It will always go through ADDA. But why do you need a volume knob pre? You can control volume through minidsp.I asked mini DSP, and as I suspected, if I want to keep phono analog, I will need to add volume knob pre such as: https://www.goldpt.com/sa1.html .
Thanks for your interest in our products.
Indeed, all our platforms are Digital Signal Processors(DSP) and will always need to convert to Digital to be able to do processing.
So to answer your question, yes, all analog input sources are always converted to Digital. Note however that the analog audio performance on the Flex HTx is really good. With audiophile performance, it would be hard to be able to hear any negative effect on the audio quality. If anything, DSP will always improve your experience as it's correcting issues in your room that no other DAC/Preamp/Analog processor could ever do.
Hoping this info helps and feel free to contact us if you have further questions.
Best Regards
miniDSP DevTeam
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
MiniDSP Flex HTx
- Thread starter MAB
- Start date
OP
MAB
Major Contributor
- Thread Starter
- #322
It is a digital device, it will certainly not 'keep phono analog'.
I think you misinterpret the next steps though. The controller you linked to is not going to allow you to keep the phono analog with he MiniDSP, or any other DSP unit, unless that unit has an 'analog passthrough'. You can build a separate system with that, it will not likely outperform the MiniDSP in any way, and of course will not have DSP so it will actually underperform in use.
Why would you want to keep phono analog? Are you concerned the A2D is poor? Or do you feel the need to keep phono analog with no A2D from an aesthetic perspective?
If one is really concerned about AD performance. Get a Cosmos ADC, which can be modded to have a spdif output. It can then be fed to HTX. But I don’t think it is meaningful in anyway because I’m pretty sure Htx’s AD performance will outperform any phono analog device.
My measurements here: https://www.audiosciencereview.com/forum/index.php?threads/minidsp-flex-htx-measurement.52585/
I think what is obvious is that minidsp does not do phono pre-amp by default (aka RIAA). I wish they added it though as a feature. Similar to RME ADI-2 PRO SE. There is though in community, projects where people have done RIAA in minidsp.
Thanks for your comments. Yes, I was concerned about the sound quality of digitizing the phono signal, but perhaps my concerns are unfounded.
If I put an analog preamplifier after the HTX, then I could theoretically switch between the two, however as you mention, I would need a preamplifier with theater bypass. I did a little searching but am not finding anything reasonably priced, especially with IR remote.
Amplifiers with HT-bypass - update 10. August 2024
Audiophile.no has made a list of stereo preamps and integrated amps that have one sort of HT-bypass. This list is updated 10. August 2024 with amplifiers from Audio Group Denmark, and will be will be continuously updated. Written by Karl Erik Sylthe, at 15.02.07 Translated to English in june...
Last edited:
OP
MAB
Major Contributor
- Thread Starter
- #328
There is a Biquad equalization. See the tutorial and Biquad calculator spreadsheet.
Advanced Biquad Programming
Flexibility really matters when it comes to innovative speaker design. As DIYers and speaker designers ourselves, we came to realize that miniDSP platforms could be used for both basic and advanced digital signal processing applications. While textbook filter implementation like Butterworth...
Looks like a nice unit for DIY doing 4-way active speakers. The price was not as high as I expected.8 channels of both analog-in and -out, both balanced and RCA.Apparently shipping in December, taking pre-orders now.
miniDSP Flex HTx
miniDSP Flex HTx, 400MHz Sharc DSP audio processor, toolbox for subwoofer or multiway speaker tuning, USB/HDMI ARC/EARC, Optional Multichannel Dirac Live license for room correction,www.minidsp.com
Hardware Description:
No manual available yet. This looks like lots of fun.
- Processor: 32-bit floating-point 400MHz Analog Devices SHARC DSP
- Multi-Core USB Audio processor (XMOS)
- HDMI EARC/ARC interface for multichannel audio in PCM. Note that Flex HTx doesn't support bitstream format (Dolby/DTS encoded).
- Optional upgrade to Dirac Live® 3.x Digital Room Correction, Multichannel (8ch) Full Range Room Correction (20~20kHz)
- Audiophile performance with very low noise floor and distortion (SNR @127dB(A), THD+N @ -120dB (0.0001%)
- Inputs: Stereo digital (SPDIF/Optical), Multichannel USB Audio streaming (up to 7.1)
- Outputs: 8 channels audio duplicated to unbalanced (RCA)
- White/Black OLED front panel controller with IR control
- Control and configuration by miniDSP Device Console (DC)
- Rack mountable with removeable rack EARS
It cuts lots of cost by using four pieces ES9017s. Unfortunately, clearly Minidsp hasn’t figured out a way to get rid of the “ESS Hump”, which is non-existent in ES9017s’ Evaluation Board.
Has anyone gotten their hands on one of these yet? I'd like to know how it performs!
I'm looking to upgrade from my Peavey VSX 48e. Application is a 4 way speaker system with built-in multiple subwoofers.... The peavey works, but I'm finding that it is a bit noisy, and I would really like 2 more inputs.
I'm looking to upgrade from my Peavey VSX 48e. Application is a 4 way speaker system with built-in multiple subwoofers.... The peavey works, but I'm finding that it is a bit noisy, and I would really like 2 more inputs.
Check my measurement thread. It performs quite well but mostly benefit from ESS’s technology. Minidsp itself doesn’t seem to get everything right at least not on my bar yet. But still, thanks to today’s technology, one has to be really sloppy to screw everything up.
Is it possible to characterize the noise output of this device in absolute terms, rather than as a ratio? I'm not sure if low THD and low signal to noise ratios always translate to low noise in the absence of a signal?
The problem I'm trying to solve is significant hiss from horns. I'm using balanced cables everywhere, and I have an aggressive L-Pad on the horn, but still the noise!
The signal chain is source -> DAC -> Processor -> Amp. If I turn the processor off, most of the noise goes away. If I connect my DAC directly to the amp, the noise [from the horn goes away, although the coax woofer is still producing the same noise]. So the DAC is essentially silent, while the [cheap] processor seems to introduce significant noise.
I saw your tests, and also Amir's review of the HT (2x4 channel version). This HTx would eliminate both my mediocre DAC and questionable processor, so it feels like a no-brainer. But then someone on the DIY audio forum said it won't change anything, and what I really need to do is increase the input levels in the processor and attenuate them before the amplifier. This would probably work, although it is adding extra components to the signal path.
Last edited:
As far as I see, folks on Diyaudio is giving you advice on Gain structure, which is most likely going to eliminate noise on measurement level but not on audible level. What I meant is that I think something in your setup is probably introducing ground loop and changing gain structure will not help.
Minidsp Ht is not a balanced device, so in terms of measurement, I doubt it will perform better than Htx. But that doesn’t mean you can just get a Htx and problem solved. I would suggest you to connect DAC directly to your AMP and figure out the ground loop first.
But the HTx is balanced.
I am pretty sure I do not have a ground loop. It doesn't sound like a ground loop. All the power for my components comes from the same outlet, all my interconnects are balanced, etc.
And I made a silly error in what said above. I swapped the horn channel (only) to the DAC, and of course I heard a lower pitch hum! Becaue the coaxial woofer was still connected to the processor! So I disconnected the speaker wire for the woofer, and repeated. No audible noise when the horn is connected to the DAC, but noise returns when the processor is in the signal path.
Last edited:
I got HTx yesterday. I have now installed it to my system, and my initial impressions about it are greatly on the positive side. I'll still need to do tunings (I have not eq'd/dirac yet) and measurements, but it shows big improvements to my previous decades old AVP, and I am not missing any decoders or other features in my use case. eARC works nicely with FeinTech VAX40101 (though I had to set it to manual mode, which I prefer too, as otherwise it seems to sometimes lose the eARC and drop to ARC) and LG UST projector. I decided to skip Genelec bass-management (the analog one), and just do it all on HTx.
Last edited:
I first used VAX40101 as eARC switch (connected to eARC on LG and eARC on HTx), but I figured out that the Apple TV on HDMI1 on LG UST doesn't enable Atmos (and I think in many cases I started to receive just L/R to HTx, not sure if this was just flaky or didn't work at all - or required clicking EDID button). Thus I changed VAX404101 to non eARC where I just connect it to LG UST normal HDMI1. And then connect LG UST eARC to HTx. This also fixed a severe lipsync issue I was having, and now you can enable Atmos on Apple TV and it works great. The CEC doesn't work anymore fully with Apple remote (volume and mute is now broken, power on/off works — but works with LG remote), but I just learned HTx infracodes and use them on Apple Remote. Just if anyone else wonders this. The https://www.minidsp.com/applications/home-theater-tuning/apple-tv-4k-minidsp doesn't show Atmos either. In Apple TV the Atmos means better quality, it is still LPCM (and I guess Dolby MAT).
Last edited:
Yes! Residual noise in uV at the speaker terminals (combined amplifier and DAC/processor noise) is a good way to compare noise levels. A lot of manufacturers don't report this value but it looks like Peavey does:
Unfortunately, -87 dBu unweighted is 35 uV which is pretty noisy. If they didn't give you noise, you could also calculate from dynamic range. They report 107 dB DR and a maximum output level of +20 dBU (7.75 V), residual noise = 7.75 x 10^(-107/20) x 10^6 = 35 uV.
Another factor in system noise is amplifier gain. Even if an amplifier has relatively low noise itself, whatever noise is present at the amplifier input will be multiplied by gain. A low gain amplifier is a big advantage with high sensitivity systems. You can also achieve similar results with passive attenuation prior the amplifier, important to note that applying attenuation in your processor / DAC won't do anything to reduce the noise level.
Here is an example, a Hypex NC252MP has DR around 100 dB at 5 W in to 4 ohm and gain of 25.6 dB. Residual noise from the amplifier = sqrt(5 x4) x 10^(-100/20) x 10^6 = 45 uV. Amplified residual noise from the processor = 35 x 10^(25.6/20) = 660 uV.
Total system noise = sqrt(660^2 + 45^2) = 661 uV. You can clearly see the amplified processor noise is dominating this calculation.
Let's say your amplifier had only 12 dB gain like a Hypex NC252MP with the input buffer removed and the amplifier residual noise was the same 45 uV. Your amplified residual noise from the processor = 35 x 10^(12/20) = 138 uV and your system noise = sqrt(138^2 + 44^2) = 145 uV. This is OK performance but probably not good enough for a high sensitivity speaker.
Looking at the HTx, miniDSP reports dynamic range of 127 dB(A), unweighted dynamic range is usually 2-3 dB lower than A weighted. From @sarieri's measurements, this looks to be about right with an unweighted dynamic range of 123-124 dB. At 4 V output, unweighted residual noise = 4 x 10^(-123/20) x 10^6 = 2.83 uV, clearly much better than the Peavey.
Amplified noise at 25.6 gain = 2.83 x 10^(25.6/20) = 54 uV, system noise = sqrt(54^2 + 45^2) = 70 uV.
Amplified noise at 12 dB gain = 2.83 x 10^(12/20) = 11 uV, system noise = sqrt(11^2 + 45^2) = 46 uV.
You definitely want to be under 100 uV system noise and ideally below 50 uV if you have high sensitivity speakers.
I also see someone recommended the SPL8 as volume control, you should be very wary of adding anything to your signal chain. The SPL8 has residual noise of 8-9 uV which means you can never achieve a truly low noise system with it.
Michael
This is the most helpful post I've seen in a long time! This answers literally all the questions I've been struggling with!
I created a spreadsheet with everything you described so I have an easy reference.
My main takeaway is that the biggest single improvement I can make is to replace my processor and DAC with this HTx. Much lower noise, more functions, and it eliminates the need for a separate DAC, not to mention a bunch of extra interconnects.
Your amplified noise estimate with the Peavey is actually a bit low, because the Emotiva XPA gen3 (class H) amp has a gain of 29dB. So I'm sitting right around 1,000µV noise.
If I upgrade to the HTx, I'll be down closer to 113dB.
If I also upgrade the amps for the horns to ICEpower's lowest noise amps, that would put me down around 55µV noise after the amplifier. Atleast I think. I can't find a gain rating, but Amir tested one that was 25dB gain.
I also learned that many of ICEpower's modules are not actually quieter than my Emotiva amp. They are about the same, except it seems the ICEpower amps have a lower gain than the Emotiva. So the Emotiva would be quieter at the same gain....
(I see you met Krivium from DIY Audio forum. It's his fault that I'm using pro audio coaxial drivers in my speakers! And no, I'm not going to buy that volume knob....)
The gain is listed in the ICEpower datasheet (see attached). One thing to watch out for is the ASX series does not have balanced inputs, which may or may not be OK depending on your system.
If you can get away with unbalanced, the 50ASXS2 SE only has 20.5 dB gain and would be great with high efficiency drivers that do not need tons of power.
Michael
Attachments
Similar threads
