets more complicated. In my case, I decided to focus on digital source (playing files, streaming), and optimize the system for this. I'm happy.
I think Mitch has said Audiolense is better in that is more automated than Acourate, but the price difference wasn't worth it for me as I had Acourate already and had learned how to use it. There are other options that are lot easier to use (higher automation), but a lot less flexible in what they can do.
Moreover, in my opinion and in the context of a $10k system and up, Acourate is not expensive either in light of the improvements it brings - provided it's the right tool for you. I purchased it 6 or 7 years ago and have used it non-stop and have gotten great support.
The learning curve is steep, though, and it requires time to learn. The e-book by Mitch Barnett is worth every penny to get a great guide and reference.
If you use analog sources the routing gets more complicated. In my case, I decided to focus on digital source (playing files, streaming), and optimize the system for this. I'm happy.
I think Mitch has said Audiolense is better in that is more automated than Acourate, but the price difference wasn't worth it for me as I had Acourate already and had learned how to use it. There are other options that are lot easier to use (higher automation), but a lot less flexible in what they can do.
I agree with you too, Acourate is great and Uli knows what he is doing. I am a bit of a cheapskate at heart and combined with a control freak I do it all manually with DRC for free. I could likely have got to the same place with Acourate in a lot less time 🙂
I already do use the timing reference, think it's key.
It's a painful affliction, the only known cure is to get it out of your system! Unfortunately doing any tweaking just fuels the fire to do even more....I'm a lost cause.
some nice options there thanks.
that's a no brainer then, no problem with getting another 1. What do you mean by interface? I was thinking you just plug it into a DAC with ADC support.
I think it's a worthwhile investment versus the cost of the system too.
I wasn't aware of Audiolense, the automation sounds interesting, I want powerful options but if I can do things in an automated way that will save a lot of time and pain. The price difference between Acourate and Audiolense doesn't seem to be that bad really, 50 EUR?
Audio Interface / Soundcard, there are very few DAC's with ADC, when you want 8 channels of outputs I don't know of anything other than an interface that has inputs that can be routed to a computer for measuring.
But you can definitely use something like the OKTO DAC and a separate interface for the ADC function together they just need to be clocked from the same source so it can be a fiddle to get it right.
If the ADC and DAC are on the same device it's not an issue.
MOTU make some really good interfaces with ESS DAC's if you like them the Ultralite Mk4 looks good to me
As an alternative as this is diyaudio and not buy it from a shop audio I have got a diyinhk ES9016 8 channel DAC that I have nearly finished but got mothballed. This has the tweakiest power supply collection I have ever made and I swapped the opamps for SMD Analog Devices ones, the DAC itself sounds pretty good and dissapointingly similar to my ES9018 that I tricked out with a fancy Crystek clock and other stuff. But of all the DAC's I own or have built I listen to the Topping D50 as I like it the most 🙄
MOTU.com - Overview
Have a look at Mitch Barnett's Audiophilestyle articles on Acourate and Audiolens.
Both do pretty much the same thing in different ways. To get the best result you need to know what you want and what you have, brute forcing a correction to a defined target creates pretty graphs and often lacklustre sound. Both have good algorithms that help to avoid correcting things that shouldn't be but they are not foolproof.
Attachments
Last edited:
Frankly, I haven't looked into this for a while. Make sure to look into the specs of the Audiolense version you are considering, as my recollection is there was a relatively inexpensive 2-channel version, but there was a significant price jump for the multichannel version. Acourate has only one version.
Dirac is further automatized and easy to use, but a lot less flexible. This is not from personal experience, but what I gathered at the time I was looking into it.
This is a good forum to check out: Audiophile Style DSP Room Correction Forum
There is also an Acourate forum hosted within the Audiovero website and Uli is very helpful.
Some 8-channel interfaces/soundcards that I have been considering, shared as examples of what's out there:
- Merging Hapi with ADA8P card (in my book, the best, and what I'll get eventually). Ethernet device, does DSD256 (Phobic said this was important to him). PCM 384kHz. About $5k in US.
- Prism Titan, connects thru USB to a computer, or can get an ethernet card, doesn't do DSD, does PCM 192kHz. About $4k in US.
- Focusrite has some units that are used by Acourate users, but don't recall the models.
- Lynx Aurora: similar to Titan, a bit cheaper.
- Lynx Hilo (what I have now): about $2.5k new in US, does PCM 192kHz, and DSD64 (I didn't try this), can be purchased with USB connection (like I have) or a newer ethernet card connection. It's a very flexible unit and can be set up to be a 6-channel DAC and has 2-channel ADC which is more than enough for a mic input. In fact Mitch Barnett has the same and he inputs from his vinyl player into the ADC, runs through the convolution, outputs a corrected signal. You can also do 8-channels with it, like I'm doing: I have an older Metrum Octave that I connected to the S/PDIF out of the Hilo and are driving the subwoofers. I configured Hilo as an 8-channel DAC. Not an ideal solution, but as mentioned before, my approach has been incremental experimentation and Hilo has allowed me to start with 2 channels and just doing digital room correction all the way to 8-channel digital xo, time-aligned, room corrected. You can get a used one and extract a lot of value!
Yet one more option for your consideration: get a Hilo, try/learn all of the above, and maybe down the road you want to buy a fancy 8-channel DAC (Nadac, e38, other) and use the Hilo for taking measurements and the DAC for playback. You need to check with Uli because even though it's not the technically perfect solution I believe he has a recent workaround for this. You'll have the very good ADC for measurements and your choice of DAC. Uli would likely tell you a Nadac is a waste of money, but that's a separate topic 😀
A lot of nice alternatives in this thread.
Stepping back a little, to look at an entire digital implementation that allows FIR,
I tend to see it having 3 major pieces.
One being the software to create the FIR file.
Second being the convolving hardware.
And third being the DACs (and ADCs if needed).
I'll leave any recommendations about the third alone, because i'm probably in the minority believing once a certain threshold of DAC quality and sample rate is met, gains are too marginal to pursue.
For instance, i can't see (or hear) any reason to go beyond a 96kHz sampling rate.
This seems especially true for FIR, because moving on to the convolving hardware, it needs to be recognized for every doubling of sample rate, the number of taps has to double as well to have the same frequency resolution.
ie for equal resolution 16k taps @ 48kHz = 65k taps @ 192kHz.
I think if someone wants a higher sampling rate than 48k and/or greater than 16k taps, there is currently nothing in the marketplace other than a PC.
Moving to FIR file creation software...
gotta love rePhase...bless POS..
lots of manual work, but sure does instill learning, and sometimes the best tool when automation isn't giving expected results
I guess the 3 programs most mentioned on DIY, are DRC, Acourate, and Audiolense.
That seems to be a list in terms of their degree of automation, but i have no experience with any so it's just the impressions gathered. When proficiency is gained, everyone seems to be pretty happy with whatever they are using.
One that is most like quite a bit more automated is FirDesigner. Audio FIR Filter Design Tools - FIR Filters for Speakers | ECLIPSE AUDIO
I've been quite happy with it, but my needs are a bit different than most because i'm constantly trying modifications to the speaker itself, or trying entirely new types of builds.
Most all have been multiways, so it really helps to be able to construct files very easily.
That said, if i had only one set of speakers i was trying to optimize, I don't know which program i would be led to.
edit: one last thought on FIR file software.
Both VituixCAD and REW are capable of making the files via impulse inversion and xover multiplication.
I just used VCAD successfully (and rather easily once i figured it out) to build files. This might well be a neat way to go and save both time and money.
Stepping back a little, to look at an entire digital implementation that allows FIR,
I tend to see it having 3 major pieces.
One being the software to create the FIR file.
Second being the convolving hardware.
And third being the DACs (and ADCs if needed).
I'll leave any recommendations about the third alone, because i'm probably in the minority believing once a certain threshold of DAC quality and sample rate is met, gains are too marginal to pursue.
For instance, i can't see (or hear) any reason to go beyond a 96kHz sampling rate.
This seems especially true for FIR, because moving on to the convolving hardware, it needs to be recognized for every doubling of sample rate, the number of taps has to double as well to have the same frequency resolution.
ie for equal resolution 16k taps @ 48kHz = 65k taps @ 192kHz.
I think if someone wants a higher sampling rate than 48k and/or greater than 16k taps, there is currently nothing in the marketplace other than a PC.
Moving to FIR file creation software...
gotta love rePhase...bless POS..
lots of manual work, but sure does instill learning, and sometimes the best tool when automation isn't giving expected results
I guess the 3 programs most mentioned on DIY, are DRC, Acourate, and Audiolense.
That seems to be a list in terms of their degree of automation, but i have no experience with any so it's just the impressions gathered. When proficiency is gained, everyone seems to be pretty happy with whatever they are using.
One that is most like quite a bit more automated is FirDesigner. Audio FIR Filter Design Tools - FIR Filters for Speakers | ECLIPSE AUDIO
I've been quite happy with it, but my needs are a bit different than most because i'm constantly trying modifications to the speaker itself, or trying entirely new types of builds.
Most all have been multiways, so it really helps to be able to construct files very easily.
That said, if i had only one set of speakers i was trying to optimize, I don't know which program i would be led to.
edit: one last thought on FIR file software.
Both VituixCAD and REW are capable of making the files via impulse inversion and xover multiplication.
I just used VCAD successfully (and rather easily once i figured it out) to build files. This might well be a neat way to go and save both time and money.
Last edited:
Epic list to Pick from.
Phobic: do you want other inputs than PC?
Is yes, I suggest you get a stand alone DSP.
For a beginner, I think Acourate and Audiolense, is way overkill.
As fluid nicely points out: you simply need to know what to correct, before just letting software help you with perfection.
Phobic: do you want other inputs than PC?
Is yes, I suggest you get a stand alone DSP.
For a beginner, I think Acourate and Audiolense, is way overkill.
As fluid nicely points out: you simply need to know what to correct, before just letting software help you with perfection.
I thought about this after posting before, if using DSD and HQPlayer is a must then the DAC solution needs to take that into account.
To me the purpose of DSD conversion is to avoid the Delta Sigma modulator in the DAC and use a DAC that does not have one or that can be bypassed.
I don't know of any multichannel DAC / Interface where the DS modulator can be bypassed. Even though the DAC may say it can deal with DSD, it is not straight DS like you might want.
Consider this when judging whether all that processing power to generate high rate DSD is worth it if you are just going to throw it in a wrapper in the DAC.
Most that want active and DSD go for analogue active to keep the DAC stereo then it can be a pure DSD DAC.
Two reasons: 1) you said you were after neutral speakers, and 2) you said you wanted to significantly outperform your current speakers.
The JBL 708Ps were designed to be neutral. They nail it. You could match them but not beat them. You have zero chance of significantly beating them.
However, you can significantly beat your current system (long term) by going multichannel. You have to wait a few years until the source material is available through streaming music services. Actually, it's available right now on Tidal and Apple Music but only a limited selection. I don't think multi-channel is going to be significantly better than your current system until a larger portion of the streaming catalogs have been converted.
If you choose to change your goals and look for a different sound field or non-neutral speakers then you can DIY speaker that outperforms the 708Ps in that context. For example, you could build a system with different directivity using wave guides or arrays. You said you don't have woodworking skills but if you did that might be another reason to DIY. There's nothing wrong at all about choosing to DIY speakers intended to look beautiful.
Personally, the only reason I DIY is to try different sound fields that aren't available in the market. My goals lead me to DIY. Your goals would lead me to buy good speakers.
Hi,
Only (pro)DAC i know implementing DSD are the 'biggest' Prism.
I seriously doubt it'll be of any interest however as last time i talked to a mastering engineer having some he explained he used this twice a year in DSD for obscur japanese's jazz label which still used reel to reel and send tape to be digitized.
One should ask if the initial record was done in DSD in digital too... for what i've seen ( and done!) even in 2021 most studio still run at 48khz.
Obviously i'm in the minority like Mark100 ( 96khz/24bit is more than enough to me).
Very nice answer and polite tone in this thread.
Mr Ba, nice to see you are happy with your system!
Only (pro)DAC i know implementing DSD are the 'biggest' Prism.
I seriously doubt it'll be of any interest however as last time i talked to a mastering engineer having some he explained he used this twice a year in DSD for obscur japanese's jazz label which still used reel to reel and send tape to be digitized.
One should ask if the initial record was done in DSD in digital too... for what i've seen ( and done!) even in 2021 most studio still run at 48khz.
Obviously i'm in the minority like Mark100 ( 96khz/24bit is more than enough to me).
Very nice answer and polite tone in this thread.
Mr Ba, nice to see you are happy with your system!
DSD from HQPlayer ends up in the Mhz range for the 512 and 1024 versions but that is to allow the ultrasonic noise to be shaped and not intrude so significantly into the audible band, it doesn't really have anything to do with high sampling rates. All my processing is done at 44.1Khz because that is 99.9% of my entire collection, no DSD though, storage is too high for native and processing is too much from converted. I know of one sensible person who swears by DSD and a NoDAC so I don't write it off though.
^ i agree Fluid, as i've said it is of interest IF the whole digital chain is or you digitize from analog source at the last stage.
There is almost zero chance for the first case so it really limit the interest of the thing. As the offer of high sampling rate streaming which is 99% upsampling of lower fs...
Mine are 96/24 because i digitize my LP at this fs/bit.
To be clear i don't discard it ( every step or try for better quality has to be encouraged in my view) just that it is mainly marketing hype in practice.
Edit: the reason you give for no DSD is the same reason it doesn't made a breakthrough in proworld... and the rarity of (pro)gear implementing it.
There is almost zero chance for the first case so it really limit the interest of the thing. As the offer of high sampling rate streaming which is 99% upsampling of lower fs...
Mine are 96/24 because i digitize my LP at this fs/bit.
To be clear i don't discard it ( every step or try for better quality has to be encouraged in my view) just that it is mainly marketing hype in practice.
Edit: the reason you give for no DSD is the same reason it doesn't made a breakthrough in proworld... and the rarity of (pro)gear implementing it.
Last edited:
This is true
This is not true without a number of qualifications. I would agree that without a lot of knowledge, skill and experience time and money it is a big hill to climb 😉
Please could you explain in what multichannel is going to be better over stereo for music? ( except for the presence of C in L-C-R which improve over phantom image)
Last edited:
Although, I guess if your intent is to stick with an existing two channel catalog of sources then you might want to rethink your goals.
Maybe you should consider the idea that you want a unique sounding system not a neutral system. Maybe you should consider you want a system that looks impressive with sound quality on a lower level of priority. Or maybe you want a system that you can constantly wrench on.
Today's speakers are so good that a technically good system is going to look pretty boring and it isn't going to cost a lot of money. You won't be tweaking it after initial setup.
The concept of a dedicated listening room is sort of outdated as well. Today, a "high end" speaker is one that does a good job removing the room. The extreme high end (Beolab 90 or a Lexicon) have adjustable sound fields intended to operate in normal rooms. Of course, that's only valuable when you're stuck with two channel sources. Multi-channel streaming will dunk on today's highest end speakers.
It's just sort of a weird time right now where the definitions of high end audio systems are changing.
Maybe you should consider the idea that you want a unique sounding system not a neutral system. Maybe you should consider you want a system that looks impressive with sound quality on a lower level of priority. Or maybe you want a system that you can constantly wrench on.
Today's speakers are so good that a technically good system is going to look pretty boring and it isn't going to cost a lot of money. You won't be tweaking it after initial setup.
The concept of a dedicated listening room is sort of outdated as well. Today, a "high end" speaker is one that does a good job removing the room. The extreme high end (Beolab 90 or a Lexicon) have adjustable sound fields intended to operate in normal rooms. Of course, that's only valuable when you're stuck with two channel sources. Multi-channel streaming will dunk on today's highest end speakers.
It's just sort of a weird time right now where the definitions of high end audio systems are changing.
I did limited experiments with DSD several years ago, when I had the Metrum Octave which did only PCM, and got an exaSound e32 and tried DSD. DSD sounded better than a non-optimized PCM chain, but a fully optimized PCM chain sounded as good as the e32 with HQPlayer doing DSD at the cost of the DSD chain being significantly more expensive so I sold the e32 and kept the Octave.
In my case DSD is not a goal. I think there is a lot more to be gained in the speakers and room camps. Plus doing 8-channel convolution at DSD256 would likely be a burden on the PC. It probably can be made, but I'd rather focus on solving other issues in my system. I just brought up DSD because the OP mentioned it.
In fact if DSD is not to be used, then there is no real need for HQPlayer in the chain as what is often mentioned as a strength of HQPlayer is the sound with DSD. You might still prefer what HQPlayer does, but Roon convolution works well and is stable and I have removed HQPlayer from my chain.
Hope I'm not coming across as a DSD promoter!
BTW, search computeraudiophile.com for 2 articles by Mitch Barnett with Acourate: basic correction, and advanced multichannel correction, from 2014 or so.
My suggestion was meant to be to get into it step by step, for example introducing room correction (DSP) into 2-channels, then making your passive speakers active, learning the process and variables, then go into DIY fully active speakers. To avoid the risk of being overwhelmed by the quantity of variable to consider if you tackle all at once.
cheers
In my case DSD is not a goal. I think there is a lot more to be gained in the speakers and room camps. Plus doing 8-channel convolution at DSD256 would likely be a burden on the PC. It probably can be made, but I'd rather focus on solving other issues in my system. I just brought up DSD because the OP mentioned it.
In fact if DSD is not to be used, then there is no real need for HQPlayer in the chain as what is often mentioned as a strength of HQPlayer is the sound with DSD. You might still prefer what HQPlayer does, but Roon convolution works well and is stable and I have removed HQPlayer from my chain.
Hope I'm not coming across as a DSD promoter!
BTW, search computeraudiophile.com for 2 articles by Mitch Barnett with Acourate: basic correction, and advanced multichannel correction, from 2014 or so.
My suggestion was meant to be to get into it step by step, for example introducing room correction (DSP) into 2-channels, then making your passive speakers active, learning the process and variables, then go into DIY fully active speakers. To avoid the risk of being overwhelmed by the quantity of variable to consider if you tackle all at once.
cheers