|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
73.229.163.4
Long PostIt begs the question, to DSP or not to DSP?
Old School
* Purist audiophiles do not use tone controls.
While being able to boost or cut bass and treble, old school tone controls in the signal path can audibly reduce the overall sound quality. I've owned a couple integrated amps with tone controls and sometimes the benefit of tone controls (or even the Loudness switch) outweigh the drawbacks. The drawback was a clearly audible reduction in overall SQ with a 'veil' over what was previously crystal clear or transparent.
A Modern Twist
Anyone here running Digital Signal Processing (DSP) to alter frequency response? I'm not talking about DSP for sample rate or PCM to DSD conversion but more along the lines of EQ to shape frequency response?
I've been running with no DSP for decades except for the very occasional "loudness bump" for low volume level listening similar to engaging the Loudness switch on older integrated amps. DSP works very nicely here w/o affecting overall sound quality.
Going Beyond the Loudness Switch
More recently I've been playing with the DSP Parametric EQ that is built into the ROON player software. HQPlayer has the same capability and more. These offer modern high quality DSP using the power of a PC or Mac.
I'm finding it very interesting that I can tailor specific bands of the frequency spectrum to help 'voice' the overall presentation without affecting overall sound quality. The music remains highly transparent or crystal clear unlike old tone controls that threw a veil over the music.
It can be very satisfying to slightly enhance or de-emphasize certain frequency ranges to compensate for speaker or room deficiencies - or to just play around with the 'voicing' while not affecting the overall SQ.
Am I breaking audiophile rules while running DSP?
My most recent DSP experiment is to give a little smooth boost to frequencies below about 30-Hz which is the bottom end range for my speakers. So I'm forcing more power at lower freqs to compensate for the reduced speaker response that low. It's certainly no substitute for running a couple big subwoofers but it appears to offer a no-cost enhancement and doesn't take up additional space on the listening room floor.
In a sense, anybody running a PC/Mac software based player is running DSP aren't they? After all how do we explain the differences in sound among the different software players? Aren't they all DSP 'voiced' to some extent by the software developer?
Just some of my rambling thoughts on DSP. What's your take?
.
.
.
.
.
.
.* True purist audiophiles swap various expensive interconnects as tone controls ;-)
Edits: 12/29/21 12/29/21 12/29/21Follow Ups:
Any rules it breaks are long obsolete. Some people are stuck in the past. I refuse to let them drag me down with them
Well, you are using DSP like I use it. I have an open baffle single driver speakers, the driver can play down to 25hz, but open baffles start to roll off at about 80hz. So I use Roons parametric EQ to boost bass after about 60hz.
Works wonders and I get usable bass now down to around 28hz or so. The key is that this driver is huge and has lots of excursion so it can handle the EQ without breaking up.
I consider any losses I get in the digital domain to be miniscule compared to using a speaker with a crossover. So I think it is actually 'less cheating' than using a speaker with passive crossover components.
Good point on the "less cheating". I like it ;-)
The rules are what the listener says they are.
Jim
Try Dirac; best upgrade I ever did. Not only removes room artifacts, but contours output as you chose.
How does one implement and run Dirac Live on a PC/Mac? It appears to be compatible with only certain playback software like Amarra, Audirvana, and JRiver Media Center to mention the ones popular with audiophiles.
I don't see it working with Roon. I suppose one can generate a room correction chart using Dirac Live then manually mimic it using DSP in Roon but I don't see a way to import actual filters from Dirac. I believe this is a Roon limitation of not wanting to integrate with Dirac Live but preferring their own DSP. I'm not sure.
The miniDSP SHD and SHD Studio devices run Dirac Live standalone; they are also "Roon-ready". Dirac Live setup does require a computer but does run on macOS Mojave or later as well as Windows.
The SHD device, (see link and picture), also provides 4-way DAC, analog input (ADC), and
is in effect a full preamp. Their SHD Studio device provides only digital outputs, leaving off the analog inputs and outputs.
Dmitri Shostakovich
Thanks for the info and the link.
I just read about the miniDSP SHD and the SHD Studio. These are Roon Ready and it appears that the SHD has its own DAC while the SHD Studio (which costs less) is meant to be used with your own external DAC(s).
I'm reading the owner's manual now.
I read that on some forum and once it sunk in I stopped being a purist. It all started with the BBE SOnic maximizer... that is such a polarizing piece but I tried it and man it is a keeper. You can adjust easily and its not exactly an eq. You can defeat it and I always come back to it. Yes I adjust for the volume...
That started it. But the main driver in the antipurist system is headphones. I needed to get the sound out the the head. Enter the Hafler HA-75 headphone amp. Has a tube input, crossfeed and a couple of bass boosts. Signal purity destroyed! Next up was the black ice soundstage expander. It also has an adjustable bass boost along with acting as a tube buffer.
Then there are the Audeze LCDi3 earbuds that are predicated on eq. So i bought a digital eq. I don't use those buds and don't use the eq. Its too much of a time sink these days to figure it out and get it right. Though one huge advantage of headphones is that there are eq settings on line so you can pretty easily get it dialed in. Some players already have the presets loaded in. Neutron player had probably 1000s of settings and its as easy as selecting your headphones in the menu.
My main cans are the Shure KSE 1200. They are electrostatic earbuds and really need a bass boost. So now the signal is TASCAM DA-3000-> Border Patrol Dac -> black ice soundstage expander -> BBE sonic Maximizer -> Hafler to the KSE1200 amp. Which is one jacked up signal chain. I prefer it over an iPhone running Neutron player with the preset eq...
Amarra has some plugins for the Audeze headphones that eq them. That might be better than the phone but I don't have the headphones that it has.
So no, nothing wrong with getting the sound to be what you want. Its all fake in the studio anyhow, unless you are listening to classical or something. Which you will be dead soon anyhow...
Cut to razor sounding violins
I eventually got a Schiit Loki for ease of adjustment. It has the right controls for my system and is easy to fiddle with both approaches made my system sound better. If it sounds better, it is better.
Gsquared
nt
Presently I'm using Foobar2000 with MathAudio's Room EQ -- it makes a very substantial improvement in the sound coming from my speakers.
Taking a step backwards, I'll admit that I haven't listened to vinyl in well over a decade simply because with good equipment the sound is better than vinyl. DSP makes it better still.
My next move is likely be the addition of a miniDSP SHD unit that will provide comprehensive EQ and room correction.
Dmitri Shostakovich
You can use DSP with freeware such as REW and Rephase, Accourate, Audiolense, or Dirac. State of the art DSP is with Accourate or Audiolense at this time. Dirac being good and user friendly compared to others. Math Audio would be my last choice and have tried it. There are several tutorials on Audio science review on setting up and learning REW. DSP is powerful and capable of making changes beyond simple EQ. Below 400hz and the room is in control, 400 and above and the speaker is in control. You fix the bottom end and hear many gains in the mid/top end as well. How far down the rabbit hole do you want to go? Mitch Barnett provides services if you do not feel like creating your own files. Also, the convolver used should apply to video, youtube, as well as music at different sample rates. Computers using DSP can handle many more samples than Mini DSP or others. In my case I have 131072 samples per channel for 2 channel music.MAK
Edits: 01/02/22
In fact I don't necessarily want to go nuts with this. At the same time I agree that MathAudio Room EQ is just simple equalization and doesn't do time domain corrections.
The Dirac on miniDSP is about as far as I'd care to go in terms of complexity. I like that miniDSP devices are standalone and require the computer only for setup.
Dmitri Shostakovich
.
if you own 3000 classical LPs.
Outboard unit, has balanced in/out, has digital out so you can use it purely in the digital domain while still using your favorite dac. Not exactly a plethora of digital in/outs but enough for me. Has volume control. There's a detailed Stereophile or TAS review from circa 2012ish(?) that convinced me to buy it.I love love love this dac, I love accurate un-wavering imaging and that it is. I put my Quad 11L2's on the floor (yes, on the floor - angled up - next to my reclining chair) ran the DSP and was in heaven shortly thereafter. Haven't moved them for years and years - deep rich sound with surefooted "elastic" bass is the only way I can describe it. When I use my Philips SHP-9500 headphones, I disable the DSP and use the DAC as is - that combo saved me a long, expensive trip down headphone lane (with Foobar as a CD transport on Windows 7, USB to DSpeaker).
I run my Technics SLQ-303 (w Vincent Audio PHO-8 + Ortofon Concorde 30th anniversary) to it's analog inputs and listen on the SHP-9500's and love it - have finally rediscovered my rekkerds. Without EQ enabled for headphones, or with EQ enabled to 11L2's.
My (decade old) unit supports 2 profiles - so you can switch between 2 sets of speakers if you use it as a source for more than one system.
I'm sure you can get computer based and/or more sophisticated systems that do a billion more things - but this unit is nice and simple if all you want is your room corrected. There are a great deal more settings if you choose to do a deep dive, I don't. HOWEVER - the unit by default does NOT EQ up to 500hz - I didn't find that out until years later, and when I set it thusly things got all the more wonderful - let the 11L2's upper midrange shine through on it's own plane. Best purchase ever. Haven't purchased new equipment in well over 5 years now!
I read this forum occasionally and am happy I will never discover all the joy (?) y'all find in new dacs - I like this one just fine - but if you review a lot of DACs, EQ'ing each will eliminate tonal (bass) differences from the comparisons. However I also love my 2008 Blue Circle USB Thingee with my Pioneer SP-BS-22-LR's (un-eq'd) and Class T Dayton DTA-120 so take my opinion with a huge grain of salt. Horses for courses...
EDIT: Final thought, IMO no EQ can make a small room "big" - it just can't. Even combined with room treatment, I doubt it. When I bought the DSpeaker, I could not emulate the large room I had just left - didn't happen. Had to rearrange my speakers to be "between rooms" so it was nowhere near any walls of either room. THEN I was in heaven (again)
Edits: 12/30/21 12/30/21
ASP
The Manley Massive Passive EQ box is the most transformative piece of gear I've ever had. It's very expensive, but they sometimes appear when a studio bites the dust. While that's always a sad affair, it's good to keep the MP in service somewhere. At least that's what I tell myself. Footnote: what they say is true - you can't make the MP sound bad, no matter how extreme the settings.
When I got it, I thought I'd play with it for a while, then pass it on. No way. With practice, I've learned how to EQ for any problem in seconds, without really thinking about it. I'm not alone - Bob Ludwig, when he's not dusting his twelve Grammys, masters recordings with his *three* Massive Passives.
DSP
The 100+ plug-ins included with the Metric Halo digital audio interfaces are phenomenal. Any combination of EQ and, say, compression is possible, especially if you use their graphic interface. Anything you can imagine can be done with this. Want to add a Corda Crossfeed to your signal, say, cross-feeding 20% of the signal with x msecs of delay? Or put a little compression on a narrow frequency range, say, for a singer with bad technique? Easy. And there's a bonus: when Metric Halo changed their architecture a few years ago, they issued field-installable upgrades for all their hardware, regardless of age. In 2023, they hope to have another field-installable upgrade that includes better converters and output sections. Nice.
Given that the MP and Metric Halo gear is used in studios the world over, I don't see why we can't use them at home. One more bonus: both companies have the best customer service on the planet.
WW
"Put on your high heeled sneakers. Baby, we''re goin'' out tonight.
Humm ... seems there is a plugin, (VST?), that simulates the analog Manley Massive Passive EQ box -- saving several thousand dollars no doubt. Additionally there are similar plugs I believe from other producers.
In any case I don't see these being the optimal solution to home hi-fi room correction. Much better to look to miniDSP devices, either setup with REW or Dirac Live. These measure your speakers' outputs and correct both frequency response and time domain issues including phase and step response.
Dmitri Shostakovich
I had the plugin, which ran on UAD hardware. It was OK, but didn't have the magic of the MP. It kinda sorta emulated the MP curves, but without some of the MP's little eccentricities, and, I thought, with different phase characteristics. (The user manual, btw, includes a master class on EQ and phase. Download it from their site.)
I don't use EQ for room correction. For room correction, fix the room; it's not hard. A new not-quite-parallel wall or two can transform a problem room into something divine, at the cost of a few inches. Take a look at Steve Addabbo's live room (pic). It's 26 x 20 x 9.5, with the long walls not quite parallel, and with the help of those little triangles, it's as neutral a room as you'll find anywhere. (The structure in the back corner is a fold-out vocal booth.)
The only room-correction system I've heard was at a demo in Yamaha's New York studio, and that was some years back. If you were in two or three seats in the middle of the room, it was pretty impressive. Anyplace else it was a mess.
The most dramatic room treatment I've seen is in one of the Kaufman-Astoria Studio's smaller sound stages. It is big enough for the Met orchestra(100+), chorus (200+) and principals. It has reflective flat panels on the walls that give it an almost 3-second decay-time. Flip the panels on their hinges so their absorptive side is exposed and the decay time is about half a second. They also have a sound stage big enough so they could film a car race with real cars racing at speed on a real track. Such great things were/are shot there, including Nurse Jackie, Orange Is the New Black, and Sesame Street. With the demise of Avatar, I suspect they're getting get a big chunk of Broadway, opera, and orchestral recordings.
It's all good.
Cheers,
WW
"Put on your high heeled sneakers. Baby, we''re goin'' out tonight.
My listening rooms is a small-medium sized, 12'x20', family living room and I don't have much scope to improve it with changes or even treatments.
Dmitri Shostakovich
could you do nearfield?
Cut to razor sounding violins
nt
Dmitri Shostakovich
Bet you could make it work. Be creative. Not nearfield like in a studio. Post a diagram of the room and I might be able to show you
Cut to razor sounding violins
My listening room is the main family living room. It's about 12'x 22' and heavily furnished and cluttered too. Here's a fairly recent photo, (from my AA Gallery)...
Dmitri Shostakovich
Got a pict from the audio wall showing the chair?
Cut to razor sounding violins
Which also begs a question for so-called 'audiophile purists'. If ASP and DSP are so commonly used in the studio why are some so opposed to it for playback in the home?
I've always shied away from applying DSP in my home audio setup maybe because early DSP didn't sound so great. I'm talking about DSP run on a home PC that also plays your music files, not additional add-on hardware.
HQPlayer offers excellent DSP, possibly the gold standard for PC based DSP in a player, but its user interface for Library management is garbage which is probably why many users run HQPlayer "behind the scenes" using another player as the front-end user interface.
I'm also finding the simpler DSP available within ROON to be outstanding for the simple Parametric EQ that I've been playing with in recent days.
I find the tradeoff useful with the HT to tame room modes and blend the subs for a more neutral bottom end response.
Fortuitous room dimensions, multiple treatments and optimized speaker placement, however, obviate the need for such in the music system.
My most recent DSP experiment is to give a little smooth boost to frequencies below about 30-Hz which is the bottom end range for my speakers.
I was able to achieve a mild tilt at the bottom and set the "brilliance" control on the backplates to about 1:00 for an opposite tilt at the top.
True purist audiophiles swap various expensive interconnects as tone controls
Never experienced such among any of the audiophiles I've known.
.
I set the RME DAC to its default settings then played with a simple one-band EQ in the DAC using the menu system but.....I'm finding Parametric EQ within Roon (on my Mac) to be much easier to deal with using a graphical display with drag and drop points and sliders for adjusting EQ.
My simple 1-band 3dB bass boost preset I created in Roon
Several adjustable EQ starting points in Roon
Many sample rate conversion possibilities in Roon. I'm not using them.
Edits: 12/31/21
You'll win on one side and you'll lose on the other side.
People who realized what a single filter (integrated on DAC)
can do to the sound might be able to imagine what a usually
quite complex sets of DSP filters can do to the sound.
And it's not just the (chain of) filters that cause issues.
It's the quality of filters. DSP doesn't equal DSP!!!
The issues related to recordings for room correction is another issue.
The vast majority of room-correctors out there do poor room recordings
with mediocre microphones and make mediocre filters from it. And these mediocre filters they apply to an attenuated signal to avoid clipping.
And on top of it come the losses that are caused by realtime processing
on your computer platform.
To me all the DSP stuff, which includes sample-rate conversion, is a NoGo. I tried several times to go that route.
I also listened to several commercial room correction systems.
In the beginning they sound impressive. After a while I realized there's something wrong with it.
I rather prefer to be an audiophile frugalist. Less is usually more.
Enjoy your toys.
-----------------------------------------------------------------
blog latest >> The Audio Streaming Series - tuning kit pCP
That's kind of where I stand too - for example, I haven't employed HQ Player's upsampling options for years now.
(I still stand by my other post though - it's OK for other folks who want to experiment with it.)
It's not as if you were changing the recording forever, so that we could never get back to the original. And it's not as if you can't undo your DSP modifications at some point whenever you want. It all seems pretty benign to me.
This is a long but very well organized and expressed overview of the more capable DSP systems applicable to audiophile 2CH and MCH systems. I learned more from this (in two passes) than I have anywhere before. It's from Mitch Barnett -- @Mitchco or @Mitchba on various forums.
https://youtu.be/yfGAUvyvdNU
Linked
Modern DSP which is integrated with quality psychoacoustic models and algorithms is great and and improvement, not a tiresome gimmick.
In particular I use Anthem ARC (limiting correction to an upper frequency bound), integrating sub with mains.
Next, the Anthem also has Dolby Volume with an absolute level-aware Fletcher-Munson correction curve (which changes with volume of course)---I don't detect any major undesirable timbre shift (unlike haphazard EQ) as opposed to Just Sounding More Right.
Probably the SOTA DSP for consumers is Dirac Live , (see link below), which provides comprehensive room correct. It can be run on a computer or devices appropriately equipped such as several miniDSp devices , the NAD M33 , and a few others. In general with such devices a computer is necessary to set up the correction but not for operation.Is this sort of comprehensive DSP breaking audiophile rule? Not if the most accurate and transparent sound is what you want. (Why do I get the impression at times that it is not what some audiophiles want?)
Dmitri Shostakovich
Edits: 12/29/21
As to the last point, of course many don't; they want a euphonic and sweetened up version. Which is fine. We each decide what sounds best to us and there are clearly different points of view on that...
Dirac is a boon for casual users, but it isn't really bleeding edge. It is a good well-rounded solution that is easy to use but is optimized for operation on embedded system hardware which is not an ideal solution. I use it myself for my 5-channel recordings and for home theater. However, for most implementations it is limited to internal processing at 24/48 and even the desktop version which is less limited is less configurable by the user than the other DSP solutions. Although I would rather my recordings be decimated to a 24/48 rate and used with Dirac than have them at full resolution without it. It does work quite well within its box of limitations..
However, for my two channel recordings I use acourate-created convolution filters at 384kHz and there is a clearly audible improvement over Dirac. Acourate is less automated than Dirac, allowing far more refinement to the phase optimization and time domain behavior. Audiolense provides the same additional levels of refinement. These are really the state of the art in DSP for hifi. My experience with them has been revelatory and I would never consider not using DSP again for my rooms.
For my old ears, 16/44.1 is adequate. Much more depends on the quality of the mastering than on the bit depth.
Dmitri Shostakovich
especially if you are using a DAC that uses an FPGA to 'correct' the output (ie filters).
I strongly suspect DACs that everyone claim produce 'great bass' may well be giving the bass frequencies a bit of a boost somewhere in the digital to analog chain.
JMNSHO and all.
...but not all DACs rely on the digital filters built into the DAC chip itself. Some use analog based filtering at the digital output rather than relying on the various digital filters in the DAC chip.... or FPGA.
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: