I set up a DEQ2496 recently to compensate for the baffle step and rising impedance of my FE127E bookshelf speakers (Fostex FE126E recommended enclosure).
Initially I felt it sounded harsh and then realised that with the +3dB of gain on some of the bands I needed to dial back the overall gain of the EQ by -3dB to prevent it from clipping.
A bit later on with some adjustments it was sounding quite good but a bit harsh again. Looking at the EQ I had the top two bands at +3.5dB just enough to cause clipping. So I pulled them back to +3.0dB and it was smooth again, but what was interesting it sounded leaner.
So I put the top two bands up to +3.5dB again and it sounded full bodied again.
I then decided to bring the entire EQ back -3.5dB and it sounded lean, so I can only conclude that the +3.5dB of high frequency gain isn't causing the full bodied sound but rather the clipping distortion.
Presuming it was just some unusual consequence of the DAC, I thought out of principle clipping wasn't good. However at a few hours on +3.0dB it just sounded lean, so I switched to +3.5dB and it was a completely different sounding setup.
On +3.5dB it sounds much more like my Electrovoice Interface C (yes).
It occurred to me that it may be possible that the distortion caused by the clipping could be having an effect on the OPT similar to Joe Rasmussen's Induced Permeability mod.
I was wondering whether anyone else has experienced this, or may have a DEQ2496 and can confirm?
The amp is a Yarland 34AIII single-ended pentode EL34 with 6dB negative feedback.
The input is 44.1kHz digital out of the CD player. On the DEQ2496 dithering is disabled as is the noise shaper.
Another thing I noticed is that with the +3.5dB clipping, the amp is much more listenable at low volumes.
Initially I felt it sounded harsh and then realised that with the +3dB of gain on some of the bands I needed to dial back the overall gain of the EQ by -3dB to prevent it from clipping.
A bit later on with some adjustments it was sounding quite good but a bit harsh again. Looking at the EQ I had the top two bands at +3.5dB just enough to cause clipping. So I pulled them back to +3.0dB and it was smooth again, but what was interesting it sounded leaner.
So I put the top two bands up to +3.5dB again and it sounded full bodied again.
I then decided to bring the entire EQ back -3.5dB and it sounded lean, so I can only conclude that the +3.5dB of high frequency gain isn't causing the full bodied sound but rather the clipping distortion.
Presuming it was just some unusual consequence of the DAC, I thought out of principle clipping wasn't good. However at a few hours on +3.0dB it just sounded lean, so I switched to +3.5dB and it was a completely different sounding setup.
On +3.5dB it sounds much more like my Electrovoice Interface C (yes).
It occurred to me that it may be possible that the distortion caused by the clipping could be having an effect on the OPT similar to Joe Rasmussen's Induced Permeability mod.
I was wondering whether anyone else has experienced this, or may have a DEQ2496 and can confirm?
The amp is a Yarland 34AIII single-ended pentode EL34 with 6dB negative feedback.
The input is 44.1kHz digital out of the CD player. On the DEQ2496 dithering is disabled as is the noise shaper.
Another thing I noticed is that with the +3.5dB clipping, the amp is much more listenable at low volumes.
Moved to Digital Line LevelHmm.. starting to think this just could be my amp.
Yesterday it was definitely a clear difference when the DEQ was clipping or not. However today it just sounds good no matter what setting is on even in bypass all mode.
My amp does have its good and bad days. I'm wonder whether HF clipping helps my amp when it is not doing so well? I wonder what component would be helped by a HF signal?
Yesterday it was definitely a clear difference when the DEQ was clipping or not. However today it just sounds good no matter what setting is on even in bypass all mode.
My amp does have its good and bad days. I'm wonder whether HF clipping helps my amp when it is not doing so well? I wonder what component would be helped by a HF signal?
With the high frequencies set to +3.5dB and that being only +0.5dB over the overall gain of -3dB it is hard to see the clipping. However it does have a more edgy sound.
In terms of seeing the clipping, when I first set it up and had some levels of +3dB with an overall gain of 0dB the red LEDs on the level indicators were flashing fairly regularly. The sound was also more edgy.
So I'm presuming that what I'm hearing is caused by clipping but I can't say for certain.
In the past when my amp doesn't sound at its best the male voice seems a bit receded whereas when it is doing better it sounds fuller and more rich. This is the first time I've been able to affect this aspect of the amp.
So I'm beginning to think this is just a problem with my amp.
The fact that what I'm seeing is intermittent I'm guessing it has nothing to do with OPT permeability as I had first thought. Perhaps it's a faulty capacitor?
On a side note this has given me an interesting idea of whether Joe Rasmussen's induced permeability could be applied from a software perspective? The signal could be upsampled to 24 bits then halved, and then a high frequency signal added before sending to the DAC.
In terms of seeing the clipping, when I first set it up and had some levels of +3dB with an overall gain of 0dB the red LEDs on the level indicators were flashing fairly regularly. The sound was also more edgy.
So I'm presuming that what I'm hearing is caused by clipping but I can't say for certain.
In the past when my amp doesn't sound at its best the male voice seems a bit receded whereas when it is doing better it sounds fuller and more rich. This is the first time I've been able to affect this aspect of the amp.
So I'm beginning to think this is just a problem with my amp.
The fact that what I'm seeing is intermittent I'm guessing it has nothing to do with OPT permeability as I had first thought. Perhaps it's a faulty capacitor?
On a side note this has given me an interesting idea of whether Joe Rasmussen's induced permeability could be applied from a software perspective? The signal could be upsampled to 24 bits then halved, and then a high frequency signal added before sending to the DAC.
Because you're using a direct digital input from a CD player you have to be very careful to avoid clipping. When you use the analogue inputs (as I do) you tend to have 10-15 dB of headroom (especially if the +22dBu switch is set on the rear) so you can generally boost without clipping.
However with a direct digital feed from your CD player, a very loud compressed CD can easily hit almost maximum 16 bit levels, which means if you try to boost beyond this, you'll inevitably get some clipping.
In other words if the recording on the CD was reaching -3dB FS (full scale) and you applied 4dB of boost to the bass, it's very likely you'll get clipping.
The solution, if you're going to stick with digital input (and I presume analogue output ?) is to either only cut rather than boost (cut the high frequencies rather than boost the low frequencies) or turn down the Gain Offset in Utility page 1 by an appropriate amount.
For example if you were applying 6dB of baffle step boost in the bass using a Parametric EQ shelf, set the Gain Offset to at least -7dB or -8dB so that the maximum possible output level is still below clipping. (Remember that different features of the DEQ are cumulative - so if you used both PEQ and GEQ the total gain/boost can be more than each on their own - take that into consideration when setting Gain Offset)
To avoid the overall levels being too low, make sure the switch on the back is set to +22dBu, and also make sure that your XLR to Phono cables are wired correctly with pin 1 linked to pin 3 to make sure unbalanced mode is activated on the DEQ, otherwise you'll see a 6dB loss in levels and may also have problems with noise and hum. (See the owners manual. Most pre-made XLR to Phono "Balanced to Unbalanced" cables do not have these pins linked)
I would also strongly suggest that you enable 16 bit dither (and noise shaping) when using a 16 bit digital input and analogue output. If you have dither set to none in this circumstance you may get digital artefacts on low level signals - a sort of gritty grainy sound. Generally the more dither you apply, the more "analogue" a digital device will sound, and the dither should never be less (greater number of bits) than the source material.
For example I use 24 bit dither (less dither than 16 bit) but I use the analogue inputs which are sampled at 24 bit, so that's ok. With 24 bit sampling I could choose to use 20 or 16 bit dither which would (in theory) give reduced digital artefacts, but increased noise floor, however when your source samples are only 16 bit as in your case with direct digital input from a CD player, 16 bit dither is the only correct choice for dither that won't result in a loss of quality and possible digital artefacts.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.