Welcome, Guest
Username: Password: Remember me
NOTE: This is a "Community" forum. Please be mindful that community members are here to help as part of a community effort. We therefore appreciate your effort in keeping this forum a happy place!

If you have a specific issue (e.g. hardware, failure) and want help from our support team, please use our tech support portal (Support menu - > Contact Us).
Thanks a lot of your help in making a better community.

TOPIC:

Target - 0d and Below or Average above and Below? 2 years 8 months ago #47515

  • Greg the Yeti
  • Greg the Yeti's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 22
  • Thank you received: 0
No - the actual "Dirac" Corrections, with adjustment for Phase/Amplifier/Speakers/Sub/Room is done ONCE, but the Tilted Target Line - which is ARGUED to be "because of" inadequacies in the Room - IS NOT and is simply an adjustment tailored to the Average Tastes of Listeners.

They ARGUE that it is for a "Techie" reason, so that Customers can have the Bass Response and lack of SCREECHING Treble that they want, without having to feel bad about it!!!

Get FACTS STRAIGHT john.reekie!

If a person ARGUES that the Tilted Target Curve is for the "Techie" reasons often given (which I won't detail since they're FALSE) then they MOST DEFINITELY are TRYING TO PUT THE THING INTO THE EQUATION TWICE!

Don't try to justify the impossible again!

Regards Greg

Please Log in or Create an account to join the conversation.

Last edit: by Greg the Yeti.

Target - 0d and Below or Average above and Below? 2 years 8 months ago #47516

  • asx77
  • asx77's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 624
  • Thank you received: 109
Struggling to understand the point of arguing.

Dirac measures the combination of source-amp-speakers-room. To its best effort it can then can give you a flat response or a house curve.

Now in my mind there is no point trying to recreate the live performance. Why? Well because how can I second guess what the sound engineer was trying to achieve? One might have assumed I listen using MP3, (I don’t and never will) another might have assumed I’ve got a £10,000 system. And over the decades as people’s listening habits and equipment changes, the way sound engineers mix has probably changed so if you have an eclectic collection good luck creating or choosing a curve that will correctly mimic the live artist.

No. You like me want a sound that you can listen to for hours. If flat floats your boat go for it but I for one will try the recommended curve since Dirac have done more listening tests and research than me and want there product to produce an enjoyable sound to get a good reputation and sell more right? And if different, I’ll also try the Harmon curve which was developed off the back of user listening studies.

The one that works for me I’ll use. I would say here each to their own and make peace but I’m enjoying the back and forth entertainment too much!!
Warning: Just because I'm a 'Platinum' member, doesn't mean I know what I'm talking about... It just means I've asked too many questions!!

Please Log in or Create an account to join the conversation.

Target - 0d and Below or Average above and Below? 2 years 8 months ago #47518

  • Greg the Yeti
  • Greg the Yeti's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 22
  • Thank you received: 0
Hi asx77,
yes - sorry for the "Shouting" (capitals) in my last post. I just hate it when "Senior Forum Members" lay into the discussion and it turns out they're saying something that's actually impossible! When they don't know as much as me - a humble "Boarder".

You're right - "Dirac measures the combination of source-amp-speakers-room. To its best effort it can then can give you a flat response or a house curve." End of! We DON'T need to compensate AGAIN in the Target Curve!

And you're right that each person may do what each person wants to with their SHD.

I'm just trying to make the case for a Totally different approach than all the one's I've seen on the Forum and different from the Dirac "Default! and the "Harmon" suggested one too.

There are good reasons for every part of the design of my Target Curve. I hate it when people bring up false "Techie" reasons why I'm wrong!

My Target Curves (four of them) you will see at the top of this discussion thread. You will see that ealier in the thread I was polite.
You will find that my Target Curves, for a 3db, 5db, 7db or 9db Bass Boost are neither Totally Flat NOR Tilted like the Dirac one.

They consist of Two Flat Plateau, with varying difference between them. They sound absolutely GREAT and on my system outperform the Standard Dirac, the Harmon and all the others I've considered and tried.

They give Great Sounding Bass, with Bass Notes equal to others , rather than with a discrepancy of volume in a Bass Line.
They give absolutely GREAT Mids and Teble. My Oberon 5's tend to be a bit bright, but the sound is Crystal Clear and "Glass like". With my curves the Treble is more controlled, but the Tonal Character of the Music and my nice speakers is retained. Mids and Treble are still Crystal Clear and "Glass Like" - but without a few of the defects (largely from my room).

I'm making a case for a different approach and resented, quite frankly, "incorrect facts" being used to counter my arguments! Sorry if that ofended.

You seem to have been entertained -fine! Enjoy your music! I enjoy mine!

Regards Greg

Please Log in or Create an account to join the conversation.

Last edit: by Greg the Yeti. Reason: Spelling again!

Target - 0d and Below or Average above and Below? 2 years 7 months ago #47611

  • Greg the Yeti
  • Greg the Yeti's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 22
  • Thank you received: 0
Hi again Entripy!
I've been following your advice to slowly increase the output gain in the SHD Plugin, watching for any clipping (or should I say listening!).

My question here is about clipping. With regard to the nature of clipping of a digital signal.

Probably like you I'm very familiar with the various sounds of clipping in terms of Guitar and Bass. Various different kinds of "Overdrive" on the Marshal Amp I have for my lead guitar, but also the rather different, more electronic and harsh sounding clipping in my Bass Amp when it's driven a bit harder.
The speakers of both are designed to take all that abuse. And used well they sound great, but I'm also well aware that Clipping in an analogue signal is the most common reason for Tweeter Failure, because Analogue clipping produces LOADS of high frequency that wasn't in the Music!

I don't suppose that this transfers to a digital signal in the same way! I can only think of two situations.

1) Where the signal goes too high it is simply and "cleanly" cut off. This MIGHT lead to a sound a bit similar to a "Compressor", just restricting the dynamic range - but only in the loud sections. This might be relatively inocuous and not cause DAMAGE to speakers - especially tweeters?

2) Where the actuall circuit elements thermselves are struglling and complaining and outputting "Non-Music" signals as a result. I imagine this kind might be just as destructive as "Analogue" Clipping?

3) Some other possiblity I haven't thought of...

Do you know which is the case with strictly "Digital" Clipping?

Regards Greg

Please Log in or Create an account to join the conversation.

Target - 0d and Below or Average above and Below? 2 years 7 months ago #47615

  • entripy
  • entripy's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 387
  • Thank you received: 130
The SHARC DSP used in the SHD has a clip instruction so the SHD likely saturates similar to analogue. Any decent circuit should be designed so that the digital extrema will be reproduced properly such that its impossible to make the ultimate analogue output circuitry clip. So 1) is possible, 2) shouldn't be possible

I suppose in general digital distortion is usually so bad that nobody goes there, you always try to maintain headroom. If you do get it, it is likely to produce higher harmonics and still risk damage to tweeters unless it is short duration. SHD has 32 bit DSP which is typically considered enough to easily maintain headroom whilst keeping signals large enough to avoid noise. So I say the SHD likely saturates because a typical system with no overall gain can't actually clip with a bounded input. Obviously as you add system gain you can potentially clip with very large input signals.

What you should be able to do is watch the plugin output meters which will give you some indication of whether you are running out of headroom and risking clipping.
The following user(s) said Thank You: Greg the Yeti

Please Log in or Create an account to join the conversation.

Target - 0d and Below or Average above and Below? 2 years 7 months ago #47616

  • Greg the Yeti
  • Greg the Yeti's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 22
  • Thank you received: 0
Hi Entripy, thanks for that reply.

To be clear, when I'm looking a the "Output Meters", there's so far nothing that goes above "0", not yet close. (I like my Dali Oberon's and don't want to damage them!)

Should I consider "0 db" the absolute amximum for "Headroom without Clipping"? Or is there clearance above that?

Regards Greg

Please Log in or Create an account to join the conversation.

Target - 0d and Below or Average above and Below? 2 years 7 months ago #47617

  • entripy
  • entripy's Avatar
  • Offline
  • Platinum Member
  • Platinum Member
  • Posts: 387
  • Thank you received: 130
0dB is absolute maximum for minidsp. This will only tell you the SHD isn't clipping but since you are trying to drive your speakers to a level you know they can handle that should be all you need to know.

Not getting near to 0dB seems common on SHD thanks to their high resolution and large dynamic range. I put that down to excellent system design. There should be no danger until you get close to 0dB but the meters are RMS which don't accurately show music levels or dynamics so you should allow a little headroom below 0dB. Maybe 10dB.

It will come down to listening tests ultimately but the meters should give you a big hint about whether adding however many dB output gain is viable.
The following user(s) said Thank You: Greg the Yeti

Please Log in or Create an account to join the conversation.

Target - 0d and Below or Average above and Below? 2 years 7 months ago #47618

  • Greg the Yeti
  • Greg the Yeti's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 22
  • Thank you received: 0
Yes, thanks Entripy. Exactly what I expected and wanted confirmed! Thanks.

I do miss the Meters that I had on my old Sony Cassette Player. It had LED meters that could simultaneously give the "Average" RMS vallue, but ALSO showed the peaks, which would hold for a few seconds and then dissapear(ready for any other peak). So I always KNEW the exact top peak value too. (it was their top of the range at that time - triple heads an'all!) Perhaps a possible improvement for the SHD. Since it's software dirven, it should be "possible" to implement (with a firmware update).

Regards Greg

Please Log in or Create an account to join the conversation.

Target - 0d and Below or Average above and Below? 2 years 7 months ago #47622

  • Greg the Yeti
  • Greg the Yeti's Avatar Topic Author
  • Offline
  • Junior Member
  • Junior Member
  • Posts: 22
  • Thank you received: 0

I don't suppose that this transfers to a digital signal in the same way! I can only think of two situations.

1) Where the signal goes too high it is simply and "cleanly" cut off. This MIGHT lead to a sound a bit similar to a "Compressor", just restricting the dynamic range - but only in the loud sections. This might be relatively inocuous and not cause DAMAGE to speakers - especially tweeters?

2) Where the actuall circuit elements thermselves are struglling and complaining and outputting "Non-Music" signals as a result. I imagine this kind might be just as destructive as "Analogue" Clipping?

3) Some other possiblity I haven't thought of...


I just thought of a problem with my analysis above - in #1, where I wondered if the sound might be Simply and "Cleanly" cut off. That was dangerous thinking!

I was thiking of the Sound being "Cut Off" as a bit like a "Flat Topped Mountain" - but it isn't. This is important, because the Clipping will take place at the Maximum Levels both ABOVE and BELOW zero. During all the time that it is clipping it will still be exactly like the clipping of analogue signals which generate a "Square Wave" for the duration. It is those square waves which are very dangerous for Tweeters! They are the "Sound Equivalent" of Sound with the Treble turned up MASSIVELY. The "YANK" from Maximum Below zero, almost instantly to Maximum above zero is a HUGE strain on a Tweeter, which handles all such fast transients! It CAN'T!

So when I said "This might be relatively inocuous" - PLEASE completely ignore that! It won't be inocuous! Trust me!

Another note while I'm here:-
I'm doing this testing of "Available Gain" with music from my collection which is mostly all very "Clean" sounds, i.e. no heavy Rock - since they already contain lots of Overdrive it will be harder to tell if distortion starts.

I've used "Grover Washington, Lighthouse Family and Madonna. It was also interesting that Madonna pushed the levels up much higher than the other two, with "Cherish" being the MAX! I listened to it at 0db on the SHD Main Gain Control. Still awesome!

Regards Greg

Please Log in or Create an account to join the conversation.

Moderators: devteam