Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Bennco]

Can someone provide some comments on the pros and cons of using the DSP features in V18 versus using external hardware e.g. Behringer DEQ 2496?  Thanks.

[glynor]

MC:

+ Flexible options
+ Powerful configuration choices
+ Always expanding
+ Extensible with third-party plugins
+ A user interface that makes sense
+ The developers are open to change, and regularly provide detailed information when asked
+ Fewer blinking LEDs
+ Included with MC along with everything else
+ Bit-perfect whenever possible, using high-quality algorithms and high bitrates everywhere else)
- Only applies to audio played through MC itself (so it can't be used on other applications, like games) (I forgot about Loopback, see below.)
- Can be more complex to set up
- No "full auto" modes for easy setup (ala Audyssey)
- Fewer pretty LEDs

Hardware:

+ More pretty blinking LEDs
+ Can be at the "end" of your audio chain, so will apply universally to the system's audio
+ Fun dials and esoteric lights
+ Manual controls (dials, faders, etc) can be easier to use
+ Some systems have nice automatic setup systems built in that are easy to use (for room correction, etc)
+ Sometimes (but not always) custom chip logic is faster, if it is built well on a modern process node
- Sometimes has analog signal loss (depending on setup and hardware)
- Often has lower-quality fixed-function processors (the Intel CPU in your machine is MUCH more advanced and capable than whatever ASIC or Flip-Chip an audio vendor sells - Intel has the best fabs in the world)
- Often less forgiving of less-than-perfect sources (worse error handling)
- What you buy is what you get (forever, usually)
- Often higher latency (sometimes much higher)
- User interface designed by an engineer from Mars with Asperger syndrome
- Expensive
- Higher power consumption
- Black box (you can't ever know exactly what is happening in there on that signal path -- frankly, with the way chip design works, they might not even quite know)
- Lots of "pretty" blinking LEDs

[glynor]

If you want details about the capabilities, try some targeted searches here in the forum.  The Parametric EQ, Convolultion, and Room Correction systems are particularly powerful in MC.  They'd be good places to start.

[mwillems]

To speak to your specific question, my best friend and I both recently finished setting up similar bi-amped systems and we both needed a lot of DSP to make it work (active crossovers, delay, compression drivers with response that needed EQing, etc.).  He went the hardware route, and I went the JRiver route, so I can tell you first hand some of the ups and downs.

He started out with a Behringer DEQ, but that wasn't really adequate for his DSP needs, and he switched over to its cousin the DCX2496 which has six output channels, more flexibility, and more DSP options.  He's quite happy with the DSP his DCX gets him, but it is limited (the DCX only supports a fixed amount of processing/number of filters, and a fixed number of channels in and out).  If he wanted to expand his set up (i.e. surround sound) or if he wanted to do much additional EQ he'd be bumping against the limits of the DCX.  Additionally depending on how you use the DCX, it may involve a A to D conversion followed by an additional D to A conversion, which (while probably inaudible given the high quality conversions the DCX does) is another two conversion steps in the signal path if you care about that kind of thing. Each of those boxes retails in the high two hundred dollar range.

By contrast JRiver supports an arbitrary amount of DSP with many more options, and is only channel limited by the number of output channels on your soundcard/HDMI output.  JRiver provides customization that's pretty hard to find in hardware, and it also retails for significantly less than either of the boxes.  There's a bit of a learning curve with using JRiver, but that's true of the boxes too.  

To summarize, for the price of JRiver and an 8-channel Asus soundcard I have DSP capability that exceeds what my friend can manage with the DCX, a comparable DAC, two fewer conversion steps in my signal path, easy integration with computer-based content (youtube, netflix, etc.), enough output channels for a 5.1 surround system (even after losing two to bi-amping) and it cost me about half what he paid.  

The DEQ (as opposed to the DCX) has some non-DSP functionality, like RTA capability, but there's free RTA software out there if you need it. You'll also need a calibrated Mic though (I'm not sure if the DEQ shipped with one or not)

If there's any specific DSP functionality you're curious about with the DEQ, the DCX, or JRiver I've got access to all three and can find out if I don't know.  

- Only applies to audio played through MC itself (so it can't be used on other applications, like games)

Also @Glynor- I use loopback to output through JRiver with Games all the time.  As I understand it the surround support isn't there, but I use JRiver to output all audio on my PC regardless of source.

[glynor]

Also @Glynor- I use loopback to output through JRiver with Games all the time.  As I understand it the surround support isn't there, but I use JRiver to output all audio on my PC regardless of source.

Yeah, I really need to get in there and play with the Loopback stuff more.  I'd read about many of the changes, but I haven't followed it that closely.  If it doesn't support 5.1 at least, though, I'd be pretty bummed out.

[mwillems]

Yeah, I really need to get in there and play with the Loopback stuff more.  I'd read about many of the changes, but I haven't followed it that closely.  If it doesn't support 5.1 at least, though, I'd be pretty bummed out.

It works flawlessly in my application, but I'm basically running a 2.1 system with bi-amped mains.  It effectively redirects everything and it sounds awesome.  It displays in audio path as a stereo source, but that may be a function of whatever is getting outputted to the loopback source (if that makes sense).  I wouldn't know (firsthand) if it supports surround because I haven't tried to get it to do so (if you see what I mean), but I heard that it did not. Play with it and see, you'll like it :-)  

Let me know if you figure out a way to make it do surround, other than just using the upmixing of course, I plan to have a surround set up one day (sooner than later) and it'd be nice to know how to do it when I get there ;-)

Sorry for the OT derailing

[glynor]

Sorry for the OT derailing

Not OT at all!  That was very relevant, and I thought your write-up was great.

[csimon]

I take it people like JRiver then? 

[mwillems]

I take it people like JRiver then? 

Compared to my old solution for PC DSP (VST Host and Virtual Audio Cable with plugins for all the different pieces using a moderately flaky M-Audio card as my output) finding JRiver was like watching the skies open after a thunderstorm.  No more crazy latency, no more mysterious static, no more having to restart the windows audio kernel periodically.  My PC is the main input for our living room television, so poor performance was a problem for the whole household.

I knew it was getting bad when my wife had to ask me to write her up an instruction sheet on how to get audio out of the computer when I wasn't around to do it myself (after she spent a few frustrating evenings unable to watch movies).  I knew it was terrible when I wrote the instructions out and, after covering all the use cases, it took five pages!  This did not promote domestic tranquility.

The switch to JRiver helped a lot. It took a while to figure out the configuration, but now that I've set up all the DSP and zoneswitch it's pretty low maintenance. I decided a few days ago to update the instruction sheet and realized we don't really need one anymore ;-)   

[Datman]

I hope I don’t derail this thread to bad. A few years ago I got a BFD  dsp 1124p to help tame my subs. I learned of it from hometheatershack.com. I also when the route of building and installing room treatments. The REW program did a great job of reading the room and programmed the BFD to help flatten the response. It works so well and my subs are so smooth I forgot how to set it up. This is the only link I could find.
http://www.hometheatershack.com/forums/home-audio-acoustics/51963-my-new-treatments.html#axzz2IGP8F5V4

My hope is as learn how to use MC18 I can use REW to help me set the DSP in MC18 to further correct the rooms response.

JFYI the room treatments by far had the biggest results.

found the other link. Just look for the screenshots of REW readings. They are actual pre and post results
http://www.hometheatershack.com/forums/electronic-processing-equalization-devices/37790-got-my-bfd.html#axzz2IGP8F5V4

[mojave]

I've owned a DCX2496 and a miniDSP in the past and JRiver is easier and more flexible than either one for most things. If using REW to generate correction filters, it is easier to update the miniDSP. I have found that JRiver has better accuracy in matching the Room EQ Wizard (REW) predicted response when using the recommended filters.

The Loopback works great even for 7.1 gaming.

My hope is as learn how to use MC18 I can use REW to help me set the DSP in MC18 to further correct the rooms response.

Use Generic as your equalizer in REW and then copy the filter settings into JRiver's PEQ. To measure the corrected response, you need to use the Loopback in JRiver or the RTA feature in REW. The Loopback will allow the REW sweeps to run through the JRiver DSP and give you the best measurement. The RTA is great for setting the correct distance setting for the subwoofer in JRiver since the acoustic distance can differ from the physical distance.

[Hipper]

I'm currently developing my computer audio set up for a headphone system but I can hear already that it will later be used on my speaker set up, therefore I am looking at this very question of JRiver DSP or external hardware.

My speaker set up is: CD transport - Behringer DEQ2496 - DAC - amp - speakers, the Behringer therefore only working in digital. I'm very happy with the Behringer and this, along with Auralex foam products are the making of my system.

The options for me are, connecting a computer to the Behringer, using a USB/SPDIF converter, and hence using my good but nine year old DAC, or, have equalising software with JRiver and still using my DAC or a newer one.

On the question of JRiver software or plug-ins, is there something that effectively does what the Behringer does, namely a 31 band graphic equaliser and 10 band parametric equaliser, and the ability to do it for each of a pair of speakers (dual mono as Behringer calls it). The latter is important to me. I am aware of REW and it's measuring abilties - the Behringer can do measurements (you have to buy a microphone, cable and stand separately). The nearest I've seen is FabFilter Pro Q.

What about the question of computer work when playing music? Will it interfere in any way with sound quality?

I currently use the JRiver equaliser to tamper with the sound for my headphone set up and would prefer a few more bands to play with. However it works well and I don't seem to have any problems associated with it.

[natehansen66]

Maybe I'm missing something but this thread seems to answer your questions already. MC's DSP can do much more then your Behr gear, aside from measurements as you noted. I use it for my 3-way active speakers, and have another zone setup for headphones...... while doing all kinds of other work on my PC with no issues. I couldn't imagine adding another box to my setup when MC handles it all quite well.

[Hipper]

Well then the question is, how?

I already have the measurements I need that give me a sound I like. I see that I could enter these using the Parametric EQ in JRiver, but it would need about 82 entries (2 speakers times 31 bands plus also 10 Para EQs for each one).

The other way, I assume, would be to use the Convolution page and add filters. How do I make filters with existing measurements?

[mwillems]

Well then the question is, how?

I already have the measurements I need that give me a sound I like. I see that I could enter these using the Parametric EQ in JRiver, but it would need about 82 entries (2 speakers times 31 bands plus also 10 Para EQs for each one).

The other way, I assume, would be to use the Convolution page and add filters. How do I make filters with existing measurements?

Without knowing what your 82 entries are, it's hard to speak with confidence, but I think you could almost certainly consolidate a large number of those entries into fewer settings as the parametric EQ is incredibly flexible.  I used to use graphic equalizers, and I found that (carefully used) a PEQ can do in one entry what it takes three or four to do in a graphic equalizer (and will do it more precisely). 

For example if you're using the graphic equalizer to create a roll off (or a rise) at the top or bottom of the range, you could replace multiple bands of your graphic equalizer with a single PEQ shelving filter.  If you have multiple overlapping EQ bands (because you're using both graphical and PEQ), you could probably simplify them into one or two PEQ entries.  I know you indicated that you applied different settings to each speaker, but if there are any of the settings which are the same, you can have a single entry for both speakers (you can decide which channels are affected on a filter by filter basis).  If you could describe your settings a little more (i.e. the general shape of the EQ applied) I could probably offer more specific suggestions.

It's going to take some paddling around (or some math), but I'm betting you could probably produce a +/- 1dB equivalent in well fewer than 20 entries (probably more like 10 or 15) unless you have some serious EQ ripples across the band.  If all that sounds a little daunting, I seem to recall that there is a 31 band graphic equalizer VST plugin that will slot right into JRiver, but I can't recall the name.  There was a forum thread about it a while back?

[Hipper]

Thanks for your advice.

You may well be right in that I can use fewer entries. I'm not ready to do it yet but am looking forward to it.

It took me a few goes at learning how to use the Behringer. What I do now is use it to calculate a flat response for each speaker, using not only its RTA but some test tones in the bass area. I then test it as a stereo pair. This is followed by a set of dB alterations to particular frequencies, and finally a few tweaks to deal with particular points of certain tracks. Once done, that's how it stays - I don't constantly fiddle with it.

I like the result but, for example, my dealer doesn't. 'Every body listens differently' was his response!

[mwillems]

Thanks for your advice.

You may well be right in that I can use fewer entries. I'm not ready to do it yet but am looking forward to it.

It took me a few goes at learning how to use the Behringer. What I do now is use it to calculate a flat response for each speaker, using not only its RTA but some test tones in the bass area. I then test it as a stereo pair. This is followed by a set of dB alterations to particular frequencies, and finally a few tweaks to deal with particular points of certain tracks. Once done, that's how it stays - I don't constantly fiddle with it.

I like the result but, for example, my dealer doesn't. 'Every body listens differently' was his response!

I also use a set and forget method, but my setting process is a little less "automated."  I take a gated logsweep measurement in Holm, and then try to estimate what combination of PEQ filters would get it flat.  Then I dial them in, figure out how wrong I got it, and tweak until things are as flat as they're getting.  Then I RTA just for a sanity check.  In my room, if I get a flat measurement response in Holm, my RTA always looks like an X-Curve, which is fine by me. 

Also, you mentioned REW above; it will do an automated measurement and EQ things flat by generating a convolution filter for your system.  You could try that out, and then make some additional manual tweaks as needed.  That might be a time saver if you don't mind running convolution.