Measurements of MiniDSP 2x4HD

[restorer-john]

My guess would be you have a hard time finding recordings that have non-noise info above 14 bits.

? All CDs bouncing on 0dBFS have 16 bit non noise related content or did I misread the context?

 

[Blumlein 88]

? All CDs bouncing on 0dBFS have 16 bit non noise related content or did I misread the context?

Not the way I see it. If anything they might only need 4 bits. Assuming the musicscope works reliably to report this random noise from microphones and every other analog step in the chain I think will increase noise above the 16th bit nearly all the time. 96 db dynamic range would have a signal, which isn't noise down in those lower bits.

 

[oivavoi]

My view on this unit overall is thst the DAC section is pretty average, the ADC is not great and the internal resampling, if using digital input, is definitely something to be avoided.

However, in the context of price and functionalty it still is good value for money. I still think it would probably beat a passive crossover.

So come on miniDSP, time to build a truly high performance unit

Seems like a reasonable assessment. Now how does one avoid the internal resampling?

 

[March Audio]

Is the video capture capability built into the Musicscope or if not what are you using for that?

I didnt use it for those captures but I would recommend OBS

https://obsproject.com/

 

[March Audio]

Seems like a reasonable assessment. Now how does one avoid the internal resampling?

If you are using the analogue input its not an issue, but if using usb or optical get your media playing software to updample to 96kHz prior to sending it to the miniDSP. J River and Roon can do this.

 

[March Audio]

Not the way I see it. If anything they might only need 4 bits. Assuming the musicscope works reliably to report this random noise from microphones and every other analog step in the chain I think will increase noise above the 16th bit nearly all the time. 96 db dynamic range would have a signal, which isn't noise down in those lower bits.

As an example in this 24bit live acoustic recording the background noise at the start is around -60 to -50 dBfs. If I am reading the bit monitor correctly is around 12 bits of noise!

This is from the 16bit release of the same album - 6bits of noise?

 

[restorer-john]

Not the way I see it. If anything they might only need 4 bits. Assuming the musicscope works reliably to report this random noise from microphones and every other analog step in the chain I think will increase noise above the 16th bit nearly all the time. 96 db dynamic range would have a signal, which isn't noise down in those lower bits.

I think we must be talking about different things.

"a hard time finding recordings that have non-noise info above 14 bits." So do you mean more than 14 bits down below the signal? As in 84.28dB below 0dBFS?

I'm scratching my head here.

 

[Blumlein 88]

Let us imagine a wonderful venue with a zero ambient noise level. Recording the music we have microphones with zero self noise as is the rest of the chain. Our musicians create nothing more than 120 db SPL so we have a wonderful 120 db dynamic range and need 20 full bits to capture everything that was there and none of the bits represent anything other than music as there is no noise.

Now lets take a step toward reality and say the ambient noise level was 24 db or roughly 4 bits. Everything else is the same. Though somewhat over simplified, with 20 bit recording the top 16 bits hold the music and the lower 4 bits are only encoding noise. So presumably dithered 16 bit would capture everything there is to capture.

Taking one more step toward reality maybe the microphones lost another bit to self noise and processing lost another bit's worth of dynamic range. Then with 20 bit recording the top 14 bits are only recording music and the bottom 6 bits are only recording noise.

So I don't know the algorithm Musicscope uses to show bits only recording noise. If it works it would give us an idea of the amount of bit depth needed to capture what is above the noise floor of the recording.

Does this explanation make sense or am I still not being clear?

 

[Krunok]

Using JRiver to perform the SRC to 96kHz prior to input into the miniDSP. I think the lesson here is perform the SRC to 96kHz first to avoid the miniDSPs somewhat sub optimal conversion!

View attachment 11767

Interesting!

Are you by any chance able to check if Volumio will do a better job performing SRC to 96kHz than miniDSP?

 

[flaviowolff]

I have a 2x4hd firmware upgraded to DDRC-24, meaning that my unit upsamples to 48khz, right? Even so, should I feed it 96khz files? Isnt 48khz enough?
Also, if one had to choose between a) a regular quality DAC fed to ddrc-24 analog inputs and b) 44.1khz files to ddrc-24 digital inputs, which should be more advised?
Thanks.

 

[Julf]

I have a 2x4hd firmware upgraded to DDRC-24, meaning that my unit upsamples to 48khz, right? Even so, should I feed it 96khz files? Isnt 48khz enough?

48 kHz is enough. No benefit from anything higher.

 

[flaviowolff]

48 kHz is enough. No benefit from anything higher.

Thank you. Do you think that the 44.1khz->48khz upsampling with normal files would be audible? Setting up an upsampler on my RaspberryPi Spotify station would be kind of a hassle

Ty!

 

[Julf]

Thank you. Do you think that the 44.1khz->48khz upsampling with normal files would be audible?

No. Upsampling has to be pretty badly implemented to be audible - just ask any record label, that is how most of the "hi res" recordings are produced.

 

[flaviowolff]

No. Upsampling has to be pretty badly implemented to be audible - just ask any record label, that is how most of the "hi res" recordings are produced.

Thanks again. I was kinda worried because of this "wiki" at reddit: https://www.reddit.com/r/audiophile/wiki/dirac , which points out to this thread stating the DDRC24 "poorly resamples" the souce to it's internal clock. I won't bother, then.

 

[Julf]

Thanks again. I was kinda worried because of this "wiki" at reddit: https://www.reddit.com/r/audiophile/wiki/dirac , which points out to this thread stating the DDRC24 "poorly resamples" the souce to it's internal clock. I won't bother, then.

I would definitely not lose any sleep over it.

 

[amirm]

As an example in this 24bit live acoustic recording the background noise at the start is around -60 to -50 dBfs. If I am reading the bit monitor correctly is around 12 bits of noise!

You have to look at the spectrum of that noise. All environments have high noise floor in bass since it penetrates through walls easily. But where our hearing is most sensitive -- 2 to 5 kHz -- rooms tend to be far quieter.

 

[oivavoi]

Just been having a quick look at the minidsp re-sampling as it converts all inputs to 48kHz or 96kHz for processing.

I input a sweep from 10Hz to 22kHz and looped it back (through USB) to see if there were any obvious SRC issues. All other DSP processing is turned off. Note this is not going through AD / DA conversion. A colourmap display is great for this as it makes the low level spuria stand out.

Original signal 44kHz - note the clean background
View attachment 11750

Loopback - Has been converted to 48kHz. The images are around -130 dB so it looks worse than its audible impact. Interesting reflection abouve about 18kHz.
View attachment 11752

48kHz input - clean
View attachment 11753

Output, no SRC required. Noise floor increases to about -146dB from around -164dB but no spuria/images
View attachment 11754

Adobe Audition conversion to 48kHz - Perfectly clean and no increase in noise floor. Much better than the miniDSP.
View attachment 11755

Video of minidsp 44 to 48kHz sweep

Small bump here. I'm wondering about the new Shd studio unit from minidsp, which by default resamples everything to 96 khz. Is there any reason to think this unit does it better, based on the hardware?
https://www.minidsp.com/products/streaming-hd-series/shd-studio
I guess you don't have any of these new shd units lying around which you can measure as well, @March Audio ? (you're the only one I know who have looked at the SRC in minidap units, which seems potentially to be important)

 

[Oukkidoukki]

Little bit out of topic......I was thinking minidsp nanodigi for some room eq.......maybe you have it, what you think....is it good option?......can it cause jitter or other problems......