At least you can try EQ but XO will require multi channel. I have heard good reviews of the Asus Zonar and daughter cards for 8 channels. But you need a big PC box. I like how miniSHARC can be stand alone and compact once set up and programmed.
The rephase program is good stuff and is in fact how minDSP and the miniSHARC work to implement FIR.
The rephase program is good stuff and is in fact how minDSP and the miniSHARC work to implement FIR.
I have used the equivalent on my FR drivers. Very nice. If using mini anything, grew is the way to go.
This is just a suggest but it makes possible to exercise such setup without high costs.
2 times UAE23 USB audio devices up to 24bit/96KHz ES9023 DAC this gives four analog or digital outputs for active FIR XO and EQ at cost (2x$42.00) $84.00 here HiFimeDIY Sabre USB DAC. Digital to Analog Converter 96khz/24bit (incl USB to optical converter feature), you just further need a desktop not too old. Not shure Foobar can use two sound devices but read somewhere JRiver can. If interested i can test if JRiver can route XO to two USB devices this weekend, just request.
I say this because i run that device myself, and prefer it even I have Xonar ESSENCE ST (PCI) and Xonar U7 (newer USB 192KHz). My thought for the better performance is that the chipset runs native Win7/8 drivers, and another thought is that a MS engineered driver have all chances for best throughput in kernel because of knowledge at lab, but just a thought. Then use the WASAPI mode in JRiver or Foobar for exclusive use of sound device bypassing windows wordlenght/samplerate lock settings.
Later i intend myself playing with the REPHASE program, but present too many plans (buzzforb me too
...
............🙂...Alpair10.2M...FAST).It's pretty much the rule that your EQ is messing with the phase unless otherwise stated. Since digital linear phase algorithms take extra work to implement they are usually labeled as such by the programmer.
EQing material on a nearly daily basis I'm still on on the fence about best usage scenarios for linear vs minimum phase. I had a love affair with linear phase for a period of about a year but then started to feel like my work was a bit digital sounding (whatever that is)... I think the pre-ringing becomes noticeable when used on multiple tracks (when mixed together) as the cumulative result seemed slightly brittle (very slight here, we're talking nuance). After that affair ended I spent more than a year working mostly with minimum phase and I feel now like bits of those mixes have extra softness at points. Lately I've been using both depending on the material and the frequency range (but more of a trial and error process not, "ohh, this snare is just asking for LP").
I'm under the impression minimum phase sounds more natural given the ringing occurs after the event like our auditory perception expects (allowing us to tune it out).
For room correction, I'd start with minimum phase and work out a profile that sounded good and then copy the profile using linear phase and a/b to see which i liked better. fwiw, EQ'ng bass seems less dangerous until your room isn't ringing like a bell after which you notice that phase shift has a pretty noticeable impact on low transients.
This article about crossovers has a good summary at the beginning:
A Bessel Filter Crossover, and Its Relation to Others
& this paper about active filters is incredibly data rich:
Active Filters
I've not done the conceptual thinking to fully understand this, but I've read that full range speakers exhibit phase shift at multiple points depending on wavelength.
It's been shown that phase shift when static isn't so noticeable (with single pure tones), so we're mostly talking about artifacts in the amplitude department... However phase shift might be noticeable when complex waves made from an instrument (and interrelated harmonics) cross a filter pole at different times... If this is noticeable, which I think it could be, I would categorize it as a "phase coherency" problem because the harmonics are related and perceptually expected to stay in time so when the fundamental cross the boundary yet the upper harmonics don't shift with it, there is now a temporal mismatch causing a softness or blur...
Last edited:
I never thought of parallel usb dac's before. I wonder if these can work?
Free shipping 2014 new product MUSE DA10 Top quality PCM2704 USB DAC Nice Mini DAC wholesale price-in Consumer Electronics on Aliexpress.com
Free shipping 2014 new product MUSE DA10 Top quality PCM2704 USB DAC Nice Mini DAC wholesale price-in Consumer Electronics on Aliexpress.com
BRYTT,
I believe you and would appreciate the test. There are some good sub dacs out there, but something like multichannel evasIOn would be ideal, if for no other reason than it plays DSD. The Oppo 105 is one of only a few co'm opponents I would be willing to purchase over building. Extraordinary sound given the price.
The FAST idea is shelved for the moment, with TL lab tweeter, AT Cquenze mid, and AS TD12 woofer being the current plan
that being said, if it does not impress like Alpairs, it will be done away with rather quickly. Always have my 10.2 and 7.3 to listen to any time🙂
Thanks to all. This has turned out to be a wonderful and educational thread
I believe you and would appreciate the test. There are some good sub dacs out there, but something like multichannel evasIOn would be ideal, if for no other reason than it plays DSD. The Oppo 105 is one of only a few co'm opponents I would be willing to purchase over building. Extraordinary sound given the price.
The FAST idea is shelved for the moment, with TL lab tweeter, AT Cquenze mid, and AS TD12 woofer being the current plan
that being said, if it does not impress like Alpairs, it will be done away with rather quickly. Always have my 10.2 and 7.3 to listen to any time🙂Thanks to all. This has turned out to be a wonderful and educational thread
Looks good it's a steal, PCM2704 though less flexibility in bit/samplerate steps than the other. Imagine must be best use same brand/type to have best sync/precision, because there is probably some buffers involved in USB chip and DAC, so as these are same sizes and routes.
I do and post suddenly thru weekend, also have another test of phase plot in a DEQ2496 to be done promised earlier.
I am envious for 7.3's, but very satisfied with 10.2.
Same here also thanks to anthonybisset for splendid inputs
I would think that the build in EQ is minimum phase. In both instances. And yes, in that case phase will be affected. The timing will be smeared with the use of that sort of EQ and it will affect tonal balance.
There are a lot of EQ plugins that claim to have linear phase response though. But not all are build the same. The better "mastering like" EQ's work the audio in 2 directions to minimize the ringing you get with fir filter solutions.
It will introduce latency though because the audio has to be processed in both directions.
PEQ plugins like Algorithmix LinearPhase Red and Orange are that way and are said to be very transparent. And also very expensive and used in professional mastering. The high quality setting can cause up to a second latency.
PLParEQ seems to operate in a similar (2 directions of audio processing) way, still not exactly cheap but also provides phase linear behavior. There used to be a free 3 PEQ version, not sure if there still is. The newest PLParEQ is optimized for Win7 and multiprocessors.
I have found some claims that linear phase PEQ is more exact than fir filters in the bass region. Not sure if its true though.
Jriver can offer both on a PC solution. I plan to try both, unless one of the two has me convinced no further research is needed 😀.
After searching a bit more about the differences in the Algorithmix Red and Orange PEQ's I found the claim that the Red is actually FIR based and the Orange uses IIR in two directions.
By this statement I was getting at the PEQ plugin and the option of using FIR filters trough convolution. But I guess I could also experiment with a FIR based PEQ to find out. But that Algorithmix plugin is way out of my price range 🙁.
I'll now stop quoting myself...
Last edited:
DMG Equilibrium is very very good, the best I've used with selectable filter types per band, similar quality and far more flexible than the overpriced algorithmix offerings. Yet, such mixing/mastering EQ's seem off target for room correction. EQ no matter how good is a band-aid/splint/surgery... Yes it can improve a situation but it's "corrective" and given that we have access to the components and designs that make up the problem we can fix it at a deeper level with more control.
fwiw, I think it's overkill/underkill for most speaker designers:
DMG Audio : VST/AU EQ, Compression & Dynamic Processing
I'm looking forward to miniDSP someday offering DEQX like functions... 🙂
Trying to tune a system with a bit of EQ is helpful, but proper convolution just gets it done.
fwiw, I think it's overkill/underkill for most speaker designers:
DMG Audio : VST/AU EQ, Compression & Dynamic Processing
I'm looking forward to miniDSP someday offering DEQX like functions... 🙂
Trying to tune a system with a bit of EQ is helpful, but proper convolution just gets it done.
Anthonybisset,
I have Sony Acid Pro 6 and found that there is a free plugin under Gnu PL called Freeverb3 VST - it claims to have an FIR synthesizer. Have you heard of this and can it work?
I have Sony Acid Pro 6 and found that there is a free plugin under Gnu PL called Freeverb3 VST - it claims to have an FIR synthesizer. Have you heard of this and can it work?
wesayso you write they very expensive so just checked what actual that was, uuuhh only worth if used in scenario to get a income out of the invest, not for normal consumer. Looks like both companies make income delivering code pieces to both hardware and software products. This makes me think of example take EQ's they are everywhere in player packages, interface software for codecs, DAWs , OS and etc, but of course in lighter version than the very expensive RED/ORANGE.
After that checked my JRiver and now wonder why there is two identical "Parametric EQ" packages which with same settings setup sounds different in high frequency area, the one have highs in your face and the other have soft highs. Later when trying to measure phase on a DEQ2496 I will try to investigate phase on those two too. First picture the two similar meny's, picture two is filters available, and is same for both.
After that checked my JRiver and now wonder why there is two identical "Parametric EQ" packages which with same settings setup sounds different in high frequency area, the one have highs in your face and the other have soft highs. Later when trying to measure phase on a DEQ2496 I will try to investigate phase on those two too. First picture the two similar meny's, picture two is filters available, and is same for both.
Attachments
Anyone interested in linear phase filters, please consider what the contributors to this thread (esp. Jay Frigoletto and Dave Collins) have to say about them:
linear phase versus minimum phase (EQ's) - Gearslutz.com
linear phase versus minimum phase (EQ's) - Gearslutz.com
No, I haven't heard of it. The vst/au/plugin space is over-saturated with literally hundreds of offerings per class of eq/filter/effect/whatever..
The linear phase process isn't terribly hard to implement but it does take some effort. The quality of their DSP programming will influence the sound, probably just as much as IIR/vs/FIR..
If it sounds good, ________,.
On the cheap you can experiment with all manner of phase filters in ddmf lp10 ($39):
DDMF: VST/AU/RTAS plugins, audio software
Nice thread on the hype of linear phase filters there Anthonybisset. For us folks without the capability to try this out now can you post a small clip with linear phase and with minimum phase EQ'ing? I want to hear this pre-ringing that everyone complains about.
Thanks
Thanks
Are you sure you need a second parametric EQ add on. Can you not arrange channels within the first? I will investigate this soon. Might order a couple of the ESS dacs to play with on this front.
With this i mean that maybe the one is IIR and the other FIR, as i expressed listening in highs the one is in your face and the other soft, and this is a comparing where the settings for the filter is at same frequency/amplitude/Q. You know I wonder why they put one named "Parametric Equalizer" and one named "Parametric Equalizer 2" when all settings and features for both can be the same. But irritating that the help bottom doesn't explain why 😕, hitting help bottom you just get routed to website with general explanation adding EQ and filters.
buzzforb ofcourse you are the commander for the decision, but if tests turns out bad performance i suggest wait till done saving the $ for other nice stuff.
This seems to be a more ideal application for what we are talking about, but is no lobnger available. The only other prodcut I have found is the USB streamer from Mini DSP. The Amanero and Wave IO do not seem to support muitlicahnnel operation. I thiink the problem may be one of synchronization. It seems that unless the different dacs are synced by an external source, they mey get out of sync with each other.
www.exadevices.com > exaU2I > Overview
USBStreamer | MiniDSP
www.exadevices.com > exaU2I > Overview
USBStreamer | MiniDSP
Is convolution with FIR filters free of pre ringing when trying for linear phase?
Example 1: Low-Pass Filtering by FFT Convolution
Not if I look at this example... the ringing is clearly visible and similar to what you'd expect from linear phase PEQ.
Wanting to know the same I came across this:
http://www.cambridge-mt.com/ms-ch11.htm
(scroll down to Pre-ringing Side-effects Of Linear-phase EQ)
Last edited:
Wesayso, glad you brought this up.
Anytime we modify a sampled signal both amplitude and phase change.
I was referencing convolution as being more *useful* for system tuning/room correction than EQ (be that IIR, FIR or 2 pass LP IIR). I didn't mean to hint that convolution is immune to ringing or other artifacts. Merely that is more able to account for the time component in the corrective processing and (either do more damage or correct more damage, depending on your digital bias 🙂)
Anytime we modify a sampled signal both amplitude and phase change.
I was referencing convolution as being more *useful* for system tuning/room correction than EQ (be that IIR, FIR or 2 pass LP IIR). I didn't mean to hint that convolution is immune to ringing or other artifacts. Merely that is more able to account for the time component in the corrective processing and (either do more damage or correct more damage, depending on your digital bias 🙂)