Meridian Unplugged - Powered by MrTechGuy.co.uk Forums Discussion Meridian - Systems +3dB "LevelPerfect" Tweak Performance Improvement Poll

I checked that with a CD that have test sequences with 0dB, means full range.
When you go over +3dB you will hear the distortions,you hear the distortions with +3.5dB but not with +3,0dB in the config file.
See my post in the original thread.

Robert

I just did another test.

I wanted to A/B the tweak and no tweak instantly.

So the closest way to do that easily was

Set my mains +2.5dB (i.e. not LevelPerfect but still very close) and compare that in Direct Preset mode on G68 with Transporter on D2 (Direct Wired).

This way I don't have to worry about level differences messing with my head so much as it's only 0.5dB apart now.

Very close sounding !! I only tested a little (wifes having a fit).

So I'm thinking that the more you offset negatively (away from +3dB) the more harm it does. It doesn't appear to be all or nothing at +3dB as I first thought. Or who knows maybe it's when you cross 0 or something. Not sure.

I'm still leaving the tweak at +3dB in front though.

Thank you very much indeed. I had the MSF file in My Documents rather than the Meridian program folder.

I hope to try the +3dB tomorrow.

George #3

Finally had time to try this out, and my choice of words is similar to Nicks's. I checked with my SPL meter and am listening at the same SPL, but it does sounds slightly more dynamic. On Alison Krauss' Stay, there was more air, presence and "snap". I voted small improvement, rather than large, only because it wasn't like upgrading to the next speaker up but like the next version of that model.

Thanks George for this free upgrade

Hi George

I'm going to give this a try. I read your recent post out to my wife and she sends her sympathy to yours (lots of eye rolling! Might this explain why I thought the SB3 sounded better direct into the speakers rather than via G68?

Now that you have this new setting and HD621, have you noticed any differences with apodizing on/off?

Cheers
Tom

I've been thinking hard about this and I think I might understand what we are dealing with.

When you convert 16bit to 24bit you can't just "stretch" 16 bits into 24 or you'd push a bunch of significant resolution down to levels you won't hear.

16 bit and 24 but are on the SAME scale starting from the TOP bits. This is done for compatibility of 16, 20, 24 bit SPDIF. This is why low noise floor is critical to hear any improvement over 16bits. It's all dynamic range added to the bottom.

On the SqueezeBox and Transporter it has 24bit bit volume control but when you set it to "Max" it just uses the top 16bits. That's why DTS can still pass through. In other words the 16bits you started with are the MOST significant ones in 24 bit. Makes me also wonder if SqueezeBox properly flags it 16bit (when set to Max) and Meridian reads it as 16bits and not 24 (more testing some day).

When Meridian says Resolution Enhancement (16 to 24bit) their may be now some new data in the low bits (from dithering and processing) but they are still VERY low level bits (most insignificant ones). The top 16 most significant should be more or less be the same.

When you adjust volume digitally you are shifting down the MOST significant bits !! And pushing the least significant ones out (the ones Meridian added and is tossing as you lower volume digitally).

So I don't think it matters if it's been resolution enhanced or not. It's still effectively 16bits that your listening to !!

So I still think you are losing .5 bits out of (roughly) 16 bits (most significant ones) at 0dB offset and 1 full bit if you were at -3dB. I was at -3dB and now I'm at +3dB (I went from 15 bits to 16 bits).

This is why the test I did (a few posts back) makes sense.

I was comparing 15.83bits vs 16bits (hard to hear even instantly). That's not quite right but it isn't 23.83bits vs 24bits (which is what we are thinking).

But my original "Tweak", for me, was 15 vs 16bits (easy to hear even if not instant).

I abandoned this post at first, thinking, no, the DSP Speaker would slide the bits back up into the same position and you should hear no difference. That would be true if the DSP speaker implemented digital volume. But it does not. It does it in analog (and we want it to be analog). Meridian does some magic their with digital too (i.e. Hybrid) but for discussion purposes we should treat it as analog because it performs as well as analog and it is trying to maintain full resolution. The DSP speaker should effectively raise the volume in analog into the same human hearing range.

But DACs and ADCs are not perfect. They have a technical format width they can accept (16bit, 24bit etc.) and an effective performance (more like 12 bit`s for 16bit DAC and more like 18 bits for a 24bit DAC). Don`t want to argue that, but it`s less than the word spec size, they are not perfect. We have oscilloscopes with 8bit ADC`s that cost $23,000.00 and their effective performance is 5.5 bits !!! DACs are not perfectly linear either.

So the point is we don`t really receive 24 bits to our ears and what is super important is the most significant bits are more important than least (that`s why they are least significant). If we push things down by .5 or 1 bit, We`ve lost .5 or 1 bit of the best performing part of the DAC (the top).

So I`m back to my same conclusion that the offset should be applied in analog section of the DSP speaker.

In other words digital attentuation is not ok because it's 24-bit word size. You have to put into the context of effective DAC performance (which is less than 24-bits).

Let's say you pay $10K for an effective 20-bit DAC. And you pay $20K for an effective 21-Bit DAC. And you hear that improvement. Both DACs ACCEPT 24bit word size. Now you shift the data down 1 bit. Your $20K DAC just became a $10K DAC.

So Meridian is right 0dB offset is the ideal (as software allows). But it's really +3dB.

I have an analogue set up, so I was expecting to hear no difference.

Front 3 speakers were at -3dB, so for comparison, I added 6dB to all speakers and listened at indicated 69dB vs 75dB with prior. Source was iPod lossless apple files through Wadia dock - Radiance - HD621 - G68ADV. Listened to both in 'Direct'. Just compared a couple of tracks - Patricia Barber 'Mourning Grace' and Diana Krall 'Almost Blue'.

I preferred the -3dB original setting. I found the +3dB to be comparatively more harsh, although soundstage appeared a little taller. I really don't see why it should be any different - maybe I'm just hearing things.

Cheers, Ian

I agree I would not expect it to sound that different on analog.

It should not sound different on DSP or Analog !!

Nope. I would have just left HD621 in the music loop and gave Meridian the benefit of doubt. And perhaps over time I would learn the differences. But since it can't pass encoded music off my music server I don't use it for music unfortunately.

I've just done some listening tests on my all analogue set-up (568.1)

I did -3dB, 0dB and +3dB for the main L & R. (spl corrected)
I did this several times on the same tracks.
The differences a small to say the least, but overall if I had to put them in order it would be;

1st -3dB
2nd 3dB
3rd 0dB

The -3dB just sounds a little more open and individual sounds are easier to make out.

The differences aren't night and day by any means, very very subtle to the point of not being valid due to expectation bias.

This may not be an issue at all with a 568.1 using the analogue outputs.
But then again Trifield width still works on the 568.1...

Steve.

Last night I watched an entire movie with my analog speaker offsets set with the highest channel at Zero and the rest just below that. My lowest channel the sub is -3. My wife an I watched the entire movie that way. At the end during the credits I pressed the type button and it put back my original settings at -5 for my left and right and my center at -4.5. Rears are at -6db and sub at -7db.

What a difference! The sound of voices in a room or in a hall sounded much more realistic like I was there in the movie. Gun fire was way more dynamic and realistic sounding. The song at the end of the movie sounded much better. Palpable imaging went way up. You could get a feel as if the instruments were in the room with you. All my speakers set close to the zero setting sounded compressed in comparison.

I have done this experiment so many times and it's always a night and day difference for me. When sound is compressed the image is flatter and more like a wall of music that isn't realistic sounding. Some people do like that type of image. They listen for detail that I call hyped detail and if you go to a real performance it just doesn't sound that way. I like classical music and I have been to the Opera house in Sydney and let me tell you that place is amazing! I've been to plenty of concerts in other halls and I strive for that real music sound. Reducing the levels to -4 or -5 in my case these days gets me the sound I remember in these music halls.

Ron

George, I wonder if this is why you don't hear apodizing at all. ??